Compare commits

...
Author SHA1 Message Date
Tyke ChenandGitHub 5c7765506c Merge pull request #3295 from xinnan-tech/platform/m0-development-plan
docs: define unified FastAPI delivery plan
2026-07-20 17:07:43 +08:00
Tyke Chen 672e51e68e docs: define unified FastAPI delivery plan 2026-07-20 17:03:11 +08:00
Tyke Chen 804ddb51f2 feat: add FastAPI manager API compatibility baseline 2026-07-20 17:00:13 +08:00
CGDandGitHub 7c58fa37b2 Merge pull request #3293 from xinnan-tech/refactor/mybatis-plus-single-source
refactor(manager-api): consolidate MyBatis-Plus 3.5.17 integration
2026-07-20 14:42:31 +08:00
Tyke Chen 10a67c772f refactor(manager-api): consolidate mybatis-plus 3.5.17 integration 2026-07-20 10:31:08 +08:00
CGDandGitHub 7e8e5d34b8 Merge pull request #3279 from xinnan-tech/fix/issue-3276-context-provider
fix: 修复 ContextProviderListTypeHandler 多版本兼容问题
2026-07-17 09:28:14 +08:00
CGDandGitHub c6bc20c05e Merge pull request #3283 from xinnan-tech/fix/device-timezone-consistency
fix: 修复跨时区场景下设备时间与在线状态异常
2026-07-16 16:58:49 +08:00
wengzhandGitHub 1998807b69 Merge pull request #3285 from xinnan-tech/py-aec
perf: 优化相关逻辑减少重复计算
2026-07-14 16:36:16 +08:00
Sakura-RanChen 55c0495661 perf: 优化相关逻辑减少重复计算 2026-07-14 15:52:30 +08:00
Tyke Chen dbeda01697 fix: 修复设备跨时区时间与在线状态 2026-07-14 10:08:23 +08:00
Tyke Chen 023dea2441 fix: 修复上下文源类型处理器兼容问题 2026-07-13 16:03:39 +08:00
CGDandGitHub ad3fb4d8c8 Merge pull request #3278 from xinnan-tech/agent-snapshot-history-fixes
fix: 完善智能体配置历史与恢复逻辑
2026-07-13 15:47:08 +08:00
3030332422 176785830a update:修改部分样式 2026-07-13 15:44:06 +08:00
Tyke Chen d78ecfce6f Merge remote-tracking branch 'origin/main' into agent-snapshot-history-fixes 2026-07-10 20:26:30 +08:00
Tyke Chen 179281e49c fix: harden agent snapshot restore flow 2026-07-10 20:21:44 +08:00
hrzandGitHub bc1aa1a089 Merge pull request #3267 from xinnan-tech/fix-agent-idor
fix: prevent agent IDOR access
2026-07-10 17:36:00 +08:00
wengzhandGitHub c5b64c3ef0 Merge pull request #3270 from xinnan-tech/fix-voiceprint-dialogue-bug
fix: 修改提示词与说话人注入逻辑,减少模型回答带着说话人信息频率
2026-07-10 15:55:58 +08:00
DaGou12138 2aa6c5ee49 fix:
1.优化context注入方式,合并为一个system,解决qwen本地部署模型多system异常400问题
2026-07-10 15:05:02 +08:00
Tyke Chen 687b6db96b fix: resolve mobile type and lint errors 2026-07-10 11:45:16 +08:00
Tyke Chen a5aee109fe feat: improve agent snapshot history 2026-07-10 11:44:35 +08:00
hrzandGitHub 8ec5851027 Merge pull request #3273 from xinnan-tech/agent-delete-name-confirm
feat: 删除智能体前需要用户进行确认
2026-07-10 10:39:47 +08:00
Tyke Chen 48d629c79c fix: localize agent deletion success message 2026-07-10 10:18:48 +08:00
wengzh 14e57a1add fix(powermem): 修复用户配置启用状态的类型判断问题
将配置的启用值统一转为小写字符串后判断是否为true,避免非布尔类型配置导致的判断失效
2026-07-10 10:06:07 +08:00
hrzandGitHub 634967b2fe Merge pull request #3274 from xinnan-tech/i18n-duplicate-check
test: 增加 i18n 重复 key 检测
2026-07-10 09:59:33 +08:00
Tyke Chen e648453c67 fix: rename i18n check command 2026-07-10 09:55:53 +08:00
Tyke Chen d11cfc6923 Merge remote-tracking branch 'origin/main' into fix-agent-idor
# Conflicts:
#	main/manager-api/src/main/java/xiaozhi/modules/agent/controller/AgentController.java
#	main/manager-api/src/main/java/xiaozhi/modules/agent/service/AgentService.java
#	main/manager-api/src/main/java/xiaozhi/modules/agent/service/impl/AgentServiceImpl.java
2026-07-09 20:21:37 +08:00
DaGou12138 32fbbf92fe Merge branch 'refs/heads/main' into fix-voiceprint-dialogue-bug 2026-07-09 16:56:22 +08:00
wengzhandGitHub bd981aac79 Merge pull request #3275 from xinnan-tech/py-audio-aec
feat: 服务端AEC功能
2026-07-09 16:24:03 +08:00
Sakura-RanChenandGitHub adee070bb0 Merge branch 'main' into py-audio-aec 2026-07-09 15:51:01 +08:00
wengzhandGitHub 03ec3bac3f Merge pull request #3271 from xinnan-tech/py-fix-bug
Py fix bug
2026-07-09 15:36:59 +08:00
Tyke Chen af5fbb5282 test: add i18n duplicate key check 2026-07-09 15:25:36 +08:00
Tyke Chen 9301c1889f feat: require agent name confirmation before delete 2026-07-09 15:03:12 +08:00
hrzandGitHub ef2ed07c1c Merge pull request #3266 from xinnan-tech/agent-snapshot-history
Add agent configuration history
2026-07-09 14:54:27 +08:00
Tyke Chen 1e2c4c6a28 fix: 修复智能体快照基线与查看逻辑 2026-07-09 11:16:07 +08:00
Sakura-RanChen a32500a6a4 fix: 消除死锁隐患 2026-07-09 10:29:41 +08:00
hrz 772af6c918 Merge branch 'agent-snapshot-history' of https://github.com/xinnan-tech/xiaozhi-esp32-server into agent-snapshot-history 2026-07-08 17:46:50 +08:00
hrz 2a917c1e62 update:对齐i18n文件行数 2026-07-08 17:44:10 +08:00
Tyke Chen c32fbe69c5 Remove wakeup word audio asset 2026-07-08 16:33:36 +08:00
DaGou12138 3b6e8f0e4b fix:
1.优化声纹称呼人提示词和注入逻辑,减少次次都称呼的频率
2026-07-08 16:14:10 +08:00
Tyke Chen aae2ef2152 fix: prevent agent IDOR access 2026-07-08 15:52:47 +08:00
Tyke Chen c82b351457 Fix agent snapshot history issues 2026-07-08 15:15:26 +08:00
DaGou12138 04c10ef225 fix:
1.修改提示词与说话人注入,减少模型每次回答都带着说话人信息
2026-07-08 14:12:16 +08:00
Sakura-RanChen 0b8bb728e2 同步方法进行包装避免阻塞 2026-07-08 11:47:04 +08:00
mie 35ec3ba803 Add agent configuration history 2026-07-07 17:49:34 +08:00
wengzhandGitHub 8808b69303 Merge pull request #3265 from xinnan-tech/ha-muisc-timeout
Ha muisc timeout
2026-07-07 16:25:12 +08:00
Sakura-RanChen fc573a72ea fix: 修复设备时间戳上报逻辑 2026-07-07 15:50:27 +08:00
Sakura-RanChen 59f6566cfb 同步工具异步化处理 2026-07-07 14:45:50 +08:00
Sakura-RanChen d8c97a9f81 fix: 修复携程死锁 2026-07-06 16:44:41 +08:00
Sakura-RanChen 3d204fb181 feat: 服务端AEC功能
ref: 入口解码为pcm音频数据,vad和asr可直接使用
2026-07-06 16:21:24 +08:00
hrzandGitHub 209fd3fbc8 Merge pull request #3260 from xinnan-tech/test-funasr
FunASR识别指定语言
2026-07-06 13:56:48 +08:00
3030332422 bc0ffd1bd3 update:修改sql文件,删去测试文件 2026-07-01 15:02:26 +08:00
hrzandGitHub f0d61442b8 Merge pull request #3259 from xinnan-tech/update-theme
顶部下拉菜单图标间距调整
2026-06-30 18:45:14 +08:00
CGDandGitHub e6c2fac7ec Merge pull request #3255 from LauraGPT/funasr-language-option
fix: allow configuring FunASR recognition language
2026-06-30 14:57:31 +08:00
zhifu gao 5a9b5af441 fix: add FunASR language config migration 2026-06-30 14:07:16 +08:00
欣南科技andGitHub 16e5dc8c78 Merge pull request #3258 from xinnan-tech/update-version
update:Bump to 0.9.5
2026-06-30 13:10:18 +08:00
hrz f77970dd7d update:Bump to 0.9.5 2026-06-30 13:10:26 +08:00
zhuoqinglian af3ca4f609 update: 顶部下拉菜单图标间距调整 2026-06-30 09:47:28 +08:00
hrzandGitHub 1ebfcd1f13 Merge pull request #3256 from xinnan-tech/update-theme
解决智能体管理查询不展示的问题&删除知识库旧版废弃文件
2026-06-30 09:36:42 +08:00
zhuoqinglian 3ca322961e fix: 解决智能体管理查询不展示的问题 2026-06-30 09:22:05 +08:00
zhuoqinglian 0eb0a600a7 update: 删除知识库旧版废弃文件 2026-06-30 09:19:26 +08:00
wengzhandGitHub 3ed06a9e34 Merge pull request #3244 from xinnan-tech/base_prompt-update
feat: 将系统提示词指令逻辑英文化,中文样本保留
2026-06-30 09:06:10 +08:00
LauraGPT 2fae96d60c fix: allow configuring FunASR recognition language 2026-06-29 20:58:51 +00:00
wengzhandGitHub 67dde7cc47 Merge pull request #3254 from xinnan-tech/update-theme
智能体管理搜索恢复查询历史功能
2026-06-29 16:58:52 +08:00
zhuoqinglian efedc18f02 update: 智能体管理搜索恢复查询历史功能 2026-06-29 16:01:14 +08:00
zhuoqinglian 0270de1844 update: 滚动条样式抽离复用 2026-06-29 15:42:52 +08:00
wengzhandGitHub 06b0ff139b Merge pull request #3247 from xinnan-tech/update-theme
Update theme
2026-06-26 17:58:18 +08:00
zhuoqinglian 799ea28d82 Merge branch 'main' into update-theme 2026-06-26 17:30:15 +08:00
zhuoqinglian 5a68252d22 update: 更新所有页面背景颜色 2026-06-26 17:28:06 +08:00
zhuoqinglian 97404e53fe update: 顶部导航栏搜索放到首页 2026-06-26 17:26:16 +08:00
zhuoqinglian b052158cee feat: 完成导航菜单组件样式优化 2026-06-26 17:10:49 +08:00
zhuoqinglian 979ef24b79 update: 音色克隆多语言配置 2026-06-26 15:32:48 +08:00
zhuoqinglian 0f4f0f01fd update: 音色克隆页面展现形式修改 2026-06-26 14:45:23 +08:00
wengzhandGitHub 18e1de4f45 Merge pull request #3245 from xinnan-tech/py-add-minimax-voice
Py add minimax voice
2026-06-26 11:42:59 +08:00
Sakura-RanChen aa846a096f 名称长度调整 2026-06-26 11:41:46 +08:00
Sakura-RanChen a679412230 增加试听音色 2026-06-26 11:40:52 +08:00
3030332422 cf3e14bb07 feat: 将系统提示词指令逻辑英文化,中文样本保留 2026-06-26 11:06:18 +08:00
wengzhandGitHub 3a73a17253 Merge pull request #3243 from xinnan-tech/py-add-minnimax-voice
调整minimax音色
2026-06-26 10:50:26 +08:00
Sakura-RanChen 3b27ebcd35 调整minimax音色 2026-06-26 10:29:07 +08:00
wengzhandGitHub fa6074d284 Merge pull request #3242 from xinnan-tech/update-theme
Update theme
2026-06-25 17:29:55 +08:00
zhuoqinglian d809cfcb01 Merge branch 'main' into update-theme 2026-06-25 17:27:40 +08:00
zhuoqinglian eb2c09bbdc feat: 完成聊天记录、修改密码页面样式改造 2026-06-25 17:26:52 +08:00
zhuoqinglian 2cf8f79482 update: 模型配置新增、编辑弹窗样式优化 2026-06-25 17:09:29 +08:00
zhuoqinglian 1c94844bb9 feat: 完成声纹管理页面样式改造 2026-06-25 09:18:38 +08:00
wengzhandGitHub 0c53b8f543 Merge pull request #3240 from xinnan-tech/update-theme
Update theme
2026-06-24 17:58:13 +08:00
zhuoqinglian 1518dfca30 feat: 完成设备管理页面样式改造 2026-06-24 17:37:37 +08:00
zhuoqinglian 7529c500a4 feat: 完成音色克隆页面样式改造 2026-06-24 11:43:36 +08:00
zhuoqinglian b6ed220e48 feat: 完成音色资源页面样式改造 2026-06-23 17:23:11 +08:00
wengzhandGitHub b152c5dc23 Merge pull request #3239 from xinnan-tech/update-theme
模型配置页面样式调整
2026-06-23 16:31:30 +08:00
zhuoqinglian 04bfddf457 update: 按钮样式风格统一 2026-06-23 14:54:45 +08:00
zhuoqinglian 9d78ea29c3 Merge branch 'main' into update-theme 2026-06-23 14:15:57 +08:00
zhuoqinglian f82f364189 feat: 完成模型配置页面样式改造 2026-06-23 14:15:32 +08:00
zhuoqinglian 9705149b2b feat: 按钮组件新增add类型 2026-06-23 14:14:29 +08:00
wengzhandGitHub 7986148013 Merge pull request #3238 from xinnan-tech/py-add-edge-config
补充edge相关配置
2026-06-23 10:40:40 +08:00
Sakura-RanChen 19f5cb5b4f 补充edge相关配置 2026-06-23 10:37:48 +08:00
zhuoqinglian ea48b948bf update: 模型配置新增弹窗样式调整 2026-06-22 17:58:35 +08:00
zhuoqinglian 3c5ca9f143 update: 模型配置编辑弹窗样式调整 2026-06-22 17:31:09 +08:00
wengzhandGitHub da72ecaa1b Merge pull request #3235 from xinnan-tech/py-fix-tts
fix: 超时情况不与activate_session同一判断条件导致的监听任务未清除问题
2026-06-22 11:21:03 +08:00
Sakura-RanChen f297e1b648 fix: 超时情况不与activate_session同一判断条件导致的监听任务未清除问题 2026-06-22 11:16:25 +08:00
wengzhandGitHub a1973e07b7 Merge pull request #3227 from xinnan-tech/update-theme
feat: 完成字典管理页面样式改造
2026-06-18 16:57:49 +08:00
zhuoqinglian cd58b2c3cb Merge branch 'main' into update-theme 2026-06-18 16:55:33 +08:00
zhuoqinglian a5be67d04b feat: 完成字典管理页面样式改造 2026-06-18 16:49:15 +08:00
wengzhandGitHub e6dc6187c8 Merge pull request #3226 from xinnan-tech/update-theme
页面样式改造
2026-06-18 16:39:41 +08:00
zhuoqinglian 7f4011f9de feat: 完成字段管理页面样式改造 2026-06-18 16:37:42 +08:00
zhuoqinglian 211cf891af Merge branch 'main' into update-theme 2026-06-18 10:26:57 +08:00
zhuoqinglian 89a1229b53 feat: 完成服务端管理页面样式改造 2026-06-18 10:26:06 +08:00
wengzhandGitHub 071ed99a32 Merge pull request #3224 from xinnan-tech/py-fix-bug
fix: 通讯录页面时间不显示
2026-06-17 14:36:48 +08:00
Sakura-RanChen ea14be27e6 fix: 通讯录页面时间不显示
fix: listen消息相关状态处理
2026-06-17 14:34:42 +08:00
zhuoqinglian a834f163b7 feat: 完成替换词页面样式改造 2026-06-17 14:20:48 +08:00
zhuoqinglian 7849e6b030 update: 模板保存添加必填校验 2026-06-17 10:36:02 +08:00
wengzhandGitHub 0d869baaa0 Merge pull request #3223 from xinnan-tech/update-theme
页面样式调整
2026-06-16 17:01:10 +08:00
zhuoqinglian 164eb5481c feat: 完成默认角色管理页面样式改造,并添加编辑保存弹窗 2026-06-16 16:58:44 +08:00
zhuoqinglian 969ee3cb12 feat: 完成字段管理页面样式改造 2026-06-16 16:18:12 +08:00
zhuoqinglian ee54d87383 feat: 完成固件管理页面样式改造 2026-06-16 14:55:08 +08:00
wengzhandGitHub 849026c015 Merge pull request #3222 from xinnan-tech/update-theme
Update theme
2026-06-16 10:42:24 +08:00
zhuoqinglian e5bb3e7791 update: 优化表格组件,增加代码的复用性 2026-06-16 10:40:44 +08:00
zhuoqinglian 1ffc40e71d feat: 完成用户管理页面样式改造 2026-06-16 10:39:52 +08:00
zhuoqinglian 45fd850757 update: 表格组件添加阴影效果 2026-06-16 09:46:25 +08:00
zhuoqinglian 8a5dac181e update: 知识库顶部标题合并 2026-06-15 17:37:53 +08:00
zhuoqinglian ecf20300ff update: 通讯录初始化加载默认选中第一项 2026-06-15 17:16:50 +08:00
zhuoqinglian 74def844f0 Merge branch 'main' into update-theme 2026-06-15 17:05:28 +08:00
zhuoqinglian a5a9bd1abe fix: 修复element组件多语言切换不生效的问题 2026-06-15 17:05:04 +08:00
zhuoqinglian 2d9b915d51 feat: 完成参数管理页面改造 2026-06-15 15:44:09 +08:00
hrzandGitHub 5f87fc8361 Merge pull request #3204 from xinnan-tech/py-fix-calling
优化相关逻辑
2026-06-15 14:06:51 +08:00
zhuoqinglian 6859fdc90e update: 系统功能配置图标大小优化 2026-06-15 11:27:12 +08:00
zhuoqinglian b8ae0568e9 update: 系统功能配置页面适配调整 2026-06-15 11:23:31 +08:00
wengzhandGitHub 1dc6f143f3 Merge pull request #3219 from xinnan-tech/update-theme
完成系统功能页面样式改造
2026-06-12 14:43:23 +08:00
zhuoqinglian 67532dc230 update: 系统功能配置完善多语言 2026-06-12 14:23:25 +08:00
zhuoqinglian b2876def99 Merge branch 'main' into update-theme 2026-06-12 14:14:59 +08:00
zhuoqinglian 0c73a8b6a4 feat: 完成系统功能页面样式改造 2026-06-12 14:13:21 +08:00
zhuoqinglian c857a3164d feat: 完成按钮组件抽离 2026-06-12 11:49:22 +08:00
hrzandGitHub a964192eb0 Merge pull request #3213 from xinnan-tech/fix-knowledge-base-ragflow-bug
#3212 修复知识库RAGFlow相关逻辑BUG
2026-06-12 11:15:35 +08:00
DaGou12138 32da03818e #3212 fix:
1.修改知识库前端卡片提示样式,优化警示效果
2026-06-12 10:35:40 +08:00
DaGou12138 7f17e3e08e Merge branch 'refs/heads/main' into fix-knowledge-base-ragflow-bug 2026-06-12 10:29:17 +08:00
DaGou12138 c237e1f323 Merge remote-tracking branch 'origin/fix-knowledge-base-ragflow-bug' into fix-knowledge-base-ragflow-bug 2026-06-12 09:51:03 +08:00
DaGou12138 087a202e31 #3212 fix:
1.优化知识库连接RAGFlow异常提示,修改前端卡片提示样式
2026-06-12 09:50:01 +08:00
hrz 4b1460d938 update:调整同步数据集错误日志的级别 2026-06-11 19:10:21 +08:00
hrz f2aec524d2 del:无用类 2026-06-11 18:57:30 +08:00
wengzhandGitHub 1a97ca137c Merge pull request #3215 from xinnan-tech/update-theme
Update theme
2026-06-11 17:45:58 +08:00
zhuoqinglian 2b6f457ae6 update: 知识库页面间距优化 2026-06-11 17:44:17 +08:00
DaGou12138 3f4e0b4aa0 #3212 fix:
1.修复RAGFlow连接异常导致级联删除问题
2.修复RAGFlow获取数量逻辑问题
3.远程失败不再删除本地影子记录,避免产生RAGFlow端孤儿文档
4.优化前端提示语
2026-06-11 17:17:59 +08:00
DaGou12138 e30ded4de9 fix:删除知识库rag相关非必要实体 2026-06-11 16:08:17 +08:00
hrzandGitHub 07242a84ff Merge pull request #3210 from xinnan-tech/fix-knowledge-base-web-bug
fix:修复知识库模块弹框组件样式全局覆盖导致显示异常问题
2026-06-10 15:32:08 +08:00
DaGou12138 a2074dc0d8 fix:修复知识库模块弹框组件样式全局覆盖导致显示异常问题 2026-06-10 15:00:29 +08:00
zhuoqinglian 28b6c85650 fix: 修复弹窗展示位置异常的问题 2026-06-10 10:47:52 +08:00
wengzhandGitHub af300138eb Merge pull request #3209 from xinnan-tech/update-theme
Update theme
2026-06-09 17:24:16 +08:00
zhuoqinglian 2bf5725048 update: 知识库页面复用弹窗、分页组件,同步修改样式 2026-06-09 17:22:18 +08:00
hrzandGitHub 3d73141788 Merge pull request #3206 from xinnan-tech/web_search-Document
docs:新增联网搜索插件配置文档
2026-06-09 17:12:16 +08:00
hrz 945ab29bc3 add:增加联网搜索的教程和自编译docker镜像的教程 2026-06-09 17:11:35 +08:00
zhuoqinglian fe0f7dcf75 feat: 完成分页组件抽离 2026-06-09 15:05:16 +08:00
zhuoqinglian 1a3ffbc8e2 feat: 完成弹窗组件抽离 2026-06-09 14:25:49 +08:00
3030332422 71278b1898 docs:新增联网搜索插件配置文档 2026-06-08 11:38:11 +08:00
zhuoqinglian 0173c7abe1 update: 知识库样式优化、添加多语言配置 2026-06-05 17:37:21 +08:00
zhuoqinglian 8a50eea4ce update: 知识库整体页面样式调整 2026-06-03 16:49:08 +08:00
Sakura-RanChen b9891bb795 优化相关逻辑 2026-06-03 16:13:03 +08:00
wengzhandGitHub b903a88cd5 Merge pull request #3203 from xinnan-tech/mac-address-mask
feat: mac地址添加马赛克功能
2026-06-03 14:41:34 +08:00
zhuoqinglian f3fcf2fedf feat: mac地址添加马赛克功能 2026-06-03 14:40:01 +08:00
欣南科技andGitHub a825f0874c Merge pull request #3202 from xinnan-tech/update-doc-video
update:更新演示视频
2026-06-03 13:58:14 +08:00
hrz 8a5ca46418 update:更新演示视频 2026-06-03 13:56:34 +08:00
欣南科技andGitHub 5fcfc529c0 Merge pull request #3201 from xinnan-tech/0.9.4
Bump to 0.9.4
2026-06-03 09:08:43 +08:00
hrz 370d148a4e Bump to 0.9.4 2026-06-03 09:07:39 +08:00
欣南科技andGitHub 3088ddce38 Merge pull request #3200 from xinnan-tech/update_doc
update:补充设备间通话开启备注
2026-06-02 18:23:49 +08:00
hrz 8e3faf20d3 update:补充设备间通话开启备注 2026-06-02 18:22:31 +08:00
欣南科技andGitHub f42088eff4 Merge pull request #3199 from xinnan-tech/py-device-calling
update:如果通讯权限 alias 为空,就用默认使用设备名称
2026-06-01 23:53:29 +08:00
hrz a0181cc723 update:如果通讯权限 alias 为空,就用默认使用设备名称 2026-06-01 23:52:38 +08:00
欣南科技andGitHub 941aaf10ce Merge pull request #3198 from xinnan-tech/py-device-calling
fix:用户别名缺失bug
2026-06-01 22:57:24 +08:00
hrz 235ec8a25d update:更换通讯录图标 2026-06-01 22:55:03 +08:00
hrz effe95ba8f fix:用户别名缺失bug 2026-06-01 22:54:34 +08:00
CGDandGitHub 4c4ba0aa08 Merge pull request #3196 from xinnan-tech/py-device-calling
Py device calling
2026-06-01 16:34:02 +08:00
3030332422 7af4b41881 update:设备呼叫工具的接听触发提示词约束 2026-06-01 16:25:52 +08:00
hrz cfcf68d4ce Merge branch 'py-device-calling' of https://github.com/xinnan-tech/xiaozhi-esp32-server into py-device-calling 2026-05-31 10:30:54 +08:00
hrz 7f2675f9df update:增加设备呼叫使用条件 2026-05-31 10:30:10 +08:00
Sakura-RanChen 69fd5a1ea8 不开启通讯录功能时不显示相关工具
关闭通讯录功能清除相关状态
2026-05-29 17:34:21 +08:00
Sakura-RanChen 4392edf27d 调整相关说明文档 2026-05-29 15:56:52 +08:00
zhuoqinglian e4ea9c8eff Merge branch 'style-device-calling' into py-device-calling 2026-05-29 14:37:55 +08:00
zhuoqinglian 2adf51ddf9 fix: 修复顶部导航栏遮挡的问题 2026-05-29 14:36:14 +08:00
CGDandGitHub 3d1b133162 Merge pull request #3195 from xinnan-tech/fix-timezone-unify
fix:修复时区导致的设备状态不同步的问题
2026-05-29 10:29:02 +08:00
DaGou12138andGitHub f539d46de3 Merge pull request #3194 from xinnan-tech/fix(model-config)
fix(model-config): 修复合并JSON时的类型异常并添加日志
2026-05-29 10:10:21 +08:00
baigao 1701bfbfca fix(model-config): 修复合并JSON时的类型异常并添加日志
1. 为mergeJson方法添加空值检查和类型校验逻辑
2. 增加日志打印便于排查问题
3. 优化代码换行和注释格式
4. 修复原有代码中潜在的类型转换异常问题
2026-05-29 10:07:11 +08:00
zhuoqinglian bb48e7efab Merge branch 'style-device-calling' into py-device-calling 2026-05-29 10:05:24 +08:00
zhuoqinglian 9ea67d4306 update: 通讯录页面样式调整 2026-05-29 10:04:54 +08:00
rainv123 3a0ce6998e fix:修复设备型号不显示的问题 2026-05-29 09:50:09 +08:00
hrz 0678521b16 update:优化addressBook配置的的存储方式 2026-05-28 19:36:26 +08:00
hrz 84187d4e68 update:优化翻译 2026-05-28 15:18:54 +08:00
hrz 38958b2d06 add:通讯录i18n行数对齐 2026-05-28 15:02:20 +08:00
hrz 1e17e5ac8a fixed:当数据库的就数据不支持新特性时,增加新特性 2026-05-28 14:41:06 +08:00
rainv123 ec5063f050 fix: 设备最后连接时间返回 UTC 时间戳,修复时区显示问题 2026-05-27 17:35:10 +08:00
Sakura-RanChen d03583f493 增加工具配置说明 2026-05-27 17:28:58 +08:00
Sakura-RanChen 4a55662e91 fix: 修复智能体在线状态获取
补充相关国际化
2026-05-27 16:57:48 +08:00
Sakura-RanChen de78e87c77 调整逻辑使用java端作为中转发送 2026-05-27 15:34:17 +08:00
wengzhandGitHub ff011d3c44 Merge pull request #3192 from xinnan-tech/fix_rag_page_size
fix:修复知识库文档切片的分页显示问题
2026-05-27 10:30:13 +08:00
rainv123 a90229a22a fix:修复知识库文档切片的分页显示问题 2026-05-27 10:26:50 +08:00
wengzhandGitHub 559d47a229 Merge pull request #3189 from xinnan-tech/digital-human_prompt
update:根据数字人 emoji 开关动态渲染系统提示词
2026-05-26 16:38:46 +08:00
Sakura-RanChen ce68c83eea feat: 新增设备互相通讯功能
feat: 新增通讯录页面
2026-05-26 16:37:39 +08:00
3030332422 7e63b7e852 update:根据数字人 emoji 开关动态渲染系统提示词 2026-05-26 11:43:18 +08:00
wengzhandGitHub 7f73dae1bf Merge pull request #3183 from xinnan-tech/digital-human_fix_emoji
fix:情绪表情默认改为false
2026-05-20 17:49:02 +08:00
3030332422 48e8839a1e fix:情绪表情默认改为false 2026-05-20 17:43:55 +08:00
wengzhandGitHub 73f20ed9b0 Merge pull request #3182 from xinnan-tech/perf-tool-add-online-search
fix:修复deepseek模型下工具名重复导致错误问题
2026-05-20 15:03:26 +08:00
DaGou12138 48a2bd5a5a fix:修复deepseek模型下工具名重复导致错误问题 2026-05-20 14:52:21 +08:00
wengzhandGitHub fb7e2e17a8 Merge pull request #3176 from xinnan-tech/perf-tool-add-online-search
Perf:新增联网搜索工具类、修复新闻描述配置获取源问题
2026-05-14 10:51:23 +08:00
欣南科技andGitHub bcce6a4ebe Merge pull request #3174 from xinnan-tech/faq-digital-human
add:faq增加数字人相关链接
2026-05-13 17:24:01 +08:00
hrz 7ef86eb3c6 add:faq增加数字人相关链接 2026-05-13 17:20:21 +08:00
wengzhandGitHub 789b207a94 Merge pull request #3173 from xinnan-tech/chore-move-multilingual-readmes-to-docs
chore:迁移多语言 README 到 docs/readme
2026-05-12 15:20:06 +08:00
DaGou12138 db14d08c7b perf:
1、新增联网搜索工具类(web_search)
2、修复新闻描述配置获取源问题,改为从连接配置获取
3、新增web_search所需配置项数据集
2026-05-12 15:16:31 +08:00
DaGou12138 46e7f8508d perf:
1、新增联网搜索工具类(web_search)
2、修复新闻描述配置获取源问题,改为从连接配置获取
3、新增web_search所需配置项数据集
2026-05-12 15:16:08 +08:00
3030332422 f3583742c7 chore:迁移多语言 README 到 docs/readme 2026-05-12 15:15:08 +08:00
wengzhandGitHub 27de4c40ea Merge pull request #3111 from xinnan-tech/py-test-wake
test测试页面唤醒功能
2026-05-12 11:00:58 +08:00
3030332422 98ba86d4a6 update:对文档的补充说明 2026-05-12 10:49:02 +08:00
3030332422 2aa1a8d590 update:补充 2026-05-12 10:24:19 +08:00
3030332422 a2c3da55cd update:关于对digital-human的markdown文档的修改 2026-05-12 09:45:43 +08:00
3030332422 41ad57ad4f update:test迁移重命名为digital-human 2026-05-12 09:27:39 +08:00
hrz 6e1ed9fa4d update:简化说明;默认开启唤醒词 2026-05-10 20:26:40 +08:00
wengzhandGitHub 1b5c00231a Merge pull request #3165 from xinnan-tech/fix-chat-total
fix: 修复聊天记录左侧会话数量统计
2026-05-09 15:56:05 +08:00
zhuoqinglian 95acba1126 fix: 修复聊天记录左侧会话数量统计 2026-05-09 15:54:53 +08:00
wengzhandGitHub 1e6f0fd1be Merge pull request #3162 from xinnan-tech/perf-tool-call-optimization
fix:修复音乐播放工具TTS上报记录问题
2026-05-08 16:40:37 +08:00
DaGou12138 eb571fae57 fix:修复音乐播放工具TTS上报记录问题 2026-05-08 16:38:23 +08:00
wengzhandGitHub 2946be7284 Merge pull request #3161 from xinnan-tech/perf-tool-call-optimization
fix:优化天气工具调用提示词
2026-05-08 14:23:11 +08:00
DaGou12138 425cb1d6bd fix:优化天气工具调用提示词,本地上下文存在当前所在地的天气信息无需再度查询工具 2026-05-08 14:20:52 +08:00
wengzhandGitHub cf9b0ec712 Merge pull request #3160 from xinnan-tech/fix-chat-title
fix: 修复聊天记录标题过长展示异常的问题
2026-05-08 11:08:23 +08:00
zhuoqinglian 1b053dc9c8 fix: 修复聊天记录标题过长展示异常的问题 2026-05-08 11:05:20 +08:00
wengzhandGitHub d7ad656bd0 Merge pull request #3159 from xinnan-tech/perf-tool-call-optimization
fix:优化对话过程中JSON闭合符号清理的问题
2026-05-08 10:48:58 +08:00
DaGou12138 7dae67a57b fix:优化对话过程中JSON闭合符号清理的问题 2026-05-08 10:46:37 +08:00
wengzhandGitHub 7b0f59a747 Merge pull request #3158 from xinnan-tech/chore/cleanup
chore: 暂时移除近期不可用或不稳定的TTS供应器及相关配置
2026-05-08 10:38:53 +08:00
DaGou12138 2eaa5bbb11 fix:优化对话过程中异常符号记录的问题 2026-05-08 10:29:50 +08:00
FAN-yeB c693ff1564 update:linkerai 近期多次测试,linkerai接口无法提供稳定服务,先暂时下架这个供应器 2026-05-08 10:18:26 +08:00
DaGou12138andGitHub e9bf006d9d Merge pull request #3154 from haixinke/Bug-get-user-prifile-null
fix:当打开用户画像开关时,修复获取用户画像为空的问题
2026-05-07 16:01:57 +08:00
wengzhandGitHub bef96f36e2 Merge pull request #3157 from xinnan-tech/fix-streaming-tts-original-text
Fix streaming tts original text
2026-05-07 15:36:56 +08:00
wengzhandGitHub bebb7eac69 Merge pull request #3156 from xinnan-tech/perf-tool-call-optimization
perf:优化新闻工具新闻源配置获取的逻辑
2026-05-07 11:43:17 +08:00
DaGou12138 a787130b74 perf:
1、优化新闻工具新闻源获取的逻辑
2、调整新闻工具描述内容
2026-05-07 11:41:17 +08:00
FAN-yeB 8b7f16b9c2 update:TTS_GizwitsTTS 近期机智云语音合成不成功,先暂时下架这个供应器 2026-05-07 11:27:32 +08:00
FAN-yeB 9ed66ec013 update:TTS_ACGNTTS 海豚配音截止5月7日未找到token入口,先暂时下架这个供应器 2026-05-07 10:54:03 +08:00
FAN-yeB 05c22cdf2a update:更新 .gitignore忽略.claude目录 2026-05-07 10:23:02 +08:00
FAN-yeB 4bc0b67cd7 update:去除无用的.claude目录 2026-05-07 10:20:48 +08:00
Sakura-RanChen 0a6f811448 fix: 修复流式情况下替换词失败 2026-05-07 10:20:10 +08:00
wengzhandGitHub 5aeb55fffd Merge pull request #3155 from xinnan-tech/perf-tool-call-optimization
perf:优化工具调用问题、调整后端智能体搜索条件逻辑
2026-05-06 17:01:15 +08:00
DaGou12138 c33ee9442e perf:
1、优化小智后端智能体搜索条件逻辑
2、优化工具调用准确率问题,修改工具提示词,去除非必要的few-shot逻辑
3、优化工具描述内容
2026-05-06 16:53:06 +08:00
minwang 4670ad2925 delete log 2026-05-01 16:03:02 +08:00
minwangandClaude Sonnet 4.6 b542388053 🐛 修复PowerMem用户画像查询返回空的问题
修复 get_user_profile() 方法在缓存未命中时主动从 SDK 获取用户画
像,支持 profile_content 和 topics 两种格式,解决新对话中用户画像无法显示的问题

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-05-01 14:57:36 +08:00
3030332422 ba68c3d423 update:修改文档示例密码 2026-04-30 19:54:09 +08:00
3030332422 c30cd5ca79 update:添加“一体机数字人”配置说明 2026-04-30 18:04:44 +08:00
rainv123 a71ce53a68 fix:修复使用替换词时,火山引擎双向流式tts的文本显示/上报问题 2026-04-30 17:33:40 +08:00
rainv123 dd88107131 Revert "fix: 修复流式TTS替换词显示/上报问题"
This reverts commit 8b847a3826.
2026-04-30 15:57:03 +08:00
rainv123 8b847a3826 fix: 修复流式TTS替换词显示/上报问题 2026-04-30 14:32:36 +08:00
wengzhandGitHub 036b89f24e Merge pull request #3151 from xinnan-tech/py-refactor-asr-tts
调整配置
2026-04-30 11:42:36 +08:00
Sakura-RanChen 354280d430 调整配置 2026-04-30 11:41:16 +08:00
hrzandGitHub e2d16ec8d4 Merge pull request #3150 from Mirrorium227/main
Closes #Issue3149 修复文档序号
2026-04-30 11:27:15 +08:00
Radar SongandGitHub 6a0192de27 Closes #Issue3149 修复文档序号 2026-04-29 20:18:17 +08:00
hrzandGitHub ac6eeb5c1f Merge pull request #3145 from haixinke/fix-bug_prompt_cache_key_wrong
Fix:【设备快速提示词缓存key错误】
2026-04-29 14:32:42 +08:00
minwang 23e213f6ea 修复设备快速提示词缓存key错误 2026-04-28 15:52:23 +08:00
wengzhandGitHub 382c948482 Merge pull request #3144 from xinnan-tech/fix-memory-config
fix:修复记忆模型配置保存字段未删除问题
2026-04-28 15:25:49 +08:00
DaGou12138 a67d938523 fix:修复记忆模型配置保存字段未删除问题 2026-04-28 15:23:57 +08:00
wengzhandGitHub 26aca54f4b Merge pull request #3141 from xinnan-tech/add-correct-word
fix修复 asyncio.gather 吞掉业务异常的问题
2026-04-27 17:31:47 +08:00
rainv123 4717b8beac fix修复 asyncio.gather 吞掉业务异常的问题 2026-04-27 17:29:41 +08:00
wengzhandGitHub cd580b6f81 Merge pull request #3139 from xinnan-tech/py-fix-mem0ai
fix: 修复Mem0AI记忆模式工具调用时保存记忆400错误
2026-04-27 16:57:26 +08:00
3030332422 70740af7a6 fix: 修复Mem0AI记忆模式工具调用时保存记忆400错误 2026-04-27 16:25:05 +08:00
wengzhandGitHub 7b30c25923 Merge pull request #3132 from xinnan-tech/replacement-word
feat: 新增替换词管理功能
2026-04-27 16:23:54 +08:00
wengzhandGitHub 7db13b8be2 Merge pull request #3130 from xinnan-tech/add-correct-word
add:替换词功能
2026-04-27 16:23:33 +08:00
wengzhandGitHub a0df41ac24 Merge pull request #3136 from xinnan-tech/performance_tester
update:解决base类修改后测速工具兼容性问题
2026-04-27 15:08:26 +08:00
FAN-yeB a432c295d3 update:解决base类修改后测速工具兼容性问题 2026-04-27 15:06:40 +08:00
wengzhandGitHub ab8a35e233 Merge pull request #3135 from xinnan-tech/perf-tool-call-optimization
Perf:优化工具调用准确性、禁用标题总结与记忆模型思考模式、拆分动态上下文
2026-04-27 14:54:16 +08:00
DaGou12138 160b67197f Merge branch 'refs/heads/main' into perf-tool-call-optimization 2026-04-27 14:23:28 +08:00
欣南科技andGitHub 3510282cd5 Merge pull request #3134 from xinnan-tech/0.9.3
Bump to 0.9.3
2026-04-27 12:46:10 +08:00
hrz bdf7c7facf Bump to 0.9.3 2026-04-27 12:44:02 +08:00
DaGou12138 151a0c1b99 perf:
1、优化工具调用相关提示词,拆分静态system-prompt和动态上下文。
2、新增few-shot示例,使用"调用→结果→回复"流程闭环,显著提高多次会话后工具调用准确性。
3、去除对话标题总结和记忆的模型思考模式。
2026-04-27 11:08:25 +08:00
zhuoqinglian 52af722463 update: 更新替换词数量限制 2026-04-24 14:54:37 +08:00
wengzhandGitHub fa91c97685 Merge pull request #3131 from xinnan-tech/fix-wifi-no-response
fix: 修复设备配网成功没有自动退出配网模式
2026-04-24 14:38:06 +08:00
zhuoqinglian a92d6ccdfb fix: 修复设备配网成功没有自动退出配网模式 2026-04-24 14:35:34 +08:00
zhuoqinglianandGitHub f5ce755fea Merge pull request #3116 from wujingquan/main
Add WiFi password translations for multiple languages
2026-04-24 14:15:27 +08:00
rainv123 606a42cfb2 update:替换词功能 2026-04-24 10:06:26 +08:00
zhuoqinglian f85e863144 feat: 新增替换词管理功能 2026-04-23 17:50:57 +08:00
wengzhandGitHub 550397de98 Merge pull request #3127 from xinnan-tech/py-fix-asr
Py fix asr
2026-04-23 10:46:09 +08:00
Sakura-RanChen ff66b815a5 fix: 在模型和工具调用期间进行打断后说完话错误退出问题 2026-04-23 10:17:43 +08:00
Sakura-RanChen 8c18e48b97 fix: 停止帧发送后与原先音频时序问题
回退退出工具打断处理(在大模型思考期间进行打断,然后大模型立马调用退出工具会造成死锁,因模型思考时间和是否调用工具不确定性,状态难管理)
2026-04-22 15:55:28 +08:00
3030332422 44587985dd fix: 优化唤醒词音频采集 2026-04-22 11:42:22 +08:00
wengzhandGitHub 4acd02e2b0 Merge pull request #3125 from xinnan-tech/py-fix-llm-chunk
fix: 修复频繁打断下生成器不会清理
2026-04-22 11:30:35 +08:00
wengzhandGitHub ddfbacc717 Merge pull request #3124 from xinnan-tech/py-add-ASRv2
使用双流优化版接口(速度提升),增加豆包流式语音识别模型2.0模型选择
2026-04-22 11:29:57 +08:00
Sakura-RanChen 2a4dbdb69b fix: 修复频繁打断下生成器不会清理 2026-04-22 11:24:43 +08:00
FAN-yeB e9a2489e48 update:测试工具更新豆包流式测试逻辑 2026-04-22 10:53:48 +08:00
3030332422 db73faeec4 fix: 优化唤醒词音频采集,使用队列解耦回调与推理处理 2026-04-22 10:35:48 +08:00
Sakura-RanChen ce13a7e3e6 使用双流优化版接口(速度提升),增加豆包流式语音识别模型2.0模型选择 2026-04-22 10:28:47 +08:00
wengzhandGitHub 8a156e99f4 Merge pull request #3122 from xinnan-tech/py-add-TTS2.0
增加豆包语音合成模型2.0专属供应器
2026-04-21 15:44:04 +08:00
Sakura-RanChen 4313216816 增加豆包语音合成模型2.0专属供应器 2026-04-21 15:33:37 +08:00
DaGou12138 8c0899415a Merge branch 'refs/heads/main' into perf-tool-call-optimization 2026-04-21 15:02:31 +08:00
DaGou12138 3736a5845a perf:
1、优化提示词context动态变化导致的缓存失效问题,拆分静态system-prompt和动态上下文
2026-04-21 15:01:57 +08:00
wengzhandGitHub dc7615c008 Merge pull request #3120 from xinnan-tech/update-theme-generator
update: 更新设备主题生成功能,添加新表情
2026-04-21 14:50:53 +08:00
wengzhandGitHub b03f9374ea Merge pull request #3119 from xinnan-tech/fix-config-slm
feat: agent-models接口返回SLM模型,LLM类型同步返回所选SLM
2026-04-21 14:18:07 +08:00
欣南科技andGitHub cacfc9bfee Merge pull request #3121 from xinnan-tech/enhanced-prompts
update:优化提示词
2026-04-21 14:01:49 +08:00
hrz ce3c8a2bd8 update:优化提示词 2026-04-21 14:00:25 +08:00
zhuoqinglian e1a7e50e1f update: 更新设备主题生成功能,添加新表情 2026-04-21 11:44:03 +08:00
DaGou12138 90888de0ba perf:
1、新增工具调用Action枚举RECORD,通过完整的三段式 assistant(tool_calls) → tool → assistant(response)让模型准确学习到工具调用的正确行为,优化play_music的工具调用枚举NONE变更为RECORD,优化解决音乐播放工具调用缺陷问题。
2、处理悬空 tool_calls问题。
3、去除工具提示词强化注入逻辑。
2026-04-21 11:36:06 +08:00
3030332422 96f0412f94 update:测试页面支持关闭情绪表情功能 2026-04-21 11:15:31 +08:00
rainv123 ad4d32a05d feat: agent-models接口返回SLM模型,LLM类型同步返回所选SLM 2026-04-21 11:12:18 +08:00
hrzandGitHub 08d6bb4056 Merge pull request #3118 from xinnan-tech/openai_nothinking
feat:默认禁用部分平台LLM思考模式
2026-04-20 17:39:20 +08:00
FAN-yeB 1f0db7dc47 feat:默认禁用部分平台LLM思考模式 2026-04-20 17:35:32 +08:00
wengzhandGitHub 6f52d004e8 Merge pull request #3115 from xinnan-tech/fix-knowledge-sync-ragflow-documents
fix: 实现知识库与RAGFlow双向同步(知识库名称/简介的修改、知识库删除、文档上传/删除)
2026-04-20 16:51:06 +08:00
rainv123 0681550bab fix: 实现知识库与RAGFlow双向同步(知识库名称/简介的修改、知识库删除、文档上传/删除),移除冷却机制确保实时生效,修复文档影子表UNIQUE约束冲突 2026-04-20 16:49:24 +08:00
wengzhandGitHub 97b9c95c2d Merge pull request #3117 from xinnan-tech/py-fix-ReportTime
fix: 修复聊天记录上报排序错乱的问题
2026-04-20 15:56:49 +08:00
wujingquan 6111c3b5a0 Add 'configuring' translation for multiple languages in device configuration 2026-04-20 15:24:53 +08:00
wujingquan 123c1e8880 Add WiFi password translations for multiple languages 2026-04-20 15:17:04 +08:00
rainv123 6ce4503b22 fix:实现知识库文档与RAGFlow双向同步,支持远端上传/删除自动对账 2026-04-20 14:49:01 +08:00
3030332422 b209de36f5 fix: 修复聊天记录上报排序错乱的问题 2026-04-20 11:00:04 +08:00
wengzhandGitHub fc3e0869fd Merge pull request #3112 from xinnan-tech/py-fix-repeated-text
Py fix repeated text
2026-04-20 10:43:09 +08:00
Sakura-RanChen 3436d50b43 只有文本内容一致的时候不进行合成 2026-04-20 10:40:08 +08:00
Sakura-RanChenandGitHub 7a58c04df3 Merge pull request #3077 from xuruiray/fix/duplicate-tts-on-tool-call
fix: LLM 流式输出文本后调用工具导致 TTS 重复播报
2026-04-20 10:08:06 +08:00
Sakura-RanChenandGitHub 150be65436 Merge branch 'py-fix-repeated-text' into fix/duplicate-tts-on-tool-call 2026-04-20 10:07:28 +08:00
3030332422 dc2094ea8b update:说明文档的补充 2026-04-17 17:30:18 +08:00
wengzhandGitHub bacf4cd5e8 Merge pull request #3110 from xinnan-tech/fix_rag_pagerank
fix: 修复更新知识库时 pagerank 为 null 导致 RAGFlow 校验失败的问题
2026-04-17 16:52:02 +08:00
rainv123 7049e8cce7 fix: 修复更新知识库时 pagerank 为 null 导致 RAGFlow 校验失败的问题 2026-04-17 16:41:42 +08:00
3030332422 b0e24732de update:更改说明文档 2026-04-17 16:12:02 +08:00
3030332422 fb469a3ee3 fix:唤醒词服务启动失败不阻塞测试页面 2026-04-17 15:09:57 +08:00
3030332422 725c562aa1 update:对齐测试页面唤醒流程与硬件设备 2026-04-17 11:33:28 +08:00
wengzhandGitHub 61f137ce4b Merge pull request #3108 from xinnan-tech/fix-role-plugin
fix: 修复智能体tooltip高度超长的问题
2026-04-17 11:25:14 +08:00
zhuoqinglian 35ea14b95b fix: 修复智能体tooltip高度超长的问题 2026-04-17 11:10:33 +08:00
wengzhandGitHub b7d07f92a2 Merge pull request #3107 from xinnan-tech/py-fix-tts-markdown
fix: 调整处理顺序保留英文空格的同时修复markdown被识别为纯ASCII 字符的错误返回
2026-04-17 11:02:53 +08:00
Sakura-RanChen d2cbec0981 fix: 调整处理顺序保留英文空格的同时修复markdown被识别为纯ASCII 字符的错误返回 2026-04-17 10:59:29 +08:00
wengzhandGitHub 6348f12d60 Merge pull request #3106 from xinnan-tech/fix-role-plugin
fix: 修复角色配置-功能管理列表没有滚动条的问题
2026-04-17 10:51:45 +08:00
zhuoqinglian 11caf6bc8a fix: 修复角色配置-功能管理列表没有滚动条的问题 2026-04-17 10:42:47 +08:00
wengzhandGitHub 3cae9e3c32 Merge pull request #3102 from xinnan-tech/fix-title-summary
fix:将标题总结从记忆流程中拆分为独立接口,支持异步执行
2026-04-16 17:24:56 +08:00
欣南科技andGitHub 0161e642d0 Merge pull request #3104 from xinnan-tech/update-default-model
Update default model
2026-04-16 17:01:34 +08:00
hrz d41938fe07 update:这些文件已经完成使命,可移除 2026-04-16 16:54:33 +08:00
hrz e20c59e3df update:移除阿里云即将下架的模型 2026-04-16 16:42:34 +08:00
rainv123 fbaf28de5c fix:移除多余try和总结标题的异步 2026-04-16 14:43:16 +08:00
rainv123 2daf492e00 fix:将标题总结从记忆流程中拆分为独立接口,支持异步执行 2026-04-16 10:36:48 +08:00
wengzhandGitHub ac9083cfc7 Merge pull request #3101 from xinnan-tech/fix-title-summary
fix:修复PowerMem和Mem0AI记忆模型不会总结会话标题的问题
2026-04-16 09:04:58 +08:00
3030332422 36ea5c5e13 fix:修复多个问题 2026-04-15 17:53:51 +08:00
rainv123 f36e605aed fix:修复PowerMem和Mem0AI记忆模型不会总结会话标题的问题 2026-04-15 15:26:52 +08:00
wengzhandGitHub c1b836aaed Merge pull request #3100 from xinnan-tech/update-checkbox-style
update: 修改checkbox背景颜色
2026-04-15 14:32:46 +08:00
zhuoqinglian 07170f09eb update: 修改checkbox背景颜色 2026-04-15 14:28:20 +08:00
hrzandGitHub 8211de242f Merge pull request #3098 from xinnan-tech/fix-title-summary
fix:修复mem_report_only记忆模型没有总结会话标题
2026-04-15 14:20:34 +08:00
欣南科技andGitHub eea6807355 Merge pull request #3099 from xinnan-tech/agent-base-prompt
update:优化提示词
2026-04-15 13:59:04 +08:00
hrz 0ba7e81438 update:优化提示词 2026-04-15 12:59:15 +08:00
rainv123 703c707b4c fix:修复mem_report_only记忆模型没有总结会话标题 2026-04-15 11:54:22 +08:00
hrzandGitHub 3ab14578c2 Merge pull request #3097 from xinnan-tech/py-fix-stream-asr
fix: 修复豆包流式asr热词层级错误
2026-04-15 10:33:24 +08:00
Sakura-RanChen a15efa3090 fix: 修复豆包流式asr热词层级错误 2026-04-15 10:31:16 +08:00
3030332422 4225f1525f fix: 移除 RuntimeInstanceLock 锁文件机制 2026-04-14 17:59:29 +08:00
3030332422 d5b883817c update:前端增加可配置的本地唤醒词运行时 2026-04-14 16:54:37 +08:00
wengzhandGitHub 8af8930997 Merge pull request #3094 from xinnan-tech/py-client-about
refactor:优化相关打断逻辑
2026-04-14 15:31:29 +08:00
Sakura-RanChen ad01b05940 打断判断优先级调整 2026-04-14 15:05:55 +08:00
Sakura-RanChen a0497fdc48 给复用型TTS增加相关过滤 2026-04-14 10:18:09 +08:00
Sakura-RanChen 9634f84a98 Merge remote-tracking branch 'origin/main' into py-client-about 2026-04-13 11:35:10 +08:00
Sakura-RanChen 27e5e1e9d7 refactor: 优化非流式音频过滤机制
fix: 修复等待中被打断队列被停止造成的死锁
2026-04-13 11:33:47 +08:00
wengzhandGitHub 8fd92729f2 Merge pull request #3090 from xinnan-tech/slm-and-title
update:添加SLM模型配置和会话标题功能
2026-04-13 09:08:11 +08:00
wengzhandGitHub 736ffd80bf Merge pull request #3086 from xinnan-tech/update-chat-history
角色配置表单标题添加Tooltip鼠标悬浮提示说明,调整小参数模型表单配置
2026-04-13 09:07:50 +08:00
wengzhandGitHub 3783351f5d Merge pull request #3085 from xinnan-tech/update-mobile-chat-history
mobile 角色配置添加小参数模型选项,历史记录标题优化,聊天详情参数记录折叠处理
2026-04-13 09:07:31 +08:00
zhuoqinglian 2cd9484c54 update: mobile 聊天记录获取最新数据 2026-04-10 16:11:04 +08:00
Sakura-RanChen 0f1e0d0d2e 音频会在队列积压等待发送,等待协程结果返回不设置超时时间 2026-04-10 14:22:36 +08:00
3030332422 705b5732a6 update:为 test 页面增加可配置的本地唤醒词运行时 2026-04-10 09:17:22 +08:00
Sakura-RanChen 0e23672be7 fix: 修复直接回复是字幕未记录 2026-04-09 17:58:19 +08:00
Sakura-RanChen a6904f6ebd refactor: vad独立时间戳,相关残余状态处理 2026-04-09 17:15:35 +08:00
zhuoqinglian 784ba359f2 update: mobile 历史记录标题优化,聊天详情参数记录折叠处理 2026-04-09 14:32:32 +08:00
Sakura-RanChen e08eb48d9a refactor: 优化相关打断处理 2026-04-09 10:46:46 +08:00
rainv123 b9a1b154b7 update:添加SLM模型配置和会话标题功能 2026-04-09 09:23:54 +08:00
zhuoqinglian 64630e1b4e feat: mobile 角色配置添加小参数模型选项 2026-04-08 16:26:48 +08:00
zhuoqinglian 7dcc014ce4 update: 调整小参数模型表单配置 2026-04-08 11:41:12 +08:00
Rui Xu 389e8559aa fix: prevent duplicate TTS when LLM streams text before tool call 2026-04-06 21:49:28 +08:00
wengzhandGitHub c5f82369e5 Merge pull request #3060 from xinnan-tech/py_test_end
Py test end
2026-04-03 17:31:56 +08:00
wengzhandGitHub 52e63706a4 Merge pull request #3058 from xinnan-tech/py_fix_powermem
fix: 优化配置读取
2026-04-03 14:37:14 +08:00
Sakura-RanChenandGitHub a42a8b2b4b Merge pull request #3069 from xinnan-tech/py_fix_wakeup
fix:停止音频发送循环(仅在流控器已初始化时调用)
2026-04-03 09:48:16 +08:00
baigao 11142104af fix:停止音频发送循环(仅在流控器已初始化时调用) 2026-04-03 09:46:55 +08:00
wengzhandGitHub a9f5f1ab81 Merge pull request #3067 from xinnan-tech/memory-model-rename
fix:从有总结记忆的模型切换至仅上报聊天记录时,只清空记忆总结内容
2026-04-02 11:56:33 +08:00
rainv123 a8f68c42fc fix:从有总结记忆的模型切换至仅上报聊天记录时,只清空记忆 2026-04-02 11:27:11 +08:00
zhuoqinglian 1d3afb3c9a feat: 角色配置表单标题添加Tooltip鼠标悬浮提示说明 2026-04-02 09:45:33 +08:00
wengzhandGitHub 943622969a Merge pull request #3063 from xinnan-tech/py-bug-fix
fix:修复intent_llm模式下工具不存在时未正确播放提示信息的问题
2026-04-01 17:46:23 +08:00
wengzhandGitHub 9d39168a35 Merge pull request #3061 from xinnan-tech/memory-model-rename
选择仅上报聊天记录时,清除历史记忆;前端优化记忆模型切换体验
2026-04-01 15:57:44 +08:00
Sakura-RanChenandGitHub 61ab7f25d7 Merge branch 'main' into py_test_end 2026-04-01 14:45:05 +08:00
Sakura-RanChen 140fcd5887 补充sql,增加错误详情上报 2026-04-01 14:32:47 +08:00
rainv123 3c2024d00d fix: 前端优化记忆模型切换体验 - 切换无记忆模型时暂存记忆、切换恢复、保存配置后才真正清除 2026-04-01 14:32:30 +08:00
rainv123 2aee1bf567 fix:选择仅上报聊天记录时,清除历史记忆 2026-04-01 11:35:26 +08:00
wengzhandGitHub d89a26889c Merge pull request #3059 from xinnan-tech/memory-model-rename
fix:重命名记忆模型名称
2026-04-01 10:23:38 +08:00
rainv123 d611b2db73 update:重命名记忆模型名称 2026-04-01 09:54:46 +08:00
Sakura-RanChenandGitHub b0e911b248 Merge pull request #2996 from lgy1027/fix-end
fix: 优化退出流程并添加超时保护
2026-04-01 09:35:34 +08:00
Sakura-RanChenandGitHub fb3b7830ee Merge branch 'py_test_end' into fix-end 2026-04-01 09:35:23 +08:00
3030332422 10eeffdf31 fix:修复intent_llm模式下工具不存在时未正确播放提示信息的问题 2026-04-01 09:23:05 +08:00
rainv123 17847188dd Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server into new 2026-04-01 09:12:39 +08:00
Sakura-RanChen 7b2bfc087d fix: 优化配置读取 2026-03-31 18:00:51 +08:00
wengzhandGitHub 24a8839753 Merge pull request #3044 from xinnan-tech/py-display
添加“工具调用日志记录”与“设备端工具调用信息显示”
2026-03-31 15:12:11 +08:00
CGDandGitHub 64d25533d9 Merge branch 'main' into py-display 2026-03-31 14:24:27 +08:00
3030332422 6f6c788cd5 update:修改ai_agent_chat_history表content字段类型为TEXT 2026-03-31 14:11:31 +08:00
欣南科技andGitHub 8aa12d0961 Merge pull request #3057 from xinnan-tech/weather-faq
update:补齐天气文档链接
2026-03-31 12:11:36 +08:00
hrz d770632152 udpate:补齐天气文档链接 2026-03-31 12:08:30 +08:00
wengzhandGitHub a30903b644 Merge pull request #3056 from xinnan-tech/update-manager-mobile
Update manager mobile
2026-03-31 11:02:40 +08:00
wengzhandGitHub 5469168d18 Merge pull request #3055 from xinnan-tech/fix-digital-human
fix: 修复数字人摄像头展示区域裁剪的问题
2026-03-31 11:02:21 +08:00
zhuoqinglian 3c31231781 fix: 修复数字人摄像头展示区域裁剪的问题 2026-03-31 10:34:38 +08:00
3030332422 cc2e30205f fix:修复play_music工具调用信息显示异常 2026-03-31 10:03:50 +08:00
3030332422 f01de027cd fix:修复 intent_llm 模式下工具调用不上报的问题 2026-03-30 16:36:30 +08:00
zhuoqinglian fb195fc57b update: mobile 切换记忆模式判断是否需要禁用历史记忆内容输入 2026-03-26 17:25:44 +08:00
zhuoqinglian cc504fe542 update: mobile 添加智能体限制调整 2026-03-26 14:21:03 +08:00
wengzhandGitHub fd93116d69 Merge pull request #3045 from xinnan-tech/fix_audioRateController
fix: 修复流式场景下的误判
2026-03-25 11:10:42 +08:00
Sakura-RanChen 5622a13735 fix: 修复流式场景下的误判 2026-03-25 11:07:43 +08:00
3030332422 079c8630d9 update:修改部分前端样式 2026-03-25 10:43:45 +08:00
wengzhandGitHub 11830b2f22 Merge pull request #3043 from xinnan-tech/update-manager-mobile
mobile 优化聊天记录音频播放功能
2026-03-25 09:34:29 +08:00
wengzhandGitHub 861b4dee4c Merge pull request #3042 from xinnan-tech/update-web-style
优化页面主体内容区域适配
2026-03-25 09:34:06 +08:00
zhuoqinglian 603fa67824 Merge branch 'main' into update-web-style 2026-03-25 09:27:30 +08:00
zhuoqinglian 4d8279d916 update: 优化页面主体内容区域适配 2026-03-25 09:26:33 +08:00
rainv123 f91c4acb8f Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server into new 2026-03-24 11:12:32 +08:00
rainv123andGitHub 1342eccf86 Merge pull request #3039 from xinnan-tech/revert-3038-agent-chat-history-audio-retention
Revert "fix:修复智能体记忆模型配置变更时错误删除历史聊天音频的问题"
2026-03-24 11:08:44 +08:00
wengzhandGitHub 87f9ab7bb4 Revert "fix:修复智能体记忆模型配置变更时错误删除历史聊天音频的问题" 2026-03-24 11:05:35 +08:00
wengzhandGitHub 94cc8564aa Merge pull request #3038 from xinnan-tech/agent-chat-history-audio-retention
fix:修复智能体记忆模型配置变更时错误删除历史聊天音频的问题
2026-03-24 10:53:51 +08:00
rainv123 8fc2d05c99 Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server into new 2026-03-24 09:51:14 +08:00
rainv123 1a6d9d7490 fix: 修复智能体配置变更时错误删除历史音频的问题 2026-03-24 09:49:57 +08:00
wengzhandGitHub 6c19f1bd26 Merge pull request #3035 from xinnan-tech/mem_report_only
fix:新增仅上报聊天记录的记忆模型供应器Mem report only
2026-03-24 09:01:31 +08:00
rainv123 6d89ae1c66 fix:增加仅上报模型说明和记忆模型排序 2026-03-23 16:58:51 +08:00
rainv123 2a49f0804c feat:新增仅上报聊天记录(mem_report_only)记忆供应器 2026-03-23 16:17:38 +08:00
zhuoqinglian d64bb36606 update: mobile 优化聊天记录音频播放功能 2026-03-23 14:18:53 +08:00
zhuoqinglian 852e097e0d fix: mobile 修复声纹管理-编辑模式下声纹向量展示异常的问题 2026-03-20 16:57:02 +08:00
zhuoqinglian 3aecb99f70 fix: 修复聊天记录播放音频重叠播放的问题 2026-03-20 16:21:58 +08:00
wengzhandGitHub 6eecf5d0cb Merge pull request #3026 from xinnan-tech/fix_audioRateController_state
fix: 修复队列恢复时音频播放时间错乱问题
2026-03-20 14:49:42 +08:00
wengzhandGitHub a4d02ca2ab Merge pull request #3029 from xinnan-tech/update-web-style
解决登录失败验证码刷新两次的问题
2026-03-20 14:48:45 +08:00
wengzhandGitHub 7ef5a67119 Merge pull request #3028 from xinnan-tech/update-manager-mobile
mobile 智能体管理多语言调整、缓存用户信息
2026-03-20 14:48:26 +08:00
zhuoqinglian 1c5dacd2d5 Merge branch 'main' into update-web-style 2026-03-20 14:45:50 +08:00
zhuoqinglian 7732b7b1b9 update: mobile 缓存用户信息 2026-03-20 14:42:47 +08:00
zhuoqinglian 6e7b938f51 fix: 解决登录失败验证码刷新两次的问题 2026-03-20 10:42:35 +08:00
zhuoqinglian 52f5859b1f update: mobile 智能体管理多语言调整 2026-03-19 17:56:27 +08:00
Sakura-RanChen 73d174bc4a fix: 修复队列恢复时音频播放时间错乱问题 2026-03-19 16:18:30 +08:00
Sakura-RanChen 9512fe642c 恢复多工具 2026-03-19 16:04:14 +08:00
Sakura-RanChen 0ef3a60a4c 调整缩进 2026-03-19 11:57:40 +08:00
Sakura-RanChen f40dc4a658 修复双重回复问题 2026-03-19 11:56:29 +08:00
wengzhandGitHub 4747e54f6c Merge pull request #3023 from xinnan-tech/update-manager-mobile
mobile 播放声纹弹窗样式调整,切换角色模板更新表单状态
2026-03-18 16:53:54 +08:00
wengzhandGitHub 9278fdb016 Merge pull request #3024 from xinnan-tech/update-web-style
解决MCP接入点输入框没有显示报错信息的问题
2026-03-18 16:53:14 +08:00
zhuoqinglian 5b23804954 Merge branch 'main' into update-manager-mobile 2026-03-18 16:50:29 +08:00
zhuoqinglian b2ed5eb2c0 fix: 解决MCP接入点输入框没有显示报错信息的问题 2026-03-18 16:47:48 +08:00
zhuoqinglian 07113dc4d2 update: mobile 播放声纹弹窗样式调整,切换角色模板更新表单状态 2026-03-18 16:44:27 +08:00
Sakura-RanChen 1a5dd2d9d9 fix: 修复日志音频上报 2026-03-18 16:40:32 +08:00
wengzhandGitHub 8a01432e9d Merge pull request #3022 from xinnan-tech/perf-userLanguage-complete
perf:补充MCP接入点国际化信息
2026-03-18 15:48:57 +08:00
DaGou12138 7a39fd0152 补充MCP接入点国际化信息 2026-03-18 09:48:11 +08:00
Sakura-RanChen 0d2cb7b3c9 增加工具调用日志记录,暂时单工具状态 2026-03-17 16:13:35 +08:00
wengzhandGitHub 9123a174da Merge pull request #3020 from xinnan-tech/update-web-style
优化表格底部按钮边距统一调整,修复刷新页面菜单、页脚闪动的问题
2026-03-17 16:11:54 +08:00
zhuoqinglian 2e165747d6 update: 参数弹窗去掉多余代码 2026-03-17 16:08:36 +08:00
zhuoqinglian 3bdf634e37 Merge branch 'main' into update-web-style 2026-03-17 16:04:22 +08:00
zhuoqinglian 20ac6f3f96 fix: 修复刷新页面菜单、页脚闪动的问题 2026-03-17 16:01:10 +08:00
zhuoqinglian 818ac1eefb style: 表格底部按钮边距统一调整 2026-03-17 10:30:43 +08:00
3030332422 980930cd42 update:在设备端显示工具调用信息 2026-03-16 13:11:45 +08:00
wengzhandGitHub 956ae37e56 Merge pull request #3015 from xinnan-tech/fix/update-sherpa-onnx
update:更新sherpa_onnx依赖至1.12.29版本
2026-03-16 11:51:01 +08:00
rainv123 15731892ad update:更新sherpa_onnx依赖至1.12.29版本 2026-03-16 11:43:07 +08:00
zhuoqinglian d6d50e0432 fix: 修复语言切换接口请求异常的问题 2026-03-16 10:37:47 +08:00
zhuoqinglian 286079c93a feat: 添加声纹管理-声纹向量音频播放功能 2026-03-16 09:50:24 +08:00
zhuoqinglian 860cbc36fd fix: 处理智能体管理全局触发下拉刷新问题 2026-03-16 09:46:43 +08:00
欣南科技andGitHub 2ac7e5be9c Merge pull request #3013 from xinnan-tech/version-num
Bump to 0.9.2
2026-03-14 08:06:31 +08:00
hrz ebfcfe148d Bump to 0.9.2 2026-03-14 08:05:57 +08:00
wengzhandGitHub 435860d339 Merge pull request #3011 from xinnan-tech/perf-userLanguage-pt-br
perf:新增葡萄牙语国际化提示
2026-03-13 17:56:46 +08:00
DaGou12138 b7cb97a373 新增葡萄牙语国际化提示 2026-03-13 17:53:07 +08:00
wengzhandGitHub e7b1daf021 Merge pull request #3010 from xinnan-tech/perf-llm-lazy-optimization
perf:优化提示词与工具调用偷懒逻辑
2026-03-13 17:28:17 +08:00
DaGou12138 bbf2febd6d 修改标签国际化相关提示词 2026-03-13 17:13:22 +08:00
DaGou12138 586520f321 Merge branch 'refs/heads/main' into perf-llm-lazy-optimization 2026-03-13 17:03:28 +08:00
wengzhandGitHub 22262be7cd Merge pull request #3007 from xinnan-tech/update-manager-mobile
mobile 更新智能体功能,解决遗留bug
2026-03-12 17:43:14 +08:00
zhuoqinglian 9021eac54e Merge branch 'main' into update-manager-mobile 2026-03-12 17:37:30 +08:00
wengzhandGitHub b72967bcc3 Merge pull request #3006 from xinnan-tech/captcha-validation-error
fix: 优化验证码验证错误提示
2026-03-12 17:37:05 +08:00
zhuoqinglian 469f836a82 fix: 修复 mobile 短信验证码错误重复提示、智能体插件保存异常问题 2026-03-12 17:36:51 +08:00
zhuoqinglian 1e8960f37c feat: mobile 添加用户协议和隐私政策 2026-03-12 17:34:59 +08:00
rainv123 7aad8da6f6 fix: 优化验证码验证错误提示 2026-03-12 17:34:04 +08:00
wengzhandGitHub dd24dc0abc Merge pull request #2995 from xinnan-tech/py_fix_HuoShanTTS
Py fix huo shan tts
2026-03-12 16:25:47 +08:00
wengzhandGitHub 4f6cdc745f Merge pull request #3001 from xinnan-tech/py_fix_clone_languages
fix: 克隆音色未显示
2026-03-12 15:51:57 +08:00
lgy1027 bcf03a0378 refactor: 优化超时配置,响应社区反馈
- 工具调用超时改为可配置项 tool_call_timeout,默认30秒
- OpenAI 超时恢复可配置机制,支持细粒度配置和单一值
- TTS 超时保护统一添加到所有流式TTS实现
- 将 ActionResponse 导入移到文件顶部
- 修复超时配置边界情况处理
2026-03-11 18:13:05 +08:00
lgy1027 4f615b3807 Merge remote-tracking branch 'origin/main' into fix-end 2026-03-11 17:15:59 +08:00
zhuoqinglian 48087582c2 update: mobile 编辑源和语速音调页面样式调整 2026-03-11 17:10:18 +08:00
Sakura-RanChen fbb0e02fdd 去除空格,最下端留空行 2026-03-11 14:59:23 +08:00
wengzhandGitHub d7ef85f6d2 Merge branch 'main' into py_fix_clone_languages 2026-03-11 14:37:55 +08:00
wengzhandGitHub 2f2bd38f1e Merge pull request #3002 from xinnan-tech/perf-rag-linqui-update
perf:新增ragflow改动相关所需数据集、调整ragflow知识库相关文档
2026-03-11 11:55:09 +08:00
DaGou12138 bc48a169f4 新增ragflow改动相关所需数据集、调整ragflow知识库相关文档 2026-03-11 11:43:29 +08:00
DaGou12138andGitHub e777db6ef3 Merge pull request #2950 from gzh246/feature/api-knowledge-dora
整理ragflow外部接口api(包括会话,知识库,文档分块,文件管理,智能体)等,基于ragflow的最新稳定的大版本tag-0.23.0。
2026-03-11 11:27:02 +08:00
Sakura-RanChen 60498b24ae fix: 克隆音色未显示 2026-03-11 10:30:29 +08:00
DaGou12138 caae839b1d pref:优化大模型工具调用偷懒问题,调整基础提示词与新增临时工具记忆加深 2026-03-11 10:13:13 +08:00
zhuoqinglian 6a14d694b0 Merge branch 'main' into update-manager-mobile 2026-03-10 16:15:51 +08:00
zhuoqinglian d86387971a fix: 修复web、h5同时打开token失效问题 2026-03-10 16:15:09 +08:00
zhuoqinglian ff1076deac update: mobile 服务端接口地址提示语优化 2026-03-10 16:14:04 +08:00
wengzhandGitHub 5b986bf57f Merge pull request #2997 from xinnan-tech/web-privacy-terms
add:增加用户协议和隐私政策
2026-03-10 14:34:48 +08:00
rainv123 f1b2080d99 add:增加用户协议和隐私政策 2026-03-10 14:28:47 +08:00
lgy1027 c2b95ca896 fix: 优化退出流程并添加超时保护
- 添加 is_exiting 标志防止退出流程被中断
- 工具调用添加 30 秒超时保护,避免流程卡死
- TTS 相关操作添加超时保护
- 优化 OpenAI 客户端超时配置
2026-03-10 11:45:38 +08:00
zhuoqinglian ed41de7b8e feat: mobile 修改设备添加逻辑,新增手动添加设备 2026-03-10 11:37:51 +08:00
Sakura-RanChen 42a132f5cd 调整新增音色时的默认填充 2026-03-10 11:29:05 +08:00
Sakura-RanChen 7b2f6c2eee fix: 单一语音流上报音频不全 2026-03-10 10:53:24 +08:00
Sakura-RanChen ee5a53d67d fix: 修复选用seed-tts-2.0文本缺失
fix: 设备端说话状态与服务端同步
2026-03-09 17:55:54 +08:00
wengzhandGitHub f691e9f0c9 Merge pull request #2993 from xinnan-tech/tag_fix
fix: 修复智能体标签顺序错乱问题
2026-03-09 14:18:13 +08:00
rainv123 7a32c208fb fix: 修复智能体标签顺序错乱问题 2026-03-09 14:15:31 +08:00
DaGou12138andGitHub 6984763ddc Merge pull request #2988 from xinnan-tech/update-manager-mobile
web端功能同步更新到mobile端
2026-03-06 15:21:18 +08:00
DaGou12138andGitHub d6c3659d70 Merge pull request #2987 from xinnan-tech/update-agent-style
update: 智能体管理样式优化
2026-03-06 15:20:57 +08:00
zhuoqinglian d51c2a241c feat: mobile端同步更新上报模式、语言设置功能 2026-03-06 09:40:04 +08:00
zhuoqinglian a1d37b61e4 Merge branch 'main' into update-manager-mobile 2026-03-06 09:23:59 +08:00
zhuoqinglian 8b24151756 update: 智能体管理样式优化 2026-03-06 09:20:15 +08:00
GZH a9854756ec ```
refactor(knowledge): 移除未完成的BotController实现

- 删除了BotController类的所有代码
2026-03-04 21:25:26 +08:00
GZH f95bed4189 Merge remote-tracking branch 'origin/feature/api-knowledge-dora' into feature/api-knowledge-dora
# Conflicts:
#	main/manager-api/src/main/java/xiaozhi/common/config/RestTemplateConfig.java
#	main/manager-api/src/main/java/xiaozhi/common/exception/ErrorCode.java
#	main/manager-api/src/main/java/xiaozhi/modules/knowledge/controller/KnowledgeBaseController.java
#	main/manager-api/src/main/java/xiaozhi/modules/knowledge/controller/KnowledgeFilesController.java
#	main/manager-api/src/main/java/xiaozhi/modules/knowledge/rag/KnowledgeBaseAdapter.java
#	main/manager-api/src/main/java/xiaozhi/modules/knowledge/rag/RAGFlowClient.java
#	main/manager-api/src/main/java/xiaozhi/modules/knowledge/rag/impl/RAGFlowAdapter.java
#	main/manager-api/src/main/resources/i18n/messages.properties
#	main/manager-api/src/main/resources/i18n/messages_de_DE.properties
#	main/manager-api/src/main/resources/i18n/messages_en_US.properties
#	main/manager-api/src/main/resources/i18n/messages_vi_VN.properties
#	main/manager-api/src/main/resources/i18n/messages_zh_TW.properties
2026-03-04 17:47:42 +08:00
GZH f7bf29a5e4 fix: 修复PR审查意见 - 新增单文档删除接口、删除BotController、优化imports、修复i18n
# Conflicts:
#	main/manager-api/src/main/resources/i18n/messages_de_DE.properties
#	main/manager-api/src/main/resources/i18n/messages_en_US.properties
#	main/manager-api/src/main/resources/i18n/messages_vi_VN.properties
2026-03-04 17:45:58 +08:00
GZH 5913dbbb7b refactor(knowledge):重构知识库模块架构并优化DTO结构
- 将BotController暂时注释以重新设计Agent集成方案
- 在多个DTO类中添加@JsonIgnoreProperties注解提升JSON序列化兼容性- 修改ChunkDTO中positions字段类型为嵌套列表并添加token字段
-为DatasetDTO和DocumentDTO添加更多时间日期字段映射
-重构KnowledgeBaseAdapter接口参数结构使用统一的请求对象
- 实现DocumentStatusSyncTask定时任务同步文档处理状态
-优化KnowledgeBaseService统计信息更新机制
- 修复数据集删除时的级联删除逻辑防止孤儿数据
- 统一本地实体ID与RAGFlow ID避免前端调用错误
2026-03-04 17:39:55 +08:00
GZH f6bb55ae90 refactor(knowledge):重构知识库模块的数据传输对象和接口适配器
- 将 BotController 类注释掉,暂时移除机器人功能实现
- 在多个 DTO 类中添加 @JsonIgnoreProperties 注解以忽略未知属性- 修改 ChunkDTO 中的 positions 字段类型为嵌套列表,并添加 token 字段
- 更新 DatasetDTO 和 DocumentDTO 中的时间日期字段格式说明-重构 KnowledgeBaseAdapter 接口方法签名,使用强类型 DTO 替代 Map 参数
- 在 KnowledgeBaseController 中注入 KnowledgeManagerService 并修改删除逻辑
- 更新数据库统计信息更新方法名称和参数结构-优化 KnowledgeBaseServiceImpl 中的数据集创建和查找逻辑
2026-03-04 17:39:55 +08:00
GZH 33e712b5fb refactor(knowledge):重构知识库模块架构并优化DTO结构
- 将BotController暂时注释以重新设计Agent集成方案
- 在多个DTO类中添加@JsonIgnoreProperties注解提升JSON序列化兼容性- 修改ChunkDTO中positions字段类型为嵌套列表并添加token字段
-为DatasetDTO和DocumentDTO添加更多时间日期字段映射
-重构KnowledgeBaseAdapter接口参数结构使用统一的请求对象
- 实现DocumentStatusSyncTask定时任务同步文档处理状态
-优化KnowledgeBaseService统计信息更新机制
- 修复数据集删除时的级联删除逻辑防止孤儿数据
- 统一本地实体ID与RAGFlow ID避免前端调用错误
2026-03-04 17:39:07 +08:00
GZH 1c26faddc4 整理ragflow外部接口api(包括会话,知识库,文档分块,文件管理,智能体)等,基于ragflow的最新稳定的大版本tag-0.23.0。
Adapter 模式落地,引入了 RAGFlowAdapter,彻底隔离了业务逻辑和底层 API 调用。现在 Service 层只管“做什么”,Adapter 层只管“怎么调用 RAGFlow ”。
代码瘦身,移除了所有手写的 HTTP 请求,全部走 Adapter统一管理。
逻辑修正,修复了重复创建、ID 冲突等 Bug,确保了事务的一致性。
在数据库中留下了拓展字段如embedding_model,chunk_method,parser_config,document_count。保证可循迹。并新增ai_rag_knowledge_document表做到ragflow远端文档信息可在本地查看,管理。后续仍需追加双端同步增加/修改/删除。

- 新增AgentDTO包含Agent管理、Webhook调试、会话和对话等完整功能
- 新增BotDTO支持SearchBot和AgentBot的检索和对话功能
- 新增ChatDTO整合对话助手、会话和消息的全套数据传输对象
- 添加ai_rag_dataset表扩展字段支持租户、头像、权限等配置
- 创建ai_rag_knowledge_document表关联RAGFlow远程文档与本地元信息
- 扩展AgentService接口添加智能体配置验证功能
- 实现validateAgentConfig方法对RAGFlow模型配置进行软同步校验
- 更新AgentServiceImpl集成新的配置验证逻辑和注释清理
2026-03-04 17:32:44 +08:00
zhuoqinglian ac887139ce feat: 新增智能体标签、上下文源功能 2026-03-04 16:33:22 +08:00
wengzhandGitHub bc2c68e89f Merge pull request #2982 from xinnan-tech/web-supplement-language
优化相关判断,补充国际化
2026-03-03 11:27:12 +08:00
Sakura-RanChen cdc2363af1 优化相关判断,补充国际化 2026-03-03 11:25:37 +08:00
DaGou12138andGitHub ba07c623df Merge pull request #2980 from xinnan-tech/fix-roleConfig
fix: 解决角色配置合并冲突异常问题
2026-03-02 17:14:22 +08:00
zhuoqinglian 344a73b87b fix: 解决角色配置合并冲突异常问题 2026-03-02 17:11:56 +08:00
DaGou12138andGitHub 9a81b6c0ee Merge pull request #2979 from xinnan-tech/milieyo-patch-1
fix:Liquibase yaml
2026-03-02 17:02:35 +08:00
wengzhandGitHub 96f65e7c35 fix:Liquibase yaml 2026-03-02 16:55:08 +08:00
wengzhandGitHub ff2cdeb483 Merge pull request #2972 from xinnan-tech/update-tts-voice-data
Update tts voice data
2026-03-02 16:50:33 +08:00
wengzhandGitHub fe8b82fcda Merge branch 'main' into update-tts-voice-data 2026-03-02 16:50:19 +08:00
wengzhandGitHub f007e6614d Merge pull request #2978 from xinnan-tech/agent-tag
feat:智能体添加标签功能
2026-03-02 16:13:59 +08:00
rainv123 0eafd37bf6 feat:智能体添加标签功能 2026-03-02 15:51:25 +08:00
wengzhandGitHub af3bc09403 Merge pull request #2977 from xinnan-tech/role-tag
新增角色标签配置
2026-03-02 15:40:59 +08:00
zhuoqinglian 90ebb8d499 feat: 语音合成-音色管理,编辑模式下添加取消操作 2026-03-02 15:35:47 +08:00
zhuoqinglian f2c6cb8641 update: 优化文本溢出悬浮窗样式 2026-03-02 15:33:38 +08:00
Sakura-RanChenandGitHub 116a6973cf Merge branch 'main' into update-tts-voice-data 2026-03-02 09:19:46 +08:00
欣南科技andGitHub 769407a5b0 Merge pull request #2976 from xinnan-tech/fix-play-music-bug
fix:重写 audio_to_opus_data_stream 方法,让音频文件处理使用独立编码器,避免与 TTS 流编码器冲突。
2026-03-01 13:00:25 +08:00
hrz a4583e273a fix:重写 audio_to_opus_data_stream 方法,让音频文件处理使用独立编码器,避免与 TTS 流编码器冲突。 2026-03-01 12:59:45 +08:00
zhuoqinglian 645462abd0 feat: 新增角色标签配置 2026-02-28 17:51:46 +08:00
Sakura-RanChen b49c4520f9 部分TTS参数调整 2026-02-27 16:37:31 +08:00
wengzhandGitHub acf45ae8ad Merge pull request #2971 from xinnan-tech/fix-manager-api-device-tools-api
fix:#2939 修复getDeviceTools()中未考虑分页问题
2026-02-27 16:36:37 +08:00
Sakura-RanChen 0d1c18f5dd Merge branch 'main' into update-tts-voice-data 2026-02-27 11:21:18 +08:00
Sakura-RanChen 7b020ba55c update: 增加智能体独立音频设置 2026-02-27 11:17:08 +08:00
DaGou12138 cc081bc9eb #2939 修复getDeviceTools()中未考虑分页问题 2026-02-26 17:46:20 +08:00
hrzandGitHub 187c2ee977 Merge pull request #2970 from ephemetra/patch-1
docs: 修复文档中的错误内容
2026-02-26 16:21:20 +08:00
ephemetraandGitHub 00ab407b04 docs: 修复文档中的错误内容
1. 修正第三部分的错误标题。
2. 修正了server.mqtt_manager_api填入`PUBLIC_IP`+`:`+`UDP_PORT`的错误内容。
2026-02-25 10:19:22 +08:00
欣南科技andGitHub bda7a90551 Merge pull request #2969 from xinnan-tech/i18n-pt_BR
I18n pt br
2026-02-24 08:05:45 +08:00
hrz 71b72f5b51 update:修正英文版翻译遗漏的chatHistory.前缀 2026-02-24 08:04:53 +08:00
hrzandGitHub 61fd8b21e0 Merge pull request #2965 from Alif254317/claude/explore-repository-YzYCv
feat: adicionar suporte a Português Brasileiro (pt_BR) no projeto
2026-02-24 07:28:45 +08:00
欣南科技andGitHub 9a4610ab04 Merge pull request #2968 from xinnan-tech/fix-vad-decoder
fix:高并发下,共享vad里decoder变量扰动
2026-02-24 02:00:29 +08:00
hrz fd8296262e fix:高并发下,共享vad里decoder变量扰动 2026-02-24 01:59:44 +08:00
Claude 63dfcf6659 feat: adicionar suporte a Português Brasileiro (pt_BR) no projeto
- Adiciona arquivo de tradução pt_BR.js para manager-web (~1332 chaves)
- Adiciona arquivo de tradução pt_BR.ts para manager-mobile (~440 chaves)
- Cria README_pt_BR.md com documentação traduzida completa
- Registra pt_BR no sistema i18n de ambos os frontends (web e mobile)
- Adiciona pt_BR nos seletores de idioma (HeaderBar, login, retrievePassword)
- Adiciona badge "Português (Brasil)" em todos os READMEs existentes
- Adiciona detecção automática de idioma pt/pt-BR no navegador
- Adiciona chave language.ptBR em todos os arquivos de tradução existentes

https://claude.ai/code/session_01XJMVWKBjy7UwqskN9BbsYM
2026-02-19 17:44:25 +00:00
hrzandGitHub 8897b4f3c6 Merge pull request #2962 from xinnan-tech/test_live2d
fix: 调整切换摄像头位置
2026-02-18 08:30:47 +08:00
GZH 8d46585101 refactor(knowledge):重构知识库模块架构并优化DTO结构
- 将BotController暂时注释以重新设计Agent集成方案
- 在多个DTO类中添加@JsonIgnoreProperties注解提升JSON序列化兼容性- 修改ChunkDTO中positions字段类型为嵌套列表并添加token字段
-为DatasetDTO和DocumentDTO添加更多时间日期字段映射
-重构KnowledgeBaseAdapter接口参数结构使用统一的请求对象
- 实现DocumentStatusSyncTask定时任务同步文档处理状态
-优化KnowledgeBaseService统计信息更新机制
- 修复数据集删除时的级联删除逻辑防止孤儿数据
- 统一本地实体ID与RAGFlow ID避免前端调用错误
2026-02-14 12:55:56 +08:00
GZH 6967b2bce2 refactor(knowledge):重构知识库模块的数据传输对象和接口适配器
- 将 BotController 类注释掉,暂时移除机器人功能实现
- 在多个 DTO 类中添加 @JsonIgnoreProperties 注解以忽略未知属性- 修改 ChunkDTO 中的 positions 字段类型为嵌套列表,并添加 token 字段
- 更新 DatasetDTO 和 DocumentDTO 中的时间日期字段格式说明-重构 KnowledgeBaseAdapter 接口方法签名,使用强类型 DTO 替代 Map 参数
- 在 KnowledgeBaseController 中注入 KnowledgeManagerService 并修改删除逻辑
- 更新数据库统计信息更新方法名称和参数结构-优化 KnowledgeBaseServiceImpl 中的数据集创建和查找逻辑
2026-02-14 12:55:56 +08:00
GZH 16a38bbfc4 refactor(knowledge):重构知识库模块架构并优化DTO结构
- 将BotController暂时注释以重新设计Agent集成方案
- 在多个DTO类中添加@JsonIgnoreProperties注解提升JSON序列化兼容性- 修改ChunkDTO中positions字段类型为嵌套列表并添加token字段
-为DatasetDTO和DocumentDTO添加更多时间日期字段映射
-重构KnowledgeBaseAdapter接口参数结构使用统一的请求对象
- 实现DocumentStatusSyncTask定时任务同步文档处理状态
-优化KnowledgeBaseService统计信息更新机制
- 修复数据集删除时的级联删除逻辑防止孤儿数据
- 统一本地实体ID与RAGFlow ID避免前端调用错误
2026-02-14 12:55:56 +08:00
GZH 861df73eab 整理ragflow外部接口api(包括会话,知识库,文档分块,文件管理,智能体)等,基于ragflow的最新稳定的大版本tag-0.23.0。
Adapter 模式落地,引入了 RAGFlowAdapter,彻底隔离了业务逻辑和底层 API 调用。现在 Service 层只管“做什么”,Adapter 层只管“怎么调用 RAGFlow ”。
代码瘦身,移除了所有手写的 HTTP 请求,全部走 Adapter统一管理。
逻辑修正,修复了重复创建、ID 冲突等 Bug,确保了事务的一致性。
在数据库中留下了拓展字段如embedding_model,chunk_method,parser_config,document_count。保证可循迹。并新增ai_rag_knowledge_document表做到ragflow远端文档信息可在本地查看,管理。后续仍需追加双端同步增加/修改/删除。

- 新增AgentDTO包含Agent管理、Webhook调试、会话和对话等完整功能
- 新增BotDTO支持SearchBot和AgentBot的检索和对话功能
- 新增ChatDTO整合对话助手、会话和消息的全套数据传输对象
- 添加ai_rag_dataset表扩展字段支持租户、头像、权限等配置
- 创建ai_rag_knowledge_document表关联RAGFlow远程文档与本地元信息
- 扩展AgentService接口添加智能体配置验证功能
- 实现validateAgentConfig方法对RAGFlow模型配置进行软同步校验
- 更新AgentServiceImpl集成新的配置验证逻辑和注释清理
2026-02-14 12:55:56 +08:00
lww155 fa9809bee1 fix: 调整切换摄像头位置 2026-02-11 17:41:52 +08:00
欣南科技andGitHub 2a93745d01 Merge pull request #2957 from xinnan-tech/test_live2d
fix:手机前后摄像头切换,调整摄像头为竖屏显示
2026-02-11 11:32:53 +08:00
lww155 85fab64588 fix: 优化h5样式 2026-02-10 09:54:39 +08:00
rainv123 b22f430e98 fix:调整摄像头显示区域为竖屏 2026-02-09 17:22:28 +08:00
lww155 c3da3e937e add: 前后摄像头切换 2026-02-09 16:50:23 +08:00
Sakura-RanChen 4ea3bea85c update: 增加TTS音色语种选择 2026-02-09 15:28:31 +08:00
DaGou12138 520e24a54e 修改统一规范ai_tts_voice语言类型数据语句 2026-02-06 17:12:47 +08:00
DaGou12138 e82cc53bfb 统一规范ai_tts_voice语言类型数据 2026-02-06 16:54:21 +08:00
wengzhandGitHub 3bddc1c722 Merge pull request #2953 from xinnan-tech/fix-modal-flicker
Fix modal flicker
2026-02-06 16:40:56 +08:00
lww155 f68fc5c90d fix: 修复点击用户框无法关闭菜单问题 2026-02-06 16:37:20 +08:00
lww155 ea8001a9c8 fix: 模态框闪烁问题 2026-02-06 14:49:17 +08:00
hrzandGitHub ca10d7290d Merge pull request #2949 from xinnan-tech/test_live2d
update:优化
2026-02-05 18:05:26 +08:00
rainv123 7b3db72285 update:优化 2026-02-05 18:02:18 +08:00
wengzhandGitHub 3639871dd2 Merge pull request #2944 from xinnan-tech/py_test_Memory_powermem
Py test memory powermem
2026-02-05 17:45:23 +08:00
wengzhandGitHub 5fb7fbe08f Merge branch 'main' into py_test_Memory_powermem 2026-02-05 17:45:07 +08:00
wengzhandGitHub 73f8a0977c Merge pull request #2948 from xinnan-tech/py_test_typing
Py test typing
2026-02-05 17:34:26 +08:00
wengzhandGitHub 87b99e0353 Merge pull request #2946 from xinnan-tech/fix_asr_Initialize
增加asr初始化ID判断
2026-02-05 17:33:35 +08:00
Sakura-RanChenandGitHub 795dcec698 Merge pull request #2890 from shengzhou1216/feature/python-typing
feat: 为多个模块添加类型注解以增强代码可读性
2026-02-05 17:15:00 +08:00
Sakura-RanChenandGitHub 4f61d21d17 Merge branch 'py_test_typing' into feature/python-typing 2026-02-05 17:14:47 +08:00
wengzhandGitHub 8a84428724 Merge pull request #2938 from xinnan-tech/test_live2d
update:测试页面增加切换live2d模型、拍照识物功能
2026-02-05 17:00:11 +08:00
wengzhandGitHub 46f1dee20e Merge pull request #2947 from xinnan-tech/fix/model-config-enable-check
新增模型配置启用开关禁用判断
2026-02-05 16:51:51 +08:00
lww155 75e24af858 新增模型配置启用开关禁用判断 2026-02-05 16:46:10 +08:00
rainv123 fae6db2198 fix:移除拨号按钮防抖,点击拨号后禁用按钮3秒 2026-02-05 16:32:05 +08:00
rainv123 32aea32e2c fix:删除工具描述输入框的多余空格 2026-02-05 16:09:52 +08:00
Sakura-RanChen 63d9dae18b 增加asr初始化ID判断 2026-02-05 15:42:23 +08:00
rainv123 041c254ea4 fix:拨号按钮和摄像头按钮增加防抖机制 2026-02-05 15:06:04 +08:00
rainv123 378d68aba9 fix: 测试页面MCP工具参数编辑样式调整 2026-02-05 14:44:37 +08:00
wengzhandGitHub 4076136fe4 Merge pull request #2945 from xinnan-tech/perf-manager-api-delTable
perf:新增数据集,删除无用业务表ai_voiceprint
2026-02-05 13:53:38 +08:00
DaGou12138 1e6e250dac 新增数据集,删除无用业务表ai_voiceprint 2026-02-05 11:28:31 +08:00
Sakura-RanChen 42b495d556 补充sql文件 2026-02-05 10:37:48 +08:00
Sakura-RanChen c88d709d56 Merge branch 'main' into py_test_Memory_powermem 2026-02-04 16:02:50 +08:00
wengzhandGitHub fa523dc615 Merge pull request #2942 from xinnan-tech/perf-manager-api-modelConfig
perf:新增关闭模型配置相关逻辑判断
2026-02-04 11:42:38 +08:00
DaGou12138 8718a07c7e 新增关闭模型配置相关逻辑判断 2026-02-04 11:32:10 +08:00
wengzhandGitHub 39124d4930 Merge pull request #2940 from xinnan-tech/py_fix_stream_asr
fix: 讯飞自动设备适应,阿里云等待时机过短
2026-02-04 10:38:29 +08:00
rainv123 8dbc0b9689 fix: 修复绑定验证码后重新连接摄像头不自动打开的问题 2026-02-04 10:29:48 +08:00
Sakura-RanChen 9fd3501605 fix: 讯飞自动设备适应,阿里云等待时机过短 2026-02-04 10:28:27 +08:00
lww155 4f72a516d3 update: 录音按钮新增防抖 2026-02-04 09:54:56 +08:00
lww155 e4cfabeb7e update: 优化代码 2026-02-04 09:54:21 +08:00
rainv123 85a83727fd feat: 摄像头验证码绑定检查 && 视觉分析URL动态获取 2026-02-03 11:52:22 +08:00
wengzhandGitHub a31e16c1d9 Merge pull request #2931 from xinnan-tech/py_test_error_response
Py test error response
2026-02-03 10:01:08 +08:00
Sakura-RanChen 3ab5dc49ac 去除无用代码,修复插入语句id缺失 2026-02-02 16:07:09 +08:00
rainv123 005dc55db3 add:测试页面增加拍照识物功能 2026-02-02 15:08:24 +08:00
Sakura-RanChenandGitHub ba1cf14ed4 Merge pull request #2888 from shengzhou1216/fix/2075
feat: 统一LLM错误处理并添加系统错误回复配置
2026-02-02 14:49:42 +08:00
Sakura-RanChenandGitHub 6da32a60c0 Merge branch 'py_test_error_response' into fix/2075 2026-02-02 14:49:26 +08:00
wengzhandGitHub ea9778f422 Merge pull request #2930 from xinnan-tech/py_test_AudioArtifacts
Py test audio artifacts
2026-02-02 14:23:09 +08:00
wengzhandGitHub f0b176768d Merge branch 'main' into py_test_AudioArtifacts 2026-02-02 14:23:01 +08:00
wengzhandGitHub 5aa671d6ff Merge pull request #2929 from xinnan-tech/py_fix_asr
fix: 模式切换时音频残余,优化相关状态管理
2026-02-02 14:05:07 +08:00
Sakura-RanChenandGitHub d02d094219 Merge pull request #2924 from shengzhou1216/refactor/asr-delete_audio
refactor(asr): 重构speech_to_text方法以接收artifacts参数
2026-02-02 11:42:59 +08:00
Sakura-RanChen f5b4bae927 fix: 模式切换时音频残余,优化相关状态管理 2026-02-02 11:29:47 +08:00
rainv123 19fdf5d917 update:测试页面切换模型和切换背景的按钮样式调整 2026-02-02 10:47:01 +08:00
欣南科技andGitHub 9682f615cf Merge pull request #2927 from xinnan-tech/update-test-bg
update:优化图片
2026-02-01 01:24:39 +08:00
hrz 8162adeca1 update:优化图片 2026-02-01 01:23:45 +08:00
rainv123 ba2749430f add:live2d增加可切换的男性角色 2026-01-30 17:59:06 +08:00
rainv123 e3088f4410 Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server into new 2026-01-30 17:21:22 +08:00
huozaimengli 614b662e3d refactor(asr): 重构speech_to_text方法以接收artifacts参数
移除各ASR提供者中重复的get_current_artifacts调用,改为通过参数传递artifacts
修改base类中process_audio方法,根据combined_pcm_data长度决定是否创建artifacts
更新所有speech_to_text方法签名,添加artifacts可选参数并更新文档字符串
2026-01-29 16:39:35 +08:00
huozaimengli f5b9995f89 feat: 添加系统错误时的默认回复配置SQL
在 sys_params 表中插入新记录,用于配置系统发生错误时返回给用户的默认回复消息。该配置允许管理员自定义错误提示文本,提升用户体验。
2026-01-29 15:58:00 +08:00
huozaimengli 0ddd9ba8ff fix: 修正系统错误响应的句子类型为MIDDLE 2026-01-29 15:57:31 +08:00
欣南科技andGitHub 9c8d614d78 Merge pull request #2918 from xinnan-tech/test-arm64-docker
fix:docker 缺失nginx文件
2026-01-28 19:02:21 +08:00
hrz 0075550229 fix:docker 缺失nginx文件 2026-01-28 19:01:37 +08:00
rainv123 4a547be882 Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server into new 2026-01-28 15:24:02 +08:00
欣南科技andGitHub d721270bed Merge pull request #2915 from xinnan-tech/test-arm64-docker
update:兼容arm架构字体
2026-01-28 11:40:26 +08:00
hrz fe2e178b19 update:兼容arm架构字体 2026-01-28 11:39:50 +08:00
欣南科技andGitHub 21337e5046 Merge pull request #2914 from xinnan-tech/test-arm64-docker
update:兼容arm架构字体
2026-01-28 11:34:54 +08:00
hrz 1c8b271e20 update:兼容arm架构字体 2026-01-28 11:34:16 +08:00
欣南科技andGitHub 3d1160b6bb Merge pull request #2913 from xinnan-tech/test-arm64-docker
update:调试编译arm镜像
2026-01-28 11:13:40 +08:00
hrz 086c4a55e6 update:调试编译arm镜像 2026-01-28 11:12:58 +08:00
欣南科技andGitHub 889f2f9a75 Merge pull request #2912 from xinnan-tech/test-arm64-docker
update:兼容arm架构字体
2026-01-28 11:03:04 +08:00
hrz 07b686be14 update:兼容arm架构字体 2026-01-28 11:02:17 +08:00
欣南科技andGitHub 4d6b453485 Merge pull request #2910 from xinnan-tech/test-arm64-docker
add:增加linux/arm64docker镜像编译
2026-01-28 09:16:25 +08:00
hrz 9909045f9d add:增加linux/arm64docker镜像编译 2026-01-28 09:15:40 +08:00
hrzandGitHub b0dc4d88db Merge pull request #2908 from xinnan-tech/py_fix_text
fix: 保存音频数据文件时文本不记录
2026-01-27 17:53:51 +08:00
Sakura-RanChen bbf3858735 fix: 保存音频数据文件时文本不记录 2026-01-27 16:22:35 +08:00
欣南科技andGitHub 73573e2f12 Merge pull request #2907 from xinnan-tech/live2d-actions
update:增加版本号,解决html引用文件缓存问题
2026-01-27 16:13:31 +08:00
hrz 99e136786e update:增加版本号,解决html引用文件缓存问题 2026-01-27 16:12:55 +08:00
欣南科技andGitHub 205be0e4ec Merge pull request #2906 from xinnan-tech/live2d-actions
add:麦克风权限的判断
2026-01-27 16:03:49 +08:00
hrz 2e64f45a31 update:优化代码排版 2026-01-27 15:59:46 +08:00
hrz 818b4f2f6d update:增加版本号,解决html引用文件缓存问题 2026-01-27 15:52:34 +08:00
hrz 5d72f9f79d update:由于现在的模型本身动作有限,使用起来体验不佳,暂时移除动作控制的代码 2026-01-27 15:45:53 +08:00
Sakura-RanChenandGitHub 22d0dcdf58 Merge pull request #2887 from shengzhou1216/refactor/asr-delete_audio
refactor(asr): 统一音频预处理逻辑并引入AudioArtifacts
2026-01-27 15:41:46 +08:00
rainv123 122d1def1d Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server into new 2026-01-27 11:59:03 +08:00
hrzandGitHub 0ee349df2d Merge pull request #2905 from xinnan-tech/py_fix_minimaxLog
补充minimax接口错误信息
2026-01-27 11:28:17 +08:00
Sakura-RanChen 4dde66339b 补充minimax接口错误信息 2026-01-27 11:22:32 +08:00
hrzandGitHub 0064f975c2 Merge pull request #2903 from spider-yamet/feature/issue-2896-clean
feat: implement Issue 2896 - Live2D Actions and Microphone Detection
2026-01-27 11:17:31 +08:00
rainv123andGitHub 43234f4798 Merge pull request #2836 from xinnan-tech/py_audio_rate
update: 自定义配置输出音频采样率
2026-01-27 10:58:32 +08:00
spider-yamet 48034fc000 Translate log messages to Chinese in recorder.js 2026-01-26 15:03:23 -08:00
spider-yamet 3fc33e9095 Optimize code density: simplify changed files while keeping comments (reduce 364 lines) 2026-01-26 14:48:08 -08:00
spider-yamet da71dce860 Optimize code : minify test files 2026-01-26 14:37:07 -08:00
spider-yamet b7e4408a0f Add browser-based unit tests for xiaozhi test modules
- Add browser-compatible test files (no npm required)

  - recorder.test.browser.js: 8 tests for microphone and HTTP detection

  - tools.test.browser.js: 5 tests for Live2D actions and error handling

- Add test runner (test-runner.html) with built-in test framework

- Add null safety checks in tools.js for DOM element access

- Add documentation (English and Chinese versions)

  - README_TESTS.md / README_TESTS_CN.md: Complete test guide

  - QUICK_START_TEST.md / QUICK_START_TEST_CN.md: Quick start guides

- Total: 13 unit tests covering microphone detection, HTTP detection, Live2D actions, and error handling
2026-01-26 10:18:40 -08:00
spider-yamet 4f4f8ca54e docs: add testing guide and fix Chinese text in controller.js 2026-01-26 08:30:39 -08:00
spider-yamet 99c12a0fed fix: translate corrupted Chinese text to proper Chinese in controller.js 2026-01-26 08:05:02 -08:00
spider-yamet dcd0ef3f40 fix: translate corrupted Chinese text to proper Chinese in controller.js
- Fix corrupted Chinese characters in addMCPTool function
- Translate placeholder and button text to proper Chinese
2026-01-26 07:57:13 -08:00
spider-yamet 2f53b921fe feat: implement Issue 2896 - Live2D Actions and Microphone Detection
- Add Live2D action tools (smile, wave, generic actions) as MCP tools
- Implement microphone availability detection
- Handle HTTP non-localhost access scenarios
- Update UI to reflect microphone availability state
- Translate all comments and messages to proper Chinese/English
- Add test files for new functionality
2026-01-26 07:44:33 -08:00
rainv123 0f16a8d31f Merge branch 'py_audio_rate' of https://github.com/xinnan-tech/xiaozhi-esp32-server into new 2026-01-26 17:39:29 +08:00
Sakura-RanChen 3051dc8327 更改默认输出采样率与设备端适配 2026-01-26 17:35:45 +08:00
Sakura-RanChen c313997a61 去除调试输出 2026-01-26 16:33:09 +08:00
欣南科技andGitHub 07c0c764d3 Merge pull request #2898 from xinnan-tech/fix-ChatHistoryConf-value
set ChatHistoryConf default value
2026-01-26 15:11:46 +08:00
hrz e339a501aa set ChatHistoryConf default value 2026-01-26 15:11:11 +08:00
hrzandGitHub f2c541b4ab Merge pull request #2897 from xinnan-tech/fix-path
fix: 修复项目访问地址
2026-01-26 15:07:47 +08:00
zhuoqinglian 206af1c074 fix: 修复项目访问地址 2026-01-26 15:06:05 +08:00
欣南科技andGitHub 344729c9a6 Merge pull request #2895 from xinnan-tech/fix-digital-human-initialize
add:响应mcp初始化请求
2026-01-26 12:51:09 +08:00
hrz 3b1c2d3b1b add:响应mcp初始化请求 2026-01-26 12:42:11 +08:00
wengzhandGitHub 3660af9619 Merge pull request #2881 from xinnan-tech/generator
feat: 添加语音盒子在线烧录功能
2026-01-26 09:41:42 +08:00
zhuoqinglian 4905ba836c fix: 修复语音盒子自定义主题表情无法跟随语气变化问题 2026-01-26 09:33:58 +08:00
huozaimengli 4b573fb4e2 feat: 为多个模块添加类型注解以增强代码可读性
为 ConnectionHandler 相关的函数参数添加类型注解,使用 TYPE_CHECKING 避免循环导入。主要修改包括:
- 在 abortHandle、textHandle 等处理模块中为 conn 参数添加 ConnectionHandler 类型注解
- 在 websocket_server、connection 等核心模块中为方法参数添加类型注解
- 在 plugins_func 下的多个功能模块中为函数参数添加类型注解
- 在 providers 相关模块中为工具执行器和方法添加类型注解
- 统一代码格式,如将单引号字符串改为双引号

Fixes #2034
2026-01-25 17:47:52 +08:00
huozaimengli 6ae0af278b feat: 统一LLM错误处理并添加系统错误回复配置
在多个LLM提供者中移除try-catch块,将错误处理统一到connection.py的流处理层
添加system_error_response配置项,支持自定义系统错误时的回复内容
在意图识别和流处理中捕获异常时返回配置的错误回复,避免硬编码错误信息

Fixes #2075
2026-01-25 16:48:01 +08:00
huozaimengli 15650e1a6c refactor(asr): 统一音频预处理逻辑并引入AudioArtifacts
重构所有ASR提供商的speech_to_text方法,将重复的音频解码、合并和文件保存逻辑提取到基类的speech_to_text_wrapper中。引入AudioArtifacts数据类封装PCM帧、字节数据、文件路径和临时路径,简化各提供商实现。移除各提供商中的冗余文件清理代码,由基类统一处理。

新增requires_file()和prefers_temp_file()方法允许提供商声明文件需求,优化内存和磁盘使用。保持接口兼容性的同时提高代码复用性和可维护性。
2026-01-25 11:27:34 +08:00
Sakura-RanChen ce00c862fe HuoshanDoubleStreamTTS增加自定义配置 2026-01-23 17:16:46 +08:00
zhuoqinglian a50229431b fix: 修复https协议导致无法上传表情的问题 2026-01-23 17:07:23 +08:00
wengzhandGitHub 275102f5b7 Merge pull request #2882 from xinnan-tech/fix-devices-loading
fix: 修复获取设备状态接口报错时, 没有执行失败回调导致loading一直存在
2026-01-23 15:36:46 +08:00
lww155 ff6286ffcc fix: 修复获取设备状态接口报错时, 没有执行失败回调导致loading一直存在 2026-01-23 15:33:44 +08:00
zhuoqinglian 974fe6c3b7 feat: 添加语音盒子在线烧录功能 2026-01-23 15:29:43 +08:00
wengzhandGitHub b78292b270 Merge pull request #2880 from xinnan-tech/perf-manager-api-business
pred:优化MCP工具名称列表顺序问题
2026-01-23 14:13:23 +08:00
DaGou12138 3559a7018c 优化MCP工具名称列表顺序问题 2026-01-23 14:03:00 +08:00
Sakura-RanChenandGitHub 9368885355 Merge pull request #2859 from wayyoungboy/add-powermem
chore: 更新 powermem 依赖至 0.3.1 版本, 增加角色过滤的能力
2026-01-23 10:24:21 +08:00
wengzhandGitHub 3222d6175d Merge pull request #2879 from xinnan-tech/fix-digital-human
update: 拨号成功后自动开启录音
2026-01-23 09:22:03 +08:00
lww155 28f9f23449 update: 拨号成功后自动开启录音 2026-01-23 09:13:19 +08:00
wengzhandGitHub 5e7d192ebc Merge pull request #2877 from xinnan-tech/fix-digital-human
update: 调整配置获取
2026-01-22 17:55:46 +08:00
lww155 d6a32416d7 update: 调整配置获取 2026-01-22 17:48:40 +08:00
wengzhandGitHub 6a2c01763e Merge pull request #2876 from xinnan-tech/fix-digital-human
Fix digital human
2026-01-22 17:27:36 +08:00
lww155 e1dd272708 update: 关闭设置框时保存配置, 阻止按钮事件冒泡, 调整配置获取方式 2026-01-22 17:23:27 +08:00
lww155 3e7bc0606a update: 优化背景加载代码 2026-01-22 16:16:09 +08:00
lww155 9d9ca1475d fix: 修复重复请求ota与重复建立连接问题 2026-01-22 15:54:26 +08:00
lww155 43ddc0135c update: 去除未使用代码 2026-01-22 15:39:40 +08:00
欣南科技andGitHub 98f907a97b Merge pull request #2874 from xinnan-tech/manager-mcp-api
update:增加mcp调用后数据类型的判断
2026-01-22 14:46:23 +08:00
hrz b2ddf060c9 update:增加mcp调用后数据类型的判断 2026-01-22 14:42:43 +08:00
欣南科技andGitHub 5e612276db Merge pull request #2873 from xinnan-tech/manager-mcp-api
自定义表情固件接口
2026-01-22 14:02:48 +08:00
hrz cbb293e7fa add:上传自定义表情固件接口 2026-01-22 13:27:10 +08:00
hrz fdb050f6ed add:获取设备的mcp工具列表 2026-01-21 22:58:58 +08:00
wengzhandGitHub f83ddf79fa Merge pull request #2868 from xinnan-tech/fix-digital-human
fix: 修复未绑定设备时状态处理
2026-01-20 17:48:35 +08:00
lww155 95b340a7a3 fix: 修复未绑定设备时状态处理 2026-01-20 17:39:06 +08:00
milieyoandGitHub d46ac89e58 Merge pull request #2867 from xinnan-tech/update-style
style: 优化部分样式,保持一致性
2026-01-20 15:45:00 +08:00
zhuoqinglian 80b0912f84 Merge branch 'main' into update-style 2026-01-20 15:36:22 +08:00
zhuoqinglian e0b4431064 style: 优化部分样式,保持一致性 2026-01-20 15:35:41 +08:00
milieyoandGitHub e3f16f91cf Merge pull request #2863 from xinnan-tech/fix-digital-human
update: 新增聊天输入框
2026-01-20 09:32:23 +08:00
milieyoandGitHub 5f39e5c308 Merge pull request #2860 from xinnan-tech/perf-manager-api-business
perf:优化字典接口、模型接口、音色资源部分业务逻辑
2026-01-20 09:17:24 +08:00
lww155 83304c9ff4 update: 新增聊天输入框 2026-01-20 09:16:45 +08:00
DaGou12138 863cfb575c 优化音色资源相关业务逻辑 2026-01-19 17:55:38 +08:00
DaGou12138 2944040648 优化音色资源相关业务逻辑 2026-01-19 17:19:45 +08:00
DaGou12138 5f19b9b0d0 优化字典接口、模型接口部分业务逻辑 2026-01-19 15:19:48 +08:00
milieyoandGitHub 2948ed091c Merge pull request #2858 from xinnan-tech/perf-manager-api-device
perf:优化知识库相关业务逻辑
2026-01-19 14:24:27 +08:00
渠磊 bfa638e17f chore: 更新 powermem 依赖至 0.3.1 版本, 增加角色过滤的能力 2026-01-19 14:16:12 +08:00
DaGou12138 97ee397b54 Merge branch 'refs/heads/main' into perf-manager-api-device 2026-01-19 12:00:07 +08:00
DaGou12138 f7fbe7efbe 优化修改知识库接口部分业务逻辑 2026-01-19 11:59:25 +08:00
hrzandGitHub ce6b0bd099 Merge pull request #2857 from xinnan-tech/build-generator
fix: 修复项目部署二级目录自定义语音盒子主题文件资源异常问题
2026-01-19 11:45:08 +08:00
zhuoqinglian 8643960104 Merge branch 'main' into build-generator 2026-01-19 11:33:11 +08:00
zhuoqinglian b4549ebc59 fix: 修复项目部署二级目录自定义语音盒子主题文件资源异常问题 2026-01-19 11:30:28 +08:00
hrzandGitHub a602cddd05 Merge pull request #2850 from xinnan-tech/build-generator
feat: 添加语音盒子自定义主题功能
2026-01-17 14:04:12 +08:00
hrzandGitHub c4278864d1 Merge pull request #2852 from xinnan-tech/fix_device-mcp
fix: 修复开机立刻唤醒导致客户端 MCP 工具列表为空的问题
2026-01-16 19:02:58 +08:00
3030332422 7094548ecf fix: 修复开机立刻唤醒导致客户端 MCP 工具列表为空的问题 2026-01-16 18:04:04 +08:00
milieyoandGitHub 363fffada8 Merge pull request #2849 from xinnan-tech/fix-multilingual-bug
fix: 语音合成-音色管理,克隆音频路径、克隆音频文本多语言异常处理
2026-01-16 17:14:28 +08:00
milieyoandGitHub c2c3eb7106 Merge pull request #2848 from xinnan-tech/perf-manager-api-device
perf:优化设备在线状态接口相关逻辑
2026-01-16 17:13:30 +08:00
zhuoqinglian 8e19a64f15 fix: 语音合成-音色管理,克隆音频路径、克隆音频文本多语言异常处理 2026-01-16 17:05:46 +08:00
linjiaqin 9042b7e6d7 优化智能体模块相关逻辑 2026-01-16 16:43:04 +08:00
linjiaqin ad1855cfb5 优化设备在线接口相关逻辑 2026-01-16 15:45:42 +08:00
zhuoqinglian efa0d1095d feat: 添加语音盒子自定义主题功能 2026-01-16 14:52:40 +08:00
hrzandGitHub a8c13d740b Merge pull request #2842 from hlleng/fix/model-remark
fix: 修复智控台模型配置中备注不能保存的问题
2026-01-15 20:49:50 +08:00
Honglin Leng 7475e9829f fix: persist model remark on update 2026-01-15 20:15:25 +08:00
milieyoandGitHub b7dde1dae2 Merge pull request #2839 from xinnan-tech/fix-digital-human
修复背景404、通话接通后自动录音、模型加载Loading
2026-01-15 14:26:50 +08:00
lww155 778c4560fd update: 新增模型加载 2026-01-15 09:20:22 +08:00
lww155 fa31096cd8 update: 拨号成功后开启自动录音 2026-01-15 09:19:25 +08:00
Sakura-RanChen fd1dd39463 update: 自定义配置输出音频采样率 2026-01-14 17:54:50 +08:00
lww9029 ecd3e12a41 fix: 切换背景404问题 2026-01-14 11:18:41 +08:00
欣南科技andGitHub 0a8121e849 Merge pull request #2831 from xinnan-tech/test_page_fix
update:修复live2d模型加在路径bug
2026-01-12 18:06:41 +08:00
hrz e93f9d7111 update:修复live2d模型加在路径bug 2026-01-12 18:05:39 +08:00
Sakura-RanChen 3bdccfea62 解析json字符串提取相关文本 2026-01-12 17:47:29 +08:00
hrzandGitHub fdf98985a1 Merge pull request #2826 from xinnan-tech/test_page_fix
update:增加live2d模型,修改测试页面样式
2026-01-12 17:44:30 +08:00
hrz 6c3609e79e update:优化图片大小 2026-01-12 17:43:30 +08:00
Sakura-RanChenandGitHub 5fa50339b4 Merge pull request #2824 from wayyoungboy/add-powermem
docs(config): 更新配置文件和集成文档
2026-01-12 15:07:08 +08:00
Sakura-RanChen f0c95bc987 为空时不做相关查询(工具调用function_call时没有相关query) 2026-01-12 15:06:38 +08:00
rainv123 a9aa709502 update:增加live2d模型,修改测试页面样式 2026-01-12 11:00:50 +08:00
渠磊 8edc23b888 docs(config): 更新配置文件和集成文档
- 添加embedding_dims配置项注释
- 同步配置文件修改到集成文档
2026-01-09 17:42:30 +08:00
Sakura-RanChen c4fc510a27 补充相关信息 2026-01-09 17:25:03 +08:00
Sakura-RanChen 134000996b query只传递相关的文本 2026-01-09 16:09:05 +08:00
Sakura-RanChen 111909c26d fix: 启用用户画像模式时,search为同步方法需包装 2026-01-09 15:46:08 +08:00
Sakura-RanChenandGitHub f66966d72f Merge pull request #2822 from wayyoungboy/add-powermem
fix: powermem sqlite支持用户画像以及修改配置模式
2026-01-09 14:46:31 +08:00
milieyoandGitHub 1f6a89ada5 Merge pull request #2823 from xinnan-tech/20260109
调整智能体管理list接口排序,根据创建时间倒序返回
2026-01-09 14:32:22 +08:00
LJH-rgsze 31a1180b79 还原 2026-01-09 14:30:27 +08:00
LJH-rgsze 77b4a416a9 调整智能体管理list接口排序,根据创建时间倒序返回 2026-01-09 14:28:01 +08:00
渠磊 06a2e6fbfa refactor(powermem): 移除未使用的内存提供者实例注册 2026-01-09 12:34:55 +08:00
渠磊 c8a28e98d6 Merge remote-tracking branch 'upstream/py_test_Memory_powermem' into add-powermem
# Conflicts:
#	main/xiaozhi-server/core/providers/memory/powermem/powermem.py
2026-01-09 12:32:59 +08:00
渠磊 564adbd851 feat(powermem): 升级配置结构并支持多数据库用户画像
- 重构配置结构为模块化设计(llm/embedder/vector_store)
- 用户画像功能支持oceanbase/seekdb/sqlite三种存储后端
- 更新文档说明和示例配置
2026-01-09 12:25:17 +08:00
Sakura-RanChen ad93f43145 去除默认初始化,补充函数签名缺失 2026-01-09 10:49:24 +08:00
wayyoungboy 2bf1123647 部分配置key修正 2026-01-08 23:52:19 +08:00
Sakura-RanChenandGitHub a6ff6426e9 Merge pull request #2819 from wayyoungboy/add-powermem
Add powermem
2026-01-08 17:54:57 +08:00
milieyoandGitHub da7a6933f5 Merge pull request #2818 from xinnan-tech/fix_wakeup_audioRate
修复设备播放音频时打断后使用唤醒词时流控器未重置造成的死锁
2026-01-08 17:27:53 +08:00
Sakura-RanChen 54cac0d2fd fix: 修复当手动设备上个会话正常结束后,使用唤醒词唤醒音频播放不完整 2026-01-08 16:17:48 +08:00
渠磊 de7aedab96 feat(memory): 添加用户画像功能支持
- 新增 `enable_user_profile` 配置项,支持用户画像模式
- 实现 `UserMemory` 类集成,自动提取用户信息
- 更新文档说明用户画像功能及配置要求
2026-01-08 14:43:48 +08:00
hrzandGitHub 7fe718c887 Merge pull request #2817 from xinnan-tech/py_TTS_text
TTS文本过滤
2026-01-08 14:21:17 +08:00
hrzandGitHub 36eaefb8b0 Merge pull request #2816 from xinnan-tech/20260105
修复音色播放按钮显示错误
2026-01-08 12:39:33 +08:00
hrzandGitHub 3a61ec84b3 移除console打印 2026-01-08 12:39:07 +08:00
渠磊 dadf05ab4e feat: 添加 PowerMem 智能记忆支持
- 新增 PowerMem 配置选项和集成文档
- 更新 README 和多语言文档
- 添加 powermem 依赖包
2026-01-08 12:02:03 +08:00
Sakura-RanChen c1bf92d243 修复设备播放音频时打断后使用唤醒词时流控器未重置造成的死锁 2026-01-08 11:39:04 +08:00
LJH-rgsze c1959e0795 优化代码逻辑 2026-01-08 10:10:39 +08:00
LJH-rgsze 40a6b7434b 调整判断逻辑 2026-01-07 18:08:43 +08:00
Sakura-RanChen 24519a2732 TTS文本过滤 2026-01-07 17:38:38 +08:00
LJH-rgsze c190abe4b1 用国际化重写硬编码提示 2026-01-07 17:03:56 +08:00
LJH-rgsze 07d5b3cd2a 补充-上个修改上传少了 2026-01-07 16:54:08 +08:00
LJH-rgsze 81e923b4f7 补充音色管理页面缺失的国际化 2026-01-07 16:52:36 +08:00
LJH-rgsze bcb51729a5 修复播放按钮的显示错误 2026-01-07 16:00:35 +08:00
hrzandGitHub a731c1d01f Merge pull request #2811 from xinnan-tech/py_fix_llm_text
修复独立LLM调用时文本未记录
2026-01-06 17:34:07 +08:00
Sakura-RanChen 0bb5a37ace 修复独立LLM调用时文本未记录 2026-01-06 10:29:56 +08:00
hrzandGitHub a711827996 Merge pull request #2810 from xinnan-tech/fix_report_connection
fix: 禁用httpx连接池并显式关闭响应防止连接泄漏,进程文件描述符耗尽
2026-01-05 18:00:41 +08:00
hrzandGitHub 8f854e6f74 Merge pull request #2809 from xinnan-tech/20260105
补充国际化
2026-01-05 17:59:12 +08:00
Sakura-RanChen f1b5afbf5c fix: 禁用httpx连接池并显式关闭响应防止连接泄漏 2026-01-05 17:50:23 +08:00
LJH-rgsze 84f6930389 补充国际化 2026-01-05 17:46:35 +08:00
hrzandGitHub f415dd4482 Merge pull request #2808 from xinnan-tech/20260105
修复音色克隆,ota管理,语音资源管理的分页错误
2026-01-05 17:30:46 +08:00
LJH-rgsze 927f65e106 修复参数名不匹配导致的翻页错误 2026-01-05 17:21:38 +08:00
hrzandGitHub 4b6b3082c7 Merge pull request #2794 from xinnan-tech/20260104SSL
拓展首页的搜索栏-支持同时搜索设备mac和智能体名字
2026-01-05 16:54:20 +08:00
LJH-rgsze 4c550245db 调整判断逻辑 2026-01-05 16:51:09 +08:00
LJH-rgsze 341ed22370 简化代码 2026-01-05 15:27:47 +08:00
LJH-rgsze 6851ccacf3 提取重复代码合并 2026-01-05 15:05:29 +08:00
hrzandGitHub 815af687ea Merge pull request #2798 from xinnan-tech/py_asr_emotion_language
Py asr emotion language
2026-01-05 15:03:50 +08:00
LJH-rgsze e9373b59c8 优化代码 2026-01-05 14:50:23 +08:00
Sakura-RanChen 3f3f3fcd57 更改Funserver使用模型 2026-01-05 14:46:34 +08:00
LJH-rgsze d003e66310 还原部分代码 2026-01-05 11:31:28 +08:00
Sakura-RanChen e6108ccbbd 增加ASR情绪和语种返回 2026-01-05 11:17:04 +08:00
LJH-rgsze a01f01198c 补全参数的返回 2026-01-05 11:13:01 +08:00
LJH-rgsze cc96eaeb8f 修复设备列表显示问题 2026-01-05 10:13:11 +08:00
LJH-rgsze e25ca3d1d3 调整搜索实现-从前端过滤改为后端查询 2026-01-05 10:10:17 +08:00
LJH-rgsze 9c111b2d5e 更新国际化提示 2026-01-04 18:04:26 +08:00
LJH-rgsze 436a01fb6e 删除依赖更新部分(还原) 2026-01-04 17:14:25 +08:00
LJH-rgsze 9b6620dd4a 搜索拓展 2026-01-04 16:53:38 +08:00
Sakura-RanChenandGitHub 85919e800e Merge pull request #2552 from Packeting1/multi_lang_tts
feat: 为阿里百炼流式TTS支持多语言音色
2026-01-04 15:02:24 +08:00
欣南科技andGitHub a2023a65c0 Merge pull request #2789 from xinnan-tech/test_page_token
Bump to 0.8.11
2026-01-02 22:46:55 +08:00
hrz 619370f36c Bump to 0.8.11 2026-01-02 22:45:53 +08:00
欣南科技andGitHub ea6c0e691f Merge pull request #2788 from xinnan-tech/test_page_token
update:删除认证令牌输入框,从ota接口获取令牌
2026-01-02 22:42:48 +08:00
hrz de0d82f1e8 update:删除认证令牌输入框,从ota接口获取令牌 2026-01-02 22:41:37 +08:00
hrzandGitHub c8a26d9c5e Merge pull request #2623 from xinnan-tech/dependabot/pip/main/xiaozhi-server/psutil-7.1.3
build(deps): bump psutil from 7.0.0 to 7.1.3 in /main/xiaozhi-server
2026-01-02 21:50:21 +08:00
hrz d89b842586 update:保留适配本项目python3.10的版本 2026-01-02 21:48:34 +08:00
hrzandGitHub 6dd3ceea7e Merge pull request #2627 from xinnan-tech/dependabot/pip/main/xiaozhi-server/mcp-1.22.0
build(deps): bump mcp from 1.20.0 to 1.22.0 in /main/xiaozhi-server
2026-01-02 21:39:03 +08:00
hrzandGitHub 5fe68030b0 Merge pull request #2626 from xinnan-tech/dependabot/pip/main/xiaozhi-server/torchaudio-2.9.1
build(deps): bump torchaudio from 2.2.2 to 2.9.1 in /main/xiaozhi-server
2026-01-02 21:38:45 +08:00
hrzandGitHub 5215105440 Merge pull request #2625 from xinnan-tech/dependabot/pip/main/xiaozhi-server/websockets-15.0.1
build(deps): bump websockets from 14.2 to 15.0.1 in /main/xiaozhi-server
2026-01-02 21:38:17 +08:00
hrzandGitHub fb69eb751a Merge pull request #2624 from xinnan-tech/dependabot/pip/main/xiaozhi-server/modelscope-1.32.0
build(deps): bump modelscope from 1.23.2 to 1.32.0 in /main/xiaozhi-server
2026-01-02 21:37:33 +08:00
hrzandGitHub b903d51592 Merge pull request #2777 from xinnan-tech/update_blasr_paraformer
Update blasr paraformer
2026-01-02 20:39:43 +08:00
Sakura-RanChen d1f6774e58 update: 增加火山流式情感字段支持 2025-12-30 16:46:04 +08:00
CGDandGitHub 9e55df4141 Merge pull request #2778 from xinnan-tech/fix_mcp_rag
fix:serch_from_ragflow调用失败返回详细错误信息
2025-12-30 11:06:50 +08:00
rainv123 b169e9413f fix:serch_from_ragflow调用失败返回详细错误信息 2025-12-30 11:02:56 +08:00
Sakura-RanChen c511766e9d update: 增加豆包流式多语种识别 2025-12-29 16:38:23 +08:00
hrzandGitHub e6f9ceb7fa Merge pull request #2767 from xinnan-tech/web_Hidden_state
update:仅在 MQTT 服务可用时显示设备在线状态列
2025-12-27 14:24:01 +08:00
3030332422 eab8b48c9c update:仅在 MQTT 服务可用时显示设备在线状态列 2025-12-26 17:39:08 +08:00
FAN-yeB 46abb1009a update:初始化阿里云百炼asr供应器 手动模式待更改 2025-12-26 16:13:14 +08:00
CGDandGitHub 2a42d86db0 Merge pull request #2581 from xinnan-tech/py_tts_huoshan
自主选择链接复用功能
2025-12-26 14:47:01 +08:00
CGDandGitHub 68b0d9654c Merge branch 'main' into py_tts_huoshan 2025-12-26 14:46:51 +08:00
hrzandGitHub 8f96cf13ac Merge pull request #2764 from xinnan-tech/fix_ServerSideManage_emit-action
fix:修复服务器管理页面WebSocket连接失败和消息丢弃问题
2025-12-26 09:04:23 +08:00
3030332422 dbd179e507 fix:修复服务器管理页面WebSocket连接失败和消息丢弃问题 2025-12-25 18:11:03 +08:00
rainv123andGitHub 9541e85600 Merge pull request #2743 from xinnan-tech/py_async_server-mcp
update:将服务端MCP的初始化MCP服务从串行改为并发执行,并添加超时机制
2025-12-23 16:50:26 +08:00
rainv123andGitHub 9ecd829bfa Merge pull request #2747 from xinnan-tech/py_asr_doubao_stream
豆包ASR流式自定义语句停止时长
2025-12-23 16:49:56 +08:00
hrzandGitHub fd4cff73ae Merge pull request #2742 from xinnan-tech/manager-web-logo-i18n
add:登录、注册、首页、忘记密码页的多语言logo显示判断
2025-12-23 15:35:20 +08:00
hrz 1d6c4751de update:股东记忆模型的token大小 2025-12-23 14:42:33 +08:00
3030332422 e066c1d6a1 update:超时时间更改为10秒,并且删去无用的模块导入 2025-12-22 16:38:28 +08:00
3030332422 85f5404b3b update:添加锁保护避免MCP并发初始化时的竞态条件 2025-12-22 15:56:40 +08:00
Sakura-RanChen 33c5893d32 豆包ASR流式自定义语句停止时长 2025-12-22 15:23:55 +08:00
hrz b597cb0686 update:移除无用代码 2025-12-20 00:30:20 +08:00
hrz c5e288ca9c Merge branch 'main' into manager-web-logo-i18n 2025-12-19 23:46:43 +08:00
欣南科技andGitHub 38d984f225 Merge pull request #2744 from xinnan-tech/fix-delete-audio-data
分批次删除音频数据,避免数据库超时
2025-12-19 23:29:40 +08:00
hrz 628c642040 分批次删除音频数据,避免数据库超时 2025-12-19 23:27:35 +08:00
hrz d217cc4dd0 update:移除无用代码 2025-12-19 22:31:20 +08:00
rainv123 4f0e54b390 调整 2025-12-19 14:51:32 +08:00
rainv123 d7f89be31c add:登录、注册、首页、忘记密码页的多语言logo显示判断 2025-12-19 14:33:35 +08:00
rainv123 53bcee7032 Merge branch 'manager-web-logo-i18n' of https://github.com/xinnan-tech/xiaozhi-esp32-server into new 2025-12-19 14:06:19 +08:00
hrz 96991ae5ef add:多语言logo 2025-12-19 13:29:06 +08:00
欣南科技andGitHub 068bcde451 Merge pull request #2741 from xinnan-tech/openrz-patch-1
Update huoshan-streamTTS-voice-cloning.md
2025-12-19 12:12:44 +08:00
hrzandGitHub 0dda4f5646 Update huoshan-streamTTS-voice-cloning.md 2025-12-19 12:12:10 +08:00
3030332422 6c57ce9dd2 update:将服务端MCP的初始化MCP服务从串行改为并发执行,并添加超时机制 2025-12-19 11:49:42 +08:00
rainv123 6ac67a7e41 fix:调整 2025-12-19 09:55:18 +08:00
hrzandGitHub a23a585ad8 Merge pull request #2737 from xinnan-tech/py_prompt_update
update:提示词上下文按模板按需获取(位置/天气/动态上下文)
2025-12-18 18:20:01 +08:00
欣南科技andGitHub eba9562e81 Merge pull request #2738 from xinnan-tech/xiaozhi-server_ota_bin
Xiaozhi server ota bin
2025-12-18 18:12:44 +08:00
hrz ce49b409ac update:优化说明 2025-12-18 18:11:53 +08:00
hrz 7222f68d4d update:添加重要说明 2025-12-18 17:57:05 +08:00
hrz f3f0d62f12 add:添加单模块部署时,使用ota接口自动升级固件的教程 2025-12-18 17:38:58 +08:00
hrz 6e7c86e159 update:从vision_url配置里读取域名和端口号 2025-12-18 17:37:20 +08:00
hrzandGitHub f5565f6700 Merge pull request #2730 from rxchen/main
add basic "real" OTA support for single server deployment.
2025-12-18 17:29:31 +08:00
3030332422 e8d0bb0c54 update:提示词上下文按模板按需获取(位置/天气/动态上下文) 2025-12-18 16:34:37 +08:00
rui chen 33d70ccc96 get OTA address from websocket address config, if failed find ota_addr, if failed again, use local address. 2025-12-18 15:48:25 +08:00
rxchenandGitHub 53313a26af Merge branch 'xinnan-tech:main' into main 2025-12-18 15:26:51 +08:00
欣南科技andGitHub 7ba180752a Merge pull request #2735 from xinnan-tech/fix-nointent-weather
fix:无意图识别时,无"plugins"配置报错的bug
2025-12-18 11:23:36 +08:00
hrz 19736e66ad fix:无意图识别时,无"plugins"配置报错的bug 2025-12-18 11:18:49 +08:00
1cccuxandGitHub a00f8e4b76 Merge pull request #2734 from xinnan-tech/py_fix_listenmessage
fix:修复唤醒状态
2025-12-18 10:12:02 +08:00
FAN-yeB 7d9895cf5b 修复唤醒状态 2025-12-18 10:11:13 +08:00
rui chen d5f804bbb3 add basic OTA support for single server deployment, remove debug 2025-12-17 16:44:54 +08:00
rui chen 33a385cfa8 add basic OTA support for single server deployment
Committer: rxchen <rchentl@hotmail.com>
2025-12-17 16:26:35 +08:00
欣南科技andGitHub 109d924591 Merge pull request #2728 from xinnan-tech/update_provider_name
Update provider name
2025-12-17 12:00:13 +08:00
hrz a825eb3d4c Merge branch 'py_websocket_activate' into update_provider_name 2025-12-17 11:59:22 +08:00
hrz 3fb40677a4 update:纠正显示名称 2025-12-17 11:58:40 +08:00
hrzandGitHub 833971cd28 Merge pull request #2722 from xinnan-tech/py_websocket_activate
update: 增加websocket心跳机制
2025-12-17 11:51:14 +08:00
hrzandGitHub d14b14bce9 Merge pull request #2709 from xinnan-tech/py_audio_change
优化
2025-12-16 22:04:02 +08:00
hrz 53e26821ad update:补回前5个包提前发送的时间,因为发送完不等于播放完 2025-12-16 22:03:12 +08:00
Sakura-RanChen 1a7c06eb81 update: 增加websocket心跳机制 2025-12-16 16:39:13 +08:00
Packeting1 6cdbe45313 feat: 为阿里百炼流式TTS添加多语言音色支持、相关 音色配置; 2025-12-16 11:35:58 +08:00
hrzandGitHub 42a5b4892d Merge pull request #2712 from xinnan-tech/fix_bind_event
fix: 等待初始化成功设置事件
2025-12-15 22:46:31 +08:00
欣南科技andGitHub 9f4508b0c7 Merge pull request #2714 from xinnan-tech/update-doc
update:优化文档说明
2025-12-15 18:16:23 +08:00
hrz b2123ff01a update:优化文档说明 2025-12-15 18:15:35 +08:00
Sakura-RanChen 43ead841a4 fix: 等待初始化成功设置事件 2025-12-15 17:31:13 +08:00
Sakura-RanChen 33b4794e83 优化 2025-12-15 16:34:50 +08:00
欣南科技andGitHub fb91e6b1dd Merge pull request #2703 from xinnan-tech/test_page_pre_buffer
update:mqtt部署更新
2025-12-14 15:29:59 +08:00
hrz 4a4dbf123e update:mqtt部署更新 2025-12-14 15:29:13 +08:00
欣南科技andGitHub 0decedd6a1 Merge pull request #2701 from xinnan-tech/test_page_pre_buffer
update:优化测试页面缓冲音频播放
2025-12-14 14:25:25 +08:00
hrz d2e3a63418 update:优化测试页面缓冲音频播放 2025-12-14 14:24:39 +08:00
hrzandGitHub 15b3f0a8f5 Merge pull request #2695 from xinnan-tech/py_device_bind
update: 未绑定设备策略优化
2025-12-13 23:30:44 +08:00
hrz 5c261528d0 update:常用音频增加缓存,抵御高并发未授权设备访问 2025-12-13 23:10:40 +08:00
hrz 8b2bbec0b9 update:audio_to_data改成异步方法 2025-12-13 22:29:10 +08:00
hrz 8b3a4ad163 update:优化丢弃消息逻辑 2025-12-13 21:45:31 +08:00
hrz b3e272281b Merge branch 'main' into py_device_bind 2025-12-13 20:27:40 +08:00
hrz 06a90d6266 update:恢复audioRateController旧版本 2025-12-13 20:27:11 +08:00
欣南科技andGitHub 6eb7acaf73 Merge pull request #2698 from xinnan-tech/ws_token_auth
Ws token auth
2025-12-13 15:42:21 +08:00
hrz 2e092a7880 update:默认开启server.auth.enabled 2025-12-13 15:34:11 +08:00
hrzandGitHub 515f669546 Merge pull request #2685 from aixiamomo/fix/ws_token_auth
fix: 为Java项目的OTA接口实现WebSocket认证token生成功能,兼容Python端
2025-12-13 14:44:12 +08:00
hrz 7699596597 update:优化 2025-12-13 14:35:55 +08:00
欣南科技andGitHub 34dd325a1d Merge pull request #2697 from xinnan-tech/test_page_fix_abort
fix:edge_tts bug
2025-12-13 00:26:32 +08:00
hrz d33cdd978d fix:edge_tts bug 2025-12-13 00:25:54 +08:00
欣南科技andGitHub 0da9b4a69c Merge pull request #2696 from xinnan-tech/test_page_fix_abort
update:发送语音消息时打断机器人说话
2025-12-13 00:20:39 +08:00
hrz 7fc2eeaaa5 update:发送语音消息时打断机器人说话 2025-12-13 00:19:38 +08:00
hrzandGitHub 508529f66e Merge pull request #2694 from qingshuiyuyu/bugfix
fix:初始化ASR时,判断ASR类型取值错误修复
2025-12-13 00:16:37 +08:00
Sakura-RanChen dc170edbc1 update: 未绑定设备策略优化
fix: 音频队列竞态问题
2025-12-12 18:58:24 +08:00
shiyin 85e65cf9e8 fix:初始化ASR时,判断ASR类型取值错误修复 2025-12-12 18:28:46 +08:00
hrzandGitHub 41887ef431 Merge pull request #2686 from xinnan-tech/py_audio_await
Py audio await
2025-12-12 16:35:10 +08:00
hrz 5dbf796fa9 update:更新版本号 2025-12-12 16:33:50 +08:00
Sakura-RanChen 20f601e607 fix: 同步方法使用线程池避免阻塞 2025-12-12 15:39:30 +08:00
Sakura-RanChen c8d2d2255d fix: 音频影响线程问题 2025-12-12 15:08:40 +08:00
rainv123 5eaaf9f01d Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server into test 2025-12-12 12:55:02 +08:00
rainv123 d6697948c2 fix:使用java端做聊天记录总结 2025-12-12 12:53:59 +08:00
Sakura-RanChen f88abd7638 fix: 聆听设备误触发 2025-12-12 09:46:15 +08:00
Sakura-RanChen fbfc408e94 update: 长按设备ASR适配 2025-12-11 17:12:56 +08:00
panjingpeng 5ac1a1d6a5 fix: 为Java项目的OTA接口实现WebSocket认证token生成功能,确保与Python端完全兼容。 2025-12-11 16:08:45 +08:00
FAN-yeB 3a74b30a0e 更新 qwen3_asr_flash.py 2025-12-10 17:57:26 +08:00
hrzandGitHub fe0a6852e9 Merge pull request #2674 from xinnan-tech/web-context-i18n
update:上下文源功能的国际化
2025-12-10 16:20:03 +08:00
3030332422 bfa04743e2 update:上下文源功能的国际化 2025-12-10 10:26:27 +08:00
Sakura-RanChen 48094f7e37 update: 音频后台队列稳定发送 2025-12-09 18:06:56 +08:00
FAN-yeB 60521b0a7e update:长按说话不走VAD直接触发ASR识别 2025-12-09 14:42:50 +08:00
hrzandGitHub 401deff2c3 Merge pull request #2665 from xinnan-tech/update_performance_tester
update:统一流式测速工具统计时间区间,新增百炼平台流式TTS测速
2025-12-08 16:05:22 +08:00
FAN-yeB 6ffa325b73 update:统一流式测速工具统计时间区间,新增百炼平台流式TTS测速 2025-12-08 15:51:04 +08:00
hrzandGitHub eeedfc8ac9 Merge pull request #2660 from xinnan-tech/py_fix_device-id
fix:修改上下文源的device_id
2025-12-08 14:28:10 +08:00
3030332422 aba7172a03 update:修改上下文源的device_id 2025-12-08 14:25:11 +08:00
hrzandGitHub 192f6f198f Merge pull request #2653 from xinnan-tech/WebMenu
fix:样式调整以及翻译补充
2025-12-05 22:02:22 +08:00
rainv123 43c9d85b8f uptate:更新配置文档,增加在页面上开启功能的教程 2025-12-05 18:33:11 +08:00
rainv123 6d5935fb27 uptate:更新配置文档,增加在页面上开启功能的教程 2025-12-05 18:29:20 +08:00
rainv123 efbef4ff5b fix:样式调整以及翻译补充 2025-12-05 18:08:26 +08:00
欣南科技andGitHub e4c99b9351 Merge pull request #2652 from xinnan-tech/update-doc
Update doc
2025-12-05 17:40:12 +08:00
hrz 11e328ea6a update:优化描述 2025-12-05 17:39:34 +08:00
hrz 8879907e38 update:修改描述 2025-12-05 17:31:37 +08:00
hrzandGitHub 885d72c4f7 Merge pull request #2651 from xinnan-tech/WebMenu
fix:修改初始状态
2025-12-05 17:06:00 +08:00
hrzandGitHub 4538542c25 Merge branch 'main' into WebMenu 2025-12-05 17:05:51 +08:00
rainv123 3d509d75dc fix:修改初始状态 2025-12-05 17:01:40 +08:00
hrzandGitHub 0ef73d7c3e Merge pull request #2648 from xinnan-tech/py_add_protocol
添加数据上下文填充功能
2025-12-05 15:47:35 +08:00
hrzandGitHub 6da90c6a97 Merge branch 'main' into py_add_protocol 2025-12-05 15:47:26 +08:00
hrzandGitHub bacdb794da Merge pull request #2646 from xinnan-tech/WebMenu
uptate:增加功能管理菜单
2025-12-05 15:41:25 +08:00
rainv123 2ece3f399b fix:修改初始状态为false 2025-12-05 15:39:05 +08:00
rainv123 68b539db15 fix:修改 2025-12-05 15:29:28 +08:00
3030332422 7a7bfa26f6 update:修改日志 2025-12-05 14:50:26 +08:00
rainv123 5f229351c8 uptate:增加功能管理菜单 2025-12-05 14:35:53 +08:00
3030332422 82125c4933 update:说明文档添加测试用例 2025-12-05 14:22:07 +08:00
3030332422 33f75d26c1 update:优化前端页面样式 2025-12-05 14:02:50 +08:00
3030332422 3c4d702bc1 update:添加数据上下文填充前端页面 2025-12-05 11:24:57 +08:00
3030332422 db8d100edb update:添加数据库表,添加数据上下文填充功能后端部分,全模块实现数据上下文填充功能 2025-12-05 10:50:44 +08:00
3030332422 6f7e8978ca update:添加数据上下文填充功能的说明文档 2025-12-05 10:40:47 +08:00
hrzandGitHub 856bf2c574 Merge pull request #2641 from qingshuiyuyu/bugfix
修复注册设备接口,验证码死循环BUG
2025-12-04 18:54:33 +08:00
hrzandGitHub b10cfa6b24 Merge pull request #2642 from xinnan-tech/py_audio_fix
fix: 状态修正
2025-12-04 18:53:00 +08:00
Sakura-RanChen d3d329bd43 fix: 状态修正 2025-12-04 18:45:49 +08:00
shiyin f358011900 修复注册设备接口,验证码死循环BUG 2025-12-04 18:32:57 +08:00
3030332422 cd6c3e4f79 update:添加数据上下文填充功能,单模块实现数据上下文填充功能 2025-12-04 11:04:16 +08:00
欣南科技andGitHub ee7342de9d Merge pull request #2639 from xinnan-tech/fix-handshake
fix:过滤8000端口使用https访问时报错日志
2025-12-03 22:50:48 +08:00
hrz 6120d49a35 fix:过滤8000端口使用https访问时报错日志 2025-12-03 22:49:32 +08:00
欣南科技andGitHub 381f8ea578 Merge pull request #2628 from xinnan-tech/py_await
Py await
2025-12-01 18:51:43 +08:00
hrz a3e2073dc6 Bump to 0.8.9 2025-12-01 18:51:24 +08:00
hrz f7bd858fea update:优化绑定间隔 2025-12-01 18:46:42 +08:00
dependabot[bot]andGitHub 80d10cf4b5 build(deps): bump mcp from 1.20.0 to 1.22.0 in /main/xiaozhi-server
Bumps [mcp](https://github.com/modelcontextprotocol/python-sdk) from 1.20.0 to 1.22.0.
- [Release notes](https://github.com/modelcontextprotocol/python-sdk/releases)
- [Changelog](https://github.com/modelcontextprotocol/python-sdk/blob/main/RELEASE.md)
- [Commits](https://github.com/modelcontextprotocol/python-sdk/compare/v1.20.0...v1.22.0)

---
updated-dependencies:
- dependency-name: mcp
  dependency-version: 1.22.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-12-01 10:28:38 +00:00
dependabot[bot]andGitHub ba1fb16f2d build(deps): bump torchaudio from 2.2.2 to 2.9.1 in /main/xiaozhi-server
Bumps [torchaudio](https://github.com/pytorch/audio) from 2.2.2 to 2.9.1.
- [Release notes](https://github.com/pytorch/audio/releases)
- [Commits](https://github.com/pytorch/audio/compare/v2.2.2...v2.9.1)

---
updated-dependencies:
- dependency-name: torchaudio
  dependency-version: 2.9.1
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-12-01 10:28:33 +00:00
dependabot[bot]andGitHub 5f574667d3 build(deps): bump websockets from 14.2 to 15.0.1 in /main/xiaozhi-server
Bumps [websockets](https://github.com/python-websockets/websockets) from 14.2 to 15.0.1.
- [Release notes](https://github.com/python-websockets/websockets/releases)
- [Commits](https://github.com/python-websockets/websockets/compare/14.2...15.0.1)

---
updated-dependencies:
- dependency-name: websockets
  dependency-version: 15.0.1
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-12-01 10:28:26 +00:00
dependabot[bot]andGitHub c470bb1db1 build(deps): bump modelscope in /main/xiaozhi-server
Bumps [modelscope](https://github.com/modelscope/modelscope) from 1.23.2 to 1.32.0.
- [Release notes](https://github.com/modelscope/modelscope/releases)
- [Commits](https://github.com/modelscope/modelscope/compare/v1.23.2...v1.32.0)

---
updated-dependencies:
- dependency-name: modelscope
  dependency-version: 1.32.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-12-01 10:28:21 +00:00
dependabot[bot]andGitHub 26ed8ca33f build(deps): bump psutil from 7.0.0 to 7.1.3 in /main/xiaozhi-server
Bumps [psutil](https://github.com/giampaolo/psutil) from 7.0.0 to 7.1.3.
- [Changelog](https://github.com/giampaolo/psutil/blob/master/HISTORY.rst)
- [Commits](https://github.com/giampaolo/psutil/compare/release-7.0.0...release-7.1.3)

---
updated-dependencies:
- dependency-name: psutil
  dependency-version: 7.1.3
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-12-01 10:28:16 +00:00
hrz 313f5b1982 Merge branch 'main' into py_await 2025-12-01 16:55:19 +08:00
hrz f3f0c24e16 update:减少日志产生 2025-12-01 16:38:10 +08:00
hrzandGitHub fe2539ac97 Merge pull request #2621 from xinnan-tech/fix_color
添加模型的输入框颜色修改
2025-12-01 15:54:44 +08:00
rainv123 dd1a430c21 fix:添加模型组件的输入框颜色修改 2025-12-01 10:56:19 +08:00
rainv123 621634697e Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server into test 2025-12-01 10:52:49 +08:00
Sakura-RanChen fb5c35e6c9 后台初始化组件,gc全局化避免每次回收触发GIL锁,音频定时发送 2025-11-28 16:12:19 +08:00
欣南科技andGitHub 228596dbe3 Merge pull request #2608 from xinnan-tech/openrz-patch-2
Update ragflow-integration.md
2025-11-26 12:21:52 +08:00
hrzandGitHub 74baa85340 Update ragflow-integration.md 2025-11-26 12:21:12 +08:00
hrzandGitHub 8f6e9ca54d Merge pull request #2604 from xinnan-tech/fix_
fix:允许ota,mcp,声纹地址为null
2025-11-26 11:31:54 +08:00
rainv123andGitHub 19420d0bb3 Merge pull request #2601 from xinnan-tech/main-upgrade
Main upgrade
2025-11-26 11:31:08 +08:00
rainv123 770c8c8a9f Merge branch 'main-upgrade' of https://github.com/xinnan-tech/xiaozhi-esp32-server into test 2025-11-26 10:38:56 +08:00
rainv123 252a34090a Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server into test 2025-11-26 10:38:01 +08:00
hrzandGitHub bde5ef4594 Merge pull request #2602 from xinnan-tech/remove_gc
update:将垃圾回收放置ws关闭处,避免频繁操作
2025-11-25 17:49:48 +08:00
rainv123 6eddf081de fix:允许ota,mcp,声纹地址为null 2025-11-25 17:38:55 +08:00
hrz 4645c50f6d update:将垃圾回收放置ws关闭处,避免频繁操作 2025-11-25 16:11:48 +08:00
hrzandGitHub 3f6e66b1ce Merge pull request #2588 from xinnan-tech/fix_
调整
2025-11-25 15:00:49 +08:00
Chingfeng Li 106889b0dd 更新默认编码 2025-11-25 13:52:18 +08:00
myifengandGitHub 785b43464b Merge pull request #2593 from xinnan-tech/dependabot/pip/main/xiaozhi-server/dashscope-1.25.2
build(deps): bump dashscope from 1.24.6 to 1.25.2 in /main/xiaozhi-server
2025-11-25 13:36:19 +08:00
myifengandGitHub 03da60dd12 Merge pull request #2594 from xinnan-tech/dependabot/pip/main/xiaozhi-server/vosk-0.3.45
build(deps): bump vosk from 0.3.44 to 0.3.45 in /main/xiaozhi-server
2025-11-25 13:35:51 +08:00
myifengandGitHub ec372f65ec Merge pull request #2596 from xinnan-tech/dependabot/pip/main/xiaozhi-server/openai-2.8.1
build(deps): bump openai from 2.7.1 to 2.8.1 in /main/xiaozhi-server
2025-11-25 13:35:18 +08:00
myifengandGitHub cfba4c45b6 Merge pull request #2597 from xinnan-tech/dependabot/pip/main/xiaozhi-server/sherpa-onnx-1.12.17
build(deps): bump sherpa-onnx from 1.12.15 to 1.12.17 in /main/xiaozhi-server
2025-11-25 13:34:37 +08:00
dependabot[bot]andGitHub 657d03f49d build(deps): bump sherpa-onnx in /main/xiaozhi-server
Bumps [sherpa-onnx](https://github.com/k2-fsa/sherpa-onnx) from 1.12.15 to 1.12.17.
- [Release notes](https://github.com/k2-fsa/sherpa-onnx/releases)
- [Changelog](https://github.com/k2-fsa/sherpa-onnx/blob/master/CHANGELOG.md)
- [Commits](https://github.com/k2-fsa/sherpa-onnx/compare/v1.12.15...v1.12.17)

---
updated-dependencies:
- dependency-name: sherpa-onnx
  dependency-version: 1.12.17
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-11-24 09:10:36 +00:00
dependabot[bot]andGitHub b4dd78b2ed build(deps): bump openai from 2.7.1 to 2.8.1 in /main/xiaozhi-server
Bumps [openai](https://github.com/openai/openai-python) from 2.7.1 to 2.8.1.
- [Release notes](https://github.com/openai/openai-python/releases)
- [Changelog](https://github.com/openai/openai-python/blob/main/CHANGELOG.md)
- [Commits](https://github.com/openai/openai-python/compare/v2.7.1...v2.8.1)

---
updated-dependencies:
- dependency-name: openai
  dependency-version: 2.8.1
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-11-24 09:10:27 +00:00
dependabot[bot]andGitHub 1ebe7b8f71 build(deps): bump vosk from 0.3.44 to 0.3.45 in /main/xiaozhi-server
Bumps [vosk](https://github.com/alphacep/vosk-api) from 0.3.44 to 0.3.45.
- [Release notes](https://github.com/alphacep/vosk-api/releases)
- [Commits](https://github.com/alphacep/vosk-api/commits/v0.3.45)

---
updated-dependencies:
- dependency-name: vosk
  dependency-version: 0.3.45
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-11-24 09:10:17 +00:00
dependabot[bot]andGitHub 30e7a71497 build(deps): bump dashscope in /main/xiaozhi-server
Bumps [dashscope](https://dashscope.aliyun.com/) from 1.24.6 to 1.25.2.

---
updated-dependencies:
- dependency-name: dashscope
  dependency-version: 1.25.2
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-11-24 09:10:14 +00:00
Sakura-RanChenandGitHub ff1de6c7e3 Merge pull request #2592 from xinnan-tech/fix-remove-openai-topk
移除OpenAI不支持的参数top_k
2025-11-24 16:50:00 +08:00
Chingfeng Li 9e39883276 移除OpenAI不支持的参数top_k 2025-11-24 16:29:46 +08:00
rainv123 0f249ae1db fix:调整 2025-11-24 10:04:40 +08:00
rainv123 31ada101b0 fix:登录页面样式调整 2025-11-24 10:03:06 +08:00
欣南科技andGitHub 1b52a7168b Merge pull request #2584 from xinnan-tech/py_test_websocket
fix: decoder 引发的内存泄漏
2025-11-23 15:15:12 +08:00
hrz 83a3d0eabe update:优化代码 2025-11-23 15:11:35 +08:00
欣南科技andGitHub 080c6b4dfb Merge pull request #2583 from xinnan-tech/fix-audio-bug
update:未授权设备会话周期管理
2025-11-23 12:31:03 +08:00
hrz ea50288c5b update:未授权设备会话周期管理 2025-11-23 12:29:18 +08:00
3030332422 5d439a4168 update:内存泄漏 2025-11-23 11:55:38 +08:00
Sakura-RanChen 3c46e16494 自主选择链接复用功能 2025-11-22 16:19:13 +08:00
hrzandGitHub 39cd0fe1ac Merge pull request #2579 from xinnan-tech/rag_fix
fix:输入问题后按回车键执行测试
2025-11-22 16:16:57 +08:00
rainv123 16c0be74c9 fix:输入问题后按回车键执行测试 2025-11-22 15:28:00 +08:00
hrzandGitHub 732828612e Merge pull request #2574 from xinnan-tech/fix_chatHistoryConf
fix:设置默认记忆不是【无记忆】时,默认chatHistoryConf=2
2025-11-22 14:51:07 +08:00
rainv123 7344b81c98 fix:模型id不能为文字或者空格 2025-11-22 11:59:34 +08:00
rainv123 1e5b03edb5 fix:设置有记忆的模型为默认时,chatHistoryConf=2 2025-11-21 18:39:57 +08:00
rainv123 b15b5af349 fix:设置默认记忆不是【无记忆】时,默认chatHistoryConf=2 2025-11-21 18:31:01 +08:00
Sakura-RanChenandGitHub 5242430b6d Merge pull request #2568 from xinnan-tech/fix-allillm-topk
FIX: Completions.create() got an unexpected keyword argument 'top_k'
2025-11-21 16:01:44 +08:00
Sakura-RanChenandGitHub c39968a5bf Merge pull request #2571 from xinnan-tech/add-mcp-callback
Add ProgressFnT LoggingFnT...
2025-11-21 15:57:49 +08:00
Chingfeng Li 8b4f9ab20e Add ProgressFnT LoggingFnT... 2025-11-21 14:27:18 +08:00
hrzandGitHub 527e6050ce Merge pull request #2569 from xinnan-tech/remove-unused-active_connections
Remove unused code
2025-11-21 13:25:21 +08:00
Chingfeng Li 1d162cb874 Remove unused code 2025-11-21 11:46:32 +08:00
Chingfeng Li c197f5c942 FIX: Completions.create() got an unexpected keyword argument 'top_k' 2025-11-21 11:19:03 +08:00
hrzandGitHub 3a4c0bb888 Merge pull request #2567 from xinnan-tech/py_fix_audio
fix: 长按设备过于灵敏
2025-11-21 10:47:27 +08:00
Sakura-RanChen 38d50f4275 fix: 长按设备过于灵敏 2025-11-21 10:11:50 +08:00
hrzandGitHub 25a3b648fc Merge pull request #2562 from xinnan-tech/rag_fix
Rag fix
2025-11-20 15:06:39 +08:00
hrzandGitHub d3638164fe Merge pull request #2543 from xinnan-tech/py_fix_LLMChat
update: 增加工具并行调用(需大模型支持,如qwen-plus)增加递归深度限制,避免循环
2025-11-20 11:20:47 +08:00
hrzandGitHub 1c2a61fd34 Merge pull request #2551 from xinnan-tech/update-client-speaking
根据语音包发送更新client_is_speaking状态
2025-11-20 11:16:59 +08:00
rainv123 f81c4aa970 fix:错误码翻译补充 2025-11-20 09:40:50 +08:00
rainv123 3caa4ec6e8 fix:知识库增加适配器 2025-11-20 09:38:31 +08:00
hrzandGitHub 5e48a2274a Merge pull request #2558 from xinnan-tech/i8n_fix_1
角色配置页面翻译调整
2025-11-19 10:35:57 +08:00
hrzandGitHub 7f7c980522 Merge pull request #2550 from xinnan-tech/py_openai_fix
参数补充
2025-11-18 22:07:01 +08:00
Chingfeng Li bbab4ebc86 语音发送的开始和结尾仅与TTS相关,不关心LLM;LLM调用MCP时异步消息会引发问题 2025-11-18 15:00:55 +08:00
Sakura-RanChen 8bae0c8ac5 参数补充 2025-11-18 14:40:17 +08:00
Chingfeng Li a5e4aa1300 根据语音包发送更新client_is_speaking状态 2025-11-18 14:37:50 +08:00
Sakura-RanChen 48313bc302 fix: 多工具时协程不阻塞,仅需等待最慢那个 2025-11-18 11:10:18 +08:00
rainv123 f0261a8236 fix:角色配置页面翻译和样式调整 2025-11-18 10:21:53 +08:00
rainv123 84c194dd34 Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server into Knowledge-Base 2025-11-18 10:21:20 +08:00
hrzandGitHub 920e24b4ff Merge pull request #2547 from xinnan-tech/yinseyemian
补充图标调整
2025-11-17 19:06:59 +08:00
LiJinHui 43d82343a3 补充图标调整 2025-11-17 18:37:40 +08:00
rainv123 320d246a83 Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server into Knowledge-Base 2025-11-17 18:34:03 +08:00
hrzandGitHub faea50f99d Merge pull request #2545 from xinnan-tech/yinseyemian
调整试听按钮大小
2025-11-17 18:31:01 +08:00
hrzandGitHub ef59798d70 Merge pull request #2544 from xinnan-tech/i8n_fix_1
fix:修改错误简写和样式调整
2025-11-17 18:13:21 +08:00
rainv123 6781eca56a Merge branch 'yinseyemian' of https://github.com/xinnan-tech/xiaozhi-esp32-server into Knowledge-Base 2025-11-17 18:05:31 +08:00
LiJinHui c1f0853271 调整 2025-11-17 18:05:10 +08:00
rainv123 d4cda5c104 Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server into Knowledge-Base 2025-11-17 18:02:34 +08:00
rainv123 f9bd5bb995 fix:删除多余空格 2025-11-17 18:00:38 +08:00
rainv123 f1c2b5cb3f fix:重新修改简写 2025-11-17 17:41:12 +08:00
rainv123 579af29a73 fix:修改错误简写和样式调整 2025-11-17 17:30:27 +08:00
Sakura-RanChen 0b0fd142a3 update: 增加工具并行调用(需大模型支持,如qwen-plus)增加递归深度限制,避免循环 2025-11-17 17:14:56 +08:00
hrzandGitHub 9e20a53ca0 Merge pull request #2530 from xinnan-tech/yinseyemian
完善音色试听页面
2025-11-17 17:10:09 +08:00
hrzandGitHub 8768240efe Merge branch 'main' into yinseyemian 2025-11-17 17:09:53 +08:00
LJH-rgszeandGitHub b429760bd5 Merge pull request #2542 from xinnan-tech/i8n_fix_1
fix:翻译及样式优化
2025-11-17 17:01:41 +08:00
rainv123 218cd35c3d fix:修改 2025-11-17 16:57:45 +08:00
rainv123 146a039c19 fix:翻译及样式优化 2025-11-17 11:53:30 +08:00
LiJinHui 9a3e1fa9c8 补充 2025-11-17 11:39:11 +08:00
LiJinHui fa194af9bb 调整 2025-11-17 10:15:53 +08:00
欣南科技andGitHub 7bdb2efd8b Merge pull request #2539 from xinnan-tech/hot-fix
fix:网页测试工具和manager-web关键缓存冲突
2025-11-16 17:42:26 +08:00
hrz 6c325f68ac fix:网页测试工具和manager-web关键缓存冲突 2025-11-16 17:40:39 +08:00
欣南科技andGitHub 3113c5cf62 Merge pull request #2538 from xinnan-tech/i18n_fix_new
I18n fix new
2025-11-16 17:20:37 +08:00
hrz c6527d7df9 update:优化翻译 2025-11-16 17:18:49 +08:00
欣南科技andGitHub e86974b7b2 Merge pull request #2537 from xinnan-tech/test-page-css
update:优化测试页面音波样式
2025-11-16 14:41:12 +08:00
hrz e37618423d update:优化测试页面音波样式 2025-11-16 14:39:53 +08:00
欣南科技andGitHub c640662663 Merge pull request #2536 from xinnan-tech/test-page-css
Test page css
2025-11-16 14:09:26 +08:00
hrz 3fe1acd153 update:优化样式代码顺序 2025-11-16 14:08:43 +08:00
hrz 9eaaec75a5 update:音频测试页面增加打断功能 2025-11-16 13:59:27 +08:00
欣南科技andGitHub 1ad86153c7 Merge pull request #2535 from xinnan-tech/fix-test-page
update:去除无用样式
2025-11-16 01:51:10 +08:00
hrz 09a58ba5bf update:去除无用样式 2025-11-16 01:50:29 +08:00
欣南科技andGitHub 867ab7aca2 Merge pull request #2534 from xinnan-tech/fix-test-page
update:统一js文件名称格式
2025-11-16 01:36:56 +08:00
hrz 5dda2f731a update:统一js文件名称格式 2025-11-16 01:34:23 +08:00
欣南科技andGitHub 71c6f04f99 Merge pull request #2533 from xinnan-tech/fix-test-page
update:优化音频测试页面的样式
2025-11-16 01:27:18 +08:00
hrz 80c7295784 update:优化音频测试页面的样式 2025-11-16 01:26:10 +08:00
欣南科技andGitHub e2869f1f50 Merge pull request #2532 from xinnan-tech/fix-test-page
update:修复测试工具页面声音卡顿问题
2025-11-15 17:54:33 +08:00
hrz dc55f8d7ab update:修复测试工具页面声音卡顿问题 2025-11-15 17:53:36 +08:00
rainv123 925ed1ec0c uptete:移动端智控台添加德语、越南语 2025-11-15 17:52:40 +08:00
rainv123 8064cdf1a2 uptate:智控台增加德语、越南语 2025-11-15 17:24:47 +08:00
LiJinHui 46b3757c13 补充 2025-11-15 16:36:59 +08:00
LiJinHui 5050ea7ac1 修复添加克隆音色后试听按钮不显示的bug 2025-11-15 16:06:58 +08:00
rainv123 bdd0ec41d5 fix:manager-api国际化优化 2025-11-15 14:55:43 +08:00
LiJinHui d121b4d0e1 删除多余代码 2025-11-15 10:48:52 +08:00
LiJinHui d45bf5f60a 补充 2025-11-15 10:45:22 +08:00
LiJinHui 94a43432fc 调整音色的前端页面,提供更方便的试听 2025-11-15 09:51:58 +08:00
欣南科技andGitHub b55e5870db Merge pull request #2528 from xinnan-tech/translation
Translation
2025-11-15 00:17:57 +08:00
hrz 2afd90f2d4 update:多语言项目自述文件 2025-11-15 00:17:11 +08:00
hrzandGitHub 4001378904 Merge pull request #2473 from Kris-YIN/claude/translate-readme-german-011CUpaqWR8pn9EPKofKGwpx
Add German translation of README
2025-11-14 23:52:53 +08:00
欣南科技andGitHub 6c313e60eb Merge pull request #2527 from xinnan-tech/update-version
Bump to 0.8.8
2025-11-14 23:18:12 +08:00
hrz 3f6fd3144b Bump to 0.8.8 2025-11-14 23:17:21 +08:00
欣南科技andGitHub e5e4aa4836 Merge pull request #2526 from xinnan-tech/KnowledgeBase_fix
uptate:根据知识库id获取rag配置,删除获取默认rag配置方法
2025-11-14 19:18:48 +08:00
rainv123 18e84a90c9 uptate:根据知识库id获取rag配置,删除获取默认rag配置方法 2025-11-14 16:56:06 +08:00
hrzandGitHub b37f87d231 Merge pull request #2523 from xinnan-tech/fix_openai_LLM
fix: 去除默认值配置
2025-11-13 21:54:43 +08:00
欣南科技andGitHub 7c592e33fe Merge pull request #2522 from xinnan-tech/KnowledgeBase_fix
Knowledge base fix
2025-11-13 17:56:01 +08:00
hrz 471749f6ff update:优化模型名称及排列顺序 2025-11-13 17:51:16 +08:00
Sakura-RanChen 637b528aad fix: 去除默认值配置 2025-11-13 17:08:21 +08:00
hrzandGitHub 502ef997d4 Merge branch 'main' into KnowledgeBase_fix 2025-11-13 16:51:01 +08:00
hrz b76e354b86 update:删除多余代码 2025-11-13 16:40:50 +08:00
hrzandGitHub 988bb90528 Merge pull request #2521 from qingshuiyuyu/bugfix
火山双流式TTS日志bug修复
2025-11-13 16:20:40 +08:00
shiyin 578ed62d18 火山双流式TTS日志bug修复 2025-11-13 14:28:21 +08:00
rainv123 0d6d44bed9 fix:还原误删的“}” 2025-11-13 14:27:58 +08:00
rainv123 918c47669c fix:删除无用方法 2025-11-13 14:08:39 +08:00
rainv123 8ad22dc082 fix:移除触发器,在DAO层添加了删除方法 2025-11-13 12:35:07 +08:00
欣南科技andGitHub c3d94a5c21 Merge pull request #2519 from xinnan-tech/openrz-patch-1
Update ragflow-integration.md
2025-11-13 12:16:48 +08:00
hrzandGitHub 48d72e373e Update ragflow-integration.md 2025-11-13 12:15:54 +08:00
hrzandGitHub c667c8ee79 Merge pull request #2514 from Harold-Guo/fix/web_audio
fix(web-audio): Fix AudioWorklet recording and AudioContext state issues
2025-11-13 11:33:56 +08:00
欣南科技andGitHub aebc1c1597 Merge pull request #2518 from xinnan-tech/update-doc
Update doc
2025-11-13 11:06:29 +08:00
hrz cd113a5351 update:插件名称前,不同颜色的点易让人产生误解 2025-11-13 11:01:28 +08:00
rainv123 b7ca7cbdb5 fix:知识库描述字数增加到300字 2025-11-13 10:54:55 +08:00
rainv123 b3b270e999 Merge branch 'KnowledgeBase_fix' of https://github.com/xinnan-tech/xiaozhi-esp32-server into Knowledge-Base 2025-11-13 10:50:29 +08:00
rainv123 ad3c420191 fix:当删除知识库记录时,自动删除相关的插件映射记录 2025-11-13 10:49:15 +08:00
hrz 8a4f786c0b update:ragflow教程升级到0.22.0 2025-11-13 10:44:00 +08:00
欣南科技andGitHub d99d122416 Merge pull request #2517 from xinnan-tech/knowledge-base-fix
Knowledge base fix
2025-11-12 19:14:05 +08:00
hrzandGitHub 5a5f670272 Merge pull request #2476 from xinnan-tech/py-bump-test
Py bump test
2025-11-12 18:45:46 +08:00
hrz be533b7814 update:调整numpy回到1.x版本 2025-11-12 18:45:15 +08:00
hrzandGitHub 9a6e5c9f33 Merge pull request #2459 from xinnan-tech/dependabot/pip/main/xiaozhi-server/numpy-2.3.4
build(deps): bump numpy from 1.26.4 to 2.3.4 in /main/xiaozhi-server
2025-11-12 17:48:26 +08:00
hrzandGitHub 9d420b4304 Merge branch 'py-bump-test' into dependabot/pip/main/xiaozhi-server/numpy-2.3.4 2025-11-12 17:48:16 +08:00
hrzandGitHub acb1fdc51d Merge pull request #2425 from xinnan-tech/dependabot/pip/main/xiaozhi-server/websockets-15.0.1
build(deps): bump websockets from 14.2 to 15.0.1 in /main/xiaozhi-server
2025-11-12 17:46:44 +08:00
hrzandGitHub c6f1fa96bb Merge branch 'py-bump-test' into dependabot/pip/main/xiaozhi-server/websockets-15.0.1 2025-11-12 17:41:35 +08:00
hrzandGitHub 7315e2e5d2 Merge pull request #2497 from xinnan-tech/dependabot/pip/main/xiaozhi-server/cozepy-0.20.0
build(deps): bump cozepy from 0.19.0 to 0.20.0 in /main/xiaozhi-server
2025-11-12 17:39:33 +08:00
hrzandGitHub 429fa0032a Merge pull request #2496 from xinnan-tech/dependabot/pip/main/xiaozhi-server/aiohttp-cors-0.8.1
build(deps): bump aiohttp-cors from 0.7.0 to 0.8.1 in /main/xiaozhi-server
2025-11-12 17:38:35 +08:00
hrzandGitHub f2a7016119 Merge branch 'py-bump-test' into dependabot/pip/main/xiaozhi-server/aiohttp-cors-0.8.1 2025-11-12 17:38:27 +08:00
hrzandGitHub 47e199a59e Merge pull request #2495 from xinnan-tech/dependabot/pip/main/xiaozhi-server/httpx-0.28.1
build(deps): bump httpx from 0.27.2 to 0.28.1 in /main/xiaozhi-server
2025-11-12 17:37:28 +08:00
hrzandGitHub 172b097a11 Merge branch 'py-bump-test' into dependabot/pip/main/xiaozhi-server/httpx-0.28.1 2025-11-12 17:37:15 +08:00
hrz 4f2693505b update:标注不升级的组件 2025-11-12 17:35:42 +08:00
rainv123 ae321b26af fix:知识库页面文本修改 2025-11-12 17:02:53 +08:00
rainv123 6de4a29dbf Merge branch 'knowledge-base-fix' of https://github.com/xinnan-tech/xiaozhi-esp32-server into Knowledge-Base 2025-11-12 16:46:57 +08:00
rainv123 37f0d2398e fix:删除空行 2025-11-12 16:46:20 +08:00
hrz 5c68be6636 update:优化引用答复 2025-11-12 16:46:04 +08:00
rainv123 ee2ac5f46f fix:优化rag配置的验证和报错 2025-11-12 16:45:26 +08:00
rainv123 99bc55f256 Merge branch 'knowledge-base-fix' of https://github.com/xinnan-tech/xiaozhi-esp32-server into Knowledge-Base 2025-11-12 15:57:51 +08:00
rainv123 5d860e5b45 fix:增加知识库操作时系统异常报错 2025-11-12 15:55:36 +08:00
hrz a2069456b2 Merge branch 'main' into knowledge-base-fix 2025-11-12 15:10:57 +08:00
rainv123 2dc0c101c6 fix:错误信息日志补充 2025-11-12 15:04:05 +08:00
Harold 4fc306bccf fix(web-audio): Fix AudioWorklet recording and AudioContext state issues 2025-11-12 15:00:39 +08:00
rainv123 4966a42ed8 fix:还原误删的提示词 2025-11-12 14:41:27 +08:00
rainv123 c3ffa9d393 uptate:增加对字符串的解析以及格式调整 2025-11-12 12:32:16 +08:00
rainv123 f16906ad9d uptate:增加知识库相关提示词 2025-11-12 12:28:08 +08:00
rainv123 1b28e8ddcb uptate:ragflow配置教程补充 2025-11-12 12:25:53 +08:00
hrzandGitHub ff3e308ce8 Merge pull request #2500 from xinnan-tech/py-test-mem0.ai
fix:修复 Mem0 搜索缺少 filters 导致的报错
2025-11-11 22:53:47 +08:00
hrzandGitHub f6219bd58e Merge pull request #2506 from xinnan-tech/main-logs-level
优化日志级别,生产环境中INFO避免打印过多日志
2025-11-11 22:53:15 +08:00
欣南科技andGitHub 92bd207ae5 Merge pull request #2505 from xinnan-tech/hot-fix
fix:mqtt配置中的websocket连接
2025-11-11 18:53:23 +08:00
hrz bbca561dd0 fix:mqtt配置中的websocket连接 2025-11-11 18:52:41 +08:00
Chingfeng Li 833f379aa1 优化日志级别,生产环境中INFO避免打印过多日志 2025-11-11 18:52:14 +08:00
hrz de0eda1b6c update:增加ragflow部署教程 2025-11-11 18:47:25 +08:00
rainv123 96da84dad4 fix:错误处理优化 2025-11-11 16:02:46 +08:00
3030332422 59f55a4187 fix:修复 Mem0 搜索缺少 filters 导致的报错 2025-11-11 09:23:39 +08:00
rainv123 ab5e91c07b fix:优化知识库页面错误信息 2025-11-10 18:41:03 +08:00
dependabot[bot]andGitHub dca5f2c837 build(deps): bump cozepy from 0.19.0 to 0.20.0 in /main/xiaozhi-server
Bumps cozepy from 0.19.0 to 0.20.0.

---
updated-dependencies:
- dependency-name: cozepy
  dependency-version: 0.20.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-11-10 08:52:49 +00:00
dependabot[bot]andGitHub 33c60ba320 build(deps): bump aiohttp-cors in /main/xiaozhi-server
Bumps [aiohttp-cors](https://github.com/aio-libs/aiohttp-cors) from 0.7.0 to 0.8.1.
- [Release notes](https://github.com/aio-libs/aiohttp-cors/releases)
- [Changelog](https://github.com/aio-libs/aiohttp-cors/blob/master/CHANGES.rst)
- [Commits](https://github.com/aio-libs/aiohttp-cors/compare/v0.7.0...v0.8.1)

---
updated-dependencies:
- dependency-name: aiohttp-cors
  dependency-version: 0.8.1
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-11-10 08:52:47 +00:00
dependabot[bot]andGitHub a4fad9e796 build(deps): bump httpx from 0.27.2 to 0.28.1 in /main/xiaozhi-server
Bumps [httpx](https://github.com/encode/httpx) from 0.27.2 to 0.28.1.
- [Release notes](https://github.com/encode/httpx/releases)
- [Changelog](https://github.com/encode/httpx/blob/master/CHANGELOG.md)
- [Commits](https://github.com/encode/httpx/compare/0.27.2...0.28.1)

---
updated-dependencies:
- dependency-name: httpx
  dependency-version: 0.28.1
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-11-10 08:52:44 +00:00
rainv123 ee0d89d8d2 fix:知识库【编辑】【删除】操作逻辑修改 2025-11-10 15:14:32 +08:00
rainv123 5bead86fd5 fix:知识库描述增加校验 2025-11-10 15:11:48 +08:00
rainv123 79eefb651f update:角色配置页面mcp图标显示优化 2025-11-10 15:10:59 +08:00
欣南科技andGitHub cbb74170ea Merge pull request #2493 from xinnan-tech/hot-fix
fix:删除知识库时,需要删除缓存数据
2025-11-10 12:25:10 +08:00
hrz c6835ddd11 fix:删除知识库时,需要删除缓存数据 2025-11-10 12:24:32 +08:00
欣南科技andGitHub c57427efbc Merge pull request #2470 from xinnan-tech/Knowledge-Base
Add:Knowledge base
2025-11-09 00:36:07 +08:00
hrz 466eae4c76 update:优化知识库输入提示 2025-11-09 00:25:12 +08:00
hrz 4b7b80aa2d update:同名知识库错误提示 2025-11-09 00:13:24 +08:00
hrz dc5ff21002 update:优化rag模型列表接口 2025-11-08 23:44:40 +08:00
hrz 75807c5845 Merge branch 'main' into Knowledge-Base 2025-11-08 16:25:32 +08:00
hrz ce1bdea54d Bump to 0.8.7 2025-11-08 16:24:16 +08:00
hrz 25801bf1d4 Merge branch 'main' into Knowledge-Base 2025-11-08 16:07:21 +08:00
hrz c02f2105d7 update:优化代码 2025-11-08 16:00:37 +08:00
欣南科技andGitHub b0ceb0b06d Merge pull request #2486 from xinnan-tech/hot-fix
fix:旧版本智控台配置文件,没有包含"prompt_template"配置会导致self.config.get("prompt_temp…
2025-11-08 10:09:41 +08:00
hrz b6ba06ec97 fix:旧版本智控台配置文件,没有包含"prompt_template"配置会导致self.config.get("prompt_template", "agent-base-prompt.txt")一直为None 2025-11-08 10:08:02 +08:00
hrz d09fa0c1f6 update:知识库自动创建function方法 2025-11-07 12:20:50 +08:00
hrz ff02c11558 Merge branch 'Knowledge-Base' of https://github.com/xinnan-tech/xiaozhi-esp32-server into Knowledge-Base 2025-11-06 22:01:06 +08:00
hrz e8f5025e4d update:意图识别插件增加知识库 2025-11-06 22:00:38 +08:00
rainv123 90980652ce update:增加文档解析状态 2025-11-06 18:20:01 +08:00
rainv123andGitHub 2527c16e18 Merge pull request #2478 from xinnan-tech/fix-mcp
fix:修复部分LLM在工具调用时数据被空字符串覆盖的问题
2025-11-06 16:57:44 +08:00
3030332422 09dee65239 fix:修复部分LLM在工具调用时数据被空字符串覆盖的问题 2025-11-06 16:42:28 +08:00
hrz eaf4f4b08e update:优化接口错误信息 2025-11-06 14:29:56 +08:00
Chingfeng Li d2b05c614c Update pip 2025-11-06 13:24:41 +08:00
Chingfeng Li 9383a44106 Merge branch 'py-bump-test' of https://github.com/xinnan-tech/xiaozhi-esp32-server 2025-11-06 13:21:33 +08:00
myifengandGitHub 3ec5ed9d22 Merge pull request #2456 from xinnan-tech/dependabot/pip/main/xiaozhi-server/torchaudio-2.9.0
build(deps): bump torchaudio from 2.2.2 to 2.9.0 in /main/xiaozhi-server
2025-11-06 13:17:34 +08:00
myifengandGitHub 482aaf453c Merge branch 'py-bump-test' into dependabot/pip/main/xiaozhi-server/torchaudio-2.9.0 2025-11-06 13:17:26 +08:00
myifengandGitHub bf9277671e Merge pull request #2457 from xinnan-tech/dependabot/pip/main/xiaozhi-server/funasr-1.2.7
build(deps): bump funasr from 1.2.3 to 1.2.7 in /main/xiaozhi-server
2025-11-06 13:15:34 +08:00
myifengandGitHub 588a391d90 Merge pull request #2458 from xinnan-tech/dependabot/pip/main/xiaozhi-server/sherpa-onnx-1.12.15
build(deps): bump sherpa-onnx from 1.12.11 to 1.12.15 in /main/xiaozhi-server
2025-11-06 13:15:11 +08:00
hrz 65f7e794bc Merge branch 'Knowledge-Base' of https://github.com/xinnan-tech/xiaozhi-esp32-server into Knowledge-Base 2025-11-06 12:28:31 +08:00
hrz 987a72a3aa update:优化样式 2025-11-06 12:27:45 +08:00
rainv123 1aab88e8df Merge branch 'Knowledge-Base' of https://github.com/xinnan-tech/xiaozhi-esp32-server into Knowledge-Base 2025-11-06 11:56:07 +08:00
rainv123 a30ef520fb fix:删除无用代码 2025-11-06 11:55:41 +08:00
hrz ea15ed39f5 update:更换更小规格的知识库图标 2025-11-06 11:10:00 +08:00
Claude 9b214279ba Add German translation of README
- Created README_de.md with German translation
- Maintains same structure and links as original
- Provides German-speaking users access to project documentation
2025-11-05 10:31:20 +00:00
rainv123 873daf093d uptate:增加知识库权限判断 2025-11-05 17:02:27 +08:00
rainv123 289f2ad199 fix:修改知识库相关错误码 2025-11-05 16:42:32 +08:00
rainv123 7e15301d2b fix:增加RAG配置验证 2025-11-05 16:06:57 +08:00
rainv123 1792bf299b fix:增加RAG配置验证 2025-11-05 16:03:19 +08:00
rainv123 4cde66e7de fix:知识库接口调整 2025-11-05 15:05:53 +08:00
rainv123 3eeace5d5f fix:修改分页传递参数 2025-11-05 14:50:52 +08:00
rainv123 7215ded7e8 fix:删除无用代码 2025-11-05 14:29:52 +08:00
rainv123 1c6ee46d61 update:删除知识库文档上传无用接口 2025-11-05 10:00:27 +08:00
rainv123 ca87b23d15 upatate:删除知识库无用接口和方法 2025-11-05 09:47:43 +08:00
rainv123 fd47ce8e69 fix:删除错误图标 2025-11-05 09:17:05 +08:00
rainv123 a82cf6e300 update:格式调整 2025-11-05 09:13:44 +08:00
rainv123 0008ca6335 uptate:删除知识库无用接口和方法 2025-11-04 18:12:19 +08:00
rainv123 f127f38cbc fix:修改知识库图标 2025-11-04 17:44:44 +08:00
rainv123 f29f4086b3 update:修改知识库图标 2025-11-04 17:19:10 +08:00
rainv123 b65ae6dc55 add:RAG 2025-11-04 16:48:19 +08:00
rainv123 11b4b06c92 Merge commit '98011272ffb66c1a4e314e496bc301e0a975bd1e' into Knowledge-Base 2025-11-04 16:22:09 +08:00
rainv123 98011272ff add:RAG 2025-11-04 16:21:01 +08:00
Chingfeng Li 222f3f1c93 Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server 2025-11-04 09:27:04 +08:00
dependabot[bot]andGitHub a2fbc07a78 build(deps): bump numpy from 1.26.4 to 2.3.4 in /main/xiaozhi-server
Bumps [numpy](https://github.com/numpy/numpy) from 1.26.4 to 2.3.4.
- [Release notes](https://github.com/numpy/numpy/releases)
- [Changelog](https://github.com/numpy/numpy/blob/main/doc/RELEASE_WALKTHROUGH.rst)
- [Commits](https://github.com/numpy/numpy/compare/v1.26.4...v2.3.4)

---
updated-dependencies:
- dependency-name: numpy
  dependency-version: 2.3.4
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-11-03 08:54:46 +00:00
dependabot[bot]andGitHub 4d728663a6 build(deps): bump sherpa-onnx in /main/xiaozhi-server
Bumps [sherpa-onnx](https://github.com/k2-fsa/sherpa-onnx) from 1.12.11 to 1.12.15.
- [Release notes](https://github.com/k2-fsa/sherpa-onnx/releases)
- [Changelog](https://github.com/k2-fsa/sherpa-onnx/blob/master/CHANGELOG.md)
- [Commits](https://github.com/k2-fsa/sherpa-onnx/compare/v1.12.11...v1.12.15)

---
updated-dependencies:
- dependency-name: sherpa-onnx
  dependency-version: 1.12.15
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-11-03 08:54:40 +00:00
dependabot[bot]andGitHub 8b1fe1dbcf build(deps): bump funasr from 1.2.3 to 1.2.7 in /main/xiaozhi-server
Bumps [funasr](https://github.com/alibaba-damo-academy/FunASR) from 1.2.3 to 1.2.7.
- [Release notes](https://github.com/alibaba-damo-academy/FunASR/releases)
- [Commits](https://github.com/alibaba-damo-academy/FunASR/commits)

---
updated-dependencies:
- dependency-name: funasr
  dependency-version: 1.2.7
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-11-03 08:54:30 +00:00
dependabot[bot]andGitHub 1c836fff39 build(deps): bump torchaudio from 2.2.2 to 2.9.0 in /main/xiaozhi-server
Bumps [torchaudio](https://github.com/pytorch/audio) from 2.2.2 to 2.9.0.
- [Release notes](https://github.com/pytorch/audio/releases)
- [Commits](https://github.com/pytorch/audio/compare/v2.2.2...v2.9.0)

---
updated-dependencies:
- dependency-name: torchaudio
  dependency-version: 2.9.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-11-03 08:54:28 +00:00
hrzandGitHub 53ee24f828 Merge pull request #2455 from xinnan-tech/py_fix_packet
fix: 等待音频过长时需修正包状态
2025-11-03 15:54:40 +08:00
Sakura-RanChen e1056f72e1 fix: 等待音频过长时需修正包状态 2025-11-03 15:17:52 +08:00
Chingfeng Li 67b89b5527 Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server 2025-11-03 14:25:39 +08:00
hrzandGitHub 06c3a8d761 Merge pull request #2441 from xinnan-tech/py_bump_test
Py bump test
2025-11-03 14:09:40 +08:00
Chingfeng Li e93f95254a Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server 2025-11-03 10:47:09 +08:00
hrzandGitHub e38609e9b8 Merge pull request #2449 from xinnan-tech/py_fix_packet
fix: 流控状态未重置
2025-11-03 09:17:33 +08:00
欣南科技andGitHub 4d631cbfe4 Merge pull request #2451 from xinnan-tech/manager-model-id
Manager model
2025-11-02 17:16:29 +08:00
hrz 838d9a3927 update:优化模型id输入框位置 2025-11-02 17:14:11 +08:00
hrzandGitHub 8e12e75b2f Merge pull request #2445 from KurobaKaitou/main
feat[Improvement]: add support for customizing model id when adding model provider configurations.
2025-11-02 15:52:04 +08:00
欣南科技andGitHub 0698a6d8ac Merge pull request #2450 from xinnan-tech/fix-longcatllm
Fix longcatllm
2025-11-02 15:50:22 +08:00
hrzandGitHub 57c2af413e Merge pull request #2433 from lsp199308/main
修复 chat chunk 缺失 'content' 导致 BadRequestError
2025-11-02 15:31:59 +08:00
Sakura-RanChen ec07088c13 fix: 流控状态未重置
update: 增加流式预缓冲
2025-11-01 17:59:16 +08:00
HanabiandGitHub 121318dc37 Merge branch 'xinnan-tech:main' into main 2025-11-01 11:55:19 +08:00
欣南科技andGitHub b1e1a29145 Merge pull request #2446 from xinnan-tech/update_version
Bump to 0.8.6
2025-11-01 11:16:33 +08:00
hrz e59df53c99 Bump to 0.8.6 2025-11-01 11:15:58 +08:00
zhilin e19fb24009 fix: checkout incorrect remove package-lock.json file 2025-11-01 10:09:11 +08:00
小雨 419e7f6dcc fix: fix incorrect remove application-dev.yml file of manager-api 2025-11-01 01:25:46 +08:00
小雨 b3f13e274f feat: Solve the problem of #2198 custom model ID 2025-11-01 01:22:21 +08:00
hrzandGitHub cdf0f91251 Merge pull request #2443 from xinnan-tech/py_fix_prompt_template
fix:修复全模块时缺少prompt_template配置的问题
2025-10-31 21:23:50 +08:00
3030332422 32ecbe55fc fix:修复全模块时缺少prompt_template配置的问题 2025-10-31 20:36:55 +08:00
Chingfeng Li 58b71f5039 更新其他依赖 2025-10-31 15:11:38 +08:00
myifengandGitHub b17a265378 Merge pull request #2424 from xinnan-tech/dependabot/pip/main/xiaozhi-server/edge-tts-7.2.3
build(deps): bump edge-tts from 7.0.0 to 7.2.3 in /main/xiaozhi-server
2025-10-31 13:22:06 +08:00
myifengandGitHub ba446d2374 Merge pull request #2426 from xinnan-tech/dependabot/pip/main/xiaozhi-server/aioconsole-0.8.2
build(deps): bump aioconsole from 0.8.1 to 0.8.2 in /main/xiaozhi-server
2025-10-31 13:21:38 +08:00
myifengandGitHub 18d66270d6 Merge pull request #2423 from xinnan-tech/dependabot/pip/main/xiaozhi-server/mcp-proxy-0.10.0
build(deps): bump mcp-proxy from 0.8.2 to 0.10.0 in /main/xiaozhi-server
2025-10-31 13:21:11 +08:00
myifengandGitHub 2d02349c55 Merge pull request #2422 from xinnan-tech/dependabot/pip/main/xiaozhi-server/ruamel-yaml-0.18.16
build(deps): bump ruamel-yaml from 0.18.15 to 0.18.16 in /main/xiaozhi-server
2025-10-31 13:20:15 +08:00
hrzandGitHub cc9d5306e6 Merge pull request #2437 from xinnan-tech/py_mcp_server_fix
update:服务端MCP新增支持Streamable HTTP传输协议
2025-10-31 10:09:02 +08:00
3030332422 5149a8049a update:服务端MCP新增支持Streamable HTTP传输协议 2025-10-29 23:14:42 +08:00
lsp199308andGitHub cf3c955b73 Update openai.py 2025-10-28 23:15:28 +08:00
欣南科技andGitHub cde093dd20 Merge pull request #2430 from xinnan-tech/mac_safe
update:默认版本不提供设备ID敏感信息,按需开启
2025-10-28 15:03:19 +08:00
hrz 455b3df5ae update:默认版本不提供设备ID敏感信息,按需开启 2025-10-28 15:02:01 +08:00
dependabot[bot]andGitHub 85741137a5 build(deps): bump aioconsole from 0.8.1 to 0.8.2 in /main/xiaozhi-server
Bumps [aioconsole](https://github.com/vxgmichel/aioconsole) from 0.8.1 to 0.8.2.
- [Release notes](https://github.com/vxgmichel/aioconsole/releases)
- [Commits](https://github.com/vxgmichel/aioconsole/compare/v0.8.1...v0.8.2)

---
updated-dependencies:
- dependency-name: aioconsole
  dependency-version: 0.8.2
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-10-27 09:28:05 +00:00
dependabot[bot]andGitHub cc2fdf1b9b build(deps): bump websockets from 14.2 to 15.0.1 in /main/xiaozhi-server
Bumps [websockets](https://github.com/python-websockets/websockets) from 14.2 to 15.0.1.
- [Release notes](https://github.com/python-websockets/websockets/releases)
- [Commits](https://github.com/python-websockets/websockets/compare/14.2...15.0.1)

---
updated-dependencies:
- dependency-name: websockets
  dependency-version: 15.0.1
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-10-27 09:05:28 +00:00
dependabot[bot]andGitHub a70542265a build(deps): bump edge-tts from 7.0.0 to 7.2.3 in /main/xiaozhi-server
Bumps [edge-tts](https://github.com/rany2/edge-tts) from 7.0.0 to 7.2.3.
- [Release notes](https://github.com/rany2/edge-tts/releases)
- [Commits](https://github.com/rany2/edge-tts/compare/7.0.0...7.2.3)

---
updated-dependencies:
- dependency-name: edge-tts
  dependency-version: 7.2.3
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-10-27 09:04:42 +00:00
dependabot[bot]andGitHub 5a9cc89b64 build(deps): bump mcp-proxy from 0.8.2 to 0.10.0 in /main/xiaozhi-server
Bumps [mcp-proxy](https://github.com/sparfenyuk/mcp-proxy) from 0.8.2 to 0.10.0.
- [Release notes](https://github.com/sparfenyuk/mcp-proxy/releases)
- [Commits](https://github.com/sparfenyuk/mcp-proxy/compare/v0.8.2...v0.10.0)

---
updated-dependencies:
- dependency-name: mcp-proxy
  dependency-version: 0.10.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-10-27 09:04:25 +00:00
dependabot[bot]andGitHub 5586515327 build(deps): bump ruamel-yaml in /main/xiaozhi-server
Bumps ruamel-yaml from 0.18.15 to 0.18.16.

---
updated-dependencies:
- dependency-name: ruamel-yaml
  dependency-version: 0.18.16
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-10-27 08:59:38 +00:00
hrzandGitHub 4d14caf0d4 Merge pull request #2416 from xinnan-tech/test-page-add-mcp
添加MCP工具测试页面
2025-10-25 09:25:16 +08:00
Chingfeng Li 47b44bb124 修复未定义的方法 2025-10-24 16:01:12 +08:00
Chingfeng Li eb743ab577 从json中加载默认工具,方便自定义tools 2025-10-24 15:58:01 +08:00
Chingfeng Li d6344bde5e 添加MCP测试功能,支持自定义添加编辑MCP Tools 2025-10-24 10:17:32 +08:00
Chingfeng Li 7ce1672d98 Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server 2025-10-24 09:02:09 +08:00
欣南科技andGitHub 4e3701e62c Merge pull request #2410 from xinnan-tech/fix-xunfei-asr
修复:讯飞asr最后一个字丢失的bug
2025-10-23 17:05:06 +08:00
hrz a035ed525a 修复:讯飞asr最后一个字丢失的bug 2025-10-23 17:04:33 +08:00
hrzandGitHub dbea7ae11e Merge pull request #2403 from xinnan-tech/main-aliyun-nls-taskid
阿里云NLS协议中message_id每次发送都必须唯一;task_id是会话id,整个请求中需要保持一致。
2025-10-23 15:12:19 +08:00
hrzandGitHub 26127e9b4a Merge pull request #2401 from xinnan-tech/py_fix_huoshan
fix: 未等待服务端残留状态问题
2025-10-23 14:32:17 +08:00
Chingfeng Li 7035a57cf1 Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server 2025-10-23 13:05:10 +08:00
hrzandGitHub aee9bdc514 Merge pull request #2399 from xinnan-tech/main-agent-base-prompt
常量可配置:允许自定义agent-base-prompt.txt文件
2025-10-23 12:26:08 +08:00
欣南科技andGitHub 9018422ecb Merge pull request #2405 from xinnan-tech/hot-fix
update:优化文档
2025-10-23 12:24:52 +08:00
hrz da8d8c53c0 update:优化文档 2025-10-23 12:24:16 +08:00
欣南科技andGitHub 2c428a380e Merge pull request #2404 from xinnan-tech/hot-fix
update:该功能易被坏人滥用,违背开发者初心
2025-10-23 11:47:22 +08:00
hrz bde50e661d update:改功能易引发坏人滥用,违背开发者初心 2025-10-23 11:43:49 +08:00
Chingfeng Li 2beeb825cf 阿里云NLS协议中message_id每次发送都必须唯一;task_id是会话id,整个请求中需要保持一致。 2025-10-22 18:21:34 +08:00
Sakura-RanChen 371dab1282 fix: 未等待服务端残留状态问题 2025-10-22 17:26:36 +08:00
Chingfeng Li 3ba437c937 常量可配置:允许自定义agent-base-prompt.txt文件 2025-10-22 14:53:05 +08:00
hrzandGitHub 4b2d7da4e5 Merge pull request #2357 from kkroid/feature_tts-audio-send-delay
feat: 添加可配置的TTS音频发送延迟功能
2025-10-22 11:48:34 +08:00
hrzandGitHub d00592020e Update config.yaml 2025-10-22 11:47:54 +08:00
hrzandGitHub e0b24f4e59 Merge pull request #2383 from xinnan-tech/docker-server-base
构建base镜像,只有当依赖发生变化时再进行重新构建,否则只使用之前的base镜像进行构建,能够使本地使用缓存的镜像
2025-10-22 11:41:35 +08:00
欣南科技andGitHub 4b7be99837 Merge pull request #2378 from xinnan-tech/py_update_vad
update: 优化vad判断逻辑
2025-10-22 11:19:33 +08:00
hrz 99db948b96 update:恢复vad判断 2025-10-22 10:46:57 +08:00
欣南科技andGitHub 2b3e205cda Merge pull request #2397 from xinnan-tech/pip_version
Pip version
2025-10-22 09:17:29 +08:00
hrz c9b7d77094 updte:python3.10最高只支持torch==2.2.2、torchaudio==2.2.2 2025-10-22 09:16:33 +08:00
hrzandGitHub f3818167f3 Merge pull request #2380 from KurobaKaitou/main
chore: 补充在 Linux 环境下进行源码部署时出现缺失动态库的解决方案以及保留检测 ffmpeg 环境时的原有报错信息
2025-10-21 17:42:52 +08:00
hrzandGitHub 55d6c2a193 Merge pull request #2387 from xinnan-tech/dependabot/pip/main/xiaozhi-server/ormsgpack-1.11.0
build(deps): bump ormsgpack from 1.7.0 to 1.11.0 in /main/xiaozhi-server
2025-10-21 17:42:09 +08:00
hrzandGitHub 6cfb6a8f43 Merge pull request #2386 from xinnan-tech/dependabot/pip/main/xiaozhi-server/dashscope-1.24.6
build(deps): bump dashscope from 1.23.1 to 1.24.6 in /main/xiaozhi-server
2025-10-21 17:41:38 +08:00
hrzandGitHub 396c3eb18b Merge pull request #2388 from xinnan-tech/dependabot/pip/main/xiaozhi-server/openai-2.5.0
build(deps): bump openai from 1.107.0 to 2.5.0 in /main/xiaozhi-server
2025-10-21 17:41:06 +08:00
hrzandGitHub d8c4c1f207 Merge branch 'pip_version' into dependabot/pip/main/xiaozhi-server/openai-2.5.0 2025-10-21 17:40:47 +08:00
hrzandGitHub 268ce9fff3 Merge pull request #2389 from xinnan-tech/dependabot/pip/main/xiaozhi-server/torchaudio-2.9.0
build(deps): bump torchaudio from 2.2.2 to 2.9.0 in /main/xiaozhi-server
2025-10-21 17:39:12 +08:00
hrzandGitHub bbc3b44336 Merge pull request #2390 from xinnan-tech/dependabot/pip/main/xiaozhi-server/mem0ai-1.0.0
build(deps): bump mem0ai from 0.1.62 to 1.0.0 in /main/xiaozhi-server
2025-10-21 17:38:48 +08:00
HanabiandGitHub fb223984d3 Merge branch 'xinnan-tech:main' into main 2025-10-21 00:52:25 +08:00
欣南科技andGitHub 9f21711cae Merge pull request #2392 from xinnan-tech/hot-fix
update:修复manager-api错误码重复问题
2025-10-20 23:39:43 +08:00
hrz c139701468 update:修复manager-api错误码重复问题 2025-10-20 23:38:59 +08:00
欣南科技andGitHub 66d3bcfa13 Merge pull request #2379 from xinnan-tech/main-fix-auth_key
如果已经配置了auth_key,则不使用uuid生成; 适配单模块启动,防止每次服务重启设备都无法连接问题
2025-10-20 22:19:10 +08:00
hrzandGitHub ff048797a1 auth_key优先级:配置文件server.auth_key > manager-api.secret > 自动生成 2025-10-20 21:25:37 +08:00
dependabot[bot]andGitHub ad6607b7f5 build(deps): bump mem0ai from 0.1.62 to 1.0.0 in /main/xiaozhi-server
Bumps [mem0ai](https://github.com/mem0ai/mem0) from 0.1.62 to 1.0.0.
- [Release notes](https://github.com/mem0ai/mem0/releases)
- [Changelog](https://github.com/mem0ai/mem0/blob/main/docs/changelog.mdx)
- [Commits](https://github.com/mem0ai/mem0/compare/0.1.62...v1.0.0)

---
updated-dependencies:
- dependency-name: mem0ai
  dependency-version: 1.0.0
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-10-20 13:19:42 +00:00
dependabot[bot]andGitHub c992ed48da build(deps): bump torchaudio from 2.2.2 to 2.9.0 in /main/xiaozhi-server
Bumps [torchaudio](https://github.com/pytorch/audio) from 2.2.2 to 2.9.0.
- [Release notes](https://github.com/pytorch/audio/releases)
- [Commits](https://github.com/pytorch/audio/compare/v2.2.2...v2.9.0)

---
updated-dependencies:
- dependency-name: torchaudio
  dependency-version: 2.9.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-10-20 13:19:35 +00:00
HanabiandGitHub bdf9cda4c5 Merge branch 'xinnan-tech:main' into main 2025-10-20 20:23:54 +08:00
dependabot[bot]andGitHub c85ffcc181 build(deps): bump openai from 1.107.0 to 2.5.0 in /main/xiaozhi-server
Bumps [openai](https://github.com/openai/openai-python) from 1.107.0 to 2.5.0.
- [Release notes](https://github.com/openai/openai-python/releases)
- [Changelog](https://github.com/openai/openai-python/blob/main/CHANGELOG.md)
- [Commits](https://github.com/openai/openai-python/compare/v1.107.0...v2.5.0)

---
updated-dependencies:
- dependency-name: openai
  dependency-version: 2.5.0
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-10-20 11:56:58 +00:00
dependabot[bot]andGitHub c53e3ed965 build(deps): bump ormsgpack from 1.7.0 to 1.11.0 in /main/xiaozhi-server
Bumps [ormsgpack](https://github.com/aviramha/ormsgpack) from 1.7.0 to 1.11.0.
- [Release notes](https://github.com/aviramha/ormsgpack/releases)
- [Changelog](https://github.com/aviramha/ormsgpack/blob/master/CHANGELOG.md)
- [Commits](https://github.com/aviramha/ormsgpack/compare/1.7.0...1.11.0)

---
updated-dependencies:
- dependency-name: ormsgpack
  dependency-version: 1.11.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-10-20 11:51:52 +00:00
dependabot[bot]andGitHub d374b77baa build(deps): bump dashscope in /main/xiaozhi-server
Bumps [dashscope](https://dashscope.aliyun.com/) from 1.23.1 to 1.24.6.

---
updated-dependencies:
- dependency-name: dashscope
  dependency-version: 1.24.6
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-10-20 11:51:47 +00:00
hrzandGitHub f3b2339803 Merge pull request #2384 from xinnan-tech/fix-dependabot
修正检测周期
2025-10-20 18:47:30 +08:00
hrzandGitHub 27ed540113 Merge pull request #2385 from xinnan-tech/fixmqtt
fix:修复单模块mqtt下发错误
2025-10-20 18:47:05 +08:00
FAN-yeB 434fba55f7 fix:修复单模块下发逻辑 2025-10-20 18:21:12 +08:00
Chingfeng Li 280c04e753 修正检测周期 2025-10-20 15:54:50 +08:00
Chingfeng Li 24832822a5 默认main分支 2025-10-20 15:11:23 +08:00
Chingfeng Li a357f0818d 依赖baseAction 2025-10-20 15:10:19 +08:00
Chingfeng Li e4bf5d2d21 构建base镜像,只有当依赖发生变化时再进行重新构建,否则只使用之前的base镜像进行构建,能够使本地使用缓存的镜像 2025-10-20 15:07:02 +08:00
HanabiandGitHub 8b261d473a fix: 补充README文档 334行的 ffmpeg 单词缺失
补充README文档 334行的 ffmpeg 单词缺失
2025-10-20 12:53:40 +08:00
HanabiandGitHub bba84382ec docs: 补充源码部署时安装 ffmpeg 环境可能会出现缺失动态库的解决方案
补充说明
2025-10-20 12:32:09 +08:00
HanabiandGitHub d081803434 chore(util): 优化项目启动时检测ffmpeg环境时在某些场景下的提示
此次变动主要针对 Linux (Ubuntu 22.04) 环境下根据官方提供的README文档以及conda安装 ffmpeg 环境后出现缺失动态库libiconv.so.2的问题 报错完整信息如下:
`ffmpeg: error while loading shared libraries: libiconv.so.2: cannot open shared object file: No such file or directory` 基于以上报错信息 优化了util.py中的check_ffmpeg_installed函数
- 保留了原有命令行的提示信息 [定位问题更加友好]
- 抓取报错中的 libiconv.so.2 文本内容 进行提示用户通过conda安装 libiconv 动态库
2025-10-20 12:28:22 +08:00
Chingfeng Li 9194ffa6b7 如果已经配置了auth_key,则不使用uuid生成; 适配单模块启动,防止每次服务重启设备都无法连接问题 2025-10-20 11:57:18 +08:00
Sakura-RanChen c06a7b1db6 update: 优化vad判断逻辑 2025-10-20 11:45:50 +08:00
欣南科技andGitHub 1736e10189 Merge pull request #2374 from xinnan-tech/myifeng-patch-1
更新 dependabot.yml
2025-10-20 11:37:55 +08:00
hrzandGitHub 43b79c50a0 每个月更新组件版本 2025-10-20 11:35:31 +08:00
欣南科技andGitHub 2e6aced1bd Merge pull request #2377 from xinnan-tech/update-doc
fix:克隆教程文字错误
2025-10-20 10:17:55 +08:00
hrz 2dc80c8432 fix:克隆教程文字错误 2025-10-20 10:16:10 +08:00
myifengandGitHub 01e7a7d3b8 更新 dependabot.yml 2025-10-19 16:06:44 +08:00
欣南科技andGitHub e5e9f2bdea Merge pull request #2373 from xinnan-tech/update-doc
update:立即复刻增加loading效果
2025-10-19 14:34:53 +08:00
hrz 06eff425f1 update:立即复刻增加loading效果 2025-10-19 14:34:05 +08:00
欣南科技andGitHub d2aae057f6 Merge pull request #2372 from xinnan-tech/update-doc
update:常见问题增加音色克隆链接
2025-10-19 13:18:39 +08:00
hrz b7b39de445 update:常见问题增加音色克隆链接 2025-10-19 13:18:07 +08:00
欣南科技andGitHub 4aedcb8419 Merge pull request #2371 from xinnan-tech/voice-clone
update:升级版本号
2025-10-19 13:14:20 +08:00
hrz 6c5bd395cb update:升级版本号 2025-10-19 13:13:44 +08:00
欣南科技andGitHub 6ad9ae8d1b Merge pull request #2353 from xinnan-tech/voice-clone
uptate:音色复刻功能完善
2025-10-19 13:11:23 +08:00
hrz 48e2369964 update:修改音色克隆教程 2025-10-19 13:10:07 +08:00
hrz e753c242ad update:将本人的克隆音色自动添加模型配置中的音色选择列表 2025-10-19 11:53:15 +08:00
hrz c1e0b5181e update:优化克隆失败显示 2025-10-19 01:45:08 +08:00
欣南科技andGitHub d3371c2875 Merge pull request #2370 from xinnan-tech/ota-test
fix:修复【启用】模型和【默认】模型时,密钥变*bug
2025-10-19 00:37:08 +08:00
hrz 2488678f34 fix:修复【启用】模型和【默认】模型时,密钥变*bug 2025-10-19 00:36:24 +08:00
3030332422 958d50c119 fix:修改音色克隆详细内容 2025-10-18 21:47:14 +08:00
欣南科技andGitHub 19c16cfcd8 Merge pull request #2369 from xinnan-tech/ota-test
Ota test
2025-10-18 21:25:27 +08:00
hrz c66c92be73 update:智控台版本ota增加token返回字段 2025-10-18 21:24:19 +08:00
hrzandGitHub 4f9ea49f8e Merge pull request #2368 from myifeng/mian-ota-test
更新test页面,适配OTA认证
2025-10-18 14:28:39 +08:00
3030332422 b776c545e3 fix:声音复刻后自动刷新页面状态和错误详情 2025-10-17 23:31:36 +08:00
3030332422 d65969a2b2 fix: 声音复刻成功后自动刷新页面状态 2025-10-17 22:29:30 +08:00
3030332422 d3eb8df7f0 update:声音复刻功能的部分修改 2025-10-17 18:22:38 +08:00
Chingfeng Li 89e299764d 调整测试认证 2025-10-17 18:06:22 +08:00
rainv123 66fb7933b7 update:增加音色克隆页面详情按钮 2025-10-17 17:55:56 +08:00
Chingfeng Li 49e1bbbf15 更新test测试页面,使用ota返回的连接信息和认证信息 2025-10-17 17:51:56 +08:00
3030332422 af02d53718 update:修复复刻失败问题 2025-10-17 17:08:41 +08:00
3030332422 f8b1e8199c fix:修复复刻失败的问题 2025-10-17 15:56:33 +08:00
rainv123 008456d7cd update:新增音色资源时判断声音id是否已经存在 2025-10-17 11:31:13 +08:00
hrzandGitHub aa7c07167c Merge pull request #2346 from xinnan-tech/mcp_add_macaddress_test
update:通过JSON-RPC参数传递设备MAC地址
2025-10-17 09:42:01 +08:00
rainv123 d451fe7a0b uptate:修改训练状态 2025-10-17 09:22:08 +08:00
will.zhang 3b8bbb5c5f feat: 添加可配置的TTS音频发送延迟功能
- 修改 sendAudio() 函数,新增 tts_audio_send_delay 配置参数(毫秒)
- 0: 使用原有逻辑
- 大于0: 使用固定延迟发送(降低首包延迟,提升响应速度)
2025-10-14 18:30:17 +08:00
hrzandGitHub 29cd69415a Merge pull request #2355 from xinnan-tech/py_fix_uuid
update:暂时恢复先前的随机生成uuid
2025-10-14 18:19:33 +08:00
欣南科技andGitHub bc99da5b28 Merge pull request #2356 from xinnan-tech/update-doc
update:优化demo
2025-10-14 18:13:39 +08:00
hrz 5b38d92cc6 update:优化demo 2025-10-14 18:12:58 +08:00
3030332422 0e1392de35 update:暂时恢复先前的随机生成uuid 2025-10-14 18:06:57 +08:00
rainv123andGitHub a033b56b8d Update:修正翻译 2025-10-14 15:51:34 +08:00
rainv123andGitHub c449d29de0 Update:更新火山双流式语音合成+音色克隆配置文档 2025-10-14 15:23:48 +08:00
rainv123 afc142b397 update:更新火山双流式语音合成+音色克隆配置文档 2025-10-14 15:16:51 +08:00
rainv123 55348ffe65 updata:更新火山双向流式TTS+声音克隆配置文档 2025-10-14 15:13:22 +08:00
rainv123 174b9f5898 updata:更新火山双向流式TTS+声音克隆配置教程 2025-10-14 15:05:02 +08:00
rainv123 3b69629a61 fix:补充键的翻译 2025-10-14 11:04:49 +08:00
rainv123 56eb25f8e2 fix:更正翻译 2025-10-14 10:59:54 +08:00
rainv123 1400e48fc7 Merge branch 'voice-clone' of https://github.com/xinnan-tech/xiaozhi-esp32-server 2025-10-14 10:27:02 +08:00
rainv123 bccda3e7b2 fix:补充音色克隆界面弹窗按钮翻译 2025-10-14 10:21:50 +08:00
rainv123 dfb0a2b213 update:修改音色编码排序逻辑 2025-10-14 10:19:44 +08:00
3030332422 f94b376f31 update:通过JSON-RPC参数传递设备MAC地址 2025-10-13 18:00:58 +08:00
hrzandGitHub 13ce67e4c7 Merge pull request #2342 from xinnan-tech/py_fix_wakeup
fix: 相关状态未重置
2025-10-13 17:57:49 +08:00
欣南科技andGitHub eece86ab71 Merge pull request #2345 from xinnan-tech/ota-auth
Ota auth
2025-10-13 17:57:18 +08:00
hrz 4fd294dc80 update:
1、把ota_auth内容合并到auth.py
2、精简auth密钥,简化部署
2025-10-13 17:52:28 +08:00
Sakura-RanChen 4952ccfbf7 fix: 相关状态未重置 2025-10-13 16:47:05 +08:00
hrz 471ad864b6 update:优化 2025-10-13 15:31:49 +08:00
hrz bec32bd085 update:把传递name改成传递speaker_id 2025-10-13 15:26:26 +08:00
rainv123 087fd5d547 update:音色克隆 2025-10-13 14:17:53 +08:00
hrzandGitHub 99f178274d Merge pull request #2339 from xinnan-tech/cc
fix:修复长按聆听设备松开没有声音的问题
2025-10-13 13:23:47 +08:00
1cccux 450560f92a fix:修复长按聆听设备松开没有声音的问题 2025-10-13 10:57:48 +08:00
hrzandGitHub 691a0529b0 Merge pull request #2330 from myifeng/ota-auth
OTA接口授权与认证
2025-10-13 10:51:34 +08:00
欣南科技andGitHub 38e63eb665 Merge pull request #2338 from xinnan-tech/uuid
UUID
2025-10-13 10:49:44 +08:00
hrzandGitHub ad8390301e Merge pull request #2337 from Yuerchu/main
使用uuid包来生成uuidv4
2025-10-13 10:35:19 +08:00
于小丘 acee7e27c6 在管理网页也修改uuid生成方法 2025-10-13 09:49:56 +08:00
于小丘 e790b668ef 删除原uuid库的导入 2025-10-13 09:42:59 +08:00
于小丘 432a5d01f8 改成 crypto.randomUUID(); 生成uuid 2025-10-13 09:35:39 +08:00
于小丘 374f2a92aa 使用uuid包来生成uuidv4 2025-10-13 09:26:16 +08:00
欣南科技andGitHub a22ad985dc Merge pull request #2336 from xinnan-tech/hot-fix
fix:回复词太短
2025-10-12 23:25:25 +08:00
hrz 44f41ebd11 fix:回复词太短 2025-10-12 23:24:43 +08:00
Chingfeng Li a10e94bef6 先认证,后创建连接,防止无效连接建立 2025-10-11 17:52:15 +08:00
hrzandGitHub e7cb67f846 Merge pull request #2331 from xinnan-tech/py_wave_fix
update:修改唤醒词
2025-10-11 15:57:11 +08:00
3030332422 b5a3cc1c78 update:修改唤醒词 2025-10-11 15:48:27 +08:00
FAN-yeBandGitHub 1d107fb39b Merge pull request #2326 from xinnan-tech/css
修复修改密码无法重复点击和恢复样式
2025-10-11 15:45:29 +08:00
LiJinHui 1ec8f639cb 补充修改 2025-10-11 15:28:53 +08:00
Chingfeng Li c15069a9b6 OTA接口授权与认证 2025-10-11 15:27:18 +08:00
LiJinHui 13ae5e3ade 补充样式 2025-10-11 15:02:01 +08:00
LiJinHui 9cffc8b0ec 修复交互bug 2025-10-11 09:26:08 +08:00
hrzandGitHub c8c5b1095b Merge pull request #2323 from xinnan-tech/py_update_document
update: 更新说明文档
2025-10-10 15:09:46 +08:00
Sakura-RanChen 9803b3c690 update: 更新说明文档 2025-10-10 14:45:30 +08:00
hrzandGitHub 341c0087b7 Merge pull request #2322 from xinnan-tech/web-i18n-updata
update:调整翻译文本顺序,删除重复翻译键
2025-10-10 14:39:56 +08:00
rainv123 0cd752be05 update:调整翻译文本顺序,删除重复翻译键 2025-10-10 14:35:35 +08:00
rainv123andGitHub 2e39b1c6f5 Merge pull request #2313 from xinnan-tech/mobile_forgot_password
添加移动端忘记密码页面
2025-10-10 11:28:02 +08:00
LiJinHui c4026ef9ac 样式的恢复和调整 2025-10-10 10:02:11 +08:00
欣南科技andGitHub 271a2e3e96 Merge pull request #2317 from xinnan-tech/hot-fix
fix:python config api no need maskSensitive
2025-10-08 22:45:22 +08:00
hrz bc3e10f09b fix:python config api no need maskSensitive 2025-10-08 22:43:45 +08:00
3030332422 fed3501bf6 update:补充 2025-10-08 10:49:02 +08:00
欣南科技andGitHub 877cb0d645 Merge pull request #2312 from xinnan-tech/voice-clone
优化:用户管理密码生成策略
2025-10-08 00:17:16 +08:00
hrz e930935202 优化:用户管理密码生成策略 2025-10-08 00:16:24 +08:00
欣南科技andGitHub 0af81fa366 Merge pull request #2311 from xinnan-tech/voice-clone
update:智控台音色克隆
2025-10-07 23:57:56 +08:00
hrz 3249970b7b update:智控台音色克隆 2025-10-07 23:54:42 +08:00
3030332422 1466806339 update:移动端忘记密码页面的前后端对接 2025-10-06 20:01:06 +08:00
3030332422 05389b1368 update:移动端忘记密码前端页面添加地区区号选择和SM2加密 2025-10-03 11:35:44 +08:00
欣南科技andGitHub bef7f79f40 Merge pull request #2307 from xinnan-tech/update_doc
update:更新docker安装提示
2025-10-02 02:28:21 +08:00
hrz 8d7394cd46 update:更新docker安装提示 2025-10-02 02:25:58 +08:00
欣南科技andGitHub 65053e98e2 Merge pull request #2306 from xinnan-tech/openrz-patch-1
Update docker-image.yml
2025-10-02 02:14:31 +08:00
hrzandGitHub c1bea43598 Update docker-image.yml 2025-10-02 02:13:45 +08:00
欣南科技andGitHub 31aeb63f89 Merge pull request #2305 from xinnan-tech/his_download
add:首页智能体添加搜索记录
2025-10-01 02:53:09 +08:00
hrz 0175a7a1d4 add:首页智能体添加搜索记录 2025-10-01 02:52:32 +08:00
欣南科技andGitHub 1417ee017e Merge pull request #2304 from xinnan-tech/his_download
update:下载智能体的聊天记录
2025-10-01 01:33:01 +08:00
hrz 5b4d79902f update:下载智能体的聊天记录 2025-10-01 01:24:13 +08:00
3030332422 f10fddc59a update:新增移动端忘记密码前端页面 2025-09-30 17:52:38 +08:00
hrzandGitHub ed07793c47 Merge pull request #2288 from xinnan-tech/sm2_test
登录、注册页面添加非对称加密
2025-09-30 11:12:50 +08:00
hrz 2ed2e9c391 update:优化移动端按钮 2025-09-30 11:10:48 +08:00
3030332422 2b9fed2609 update:新增SM2解密工具类,优化登录注册密码找回功能 2025-09-30 09:07:06 +08:00
3030332422 848ce65d13 update:web端重置密码页面加密 2025-09-29 17:57:53 +08:00
hrzandGitHub f0512ce1cc Merge pull request #2302 from xinnan-tech/CSS_update
调整底框位置
2025-09-29 17:57:24 +08:00
LiJinHui 313fad1f81 样式调整 2025-09-29 17:56:19 +08:00
3030332422 ff26a84988 fix:完善重置密码页面多语言支持 2025-09-29 16:51:23 +08:00
3030332422 f7ee54598e fix:修复移动端创建智能体时显示问题 2025-09-29 16:09:28 +08:00
LiJinHui 7a62462c1b 调整底框位置 2025-09-29 15:18:46 +08:00
3030332422 a4433ebe68 fix:修复移动端注册页面的一些bug 2025-09-29 14:21:41 +08:00
3030332422 48feea55fe update:移动端注册页面加密 2025-09-29 10:29:28 +08:00
3030332422 6e5f708f5a update:移动端登录页面加密 2025-09-28 22:40:05 +08:00
hrzandGitHub f2ed645f9c Merge pull request #2290 from xinnan-tech/jia_mi_miyao
模型秘钥加密
2025-09-28 18:32:53 +08:00
hrz d47664e670 update:优化输入框提示 2025-09-28 18:25:29 +08:00
LiJinHui 8f3229ddda 调整 input onfocus 判断 2025-09-28 17:37:06 +08:00
LiJinHui 4f2b7524aa 改用精准匹配生成提示词 2025-09-28 17:20:09 +08:00
LiJinHui 4e99e176e9 删除多余样式 2025-09-28 17:05:17 +08:00
LiJinHui cfc7047967 页面优化 2025-09-28 16:52:28 +08:00
LiJinHui 776b5fb3b4 修复加密的bug 2025-09-28 15:27:46 +08:00
3030332422 f16d42d2b5 Merge branch 'sm2_test' of https://github.com/xinnan-tech/xiaozhi-esp32-server into sm2_test 2025-09-28 14:18:44 +08:00
3030332422 344d183b6f fix:修复web端手机号码注册时遇到的问题 2025-09-28 14:06:13 +08:00
LiJinHui feb668c96c 还原Python的端口 2025-09-28 10:28:01 +08:00
LiJinHui f345568675 修复秘钥污染的bug 2025-09-28 09:32:54 +08:00
hrz 11207f793d update:登录类型按钮居中显示 2025-09-26 23:56:40 +08:00
hrz ce09eb48c4 update:共钥信息从旧接口获取,去除getServerPublicKey方法 2025-09-26 23:39:27 +08:00
hrzandGitHub 6925732f72 Merge pull request #2285 from myifeng/prompt-template-args
提示词模板支持更多参数
2025-09-26 19:07:41 +08:00
hrzandGitHub dd9d08c8c3 Merge pull request #2291 from xinnan-tech/performance
update:更新测试工具以测试讯飞流式ASR、TTS首词响应时间
2025-09-26 19:06:28 +08:00
hrzandGitHub 1443c883f1 Merge pull request #2292 from xinnan-tech/mobile-i18n
updata:移动端添加语言切换功能
2025-09-26 19:05:21 +08:00
hrzandGitHub 30676e2170 Merge branch 'main' into sm2_test 2025-09-26 19:03:23 +08:00
rainv123 e41e6d4284 updata:删除无用功能 2025-09-26 17:10:47 +08:00
rainv123 1a7b1d24fa updata:移动端添加语言切换功能 2025-09-26 17:03:17 +08:00
rainv123 99443d1c9f updata:移动端添加语言切换功能 2025-09-26 14:54:56 +08:00
rainv123 a84b8d818e updata:移动端添加语言切换功能 2025-09-26 14:53:16 +08:00
LiJinHui ef4abf08f0 格式调整 2025-09-26 14:44:20 +08:00
rainv123 dfcac51312 updata:移动端添加语言切换功能 2025-09-26 14:33:06 +08:00
3030332422 c963a2de18 fix:修复验证图形验证码 2025-09-26 12:58:28 +08:00
3030332422 9ec8ecf7f4 update:SM2非对称加密的优化 2025-09-26 11:10:42 +08:00
FAN-yeB 5d77c96d76 update:更新测试工具以测试讯飞流式ASR、TTS首词响应时间 2025-09-26 11:05:00 +08:00
LiJinHui a8932f3743 秘钥加密 2025-09-26 10:04:43 +08:00
欣南科技andGitHub d022dd385a Merge pull request #2286 from xinnan-tech/Data_display_update
Data display update
2025-09-25 16:12:42 +08:00
hrz b2dfbb7973 update:优化判断 2025-09-25 16:12:02 +08:00
LiJinHui 7fee922796 更新设备数量显示(实时) 2025-09-25 15:56:51 +08:00
3030332422 9b6acf1ffc fix:修复SM2非对称加密的各种问题 2025-09-25 15:50:32 +08:00
hrzandGitHub a0e7cb56dc Merge pull request #2284 from xinnan-tech/mqtt_doc
update:更新mqtt网关集成文档
2025-09-25 14:54:51 +08:00
hrz f8447033b9 update:优化 2025-09-25 14:54:05 +08:00
hrz bb07015feb update:抽离公共教程到最前面 2025-09-25 14:51:34 +08:00
Chingfeng Li 111d945a9a 提示词模板支持更多参数 2025-09-25 14:35:01 +08:00
FAN-yeB d1587cc645 update:更新mqtt网关集成文档 2025-09-25 14:31:23 +08:00
欣南科技andGitHub e50e21c240 Merge pull request #2279 from xinnan-tech/update_doc
update:更新文档
2025-09-25 12:26:14 +08:00
hrz 9bbcf7f31a update:更新文档 2025-09-25 12:25:36 +08:00
hrzandGitHub 4f29e375c3 Merge pull request #2277 from xinnan-tech/py_fix_tts
fix: Markdown调整
2025-09-25 11:05:12 +08:00
Sakura-RanChen e15c784bdb fix: 文档链接 2025-09-25 10:57:12 +08:00
Sakura-RanChen e77a3b79cd fix: Markdown调整 2025-09-25 10:17:56 +08:00
3030332422 7ac5113fef update:非对称加密注册功能 2025-09-25 09:15:41 +08:00
hrzandGitHub ae58537979 Merge pull request #2275 from xinnan-tech/py_indexTTS
update: 更新文档
2025-09-24 17:01:02 +08:00
Sakura-RanChen ebf10828f0 update: 更新文档 2025-09-24 16:35:26 +08:00
3030332422 33bad7ed40 update:前端SM2非对称加密登录功能 2025-09-23 20:22:12 +08:00
hrzandGitHub af9ed0cd86 Merge pull request #2266 from xinnan-tech/py_mqtt
update:增加单模块mqtt-gateway支持
2025-09-23 11:51:41 +08:00
FAN-yeB 14a0829190 更新ota下发mqtt协议逻辑 2025-09-23 11:37:28 +08:00
FAN-yeB ca400d8ea7 Revert "更改下发配置"
This reverts commit 8891a96731.
2025-09-23 11:29:05 +08:00
hrzandGitHub d28750dae0 Merge pull request #2264 from xinnan-tech/py_xunfei_tts
update:添加讯飞语音识别、大模型、语音合成供应器
2025-09-23 11:06:00 +08:00
hrzandGitHub ef81a17311 Merge pull request #2273 from xinnan-tech/i18n-fix
fix:修改用户管理界面用户的密码重置、禁用、删除的逻辑
2025-09-23 11:05:17 +08:00
FAN-yeB 8891a96731 更改下发配置 2025-09-23 10:45:25 +08:00
rainv123 49e69ecad4 fix:修改用户管理界面用户的密码重置、禁用、删除的逻辑 2025-09-23 10:16:17 +08:00
rainv123 1db2483557 Merge branch 'main' of https://github.com/xinnan-tech/xiaozhi-esp32-server 2025-09-23 10:08:55 +08:00
rainv123 7067050006 Revert "fix:修改用户管理界面的用户密码重置、禁用、删除的逻辑"
This reverts commit 9f303d840a.
2025-09-23 10:03:43 +08:00
FAN-yeBandGitHub 55b455216c Merge pull request #2267 from xinnan-tech/i18n-fix
i18n-fix
2025-09-23 09:58:36 +08:00
3030332422 ffba2f4fa3 update:添加SM2国密加密的登录接口 2025-09-23 09:58:27 +08:00
rainv123 9f303d840a fix:修改用户管理界面的用户密码重置、禁用、删除的逻辑 2025-09-23 09:49:13 +08:00
FAN-yeB 86b4e61c1c 添加讯飞星火大模型的配置示例 2025-09-23 09:06:27 +08:00
欣南科技andGitHub 3232b36e15 Merge pull request #2271 from xinnan-tech/fix-docker-web
update:优化编译
2025-09-22 21:35:04 +08:00
hrz 83329cf9db update:优化编译 2025-09-22 21:34:09 +08:00
rainv123 4608c81ea6 updata:补充国际化键名 2025-09-22 15:04:49 +08:00
rainv123 41679f4e1d updata:调整文本顺序 2025-09-22 14:55:22 +08:00
FAN-yeB 559699ec19 update:增加单模块mqtt-gateway支持 2025-09-22 14:18:03 +08:00
hrzandGitHub f785219a43 Merge pull request #2265 from xinnan-tech/LJH
update:修改排序
2025-09-22 11:24:41 +08:00
FAN-yeB bf2b6d7554 还原错误更改 2025-09-22 10:10:46 +08:00
FAN-yeB 9a6348c71a 更新 202509220958.sql 2025-09-22 10:01:30 +08:00
FAN-yeB 646f4cbfe9 更新讯飞星火认知大模型配置的说明文档 2025-09-22 10:00:29 +08:00
LiJinHui d4c6b6840e 增加排序更新逻辑 2025-09-22 09:52:21 +08:00
LiJinHui 758aa1c78c 修改排序的逻辑 2025-09-22 09:41:18 +08:00
FAN-yeB ec2b52ccac 优化讯飞asr参数 2025-09-22 09:34:05 +08:00
FAN-yeB 38e1eb0c18 更新 202509191732.sql 2025-09-22 09:25:43 +08:00
hrzandGitHub 00e8478f61 Merge pull request #2251 from xinnan-tech/LJH
update:新增调整角色模版的功能,添加新功能的国际化
2025-09-21 18:10:05 +08:00
hrzandGitHub 0c556b1424 Merge branch 'main' into LJH 2025-09-21 18:09:48 +08:00
hrz 9987555a72 Bump to v0.8.3 2025-09-21 18:07:48 +08:00
hrz ebb4ba5595 update:调整功能顺序 2025-09-21 18:04:56 +08:00
LiJinHui 7e338b28df 删除多余定义 2025-09-21 16:44:18 +08:00
LiJinHui c4b3e30c5d 删除多的定义 2025-09-21 16:41:53 +08:00
LiJinHui 4d283c3504 去掉调试代码 2025-09-21 01:51:35 +08:00
LiJinHui ffc1541343 删除多余代码 2025-09-21 00:25:23 +08:00
LiJinHui 588d75763c 删除多余样式 2025-09-20 23:43:58 +08:00
LiJinHui 9e218281da 调整代码 2025-09-20 17:30:00 +08:00
欣南科技andGitHub 6fa58c7948 Merge pull request #2257 from xinnan-tech/update_doc
update:优化图片
2025-09-20 08:13:51 +08:00
hrz 7fe40579b9 update:优化图片 2025-09-20 08:13:04 +08:00
欣南科技andGitHub 7e2782ce22 Merge pull request #2256 from xinnan-tech/update_doc
update:架构图增加mqtt
2025-09-20 08:04:30 +08:00
hrz 1fbe8dab77 update:架构图增加mqtt 2025-09-20 08:04:00 +08:00
LiJinHui 0aec742bd3 格式统一 2025-09-19 18:33:54 +08:00
LiJinHui a6f5d868dd 格式调整 2025-09-19 18:18:18 +08:00
FAN-yeB 966d8a8a84 update:xunfei asr 2025-09-19 17:54:52 +08:00
LiJinHui 611d1a65e7 Update index.js
权限调整
2025-09-19 17:53:43 +08:00
LiJinHui b23e31bb4b Delete DefaultRoleDialog.vue
删除多余文件
2025-09-19 17:41:11 +08:00
Sakura-RanChen bec1c7b88f update: 增加讯飞双流式TTS 2025-09-19 17:29:09 +08:00
LiJinHui 5d407b9227 调整模版排序的漏洞 2025-09-19 17:18:24 +08:00
LiJinHui 608abb7dc3 调整页面 2025-09-19 15:41:28 +08:00
LiJinHui 5fcc7592c9 优化代码,调整权限设置( 非管理员打开提示拒绝访问并自动跳转到首页),修复bug; 2025-09-19 15:34:43 +08:00
LiJinHui 09d5d37867 完善导航栏的语言切换 2025-09-18 16:54:00 +08:00
LiJinHui 6a8c4ccbf8 添加国际化的功能,调整页面 2025-09-18 16:43:52 +08:00
LiJinHui a199e47701 页面布局调整 2025-09-18 09:35:03 +08:00
LiJinHui c331b8322c 新增调整角色模版的功能 2025-09-17 15:38:23 +08:00
1195 changed files with 219989 additions and 19667 deletions
+19
View File
@@ -5,3 +5,22 @@ __pycache__
Dockerfile Dockerfile
tmp/ tmp/
data/ data/
# Repository-local runtimes and test/build products are several gigabytes and
# are never inputs to either manager-api-fastapi image.
.runtime/
.venv-*/
**/.venv/
**/.test-runtime/
**/.pytest_cache/
**/.mypy_cache/
**/.ruff_cache/
**/node_modules/
**/dist/
**/target/
**/uploadfile/
# Runtime state and local model assets must not enter the Docker build context.
main/xiaozhi-server/mysql/
main/xiaozhi-server/models/
main/xiaozhi-server/data/
@@ -0,0 +1,63 @@
## 计划工作包
- 里程碑:M?
- 计划 PRP??
- 负责人角色:
- 依赖 PR/Commit
- 目标分支:`refactor/unified-fastapi-platform`
## 用户可见结果
<!-- 描述完成后用户、设备或运维可以观察到的结果。 -->
## 范围
- 包含:
- 不包含:
- 临时文件所有权:
## 契约影响
- [ ] 不改变公开契约
- [ ] REST API
- [ ] OTA/设备鉴权
- [ ] WebSocket/音频/MQTT 帧
- [ ] 数据库/Redis/文件
- [ ] Provider 配置或结果
说明:
## 验证证据
```text
# 列出可重复执行的命令及结果,不要只写“测试通过”。
```
- [ ] 相关 Tier A 门禁通过
- [ ] 相关隔离集成/协议/差分通过
- [ ] 文档和生成产物一致
- [ ] 没有未解释的 skip、xfail 或 warning 增长
## 风险与回滚
- 风险登记项:
- 失败表现:
- 回滚方式:
- 未执行项目及原因:
## 审阅重点
- [ ] 兼容性
- [ ] 安全/密钥
- [ ] 并发/取消/资源释放
- [ ] 数据迁移
- [ ] 部署/可观测性
- [ ] 测试证据完整性
## 合并检查
- [ ] 已同步最新目标分支
- [ ] 本 PR 的目标不是 `main`
- [ ] 实现者与 Reviewer 不是同一角色
- [ ] PR 只包含一个计划工作包或可独立回滚切片
- [ ] 路线图状态和证据链接将在合并后更新
+6
View File
@@ -0,0 +1,6 @@
version: 2
updates:
- package-ecosystem: pip
directory: /main/xiaozhi-server
schedule:
interval: weekly
+49
View File
@@ -0,0 +1,49 @@
name: Build Base Image
on:
push:
branches:
- main
paths:
- 'main/xiaozhi-server/requirements.txt'
- 'Dockerfile-server-base'
- '.github/workflows/build-base-image.yml'
jobs:
build-base:
name: Build and push server base image
runs-on: ubuntu-latest
permissions:
packages: write
contents: read
steps:
- name: Checkout code
uses: actions/checkout@v4
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@v3
- name: Login to GitHub Container Registry
uses: docker/login-action@v3
with:
registry: ghcr.io
username: ${{ github.actor }}
password: ${{ secrets.TOKEN }}
- name: Build and push server-base
uses: docker/build-push-action@v6
with:
context: .
file: Dockerfile-server-base
push: true
tags: ghcr.io/${{ github.repository }}:server-base
platforms: linux/amd64,linux/arm64
cache-from: type=gha,scope=server-base
cache-to: type=gha,mode=max,scope=server-base
build-args: |
BUILDKIT_PROGRESS=plain
- name: Output image info
run: |
echo "✅ Base image built and pushed successfully!"
echo "📦 Tag: ghcr.io/${{ github.repository }}:server-base"
+4
View File
@@ -5,6 +5,10 @@ on:
tags: tags:
- 'v*.*.*' # 只在以 v 开头的标签推送时触发,例如 v1.0.0 - 'v*.*.*' # 只在以 v 开头的标签推送时触发,例如 v1.0.0
workflow_dispatch: workflow_dispatch:
workflow_run:
workflows: ["Build Base Image"]
types:
- completed
jobs: jobs:
release: release:
+14
View File
@@ -3,6 +3,9 @@ __pycache__/
.idea/ .idea/
*.py[cod] *.py[cod]
*$py.class *$py.class
.vscode
.claude
AGENTS.md
# C extensions # C extensions
*.so *.so
@@ -109,12 +112,17 @@ celerybeat.pid
# Environments # Environments
.env .env
.venv .venv
/.venv-*/
env/ env/
venv/ venv/
ENV/ ENV/
env.bak/ env.bak/
venv.bak/ venv.bak/
# Repository-local runtimes and package-manager caches
/.runtime/
/main/manager-web/.npm-cache/
# Spyder project settings # Spyder project settings
.spyderproject .spyderproject
.spyproject .spyproject
@@ -146,6 +154,7 @@ tmp
.history .history
.DS_Store .DS_Store
main/xiaozhi-server/data main/xiaozhi-server/data
main/xiaozhi-server/.claude
main/manager-web/node_modules main/manager-web/node_modules
.config.yaml .config.yaml
@@ -183,3 +192,8 @@ uploadfile
!main/manager-mobile/**/env/ !main/manager-mobile/**/env/
!main/manager-mobile/**/.env* !main/manager-mobile/**/.env*
!main/manager-mobile/**/*.json !main/manager-mobile/**/*.json
!main/xiaozhi-server/**/*.json
!main/digital-human/**/*.json
main/digital-human/wakeword_runtime/models
main/digital-human/wakeword_runtime/sherpa-onnx-kws-zipformer-wenetspeech-3.3M-2024-01-01
sherpa-onnx-kws-zipformer-wenetspeech-3.3M-2024-01-01.tar.bz2
+2 -32
View File
@@ -1,36 +1,6 @@
# 第一阶段:构建Python依赖 # 生产镜像,仅包含应用代码
FROM python:3.10-slim AS builder FROM ghcr.io/xinnan-tech/xiaozhi-esp32-server:server-base
WORKDIR /app
# 配置pip使用国内镜像源(阿里云)并设置超时和重试
RUN pip config set global.index-url https://mirrors.aliyun.com/pypi/simple/ && \
pip config set global.trusted-host mirrors.aliyun.com && \
pip config set global.timeout 120 && \
pip config set install.retries 5
COPY main/xiaozhi-server/requirements.txt .
# 安装Python依赖,使用并行下载
RUN pip install --no-cache-dir --upgrade pip setuptools wheel && \
pip install --no-cache-dir -r requirements.txt --default-timeout=120 --retries 5
# 第二阶段:生产镜像
FROM python:3.10-slim
WORKDIR /opt/xiaozhi-esp32-server
# 安装系统依赖
RUN apt-get update && \
apt-get install -y --no-install-recommends libopus0 ffmpeg && \
apt-get clean && \
rm -rf /var/lib/apt/lists/*
# 从构建阶段复制Python包和前端构建产物
COPY --from=builder /usr/local/lib/python3.10/site-packages /usr/local/lib/python3.10/site-packages
COPY --from=builder /usr/local/bin/mcp-proxy /usr/local/bin/mcp-proxy
# 复制应用代码
COPY main/xiaozhi-server . COPY main/xiaozhi-server .
# 启动应用 # 启动应用
+32
View File
@@ -0,0 +1,32 @@
# Dockerfile-server-base
# 基础镜像,包含系统依赖和Python包
FROM python:3.10-slim
# 安装系统依赖
RUN apt-get update && \
apt-get install -y --no-install-recommends libopus0 ffmpeg locales && \
sed -i '/zh_CN.UTF-8/s/^# //g' /etc/locale.gen && \
locale-gen && \
apt-get clean && \
rm -rf /var/lib/apt/lists/*
# 配置pip使用国内镜像源(阿里云)并设置超时和重试
RUN pip config set global.index-url https://mirrors.aliyun.com/pypi/simple/ && \
pip config set global.trusted-host mirrors.aliyun.com && \
pip config set global.timeout 120 && \
pip config set install.retries 5
# 设置环境变量以确保正确的字符编码
ENV LANG=zh_CN.UTF-8 \
LC_ALL=zh_CN.UTF-8 \
LANGUAGE=zh_CN:zh \
PYTHONIOENCODING=utf-8
WORKDIR /opt/xiaozhi-esp32-server
# 复制requirements.txt
COPY main/xiaozhi-server/requirements.txt .
# 安装Python依赖
RUN pip install --no-cache-dir --upgrade pip setuptools wheel && \
pip install --no-cache-dir -r requirements.txt --default-timeout=120 --retries 5
+13 -6
View File
@@ -1,5 +1,5 @@
# 第一阶段:构建Vue前端 # 第一阶段:构建Vue前端
FROM node:18 as web-builder FROM node:18 AS web-builder
WORKDIR /app WORKDIR /app
COPY main/manager-web/package*.json ./ COPY main/manager-web/package*.json ./
RUN npm install RUN npm install
@@ -7,7 +7,7 @@ COPY main/manager-web .
RUN npm run build RUN npm run build
# 第二阶段:构建Java后端 # 第二阶段:构建Java后端
FROM maven:3.9.4-eclipse-temurin-21 as api-builder FROM maven:3.9.4-eclipse-temurin-21 AS api-builder
WORKDIR /app WORKDIR /app
COPY main/manager-api/pom.xml . COPY main/manager-api/pom.xml .
COPY main/manager-api/src ./src COPY main/manager-api/src ./src
@@ -18,12 +18,19 @@ FROM bellsoft/liberica-runtime-container:jre-21-glibc
# 安装Nginx和字体库 # 安装Nginx和字体库
RUN apk update && \ RUN apk update && \
apk add --no-cache nginx bash fontconfig ttf-dejavu && \ apk add --no-cache --no-scripts \
apk add --no-cache --repository=http://dl-cdn.alpinelinux.org/alpine/edge/testing/ msttcorefonts-installer || true && \ nginx \
rm -rf /var/cache/apk/* bash \
fontconfig \
ttf-dejavu \
&& rm -rf /var/cache/apk/* \
&& mkdir -p /run/nginx /var/log/nginx /var/tmp/nginx /etc/nginx/conf.d
# 复制项目自带的中文字体
COPY main/manager-web/public/generator/static/fonts/*.ttf /usr/share/fonts/
# 更新字体缓存 # 更新字体缓存
RUN (printf 'YES\n' | update-ms-fonts || true) && fc-cache -f -v RUN fc-cache -f -v
# 配置Nginx # 配置Nginx
COPY docs/docker/nginx.conf /etc/nginx/nginx.conf COPY docs/docker/nginx.conf /etc/nginx/nginx.conf
+80 -94
View File
@@ -6,29 +6,25 @@
本项目基于人机共生智能理论和技术研发智能终端软硬件体系<br/>为开源智能硬件项目 本项目基于人机共生智能理论和技术研发智能终端软硬件体系<br/>为开源智能硬件项目
<a href="https://github.com/78/xiaozhi-esp32">xiaozhi-esp32</a>提供后端服务<br/> <a href="https://github.com/78/xiaozhi-esp32">xiaozhi-esp32</a>提供后端服务<br/>
根据<a href="https://ccnphfhqs21z.feishu.cn/wiki/M0XiwldO9iJwHikpXD5cEx71nKh">小智通信协议</a>使用Python、Java、Vue实现<br/> 根据<a href="https://ccnphfhqs21z.feishu.cn/wiki/M0XiwldO9iJwHikpXD5cEx71nKh">小智通信协议</a>使用Python、Java、Vue实现<br/>
支持MQTT+UDP协议、Websocket协议、MCP接入点、声纹识别 支持MQTT+UDP协议、Websocket协议、MCP接入点、声纹识别、知识库
</p> </p>
<p align="center"> <p align="center">
<a href="./README_en.md">English</a> <a href="./docs/FAQ.md">常见问题</a>
· <a href="./docs/FAQ.md">常见问题</a>
· <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/issues">反馈问题</a> · <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/issues">反馈问题</a>
· <a href="./README.md#%E9%83%A8%E7%BD%B2%E6%96%87%E6%A1%A3">部署文档</a> · <a href="./README.md#%E9%83%A8%E7%BD%B2%E6%96%87%E6%A1%A3">部署文档</a>
· <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/releases">更新日志</a> · <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/releases">更新日志</a>
</p> </p>
<p align="center"> <p align="center">
<a href="./README.md"><img alt="简体中文版自述文件" src="https://img.shields.io/badge/简体中文-DBEDFA"></a>
<a href="./docs/readme/README_en.md"><img alt="README in English" src="https://img.shields.io/badge/English-DFE0E5"></a>
<a href="./docs/readme/README_vi.md"><img alt="Tiếng Việt" src="https://img.shields.io/badge/Tiếng Việt-DFE0E5"></a>
<a href="./docs/readme/README_de.md"><img alt="Deutsch" src="https://img.shields.io/badge/Deutsch-DFE0E5"></a>
<a href="./docs/readme/README_pt_BR.md"><img alt="Português (Brasil)" src="https://img.shields.io/badge/Português (Brasil)-DFE0E5"></a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/releases"> <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/releases">
<img alt="GitHub Contributors" src="https://img.shields.io/github/v/release/xinnan-tech/xiaozhi-esp32-server?logo=docker" /> <img alt="GitHub Contributors" src="https://img.shields.io/github/v/release/xinnan-tech/xiaozhi-esp32-server?logo=docker" />
</a> </a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/graphs/contributors">
<img alt="GitHub Contributors" src="https://img.shields.io/github/contributors/xinnan-tech/xiaozhi-esp32-server?logo=github" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/issues">
<img alt="Issues" src="https://img.shields.io/github/issues/xinnan-tech/xiaozhi-esp32-server?color=0088ff" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/pulls">
<img alt="GitHub pull requests" src="https://img.shields.io/github/issues-pr/xinnan-tech/xiaozhi-esp32-server?color=0088ff" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/blob/main/LICENSE"> <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/blob/main/LICENSE">
<img alt="GitHub pull requests" src="https://img.shields.io/badge/license-MIT-white?labelColor=black" /> <img alt="GitHub pull requests" src="https://img.shields.io/badge/license-MIT-white?labelColor=black" />
</a> </a>
@@ -57,112 +53,97 @@ Spearheaded by Professor Siyuan Liu's Team (South China University of Technology
<table> <table>
<tr> <tr>
<td> <td>
<a href="https://www.bilibili.com/video/BV1C1tCzUEZh" target="_blank"> <a href="https://www.bilibili.com/video/BV1FMFyejExX" target="_blank">
<picture> <picture>
<img alt="小智医疗救护车场景" src="docs/images/demo1.png" /> <img alt="响应速度感受" src="docs/images/demo9.png" /></picture>
</picture>
</a> </a>
</td> </td>
<td>
<a href="https://www.bilibili.com/video/BV1CDKWemEU6" target="_blank">
<picture>
<img alt="自定义音色" src="docs/images/demo2.png" />
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV12yA2egEaC" target="_blank">
<picture>
<img alt="使用粤语交流" src="docs/images/demo3.png" />
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1pNXWYGEx1" target="_blank">
<picture>
<img alt="控制家电开关" src="docs/images/demo5.png" />
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1kgA2eYEQ9" target="_blank">
<picture>
<img alt="成本最低配置" src="docs/images/demo4.png" />
</picture>
</a>
</td>
</tr>
<tr>
<td> <td>
<a href="https://www.bilibili.com/video/BV1vchQzaEse" target="_blank"> <a href="https://www.bilibili.com/video/BV1vchQzaEse" target="_blank">
<picture> <picture>
<img alt="自定义音色" src="docs/images/demo6.png" /> <img alt="速度优化秘诀" src="docs/images/demo6.png" /></picture>
</picture>
</a> </a>
</td> </td>
<td> <td>
<a href="https://www.bilibili.com/video/BV1VC96Y5EMH" target="_blank"> <a href="https://www.bilibili.com/video/BV1WEcxzFEAT" target="_blank">
<picture> <picture>
<img alt="播放音乐" src="docs/images/demo7.png" /> <img alt="小智数字人 支持语音唤醒" src="docs/images/demo8.png" /></picture>
</picture>
</a> </a>
</td> </td>
<td> <td>
<a href="https://www.bilibili.com/video/BV1Z8XuYZEAS" target="_blank"> <a href="https://www.bilibili.com/video/BV1CKVz6UEuB" target="_blank">
<picture> <picture>
<img alt="天气插件" src="docs/images/demo8.png" /> <img alt="设备呼叫设备,打电话" src="docs/images/demo0.png" /></picture>
</picture>
</a> </a>
</td> </td>
<td> <td>
<a href="https://www.bilibili.com/video/BV178XuYfEpi" target="_blank"> <a href="https://www.bilibili.com/video/BV1C1tCzUEZh" target="_blank">
<picture> <picture>
<img alt="IOT指令控制设备" src="docs/images/demo9.png" /> <img alt="复杂医疗场景" src="docs/images/demo1.png" /></picture>
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV17LXWYvENb" target="_blank">
<picture>
<img alt="播报新闻" src="docs/images/demo0.png" />
</picture>
</a> </a>
</td> </td>
</tr> </tr>
<tr> <tr>
<td>
<a href="https://www.bilibili.com/video/BV1VC96Y5EMH" target="_blank">
<picture>
<img alt="播放音乐查询天气播报新闻" src="docs/images/demo7.png" /></picture>
</a>
</td>
<td> <td>
<a href="https://www.bilibili.com/video/BV12J7WzBEaH" target="_blank"> <a href="https://www.bilibili.com/video/BV12J7WzBEaH" target="_blank">
<picture> <picture>
<img alt="实时打断" src="docs/images/demo10.png" /> <img alt="实时打断" src="docs/images/demo10.png" /></picture>
</picture>
</a> </a>
</td> </td>
<td> <td>
<a href="https://www.bilibili.com/video/BV1Co76z7EvK" target="_blank"> <a href="https://www.bilibili.com/video/BV1Co76z7EvK" target="_blank">
<picture> <picture>
<img alt="拍照识物品" src="docs/images/demo12.png" /> <img alt="拍照识物品" src="docs/images/demo12.png" /></picture>
</picture> </a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1pNXWYGEx1" target="_blank">
<picture>
<img alt="控制家电开关" src="docs/images/demo5.png" /></picture>
</a> </a>
</td> </td>
<td> <td>
<a href="https://www.bilibili.com/video/BV1TJ7WzzEo6" target="_blank"> <a href="https://www.bilibili.com/video/BV1TJ7WzzEo6" target="_blank">
<picture> <picture>
<img alt="多指令任务" src="docs/images/demo11.png" /> <img alt="多指令任务" src="docs/images/demo11.png" /></picture>
</picture>
</a> </a>
</td> </td>
</tr>
<tr>
<td> <td>
<a href="https://www.bilibili.com/video/BV1ZQKUzYExM" target="_blank"> <a href="https://www.bilibili.com/video/BV1ZQKUzYExM" target="_blank">
<picture> <picture>
<img alt="MCP接入点" src="docs/images/demo13.png" /> <img alt="MCP接入点" src="docs/images/demo13.png" /></picture>
</picture> </a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1zUW5zJEkq" target="_blank">
<picture>
<img alt="MQTT指令下发" src="docs/images/demo4.png" /></picture>
</a> </a>
</td> </td>
<td> <td>
<a href="https://www.bilibili.com/video/BV1Exu3zqEDe" target="_blank"> <a href="https://www.bilibili.com/video/BV1Exu3zqEDe" target="_blank">
<picture> <picture>
<img alt="声纹识别" src="docs/images/demo14.png" /> <img alt="声纹识别" src="docs/images/demo14.png" /></picture>
</picture> </a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1CDKWemEU6" target="_blank">
<picture>
<img alt="自定义音色" src="docs/images/demo2.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV12yA2egEaC" target="_blank">
<picture>
<img alt="使用粤语交流" src="docs/images/demo3.png" /></picture>
</a> </a>
</td> </td>
</tr> </tr>
@@ -188,8 +169,8 @@ Spearheaded by Professor Siyuan Liu's Team (South China University of Technology
#### 🚀 部署方式选择 #### 🚀 部署方式选择
| 部署方式 | 特点 | 适用场景 | 部署文档 | 配置要求 | 视频教程 | | 部署方式 | 特点 | 适用场景 | 部署文档 | 配置要求 | 视频教程 |
|---------|------|---------|---------|---------|---------| |---------|------|---------|---------|---------|---------|
| **最简化安装** | 智能对话、IOT、MCP、视觉感知 | 低配置环境,数据存储在配置文件,无需数据库 | [①Docker版](./docs/Deployment.md#%E6%96%B9%E5%BC%8F%E4%B8%80docker%E5%8F%AA%E8%BF%90%E8%A1%8Cserver) / [②源码部署](./docs/Deployment.md#%E6%96%B9%E5%BC%8F%E4%BA%8C%E6%9C%AC%E5%9C%B0%E6%BA%90%E7%A0%81%E5%8F%AA%E8%BF%90%E8%A1%8Cserver)| 如果使用`FunASR`要2核4G,如果全API,要2核2G | - | | **最简化安装** | 智能对话、单智能体管理 | 低配置环境,数据存储在配置文件,无需数据库 | [①Docker版](./docs/Deployment.md#%E6%96%B9%E5%BC%8F%E4%B8%80docker%E5%8F%AA%E8%BF%90%E8%A1%8Cserver) / [②源码部署](./docs/Deployment.md#%E6%96%B9%E5%BC%8F%E4%BA%8C%E6%9C%AC%E5%9C%B0%E6%BA%90%E7%A0%81%E5%8F%AA%E8%BF%90%E8%A1%8Cserver)| 如果使用`FunASR`要2核4G,如果全API,要2核2G | - |
| **全模块安装** | 智能对话、IOT、MCP接入点、声纹识别、视觉感知、OTA、智控台 | 完整功能体验,数据存储在数据库 |[①Docker版](./docs/Deployment_all.md#%E6%96%B9%E5%BC%8F%E4%B8%80docker%E8%BF%90%E8%A1%8C%E5%85%A8%E6%A8%A1%E5%9D%97) / [②源码部署](./docs/Deployment_all.md#%E6%96%B9%E5%BC%8F%E4%BA%8C%E6%9C%AC%E5%9C%B0%E6%BA%90%E7%A0%81%E8%BF%90%E8%A1%8C%E5%85%A8%E6%A8%A1%E5%9D%97) / [③源码部署自动更新教程](./docs/dev-ops-integration.md) | 如果使用`FunASR`要4核8G,如果全API,要2核4G| [本地源码启动视频教程](https://www.bilibili.com/video/BV1wBJhz4Ewe) | | **全模块安装** | 智能对话、多用户管理、多智能体管理、智控台界面操作 | 完整功能体验,数据存储在数据库 |[①Docker版](./docs/Deployment_all.md#%E6%96%B9%E5%BC%8F%E4%B8%80docker%E8%BF%90%E8%A1%8C%E5%85%A8%E6%A8%A1%E5%9D%97) / [②源码部署](./docs/Deployment_all.md#%E6%96%B9%E5%BC%8F%E4%BA%8C%E6%9C%AC%E5%9C%B0%E6%BA%90%E7%A0%81%E8%BF%90%E8%A1%8C%E5%85%A8%E6%A8%A1%E5%9D%97) / [③源码部署自动更新教程](./docs/dev-ops-integration.md) | 如果使用`FunASR`要4核8G,如果全API,要2核4G| [本地源码启动视频教程](https://www.bilibili.com/video/BV1wBJhz4Ewe) |
常见问题及相关教程,可参考[这个链接](./docs/FAQ.md) 常见问题及相关教程,可参考[这个链接](./docs/FAQ.md)
@@ -216,10 +197,10 @@ Websocket接口地址: wss://2662r3426b.vicp.fun/xiaozhi/v1/
| 模块名称 | 入门全免费设置 | 流式配置 | | 模块名称 | 入门全免费设置 | 流式配置 |
|:---:|:---:|:---:| |:---:|:---:|:---:|
| ASR(语音识别) | FunASR(本地) | 👍FunASR(本地GPU模式) | | ASR(语音识别) | FunASR(本地) | 👍XunfeiStreamASR(讯飞流式) |
| LLM(大模型) | ChatGLMLLM(智谱glm-4-flash) | 👍AliLLM(qwen3-235b-a22b-instruct-2507) 或 👍DoubaoLLM(doubao-1-5-pro-32k-250115) | | LLM(大模型) | glm-4-flash(智谱) | 👍qwen-flash(阿里百炼) |
| VLLM(视觉大模型) | ChatGLMVLLM(智谱glm-4v-flash) | 👍QwenVLVLLM(千问qwen2.5-vl-3b-instructh) | | VLLM(视觉大模型) | glm-4v-flash(智谱) | 👍qwen3.5-flash(阿里百炼) |
| TTS(语音合成) | ✅LinkeraiTTS(灵犀流式) | 👍HuoshanDoubleStreamTTS(火山流式语音合成) 或 👍AliyunStreamTTS(阿里云流式语音合成) | | TTS(语音合成) | EdgeTTS(微软) | 👍HuoshanDoubleStreamTTS(火山流式) |
| Intent(意图识别) | function_call(函数调用) | function_call(函数调用) | | Intent(意图识别) | function_call(函数调用) | function_call(函数调用) |
| Memory(记忆功能) | mem_local_short(本地短期记忆) | mem_local_short(本地短期记忆) | | Memory(记忆功能) | mem_local_short(本地短期记忆) | mem_local_short(本地短期记忆) |
@@ -230,7 +211,7 @@ Websocket接口地址: wss://2662r3426b.vicp.fun/xiaozhi/v1/
| 工具名称 | 位置 | 使用方法 | 功能说明 | | 工具名称 | 位置 | 使用方法 | 功能说明 |
|:---:|:---|:---:|:---:| |:---:|:---|:---:|:---:|
| 音频交互测试工具 | main》xiaozhi-server》test》test_page.html | 使用谷歌浏览器直接打开 | 测试音频播放和接收功能,验证Python端音频处理是否正常 | | 音频交互测试工具 | main》digital-human》index.html | 在 `main/digital-human` 执行 `python start.py` 后访问 `http://127.0.0.1:8006/index.html` | 测试音频播放和接收功能,验证Python端音频处理是否正常 |
| 模型响应测试工具 | main》xiaozhi-server》performance_tester.py | 执行 `python performance_tester.py` | 测试ASR(语音识别)、LLM(大模型)、VLLM(视觉模型)、TTS(语音合成)三个核心模块的响应速度 | | 模型响应测试工具 | main》xiaozhi-server》performance_tester.py | 执行 `python performance_tester.py` | 测试ASR(语音识别)、LLM(大模型)、VLLM(视觉模型)、TTS(语音合成)三个核心模块的响应速度 |
> 💡 提示:测试模型速度时,只会测试配置了密钥的模型。 > 💡 提示:测试模型速度时,只会测试配置了密钥的模型。
@@ -241,15 +222,17 @@ Websocket接口地址: wss://2662r3426b.vicp.fun/xiaozhi/v1/
![请参考-全模块安装架构图](docs/images/deploy2.png) ![请参考-全模块安装架构图](docs/images/deploy2.png)
| 功能模块 | 描述 | | 功能模块 | 描述 |
|:---:|:---| |:---:|:---|
| 核心架构 | 基于MQTT+UDP网关、WebSocket、HTTP服务器,提供完整的控制台管理和认证系统 | | 核心架构 | 基于[MQTT+UDP网关](https://github.com/xinnan-tech/xiaozhi-esp32-server/blob/main/docs/mqtt-gateway-integration.md)、WebSocket、HTTP服务器,提供完整的控制台管理和认证系统 |
| 语音交互 | 支持流式ASR(语音识别)、流式TTS(语音合成)、VAD(语音活动检测),支持多语言识别和语音处理 | | 语音交互 | 支持流式ASR(语音识别)、流式TTS(语音合成)、VAD(语音活动检测),支持多语言识别和语音处理 |
| 声纹识别 | 支持多用户声纹注册、管理和识别,与ASR并行处理,实时识别说话人身份并传递给LLM进行个性化回应 | | 声纹识别 | 支持多用户声纹注册、管理和识别,与ASR并行处理,实时识别说话人身份并传递给LLM进行个性化回应 |
| 智能对话 | 支持多种LLM(大语言模型),实现智能对话 | | 智能对话 | 支持多种LLM(大语言模型),实现智能对话 |
| 视觉感知 | 支持多种VLLM(视觉大模型),实现多模态交互 | | 视觉感知 | 支持多种VLLM(视觉大模型),实现多模态交互 |
| 意图识别 | 支持LLM意图识别、Function Call函数调用,提供插件化意图处理机制 | | 意图识别 | 支持外挂的大模型意图识别、大模型自主函数调用,提供插件化意图处理机制 |
| 记忆系统 | 支持本地短期记忆、mem0ai接口记忆,具备记忆总结功能 | | 记忆系统 | 支持本地短期记忆、mem0ai接口记忆、PowerMem智能记忆,具备记忆总结功能 |
| 知识库 | 支持RAGFlow知识库,让大模型判断需要调度知识库后再回答 |
| 工具调用 | 支持客户端IOT协议、客户MCP协议、服务端MCP协议、MCP接入点协议、自定义工具函数 | | 工具调用 | 支持客户端IOT协议、客户MCP协议、服务端MCP协议、MCP接入点协议、自定义工具函数 |
| 管理后台 | 提供Web管理界面,支持用户管理、系统配置和设备管理 | | 指令下发 | 依托MQTT协议,支持从智控台将MCP指令下发到ESP32设备 |
| 管理后台 | 提供Web管理界面,支持用户管理、系统配置和设备管理;界面支持中文简体、中文繁体、英文显示 |
| 测试工具 | 提供性能测试工具、视觉模型测试工具和音频交互测试工具 | | 测试工具 | 提供性能测试工具、视觉模型测试工具和音频交互测试工具 |
| 部署支持 | 支持Docker部署和本地部署,提供完整的配置文件管理 | | 部署支持 | 支持Docker部署和本地部署,提供完整的配置文件管理 |
| 插件系统 | 支持功能插件扩展、自定义插件开发和插件热加载 | | 插件系统 | 支持功能插件扩展、自定义插件开发和插件热加载 |
@@ -263,13 +246,7 @@ Websocket接口地址: wss://2662r3426b.vicp.fun/xiaozhi/v1/
--- ---
## 产品生态 👬 ## 产品生态 👬
小智是一个生态,当你使用这个产品时,也可以看看其他在这个生态圈的优秀项目 小智是一个生态,当你使用这个产品时,也可以看看其他在这个生态圈的[优秀项目](https://github.com/78/xiaozhi-esp32/blob/main/README_zh.md#%E7%9B%B8%E5%85%B3%E5%BC%80%E6%BA%90%E9%A1%B9%E7%9B%AE)
| 项目名称 | 项目地址 | 项目描述 |
|:---------------------|:--------|:--------|
| 小智安卓客户端 | [xiaozhi-android-client](https://github.com/TOM88812/xiaozhi-android-client) | 一个基于xiaozhi-server的Android、IOS语音对话应用,支持实时语音交互和文字对话。<br/>现在是flutter版本,打通IOS、Android端。 |
| 小智电脑客户端 | [py-xiaozhi](https://github.com/Huang-junsen/py-xiaozhi) | 该项目提供了一个基于 Python 实现的小白 AI 客户端,使得在不具备实体硬件条件的情况下,<br/>依然能够体过代码体验小智 AI 的功能。 |
| 小智Java服务端 | [xiaozhi-esp32-server-java](https://github.com/joey-zhou/xiaozhi-esp32-server-java) | 小智开源后端服务 Java 版本是一个基于 Java 的开源项目。<br/>它包括前后端的服务,旨在为用户提供一个完整的后端服务解决方案。 |
--- ---
@@ -278,7 +255,7 @@ Websocket接口地址: wss://2662r3426b.vicp.fun/xiaozhi/v1/
| 使用方式 | 支持平台 | 免费平台 | | 使用方式 | 支持平台 | 免费平台 |
|:---:|:---:|:---:| |:---:|:---:|:---:|
| openai 接口调用 | 阿里百炼、火山引擎豆包、深度求索、智谱ChatGLM、Gemini | 智谱ChatGLM、Gemini | | openai 接口调用 | 阿里百炼、火山引擎、DeepSeek、智谱、Gemini、科大讯飞 | 智谱、Gemini |
| ollama 接口调用 | Ollama | - | | ollama 接口调用 | Ollama | - |
| dify 接口调用 | Dify | - | | dify 接口调用 | Dify | - |
| fastgpt 接口调用 | Fastgpt | - | | fastgpt 接口调用 | Fastgpt | - |
@@ -304,7 +281,7 @@ Websocket接口地址: wss://2662r3426b.vicp.fun/xiaozhi/v1/
| 使用方式 | 支持平台 | 免费平台 | | 使用方式 | 支持平台 | 免费平台 |
|:---:|:---:|:---:| |:---:|:---:|:---:|
| 接口调用 | EdgeTTS、火山引擎豆包TTS、腾讯云、阿里云TTS、阿里云流式TTS、CosyVoiceSiliconflow、TTS302AI、CozeCnTTS、GizwitsTTS、ACGNTTS、OpenAITTS、灵犀流式TTS、MinimaxTTS、火山双流式TTS | 灵犀流式TTS、EdgeTTS、CosyVoiceSiliconflow(部分) | | 接口调用 | EdgeTTS、科大讯飞、火山引擎、腾讯云、阿里云及百炼、CosyVoiceSiliconflow、TTS302AI、CozeCnTTS、GizwitsTTS、ACGNTTS、OpenAITTS、灵犀流式TTS、MinimaxTTS | 灵犀流式TTS、EdgeTTS、CosyVoiceSiliconflow(部分) |
| 本地服务 | FishSpeech、GPT_SOVITS_V2、GPT_SOVITS_V3、Index-TTS、PaddleSpeech | Index-TTS、PaddleSpeech、FishSpeech、GPT_SOVITS_V2、GPT_SOVITS_V3 | | 本地服务 | FishSpeech、GPT_SOVITS_V2、GPT_SOVITS_V3、Index-TTS、PaddleSpeech | Index-TTS、PaddleSpeech、FishSpeech、GPT_SOVITS_V2、GPT_SOVITS_V3 |
--- ---
@@ -322,7 +299,7 @@ Websocket接口地址: wss://2662r3426b.vicp.fun/xiaozhi/v1/
| 使用方式 | 支持平台 | 免费平台 | | 使用方式 | 支持平台 | 免费平台 |
|:---:|:---:|:---:| |:---:|:---:|:---:|
| 本地使用 | FunASR、SherpaASR | FunASR、SherpaASR | | 本地使用 | FunASR、SherpaASR | FunASR、SherpaASR |
| 接口调用 | DoubaoASR、Doubao流式ASR、FunASRServer、TencentASR、AliyunASR、Aliyun流式ASR、百度ASR、OpenAI ASR | FunASRServer | | 接口调用 | FunASRServer、火山引擎、科大讯飞、腾讯云、阿里云、百度云、OpenAI ASR | FunASRServer |
--- ---
@@ -339,6 +316,7 @@ Websocket接口地址: wss://2662r3426b.vicp.fun/xiaozhi/v1/
| 类型 | 平台名称 | 使用方式 | 收费模式 | 备注 | | 类型 | 平台名称 | 使用方式 | 收费模式 | 备注 |
|:------:|:---------------:|:----:|:---------:|:--:| |:------:|:---------------:|:----:|:---------:|:--:|
| Memory | mem0ai | 接口调用 | 1000次/月额度 | | | Memory | mem0ai | 接口调用 | 1000次/月额度 | |
| Memory | [powermem](./docs/powermem-integration.md) | 本地总结 | 取决于LLM和DB | OceanBase开源,支持智能检索 |
| Memory | mem_local_short | 本地总结 | 免费 | | | Memory | mem_local_short | 本地总结 | 免费 | |
| Memory | nomem | 无记忆模式 | 免费 | | | Memory | nomem | 无记忆模式 | 免费 | |
@@ -354,6 +332,14 @@ Websocket接口地址: wss://2662r3426b.vicp.fun/xiaozhi/v1/
--- ---
### Rag 检索增强生成
| 类型 | 平台名称 | 使用方式 | 收费模式 | 备注 |
|:------:|:-------------:|:----:|:-------:|:---------------------:|
| Rag | ragflow | 接口调用 | 根据切片、分词消耗的token收费 | 借助RagFlow的检索增强生成功能,提供更准确的对话回复 |
---
## 鸣谢 🙏 ## 鸣谢 🙏
| Logo | 项目/公司 | 说明 | | Logo | 项目/公司 | 说明 |
-369
View File
@@ -1,369 +0,0 @@
[![Banners](docs/images/banner1.png)](https://github.com/xinnan-tech/xiaozhi-esp32-server)
<h1 align="center">Xiaozhi Backend Service xiaozhi-esp32-server</h1>
<p align="center">
This project is based on human-machine symbiotic intelligence theory and technology to develop intelligent terminal hardware and software systems<br/>providing backend services for the open-source intelligent hardware project
<a href="https://github.com/78/xiaozhi-esp32">xiaozhi-esp32</a><br/>
Implemented using Python, Java, and Vue according to the <a href="https://ccnphfhqs21z.feishu.cn/wiki/M0XiwldO9iJwHikpXD5cEx71nKh">Xiaozhi Communication Protocol</a><br/>
Supports MCP endpoints and voiceprint recognition
</p>
<p align="center">
<a href="./README.md">中文</a>
· <a href="./docs/FAQ.md">FAQ</a>
· <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/issues">Report Issues</a>
· <a href="./README.md#%E9%83%A8%E7%BD%B2%E6%96%87%E6%A1%A3">Deployment Docs</a>
· <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/releases">Release Notes</a>
</p>
<p align="center">
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/releases">
<img alt="GitHub Contributors" src="https://img.shields.io/github/v/release/xinnan-tech/xiaozhi-esp32-server?logo=docker" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/graphs/contributors">
<img alt="GitHub Contributors" src="https://img.shields.io/github/contributors/xinnan-tech/xiaozhi-esp32-server?logo=github" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/issues">
<img alt="Issues" src="https://img.shields.io/github/issues/xinnan-tech/xiaozhi-esp32-server?color=0088ff" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/pulls">
<img alt="GitHub pull requests" src="https://img.shields.io/github/issues-pr/xinnan-tech/xiaozhi-esp32-server?color=0088ff" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/blob/main/LICENSE">
<img alt="GitHub pull requests" src="https://img.shields.io/badge/license-MIT-white?labelColor=black" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server">
<img alt="stars" src="https://img.shields.io/github/stars/xinnan-tech/xiaozhi-esp32-server?color=ffcb47&labelColor=black" />
</a>
</p>
<p align="center">
Spearheaded by Professor Siyuan Liu's Team (South China University of Technology)
</br>
刘思源教授团队主导研发(华南理工大学)
</br>
<img src="./docs/images/hnlg.jpg" alt="South China University of Technology" width="50%">
</p>
---
## Target Users 👥
This project requires ESP32 hardware devices to work. If you have purchased ESP32-related hardware, successfully connected to Brother Xia's deployed backend service, and want to build your own `xiaozhi-esp32` backend service independently, then this project is perfect for you.
Want to see the usage effects? Click the videos below 🎥
<table>
<tr>
<td>
<a href="https://www.bilibili.com/video/BV1FMFyejExX" target="_blank">
<picture>
<img alt="Xiaozhi ESP32 connecting to own backend model" src="docs/images/demo1.png" />
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1CDKWemEU6" target="_blank">
<picture>
<img alt="Custom voice timbre" src="docs/images/demo2.png" />
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV12yA2egEaC" target="_blank">
<picture>
<img alt="Using Cantonese for communication" src="docs/images/demo3.png" />
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1pNXWYGEx1" target="_blank">
<picture>
<img alt="Controlling home appliances" src="docs/images/demo5.png" />
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1kgA2eYEQ9" target="_blank">
<picture>
<img alt="Lowest cost configuration" src="docs/images/demo4.png" />
</picture>
</a>
</td>
</tr>
<tr>
<td>
<a href="https://www.bilibili.com/video/BV1vchQzaEse" target="_blank">
<picture>
<img alt="Custom voice timbre" src="docs/images/demo6.png" />
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1VC96Y5EMH" target="_blank">
<picture>
<img alt="Playing music" src="docs/images/demo7.png" />
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1Z8XuYZEAS" target="_blank">
<picture>
<img alt="Weather plugin" src="docs/images/demo8.png" />
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV178XuYfEpi" target="_blank">
<picture>
<img alt="IOT command control devices" src="docs/images/demo9.png" />
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV17LXWYvENb" target="_blank">
<picture>
<img alt="News broadcasting" src="docs/images/demo0.png" />
</picture>
</a>
</td>
</tr>
<tr>
<td>
<a href="https://www.bilibili.com/video/BV12J7WzBEaH" target="_blank">
<picture>
<img alt="Real-time interruption" src="docs/images/demo10.png" />
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1Co76z7EvK" target="_blank">
<picture>
<img alt="Photo recognition of objects" src="docs/images/demo12.png" />
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1TJ7WzzEo6" target="_blank">
<picture>
<img alt="Multi-command tasks" src="docs/images/demo11.png" />
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1ZQKUzYExM" target="_blank">
<picture>
<img alt="MCP endpoint" src="docs/images/demo13.png" />
</picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1Exu3zqEDe" target="_blank">
<picture>
<img alt="Voiceprint recognition" src="docs/images/demo14.png" />
</picture>
</a>
</td>
</tr>
</table>
---
## Warnings ⚠️
1. This project is open-source software. This software has no commercial partnership with any third-party API service providers (including but not limited to speech recognition, large models, speech synthesis, and other platforms) that it interfaces with, and does not provide any form of guarantee for their service quality or financial security. It is recommended that users prioritize service providers with relevant business licenses and carefully read their service agreements and privacy policies. This software does not host any account keys, does not participate in fund flows, and does not bear the risk of recharge fund losses.
2. The functionality of this project is not complete and has not passed network security assessment. Please do not use it in production environments. If you deploy this project for learning purposes in a public network environment, please ensure necessary protection measures are in place.
---
## Deployment Documentation
![Banners](docs/images/banner2.png)
This project provides two deployment methods. Please choose based on your specific needs:
#### 🚀 Deployment Method Selection
| Deployment Method | Features | Applicable Scenarios | Deployment Docs | Configuration Requirements | Video Tutorials |
|---------|------|---------|---------|---------|---------|
| **Simplified Installation** | Intelligent dialogue, IOT, MCP, visual perception | Low-configuration environments, data stored in config files, no database required | [①Docker Version](./docs/Deployment.md#%E6%96%B9%E5%BC%8F%E4%B8%80docker%E5%8F%AA%E8%BF%90%E8%A1%8Cserver) / [②Source Code Deployment](./docs/Deployment.md#%E6%96%B9%E5%BC%8F%E4%BA%8C%E6%9C%AC%E5%9C%B0%E6%BA%90%E7%A0%81%E5%8F%AA%E8%BF%90%E8%A1%8Cserver)| 2 cores 4GB if using `FunASR`, 2 cores 2GB if all APIs | - |
| **Full Module Installation** | Intelligent dialogue, IOT, MCP endpoints, voiceprint recognition, visual perception, OTA, intelligent control console | Complete functionality experience, data stored in database |[①Docker Version](./docs/Deployment_all.md#%E6%96%B9%E5%BC%8F%E4%B8%80docker%E8%BF%90%E8%A1%8C%E5%85%A8%E6%A8%A1%E5%9D%97) / [②Source Code Deployment](./docs/Deployment_all.md#%E6%96%B9%E5%BC%8F%E4%BA%8C%E6%9C%AC%E5%9C%B0%E6%BA%90%E7%A0%81%E8%BF%90%E8%A1%8C%E5%85%A8%E6%A8%A1%E5%9D%97) / [③Source Code Deployment Auto-Update Tutorial](./docs/dev-ops-integration.md) | 4 cores 8GB if using `FunASR`, 2 cores 4GB if all APIs| [Local Source Code Startup Video Tutorial](https://www.bilibili.com/video/BV1wBJhz4Ewe) |
> 💡 Note: Below is a test platform deployed with the latest code. You can burn and test if needed. Concurrent users: 6, data will be cleared daily.
```
Intelligent Control Console Address: https://2662r3426b.vicp.fun
Intelligent Control Console Address (H5): https://2662r3426b.vicp.fun/h5/index.html
Service Test Tool: https://2662r3426b.vicp.fun/test/
OTA Interface Address: https://2662r3426b.vicp.fun/xiaozhi/ota/
Websocket Interface Address: wss://2662r3426b.vicp.fun/xiaozhi/v1/
```
#### 🚩 Configuration Description and Recommendations
> [!Note]
> This project provides two configuration schemes:
>
> 1. `Entry Level Free Settings`: Suitable for personal and home use, all components use free solutions, no additional payment required.
>
> 2. `Streaming Configuration`: Suitable for demonstrations, training, scenarios with more than 2 concurrent users, etc. Uses streaming processing technology for faster response speed and better experience.
>
> Starting from version `0.5.2`, the project supports streaming configuration. Compared to earlier versions, response speed is improved by approximately `2.5 seconds`, significantly improving user experience.
| Module Name | Entry Level Free Settings | Streaming Configuration |
|:---:|:---:|:---:|
| ASR(Speech Recognition) | FunASR(Local) | 👍FunASRServer or 👍DoubaoStreamASR |
| LLM(Large Model) | ChatGLMLLM(Zhipu glm-4-flash) | 👍DoubaoLLM(Volcano doubao-1-5-pro-32k-250115) |
| VLLM(Vision Large Model) | ChatGLMVLLM(Zhipu glm-4v-flash) | 👍QwenVLVLLM(Qwen qwen2.5-vl-3b-instructh) |
| TTS(Speech Synthesis) | ✅LinkeraiTTS(Lingxi streaming) | 👍HuoshanDoubleStreamTTS(Volcano dual-stream speech synthesis) |
| Intent(Intent Recognition) | function_call(Function calling) | function_call(Function calling) |
| Memory(Memory function) | mem_local_short(Local short-term memory) | mem_local_short(Local short-term memory) |
#### 🔧 Testing Tools
This project provides the following testing tools to help you verify the system and choose suitable models:
| Tool Name | Location | Usage Method | Function Description |
|:---:|:---|:---:|:---:|
| Audio Interaction Test Tool | main》xiaozhi-server》test》test_page.html | Open directly with Google Chrome | Tests audio playback and reception functions, verifies if Python-side audio processing is normal |
| Model Response Test Tool 1 | main》xiaozhi-server》performance_tester.py | Execute `python performance_tester.py` | Tests response speed of three core modules: ASR(speech recognition), LLM(large model), TTS(speech synthesis) |
| Model Response Test Tool 2 | main》xiaozhi-server》performance_tester_vllm.py | Execute `python performance_tester_vllm.py` | Tests VLLM(vision model) response speed |
> 💡 Note: When testing model speed, only models with configured keys will be tested.
---
## Feature List ✨
### Implemented ✅
![请参考-全模块安装架构图](docs/images/deploy2.png)
| Feature Module | Description |
|:---:|:---|
| Core Architecture | Based on WebSocket and HTTP servers, provides complete console management and authentication system |
| Voice Interaction | Supports streaming ASR(speech recognition), streaming TTS(speech synthesis), VAD(voice activity detection), supports multi-language recognition and voice processing |
| Voiceprint Recognition | Supports multi-user voiceprint registration, management, and recognition, processes in parallel with ASR, real-time speaker identity recognition and passes to LLM for personalized responses |
| Intelligent Dialogue | Supports multiple LLM(large language models), implements intelligent dialogue |
| Visual Perception | Supports multiple VLLM(vision large models), implements multimodal interaction |
| Intent Recognition | Supports LLM intent recognition, Function Call function calling, provides plugin-based intent processing mechanism |
| Memory System | Supports local short-term memory, mem0ai interface memory, with memory summarization functionality |
| Tool Calling | Supports client IOT protocol, client MCP protocol, server MCP protocol, MCP endpoint protocol, custom tool functions |
| Management Backend | Provides Web management interface, supports user management, system configuration, and device management |
| Testing Tools | Provides performance testing tools, vision model testing tools, and audio interaction testing tools |
| Deployment Support | Supports Docker deployment and local deployment, provides complete configuration file management |
| Plugin System | Supports functional plugin extensions, custom plugin development, and plugin hot-loading |
### Under Development 🚧
To learn about specific development plan progress, [click here](https://github.com/users/xinnan-tech/projects/3)
If you are a software developer, here is an [Open Letter to Developers](docs/contributor_open_letter.md). Welcome to join!
---
## Product Ecosystem 👬
Xiaozhi is an ecosystem. When using this product, you can also check out other excellent projects in this ecosystem
| Project Name | Project Address | Project Description |
|:---------------------|:--------|:--------|
| Xiaozhi Android Client | [xiaozhi-android-client](https://github.com/TOM88812/xiaozhi-android-client) | An Android and iOS voice dialogue application based on xiaozhi-server, supporting real-time voice interaction and text dialogue.<br/>Currently a Flutter version, connecting iOS and Android platforms. |
| Xiaozhi Desktop Client | [py-xiaozhi](https://github.com/Huang-junsen/py-xiaozhi) | This project provides a Python-based AI client for beginners, allowing users to experience Xiaozhi AI functionality through code even without physical hardware conditions. |
| Xiaozhi Java Server | [xiaozhi-esp32-server-java](https://github.com/joey-zhou/xiaozhi-esp32-server-java) | Xiaozhi open-source backend service Java version is a Java-based open-source project.<br/>It includes frontend and backend services, aiming to provide users with a complete backend service solution. |
---
## Supported Platforms/Components List 📋
### LLM Language Models
| Usage Method | Supported Platforms | Free Platforms |
|:---:|:---:|:---:|
| OpenAI interface calls | Alibaba Bailian, Volcano Engine Doubao, DeepSeek, Zhipu ChatGLM, Gemini | Zhipu ChatGLM, Gemini |
| Ollama interface calls | Ollama | - |
| Dify interface calls | Dify | - |
| FastGPT interface calls | FastGPT | - |
| Coze interface calls | Coze | - |
In fact, any LLM that supports OpenAI interface calls can be integrated and used.
---
### VLLM Vision Models
| Usage Method | Supported Platforms | Free Platforms |
|:---:|:---:|:---:|
| OpenAI interface calls | Alibaba Bailian, Zhipu ChatGLMVLLM | Zhipu ChatGLMVLLM |
In fact, any VLLM that supports OpenAI interface calls can be integrated and used.
---
### TTS Speech Synthesis
| Usage Method | Supported Platforms | Free Platforms |
|:---:|:---:|:---:|
| Interface calls | EdgeTTS, Volcano Engine Doubao TTS, Tencent Cloud, Alibaba Cloud TTS, CosyVoiceSiliconflow, TTS302AI, CozeCnTTS, GizwitsTTS, ACGNTTS, OpenAITTS, Lingxi Streaming TTS | Lingxi Streaming TTS, EdgeTTS, CosyVoiceSiliconflow(partial) |
| Local services | FishSpeech, GPT_SOVITS_V2, GPT_SOVITS_V3, MinimaxTTS | FishSpeech, GPT_SOVITS_V2, GPT_SOVITS_V3, MinimaxTTS |
---
### VAD Voice Activity Detection
| Type | Platform Name | Usage Method | Pricing Model | Notes |
|:---:|:---------:|:----:|:----:|:--:|
| VAD | SileroVAD | Local use | Free | |
---
### ASR Speech Recognition
| Usage Method | Supported Platforms | Free Platforms |
|:---:|:---:|:---:|
| Local use | FunASR, SherpaASR | FunASR, SherpaASR |
| Interface calls | DoubaoASR, FunASRServer, TencentASR, AliyunASR | FunASRServer |
---
### Voiceprint Recognition
| Usage Method | Supported Platforms | Free Platforms |
|:---:|:---:|:---:|
| Local use | 3D-Speaker | 3D-Speaker |
---
### Memory Storage
| Type | Platform Name | Usage Method | Pricing Model | Notes |
|:------:|:---------------:|:----:|:---------:|:--:|
| Memory | mem0ai | Interface calls | 1000 times/month quota | |
| Memory | mem_local_short | Local summarization | Free | |
---
### Intent Recognition
| Type | Platform Name | Usage Method | Pricing Model | Notes |
|:------:|:-------------:|:----:|:-------:|:---------------------:|
| Intent | intent_llm | Interface calls | Based on LLM pricing | Recognizes intent through large models, strong generalization |
| Intent | function_call | Interface calls | Based on LLM pricing | Completes intent through large model function calling, fast speed, good effect |
---
## Acknowledgments 🙏
| Logo | Project/Company | Description |
|:---:|:---:|:---|
| <img src="./docs/images/logo_bailing.png" width="160"> | [Bailing Voice Dialogue Robot](https://github.com/wwbin2017/bailing) | This project is inspired by [Bailing Voice Dialogue Robot](https://github.com/wwbin2017/bailing) and implemented on its basis |
| <img src="./docs/images/logo_tenclass.png" width="160"> | [Tenclass](https://www.tenclass.com/) | Thanks to [Tenclass](https://www.tenclass.com/) for formulating standard communication protocols, multi-device compatibility solutions, and high-concurrency scenario practice demonstrations for the Xiaozhi ecosystem; providing full-link technical documentation support for this project |
| <img src="./docs/images/logo_xuanfeng.png" width="160"> | [Xuanfeng Technology](https://github.com/Eric0308) | Thanks to [Xuanfeng Technology](https://github.com/Eric0308) for contributing function calling framework, MCP communication protocol, and plugin-based calling mechanism implementation code. Through standardized instruction scheduling system and dynamic expansion capabilities, it significantly improves the interaction efficiency and functional extensibility of frontend devices (IoT) |
| <img src="./docs/images/logo_junsen.png" width="160"> | [huangjunsen](https://github.com/huangjunsen0406) | Thanks to [huangjunsen](https://github.com/huangjunsen0406) for contributing the `Smart Control Console Mobile` module, which enables efficient control and real-time interaction across mobile devices, significantly enhancing the system's operational convenience and management efficiency in mobile scenarios. |
| <img src="./docs/images/logo_huiyuan.png" width="160"> | [Huiyuan Design](http://ui.kwd988.net/) | Thanks to [Huiyuan Design](http://ui.kwd988.net/) for providing professional visual solutions for this project, using their design practical experience serving over a thousand enterprises to empower this project's product user experience |
| <img src="./docs/images/logo_qinren.png" width="160"> | [Xi'an Qinren Information Technology](https://www.029app.com/) | Thanks to [Xi'an Qinren Information Technology](https://www.029app.com/) for deepening this project's visual system, ensuring consistency and extensibility of overall design style in multi-scenario applications |
| <img src="./docs/images/logo_contributors.png" width="160"> | [Code Contributors](https://github.com/xinnan-tech/xiaozhi-esp32-server/graphs/contributors) | Thanks to [all code contributors](https://github.com/xinnan-tech/xiaozhi-esp32-server/graphs/contributors), your efforts have made the project more robust and powerful. |
<a href="https://star-history.com/#xinnan-tech/xiaozhi-esp32-server&Date">
<picture>
<source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=xinnan-tech/xiaozhi-esp32-server&type=Date&theme=dark" />
<source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=xinnan-tech/xiaozhi-esp32-server&type=Date" />
<img alt="Star History Chart" src="https://api.star-history.com/svg?repos=xinnan-tech/xiaozhi-esp32-server&type=Date" />
</picture>
</a>
+14 -8
View File
@@ -2,7 +2,7 @@
![请参考-最简化架构图](../docs/images/deploy1.png) ![请参考-最简化架构图](../docs/images/deploy1.png)
# 方式一:Docker只运行Server # 方式一:Docker只运行Server
docker镜像支持x86架构arm64架构的CPU,支持在国产操作系统上运行 `0.8.2`版本开始,本项目发行的docker镜像支持`x86架构`,如果需要在`arm64架构`的CPU上部署,可按照[这个教程](docker-build.md)在本机编译`arm64的镜像`
## 1. 安装docker ## 1. 安装docker
@@ -80,7 +80,7 @@ xiaozhi-server
打开命令行工具,使用`终端``命令行`工具 进入到你的`xiaozhi-server`,执行以下命令 打开命令行工具,使用`终端``命令行`工具 进入到你的`xiaozhi-server`,执行以下命令
``` ```
docker-compose up -d docker compose up -d
``` ```
执行完后,再执行以下命令,查看日志信息。 执行完后,再执行以下命令,查看日志信息。
@@ -139,6 +139,9 @@ conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/
conda install libopus -y conda install libopus -y
conda install ffmpeg -y conda install ffmpeg -y
# 在 Linux 环境下进行部署时,如出现类似缺失 libiconv.so.2 动态库的报错 请通过以下命令进行安装
conda install libiconv -y
``` ```
请注意,以上命令,不是一股脑执行就成功的,你需要一步步执行,每一步执行完后,都检查一下输出的日志,查看是否成功。 请注意,以上命令,不是一股脑执行就成功的,你需要一步步执行,每一步执行完后,都检查一下输出的日志,查看是否成功。
@@ -238,7 +241,7 @@ LLM:
250427 13:04:20[0.3.11_SiFuChTTnofu][__main__]-INFO-OTA接口是 http://192.168.4.123:8003/xiaozhi/ota/ 250427 13:04:20[0.3.11_SiFuChTTnofu][__main__]-INFO-OTA接口是 http://192.168.4.123:8003/xiaozhi/ota/
250427 13:04:20[0.3.11_SiFuChTTnofu][__main__]-INFO-Websocket地址是 ws://192.168.4.123:8000/xiaozhi/v1/ 250427 13:04:20[0.3.11_SiFuChTTnofu][__main__]-INFO-Websocket地址是 ws://192.168.4.123:8000/xiaozhi/v1/
250427 13:04:20[0.3.11_SiFuChTTnofu][__main__]-INFO-=======上面的地址是websocket协议地址,请勿用浏览器访问======= 250427 13:04:20[0.3.11_SiFuChTTnofu][__main__]-INFO-=======上面的地址是websocket协议地址,请勿用浏览器访问=======
250427 13:04:20[0.3.11_SiFuChTTnofu][__main__]-INFO-如想测试websocket请用谷歌浏览器打开test目录下的test_page.html 250427 13:04:20[0.3.11_SiFuChTTnofu][__main__]-INFO-如想测试websocket请启动digital-human模块,打开浏览器交互测试
250427 13:04:20[0.3.11_SiFuChTTnofu][__main__]-INFO-======================================================= 250427 13:04:20[0.3.11_SiFuChTTnofu][__main__]-INFO-=======================================================
``` ```
@@ -267,7 +270,8 @@ LLM:
6、[我说话很慢,停顿时小智老是抢话](./FAQ.md)<br/> 6、[我说话很慢,停顿时小智老是抢话](./FAQ.md)<br/>
## 部署相关教程 ## 部署相关教程
1、[如何自动拉取本项目最新代码自动编译和启动](./dev-ops-integration.md)<br/> 1、[如何自动拉取本项目最新代码自动编译和启动](./dev-ops-integration.md)<br/>
2、[如何与Nginx集成](https://github.com/xinnan-tech/xiaozhi-esp32-server/issues/791)<br/> 2、[如何部署MQTT网关开启MQTT+UDP协议](./mqtt-gateway-integration.md)<br/>
3、[如何与Nginx集成](https://github.com/xinnan-tech/xiaozhi-esp32-server/issues/791)<br/>
## 拓展相关教程 ## 拓展相关教程
1、[如何开启手机号码注册智控台](./ali-sms-integration.md)<br/> 1、[如何开启手机号码注册智控台](./ali-sms-integration.md)<br/>
2、[如何集成HomeAssistant实现智能家居控制](./homeassistant-integration.md)<br/> 2、[如何集成HomeAssistant实现智能家居控制](./homeassistant-integration.md)<br/>
@@ -275,11 +279,13 @@ LLM:
4、[如何部署MCP接入点](./mcp-endpoint-enable.md)<br/> 4、[如何部署MCP接入点](./mcp-endpoint-enable.md)<br/>
5、[如何接入MCP接入点](./mcp-endpoint-integration.md)<br/> 5、[如何接入MCP接入点](./mcp-endpoint-integration.md)<br/>
6、[如何开启声纹识别](./voiceprint-integration.md)<br/> 6、[如何开启声纹识别](./voiceprint-integration.md)<br/>
10、[新闻插件源配置指南](./newsnow_plugin_config.md)<br/> 7、[新闻插件源配置指南](./newsnow_plugin_config.md)<br/>
8、[天气插件使用指南](./weather-integration.md)<br/>
## 语音克隆、本地语音部署相关教程 ## 语音克隆、本地语音部署相关教程
1、[如何部署集成index-tts本地语音](./index-stream-integration.md)<br/> 1、[如何在智控台克隆音色](./huoshan-streamTTS-voice-cloning.md)<br/>
2、[如何部署集成fish-speech本地语音](./fish-speech-integration.md)<br/> 2、[如何部署集成index-tts本地语音](./index-stream-integration.md)<br/>
3、[如何部署集成PaddleSpeech本地语音](./paddlespeech-deploy.md)<br/> 3、[如何部署集成fish-speech本地语音](./fish-speech-integration.md)<br/>
4、[如何部署集成PaddleSpeech本地语音](./paddlespeech-deploy.md)<br/>
## 性能测试教程 ## 性能测试教程
1、[各组件速度测试指南](./performance_tester.md)<br/> 1、[各组件速度测试指南](./performance_tester.md)<br/>
2、[定期公开测试结果](https://github.com/xinnan-tech/xiaozhi-performance-research)<br/> 2、[定期公开测试结果](https://github.com/xinnan-tech/xiaozhi-performance-research)<br/>
+13 -8
View File
@@ -1,7 +1,7 @@
# 部署架构图 # 部署架构图
![请参考-全模块安装架构图](../docs/images/deploy2.png) ![请参考-全模块安装架构图](../docs/images/deploy2.png)
# 方式一:Docker运行全模块 # 方式一:Docker运行全模块
docker镜像支持x86架构arm64架构的CPU,支持在国产操作系统上运行 `0.8.2`版本开始,本项目发行的docker镜像支持`x86架构`,如果需要在`arm64架构`的CPU上部署,可按照[这个教程](docker-build.md)在本机编译`arm64的镜像`
## 1. 安装docker ## 1. 安装docker
@@ -198,7 +198,7 @@ docker logs -f xiaozhi-esp32-server
``` ```
25-02-23 12:01:09[core.websocket_server] - INFO - Websocket地址是 ws://xxx.xx.xx.xx:8000/xiaozhi/v1/ 25-02-23 12:01:09[core.websocket_server] - INFO - Websocket地址是 ws://xxx.xx.xx.xx:8000/xiaozhi/v1/
25-02-23 12:01:09[core.websocket_server] - INFO - =======上面的地址是websocket协议地址,请勿用浏览器访问======= 25-02-23 12:01:09[core.websocket_server] - INFO - =======上面的地址是websocket协议地址,请勿用浏览器访问=======
25-02-23 12:01:09[core.websocket_server] - INFO - 如想测试websocket请用谷歌浏览器打开test目录下的test_page.html 25-02-23 12:01:09[core.websocket_server] - INFO - 如想测试websocket请启动digital-human模块,打开浏览器交互测试
25-02-23 12:01:09[core.websocket_server] - INFO - ======================================================= 25-02-23 12:01:09[core.websocket_server] - INFO - =======================================================
``` ```
@@ -355,6 +355,9 @@ conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/
conda install libopus -y conda install libopus -y
conda install ffmpeg -y conda install ffmpeg -y
# 在 Linux 环境下进行部署时,如出现类似缺失 libiconv.so.2 动态库的报错 请通过以下命令进行安装
conda install libiconv -y
``` ```
请注意,以上命令,不是一股脑执行就成功的,你需要一步步执行,每一步执行完后,都检查一下输出的日志,查看是否成功。 请注意,以上命令,不是一股脑执行就成功的,你需要一步步执行,每一步执行完后,都检查一下输出的日志,查看是否成功。
@@ -415,7 +418,7 @@ manager-api:
secret: 12345678-xxxx-xxxx-xxxx-123456789000 secret: 12345678-xxxx-xxxx-xxxx-123456789000
``` ```
## 5.运行项目 ## 9.运行项目
``` ```
# 确保在xiaozhi-server目录下执行 # 确保在xiaozhi-server目录下执行
@@ -428,7 +431,7 @@ python app.py
``` ```
25-02-23 12:01:09[core.websocket_server] - INFO - Server is running at ws://xxx.xx.xx.xx:8000/xiaozhi/v1/ 25-02-23 12:01:09[core.websocket_server] - INFO - Server is running at ws://xxx.xx.xx.xx:8000/xiaozhi/v1/
25-02-23 12:01:09[core.websocket_server] - INFO - =======上面的地址是websocket协议地址,请勿用浏览器访问======= 25-02-23 12:01:09[core.websocket_server] - INFO - =======上面的地址是websocket协议地址,请勿用浏览器访问=======
25-02-23 12:01:09[core.websocket_server] - INFO - 如想测试websocket请用谷歌浏览器打开test目录下的test_page.html 25-02-23 12:01:09[core.websocket_server] - INFO - 如想测试websocket请启动digital-human模块,打开浏览器交互测试
25-02-23 12:01:09[core.websocket_server] - INFO - ======================================================= 25-02-23 12:01:09[core.websocket_server] - INFO - =======================================================
``` ```
@@ -468,7 +471,8 @@ ws://你电脑局域网的ip:8000/xiaozhi/v1/
6、[我说话很慢,停顿时小智老是抢话](./FAQ.md)<br/> 6、[我说话很慢,停顿时小智老是抢话](./FAQ.md)<br/>
## 部署相关教程 ## 部署相关教程
1、[如何自动拉取本项目最新代码自动编译和启动](./dev-ops-integration.md)<br/> 1、[如何自动拉取本项目最新代码自动编译和启动](./dev-ops-integration.md)<br/>
2、[如何与Nginx集成](https://github.com/xinnan-tech/xiaozhi-esp32-server/issues/791)<br/> 2、[如何部署MQTT网关开启MQTT+UDP协议](./mqtt-gateway-integration.md)<br/>
3、[如何与Nginx集成](https://github.com/xinnan-tech/xiaozhi-esp32-server/issues/791)<br/>
## 拓展相关教程 ## 拓展相关教程
1、[如何开启手机号码注册智控台](./ali-sms-integration.md)<br/> 1、[如何开启手机号码注册智控台](./ali-sms-integration.md)<br/>
2、[如何集成HomeAssistant实现智能家居控制](./homeassistant-integration.md)<br/> 2、[如何集成HomeAssistant实现智能家居控制](./homeassistant-integration.md)<br/>
@@ -479,9 +483,10 @@ ws://你电脑局域网的ip:8000/xiaozhi/v1/
7、[新闻插件源配置指南](./newsnow_plugin_config.md)<br/> 7、[新闻插件源配置指南](./newsnow_plugin_config.md)<br/>
8、[天气插件使用指南](./weather-integration.md)<br/> 8、[天气插件使用指南](./weather-integration.md)<br/>
## 语音克隆、本地语音部署相关教程 ## 语音克隆、本地语音部署相关教程
1、[如何部署集成index-tts本地语音](./index-stream-integration.md)<br/> 1、[如何在智控台克隆音色](./huoshan-streamTTS-voice-cloning.md)<br/>
2、[如何部署集成fish-speech本地语音](./fish-speech-integration.md)<br/> 2、[如何部署集成index-tts本地语音](./index-stream-integration.md)<br/>
3、[如何部署集成PaddleSpeech本地语音](./paddlespeech-deploy.md)<br/> 3、[如何部署集成fish-speech本地语音](./fish-speech-integration.md)<br/>
4、[如何部署集成PaddleSpeech本地语音](./paddlespeech-deploy.md)<br/>
## 性能测试教程 ## 性能测试教程
1、[各组件速度测试指南](./performance_tester.md)<br/> 1、[各组件速度测试指南](./performance_tester.md)<br/>
2、[定期公开测试结果](https://github.com/xinnan-tech/xiaozhi-performance-research)<br/> 2、[定期公开测试结果](https://github.com/xinnan-tech/xiaozhi-performance-research)<br/>
+29 -15
View File
@@ -38,10 +38,10 @@ conda install conda-forge::ffmpeg
| 模块名称 | 入门全免费设置 | 流式配置 | | 模块名称 | 入门全免费设置 | 流式配置 |
|:---:|:---:|:---:| |:---:|:---:|:---:|
| ASR(语音识别) | FunASR(本地) | 👍FunASR(本地GPU模式) | | ASR(语音识别) | FunASR(本地) | 👍XunfeiStreamASR(讯飞流式) |
| LLM(大模型) | ChatGLMLLM(智谱glm-4-flash) | 👍AliLLM(qwen3-235b-a22b-instruct-2507) 或 👍DoubaoLLM(doubao-1-5-pro-32k-250115) | | LLM(大模型) | glm-4-flash(智谱) | 👍qwen-flash(阿里百炼) |
| VLLM(视觉大模型) | ChatGLMVLLM(智谱glm-4v-flash) | 👍QwenVLVLLM(千问qwen2.5-vl-3b-instructh) | | VLLM(视觉大模型) | glm-4v-flash(智谱) | 👍qwen3.5-flash(阿里百炼) |
| TTS(语音合成) | ✅LinkeraiTTS(灵犀流式) | 👍HuoshanDoubleStreamTTS(火山流式语音合成) 或 👍AliyunStreamTTS(阿里云流式语音合成) | | TTS(语音合成) | EdgeTTS(微软) | 👍HuoshanDoubleStreamTTS(火山流式) |
| Intent(意图识别) | function_call(函数调用) | function_call(函数调用) | | Intent(意图识别) | function_call(函数调用) | function_call(函数调用) |
| Memory(记忆功能) | mem_local_short(本地短期记忆) | mem_local_short(本地短期记忆) | | Memory(记忆功能) | mem_local_short(本地短期记忆) | mem_local_short(本地短期记忆) |
@@ -62,29 +62,43 @@ VAD:
### 7、部署相关教程 ### 7、部署相关教程
1、[如何进行最简化部署](./Deployment.md)<br/> 1、[如何进行最简化部署](./Deployment.md)<br/>
2、[如何进行全模块部署](./Deployment_all.md)<br/> 2、[如何进行全模块部署](./Deployment_all.md)<br/>
2、[如何部署MQTT网关开启MQTT+UDP协议](./mqtt-gateway-integration.md)<br/> 3、[如何部署MQTT网关开启MQTT+UDP协议](./mqtt-gateway-integration.md)<br/>
3、[如何自动拉取本项目最新代码自动编译和启动](./dev-ops-integration.md)<br/> 4、[如何自动拉取本项目最新代码自动编译和启动](./dev-ops-integration.md)<br/>
4、[如何与Nginx集成](https://github.com/xinnan-tech/xiaozhi-esp32-server/issues/791)<br/> 5、[如何与Nginx集成](https://github.com/xinnan-tech/xiaozhi-esp32-server/issues/791)<br/>
6、[修改代码后怎么编译自己的Docker镜像](./docker-build.md)<br/>
### 8、编译固件相关教程 ### 8、编译固件相关教程
1、[如何自己编译小智固件](./firmware-build.md)<br/> 1、[如何自己编译小智固件](./firmware-build.md)<br/>
2、[如何基于虾哥编译好的固件修改OTA地址](./firmware-setting.md)<br/> 2、[如何基于虾哥编译好的固件修改OTA地址](./firmware-setting.md)<br/>
3、[单模块部署如何配置固件OTA自动升级](./ota-upgrade-guide.md)<br/>
### 8、拓展相关教程 ### 9、拓展相关教程
1、[如何开启手机号码注册智控台](./ali-sms-integration.md)<br/> 1、[如何开启手机号码注册智控台](./ali-sms-integration.md)<br/>
2、[如何集成HomeAssistant实现智能家居控制](./homeassistant-integration.md)<br/> 2、[如何集成HomeAssistant实现智能家居控制](./homeassistant-integration.md)<br/>
3、[如何开启视觉模型实现拍照识物](./mcp-vision-integration.md)<br/> 3、[如何开启视觉模型实现拍照识物](./mcp-vision-integration.md)<br/>
4、[如何部署MCP接入点](./mcp-endpoint-enable.md)<br/> 4、[如何部署MCP接入点](./mcp-endpoint-enable.md)<br/>
5、[如何接入MCP接入点](./mcp-endpoint-integration.md)<br/> 5、[如何接入MCP接入点](./mcp-endpoint-integration.md)<br/>
6、[如何开启声纹识别](./voiceprint-integration.md)<br/> 6、[MCP方法如何获取设备信息](./mcp-get-device-info.md)<br/>
10、[新闻插件源配置指南](./newsnow_plugin_config.md)<br/> 7、[如何开启声纹识别](./voiceprint-integration.md)<br/>
8、[新闻插件源配置指南](./newsnow_plugin_config.md)<br/>
9、[知识库ragflow集成指南](./ragflow-integration.md)<br/>
10、[如何部署上下文源](./context-provider-integration.md)<br/>
11、[如何集成PowerMem智能记忆](./powermem-integration.md)<br/>
12、[如何配置天气插件查询天气](./weather-integration.md)<br/>
13、[如何开启设备呼叫插件](./device-call-guide.md)<br/>
14、[如何开启联网搜索功能](./web-search-integration.md)<br/>
### 9、语音克隆、本地语音部署相关教程 ### 10、数字人相关教程
1、[如何部署集成index-tts本地语音](./index-stream-integration.md)<br/> 1、[数字人digital-human启动方法](./digital-human-wakeword.md)<br/>
2、[如何部署集成fish-speech本地语音](./fish-speech-integration.md)<br/> 2、[如何在N100迷你主机上部署数字人digital-human](./all-in-one-digital-human-setup.md)<br/>
3、[如何部署集成PaddleSpeech本地语音](./paddlespeech-deploy.md)<br/>
### 10、性能测试教程 ### 11、语音克隆、本地语音部署相关教程
1、[如何在智控台克隆音色](./huoshan-streamTTS-voice-cloning.md)<br/>
2、[如何部署集成index-tts本地语音](./index-stream-integration.md)<br/>
3、[如何部署集成fish-speech本地语音](./fish-speech-integration.md)<br/>
4、[如何部署集成PaddleSpeech本地语音](./paddlespeech-deploy.md)<br/>
### 12、性能测试教程
1、[各组件速度测试指南](./performance_tester.md)<br/> 1、[各组件速度测试指南](./performance_tester.md)<br/>
2、[定期公开测试结果](https://github.com/xinnan-tech/xiaozhi-performance-research)<br/> 2、[定期公开测试结果](https://github.com/xinnan-tech/xiaozhi-performance-research)<br/>
+442
View File
@@ -0,0 +1,442 @@
# 一体机数字人配置指南
本项目用于在 x86 架构设备(如迷你主机、工控机、普通电脑等)上部署一套完整的数字人展示系统,实现以下功能:
- 开机自动进入 Kiosk 全屏浏览器,展示指定网页
- 后台运行唤醒词检测服务,支持语音交互
> **说明**:本文档以 **Intel N100 迷你主机(天虹 QN10-100B4)** 为例进行部署演示,其他 x86 设备可参考调整(注意网络配置和声卡设备差异)。
## 适用环境
| 项目 | 说明 |
|------|------|
| 示例硬件 | 天虹 QN10-100B4Intel N100 |
| 操作系统 | Ubuntu 24.04 LTS (Noble Numbat) |
| 示例用户 | xz(请根据实际情况替换) |
| 网络 | Wi-Fi 连接,固定 IP(可按需改为有线) |
## 部署流程
1. 系统初始化(换源、连网)
2. 安装图形组件与 Kiosk 浏览器
3. 配置自动登录与图形界面
4. 部署唤醒词服务(Python 环境 + 麦克风)
5. 优化开机速度与隐藏启动信息
---
### 系统初始化(换源、连网)
```
sudo cp /etc/apt/sources.list /etc/apt/sources.list.bak
sudo tee /etc/apt/sources.list > /dev/null <<EOF
deb http://mirrors.aliyun.com/ubuntu/ noble main restricted universe multiverse
# deb-src http://mirrors.aliyun.com/ubuntu/ noble main restricted universe multiverse
deb http://mirrors.aliyun.com/ubuntu/ noble-security main restricted universe multiverse
# deb-src http://mirrors.aliyun.com/ubuntu/ noble-security main restricted universe multiverse
deb http://mirrors.aliyun.com/ubuntu/ noble-updates main restricted universe multiverse
# deb-src http://mirrors.aliyun.com/ubuntu/ noble-updates main restricted universe multiverse
deb http://mirrors.aliyun.com/ubuntu/ noble-proposed main restricted universe multiverse
# deb-src http://mirrors.aliyun.com/ubuntu/ noble-proposed main restricted universe multiverse
deb http://mirrors.aliyun.com/ubuntu/ noble-backports main restricted universe multiverse
# deb-src http://mirrors.aliyun.com/ubuntu/ noble-backports main restricted universe multiverse
EOF
echo 'Acquire::ForceIPv4 "true";' | sudo tee /etc/apt/apt.conf.d/99force-ipv4
```
安装网络管理工具(若已存在可忽略)
Bash
```
sudo apt update
sudo apt install network-manager -y
sudo systemctl start NetworkManager
sudo systemctl enable NetworkManager
```
设置wifi密码,固定ip
> **提醒**:以下命令中的 Wi-Fi 名称、密码、IP 地址均为示例,请务必替换为你自己的实际信息。
Bash
```
sudo nmcli device wifi connect "MERCURY_1812" password "12345678"
sudo nmcli connection modify "MERCURY_1812" ipv4.addresses "192.168.0.86/24" ipv4.gateway "192.168.0.1" ipv4.dns "8.8.8.8,114.114.114.114" ipv4.method "manual"
sudo nmcli connection up "MERCURY_1812"
```
### 第一步:安装核心图形组件与浏览器
这里我们坚持“极简主义”,坚决不装多余的桌面环境(如 GNOME/KDE),只装底层驱动、最轻量的窗口管理器(Openbox)、隐藏鼠标工具以及 Chromium 浏览器。
Bash
```
sudo timedatectl set-timezone Asia/Shanghai
sudo apt install net-tools vim fonts-wqy-microhei fonts-wqy-zenhei alsa-utils pulseaudio -y
sudo apt install --no-install-recommends xserver-xorg x11-xserver-utils xinit openbox unclutter -y
wget https://dl.google.com/linux/direct/google-chrome-stable_current_amd64.deb
sudo apt install ./google-chrome-stable_current_amd64.deb -y
rm google-chrome-stable_current_amd64.deb
sudo apt purge snapd -y
```
### 第二步:配置 TTY1 开机免密自动登录
为了避免手动输入账号密码的尴尬,我们通过修改 systemd 服务,让系统一开机就自动以 `xz` 的身份登录。这里使用一键写入命令,彻底避开由于 `nano``vi` 操作不当导致的保存失败问题。
**1. 创建配置目录:**
Bash
```
sudo mkdir -p /etc/systemd/system/getty@tty1.service.d/
```
**2. 写入自动登录规则:**
Bash
```
echo -e "[Service]\nExecStart=\nExecStart=-/sbin/agetty --autologin xz --noclear %I \$TERM" | sudo tee /etc/systemd/system/getty@tty1.service.d/override.conf
```
**3. 重载服务并设置默认启动目标:**
Bash
```
sudo systemctl daemon-reload
sudo systemctl set-default multi-user.target
```
### 第三步:配置登录后自动启动图形界面
系统自动登录后默认停留在黑底白字的命令行,我们需要配置脚本,让它一旦登录立马启动 X11 图形环境。
**1. 触发 `startx` 的启动逻辑:**
直接将触发代码追加写入你的个人环境配置文件中:
Bash
```
cat << 'EOF' >> ~/.bash_profile
if [ -z "$DISPLAY" ] && [ "$(fgconsole)" -eq 1 ]; then
exec startx
fi
EOF
```
**2. 告诉 `startx` 去启动 Openbox**
Bash
```
echo "exec openbox-session" > ~/.xinitrc
```
### 第四步:配置“铜墙铁壁”的 Openbox 与浏览器
这是最核心的一步:关闭屏幕休眠、隐藏鼠标、全屏锁定浏览器,并写一个“死循环”保证浏览器被意外关闭后也能瞬间复活。
**1. 创建 Openbox 配置目录:**
Bash
```
mkdir -p ~/.config/openbox
```
**2. 写入自启动脚本 (`autostart`)**
复制以下整段代码并回车(这会自动把所有保护规则写进文件里):
Bash
```
cat << 'EOF' > ~/.config/openbox/autostart
# 关闭屏幕保护
xset -dpms
xset s noblank
xset s off
# 隐藏鼠标
unclutter -idle 0.1 -root &
# 死循环启动 Chromium(崩溃或被关也能秒重启)
while true; do
google-chrome \
--kiosk \
--no-first-run \
--no-default-browser-check \
--disable-infobars \
--disable-session-crashed-bubble \
--disable-translate \
--disable-external-intent-requests \
--autoplay-policy=no-user-gesture-required \
--use-fake-ui-for-media-stream \
"https://www.douyin.com"
sleep 2
done &
EOF
```
**3. 屏蔽 `Alt+F4` 退出快捷键:**
为了防止别人插上键盘强行关掉窗口,我们把 Openbox 默认的系统快捷键干掉。
Bash
```
cp /etc/xdg/openbox/rc.xml ~/.config/openbox/
sed -i '/<keybind key="A-F4">/,/<\/keybind>/d' ~/.config/openbox/rc.xml
```
### 第五步:重启验收成果
如果拔掉网线或不需要等待所有网络上线,可禁用网络等待服务,避免开机卡顿
Bash
```
sudo systemctl mask systemd-networkd-wait-online.service
sudo systemctl mask NetworkManager-wait-online.service
```
隐藏开机信息(GRUB
Bash
```
sudo sed -i 's/GRUB_CMDLINE_LINUX_DEFAULT=.*/GRUB_CMDLINE_LINUX_DEFAULT="quiet loglevel=3 systemd.show_status=false vt.global_cursor_default=0"/g' /etc/default/grub
echo 'GRUB_TIMEOUT_STYLE="hidden"' | sudo tee -a /etc/default/grub
echo 'GRUB_RECORDFAIL_TIMEOUT=0' | sudo tee -a /etc/default/grub
sudo update-grub
```
声音设置成100%,然后 重启:
Bash
```
amixer -q sset Master 100% unmute
sudo reboot
```
### 部署唤醒词服务
在一体机上部署唤醒词检测服务,需要安装 Python 环境、上传项目文件、配置 Camera 麦克风和开机自启。
#### 1. 安装 Miniconda
```bash
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
bash Miniconda3-latest-Linux-x86_64.sh -b -p $HOME/miniconda3
~/miniconda3/bin/conda init bash
source ~/.bashrc
rm Miniconda3-latest-Linux-x86_64.sh
```
确保登录时自动进入conda环境
```bash
if ! grep -q '.bashrc' ~/.bash_profile; then
cat << 'EOF' >> ~/.bash_profile
if [ -f ~/.bashrc ]; then
. ~/.bashrc
fi
EOF
fi
```
#### 2. 创建 Python 虚拟环境
```bash
conda create -n test python=3.10 -y
conda activate test
```
若出现 Terms of Service have not been accepted 错误,执行:
```bash
conda tos accept --override-channels --channel https://repo.anaconda.com/pkgs/main
conda tos accept --override-channels --channel https://repo.anaconda.com/pkgs/r
```
#### 3. 上传项目文件
将开发机上的 `main/digital-human/` 整个目录上传到一体机的 `~/digital-human/` 目录:
```bash
# 在开发机上执行(将 <一体机IP> 替换为实际 IP
scp -r main/digital-human/ xz@<一体机IP>:~/digital-human/
```
#### 4. 安装系统依赖
唤醒词服务需要音频采集库和 ALSA PulseAudio 插件:
```bash
sudo apt install libportaudio2 portaudio19-dev libasound2-plugins -y
```
#### 5. 安装 Python 依赖
```bash
cd ~/digital-human/wakeword_runtime
pip install numpy
pip install -r requirements.txt
```
#### 6. 下载唤醒词模型
模型文件不包含在项目中,需要单独下载配置,详见 [docs/digital-human-wakeword.md](digital-human-wakeword.md) 中的“模型下载”章节。
#### 7. 修改 Openbox 自启动脚本
需要在 autostart 中加上 PulseAudio 和 Camera 麦克风配置,并将 Chrome 地址改为测试页面。
先确认 Camera 麦克风在 PulseAudio 中的设备名:
```bash
pulseaudio --start
pactl list sources short
```
找到包含 `USB_Camera` 的那一行,记下完整名称,例如:
```
alsa_input.usb-SN0002_2K_USB_Camera_46435000_P030D00_SN0002-02.mono-fallback
```
然后用完整内容覆盖 autostart(将 `TARGET_MIC` 替换为你的实际设备名):
```bash
cat << 'EOF' > ~/.config/openbox/autostart
# 1. 启动声音服务并稍作等待
pulseaudio --start
sleep 1
# 2. 锁定 Camera 的麦克风(请替换为你的实际设备名)
TARGET_MIC="alsa_input.usb-SN0002_2K_USB_Camera_46435000_P030D00_SN0002-02.mono-fallback"
# 3. 设为系统默认麦克风
pactl set-default-source "$TARGET_MIC"
# 4. 解除静音
pactl set-source-mute "$TARGET_MIC" 0
# 5. 音量拉到 100%
pactl set-source-volume "$TARGET_MIC" 100%
# --- 极简桌面与浏览器环境配置 ---
# 关闭屏幕保护
xset -dpms
xset s noblank
xset s off
# 隐藏鼠标
unclutter -idle 0.1 -root &
# 死循环启动浏览器(崩溃或被关也能秒重启)
while true; do
google-chrome \
--kiosk \
--no-first-run \
--no-default-browser-check \
--disable-infobars \
--disable-session-crashed-bubble \
--disable-translate \
--disable-external-intent-requests \
--autoplay-policy=no-user-gesture-required \
--use-fake-ui-for-media-stream \
"http://127.0.0.1:8006/index.html"
sleep 2
done &
EOF
```
#### 8. 配置唤醒词服务开机自启
创建 systemd 服务文件,让唤醒词服务开机自动运行。
先确认当前用户的 UID
```bash
id -u $(whoami)
```
然后用查到的 UID 替换下面 `1000`(通常第一个用户就是 1000):
```bash
sudo tee /etc/systemd/system/digital-human.service << 'EOF'
[Unit]
Description=Digital Human Runtime
After=network.target sound.target
[Service]
Type=simple
User=xz
Environment=XDG_RUNTIME_DIR=/run/user/1000
Environment=PULSE_SERVER=unix:/run/user/1000/pulse/native
WorkingDirectory=/home/xz/digital-human
ExecStartPre=/bin/sleep 10
ExecStart=/home/xz/miniconda3/envs/test/bin/python start.py
Restart=on-failure
RestartSec=10
[Install]
WantedBy=multi-user.target
EOF
```
> **重要说明**
> - `User=xz` — 替换为你的实际用户名
> - `/run/user/1000` — 替换为你实际的 UID
> - `WorkingDirectory` 和 `ExecStart` 中的路径 — 替换为你的实际部署路径
> - `Environment` 中的 PulseAudio 环境变量**必须保留**,否则唤醒词服务和浏览器无法同时使用 Camera 麦克风
启用并启动服务:
```bash
sudo systemctl daemon-reload
sudo systemctl enable digital-human
sudo systemctl start digital-human
```
#### 9. 常用服务管理命令
```bash
sudo systemctl start digital-human # 立即启动
sudo systemctl stop digital-human # 停止
sudo systemctl restart digital-human # 重启
sudo systemctl status digital-human # 查看状态
journalctl -u digital-human -f # 查看实时日志
```
+224
View File
@@ -0,0 +1,224 @@
# 上下文源使用教程
## 概述
`上下文源`,就是为小智系统提示词的上下文添加【数据源】。
`上下文源` 在小智在唤醒那一刻,获取外部系统的数据,并将其动态注入到大模型的系统提示词(System Prompt)中。
让其做到唤醒时感知世界某个事物的状态。
它和MCP、记忆有本质的区别:`上下文源`是强制让小智感知世界的数据;`记忆(Mem)`是让他知道之前聊了什么内容;`MCP(functionc all)`是当需要调用某项能力/知识的时候使用调用。
通过这个功能,在小智唤醒的一刹那,“感知”到:
- 人体健康传感器状态(体温、血压、血氧状态等)
- 业务系统的实时数据(服务器负载、待办数据、股票信息等)
- 任何可以通过 HTTP API 获取的文本信息
**注意**:该功能只是方便小智在唤醒的时候感知事物的状态,而如果想要小智唤醒后实时获取事物的状态,建议在此功能上再结合MCP工具的调用。
## 工作原理
1. **配置源**:用户配置一个或多个 HTTP API 地址。
2. **触发请求**:当系统构建 Prompt 时,如果发现模板中包含 `{{ dynamic_context }}` 占位符,会请求所有配置的 API。
3. **自动注入**:系统会自动将 API 返回的数据格式化为 Markdown 列表,替换 `{{ dynamic_context }}` 占位符。
## 接口规范
为了让小智正确解析数据,您的 API 需要满足以下规范:
- **请求方式**`GET`
- **请求头**:系统会自动添加 `device-id` 字段到 Request Header。
- **响应格式**:必须返回 JSON 格式,且包含 `code``data` 字段。
### 响应示例
**情况 1:返回键值对**
```json
{
"code": 0,
"msg": "success",
"data": {
"客厅温度": "26℃",
"客厅湿度": "45%",
"大门状态": "已关闭"
}
}
```
*注入效果:*
```markdown
<context>
- **客厅温度:** 26℃
- **客厅湿度:** 45%
- **大门状态:** 已关闭
</context>
```
**情况 2:返回列表**
```json
{
"code": 0,
"data": [
"您有10个待办事项",
"当前汽车的行驶速度是100km每小时"
]
}
```
*注入效果:*
```markdown
<context>
- 您有10个待办事项
- 当前汽车的行驶速度是100km每小时
</context>
```
## 配置指南
### 方式 1:智控台配置(全模块部署)
1. 登录智控台,进入**角色配置**页面。
2. 找到**上下文源**配置项(点击“编辑源”按钮)。
3. 点击**添加**,输入您的 API 地址。
4. 如果 API 需要鉴权,可以在**请求头**部分添加 `Authorization` 或其他 Header。
5. 保存配置。
### 方式 2:配置文件配置(单模块部署)
编辑 `xiaozhi-server/data/.config.yaml` 文件,添加 `context_providers` 配置段:
```yaml
# 上下文源配置
context_providers:
- url: "http://api.example.com/data"
headers:
Authorization: "Bearer your-token"
- url: "http://another-api.com/data"
```
## 启用功能
默认情况下,系统的提示词模板文件(`data/.agent-base-prompt.txt`)中已经预置了 `{{ dynamic_context }}` 占位符,您无需手动添加。
**示例:**
```markdown
<context>
【重要!以下信息已实时提供,无需调用工具查询,请直接使用:】
- **设备ID** {{device_id}}
- **当前时间:** {{current_time}}
...
{{ dynamic_context }}
</context>
```
**注意**:如果您不需要使用此功能,可以选择**不配置任何上下文源**,也可以从提示词模板文件中**删除** `{{ dynamic_context }}` 占位符。
## 附录:Mock 测试服务示例
为了方便您测试和开发,我们提供了一个简单的 Python Mock Server 脚本。您可以运行此脚本在本地模拟 API 接口。
**mock_api_server.py**
```python
import http.server
import socketserver
import json
from urllib.parse import urlparse, parse_qs
# 设置端口号
PORT = 8081
class MockRequestHandler(http.server.SimpleHTTPRequestHandler):
def do_GET(self):
# 解析路径和参数
parsed_path = urlparse(self.path)
path = parsed_path.path
query = parse_qs(parsed_path.query)
response_data = {}
status_code = 200
print(f"收到请求: {path}, 参数: {query}")
# Case 1: 模拟健康数据 (返回字典 Dict)
# 路径参数风格: /health
# device_id 从 Header 获取
if path == "/health":
device_id = self.headers.get("device-id", "unknown_device")
print(f"device_id: {device_id}")
response_data = {
"code": 0,
"msg": "success",
"data": {
"测试设备ID": device_id,
"心率": "80 bpm",
"血压": "120/80 mmHg",
"状态": "良好"
}
}
# Case 2: 模拟新闻列表 (返回列表 List)
# 无参数: /news/list
elif path == "/news/list":
response_data = {
"code": 0,
"msg": "success",
"data": [
"今日头条:Python 3.14 发布",
"科技新闻:AI 助手改变生活",
"本地新闻:明日有大雨,记得带伞"
]
}
# Case 3: 模拟天气简报 (返回字符串 String)
# 无参数: /weather/simple
elif path == "/weather/simple":
response_data = {
"code": 0,
"msg": "success",
"data": "今日晴转多云,气温 20-25 度,空气质量优,适合出行。"
}
# Case 4: 模拟设备详情 (Query参数风格)
# 参数风格: /device/info
# device_id 从 Header 获取
elif path == "/device/info":
device_id = self.headers.get("device-id", "unknown_device")
response_data = {
"code": 0,
"msg": "success",
"data": {
"查询方式": "Header参数",
"设备ID": device_id,
"电量": "85%",
"固件": "v2.0.1"
}
}
# Case 5: 404 Not Found
else:
status_code = 404
response_data = {"error": "接口不存在"}
# 发送响应
self.send_response(status_code)
self.send_header('Content-type', 'application/json; charset=utf-8')
self.end_headers()
self.wfile.write(json.dumps(response_data, ensure_ascii=False).encode('utf-8'))
# 启动服务
# 允许地址重用,防止快速重启报错
socketserver.TCPServer.allow_reuse_address = True
with socketserver.TCPServer(("", PORT), MockRequestHandler) as httpd:
print(f"==================================================")
print(f"Mock API Server 已启动: http://localhost:{PORT}")
print(f"可用接口列表:")
print(f"1. [字典] http://localhost:{PORT}/health")
print(f"2. [列表] http://localhost:{PORT}/news/list")
print(f"3. [文本] http://localhost:{PORT}/weather/simple")
print(f"4. [参数] http://localhost:{PORT}/device/info")
print(f"==================================================")
try:
httpd.serve_forever()
except KeyboardInterrupt:
print("\n服务已停止")
```
+148
View File
@@ -0,0 +1,148 @@
# 设备间相互呼叫插件使用指南
## 概述
设备呼叫功能允许两个已配置设备之间通过语音/数据通道进行双向通信。设备A呼叫设备B时,系统通过以下流程实现:
```
设备A → 授权校验 → MQTT网关 → 设备B远程唤醒 → 建立连接 → 通话建立
```
## 使用这个功能的前提条件
1. 你必须要有至少两个设备,每个设备型号必须是`ESP32-S3`,因为只有`ESP32-S3`才支持远程唤醒功能。
2. 你的设备必须要有`两个麦克风`。但是如果你的设备只有`单个麦克风`,你只是想体验一下这个功能,也是可以的,但是会有强烈的卡顿感。
3. 你必须使用[全模块部署](Deployment_all.md)本项目,因为你需要`智控台`来管理设备的权限和通信。
4. 你必须安装并配置好`2026年5月27日`以后的[MQTT网关服务](mqtt-gateway-integration.md),如果你已经部署了MQTT网关服务,请确认代码的版本是`2026年5月27日`之后的版本。
以上是使用这个功能的硬性条件,接下来会详细介绍。
## 配置步骤
### 第一步:开启通讯录功能
1. 确认你的智控台版本是`0.9.4`或以上版本。
2. 登录智控台后台
3. 进入 **系统功能配置**
4. 在左侧功能列表中勾选 **通讯录**
5. 点击 **保存配置** 确认
### 第二步:配置设备间呼叫权限
1. 在智控台顶部菜单点击 **通讯录**
2. 在左侧智能体下,设备列表中选择你的设备A(支持按 MAC地址 或 备注名 搜索)
3. 在右侧详情面板中,找到目标设备B的称呼设置,例如 **"小王"**
4. 勾选设备B的 **呼叫权限** 复选框
5. 点击 **保存**
**双向授权说明:** 如需设备A和设备B互相通信,必须在两侧智控台分别配置对方权限。例如:
- 在设备A的配置中勾选设备B → 设备A可与设备B通信
- 在设备B的配置中勾选设备A → 设备B可与设备A通信
### 第三步:在智能体配置添加呼叫工具
1. 在智控台顶部菜单点击 **智能体管理**
2. 在刚刚配置设备联系人的相关智能体中点击 **编辑角色**
3. 在右侧详情面板中,点击 **编辑功能**
4. 勾选 **设备呼叫设备** 工具
5. 点击 **保存配置** 确认
6. 在外侧再次点击 **保存配置** ,随即重启设备
### 第四步:固件端添加远程唤醒工具
1. 在[xiaozhi-esp32](https://github.com/78/xiaozhi-esp32) 代码的基础上增加远程唤醒工具MCP,版本支持为2.1.0至2.2.62026年5月29日的版本)
2. 在application.h文件中添加远程唤醒函数声明
```cpp
void RemoteWakeup(const std::string& reason);
```
3. 在application.cc文件中添加远程唤醒函数
```cpp
void Application::RemoteWakeup(const std::string& reason){
if (!protocol_) {
return;
}
auto state = GetDeviceState();
if (state == kDeviceStateIdle) {
audio_service_.EncodeWakeWord();
if (!protocol_->IsAudioChannelOpened()) {
SetDeviceState(kDeviceStateConnecting);
if (!protocol_->OpenAudioChannel()) {
audio_service_.EnableWakeWordDetection(true);
return;
}
}
std::string wake_word = reason;
#if CONFIG_USE_AFE_WAKE_WORD || CONFIG_USE_CUSTOM_WAKE_WORD
// Encode and send the wake word data to the server
while (auto packet = audio_service_.PopWakeWordPacket()) {
protocol_->SendAudio(std::move(packet));
}
// Set the chat state to wake word detected
protocol_->SendWakeWordDetected(wake_word);
SetListeningMode(aec_mode_ == kAecOff ? kListeningModeAutoStop : kListeningModeRealtime);
#else
// Set flag to play popup sound after state changes to listening
// (PlaySound here would be cleared by ResetDecoder in EnableVoiceProcessing)
play_popup_on_listening_ = true;
SetListeningMode(aec_mode_ == kAecOff ? kListeningModeAutoStop : kListeningModeRealtime);
#endif
} else if (state == kDeviceStateSpeaking) {
AbortSpeaking(kAbortReasonWakeWordDetected);
SetDeviceState(kDeviceStateIdle);
} else if (state == kDeviceStateActivating) {
SetDeviceState(kDeviceStateIdle);
}
}
```
4. 在mcp_server.cc文件中添加远程唤醒工具
```cpp
AddUserOnlyTool("self.remote_wakeup", "Remote wakeup function with configurable parameters",
PropertyList({
Property("reason", kPropertyTypeString, "Wakeup reason"),
}),
[this](const PropertyList& properties) -> ReturnValue {
std::string reason = properties["reason"].value<std::string>();
ESP_LOGI(TAG, "Wakeup reason=%s", reason.c_str());
auto& app = Application::GetInstance();
app.RemoteWakeup(reason);
return true;
```
5. 按照 [固件编译烧录指南](firmware-build.md) 完成固件烧录
6. 无论你的设备是单麦还是双麦,请在编译环节,勾选开启AEC功能!
7. 无论你的设备是单麦还是双麦,请在编译环节,勾选开启AEC功能!
8. 无论你的设备是单麦还是双麦,请在编译环节,勾选开启AEC功能!
### 第五步:配置MQTT网关服务
1. 部署MQTT网关服务,参考 [MQTT网关集成文档](mqtt-gateway-integration.md)
2. 如果已经部署请确认代码的版本是2026年5月27日的版本
## 呼叫流程说明
准备两个设备,在智控台上面配置好通讯权限和在智能体中添加呼叫工具之后,在其中一个小智对话那里对他说:”呼叫XXX“,观察设备B是否响应。
## 常见问题
### Q: 设备B没有响应呼叫?
- 检查设备B是否在线(智控台设备状态)
- 确认设备B的固件已正确集成远程唤醒工具
- 检查MQTT网关连接是否正常
- 验证双向权限配置是否完整
### Q: 提示"无呼叫权限"
- 在智控台确认设备A已勾选设备B的呼叫权限
- 确认配置已保存(非仅修改未保存)
### Q: 如何确认通讯录功能已开启?
- 智控台顶部菜单如显示"通讯录"入口,则表示已开启
### Q: 我叫他呼叫"张山",但是他老是识别成"张三",怎么办?
- 可以查阅你使用的asr服务的文档,确认是否支持热词识别。
- 如果你用的是`FunASRServer`,可以在容器里的`热词文件`里添加"张山",然后重启容器。
- 如果你用的是`火山引擎`的服务,可以在`火山引擎的控制台`里添加`热词文件`,然后回到智控台的`模型配置页面`,把`热词文件名称`配置在`火山引擎的tts`上去。
+220
View File
@@ -0,0 +1,220 @@
# 数字人digital-human启动方法
## 概述
测试页面集成了基于 **Sherpa-ONNX** 的高精度语音唤醒功能,支持自定义唤醒词和实时检测。使用轻量级关键词检测模型,提供毫秒级响应速度。
## 唤醒词模型
### 模型下载(必需)
**重要说明**: 项目不包含模型文件,需要提前下载配置。
### 官方模型下载地址
- **官方模型列表**: <https://csukuangfj.github.io/sherpa/onnx/kws/pretrained_models/index.html>
- **推荐模型**: `sherpa-onnx-kws-zipformer-wenetspeech-3.3M-2024-01-01`
### 下载和配置步骤
#### 1. 下载模型包
```bash
# 方法1:直接下载(推荐)
cd main/digital-human/wakeword_runtime/
wget https://github.com/k2-fsa/sherpa-onnx/releases/download/kws-models/sherpa-onnx-kws-zipformer-wenetspeech-3.3M-2024-01-01.tar.bz2
# 解压
tar xvf sherpa-onnx-kws-zipformer-wenetspeech-3.3M-2024-01-01.tar.bz2
# 方法2:使用ModelScope
pip install modelscope
python -c "
from modelscope import snapshot_download
snapshot_download('pkufool/sherpa-onnx-kws-zipformer-wenetspeech-3.3M-2024-01-01', cache_dir='./models')
"
```
#### 2. 配置模型文件
模型包下载后包含以下文件:
```
sherpa-onnx-kws-zipformer-wenetspeech-3.3M-2024-01-01/
├── encoder-epoch-12-avg-2-chunk-16-left-64.int8.onnx # 速度优先
├── encoder-epoch-12-avg-2-chunk-16-left-64.onnx
├── encoder-epoch-99-avg-1-chunk-16-left-64.int8.onnx # 速度优先
├── encoder-epoch-99-avg-1-chunk-16-left-64.onnx # 精度优先
├── decoder-epoch-12-avg-2-chunk-16-left-64.onnx
├── decoder-epoch-99-avg-1-chunk-16-left-64.onnx # 精度优先
├── joiner-epoch-12-avg-2-chunk-16-left-64.int8.onnx # 速度优先
├── joiner-epoch-12-avg-2-chunk-16-left-64.onnx
├── joiner-epoch-99-avg-1-chunk-16-left-64.int8.onnx # 速度优先
├── joiner-epoch-99-avg-1-chunk-16-left-64.onnx # 精度优先
├── tokens.txt # Token映射表(必需)
├── keywords_raw.txt # 模型包里可能附带(可选,runtime 不依赖)
├── keywords.txt # 现成的
├── test_wavs/ # 测试音频(可选)
├── configuration.json # 模型元信息(可选)
└── README.md # 说明文档(可选)
```
#### 3. 选择配置方案
**方案一:精度优先(推荐)**
```bash
cd sherpa-onnx-kws-zipformer-wenetspeech-3.3M-2024-01-01
# 创建模型目录
mkdir -p ../models
# 复制精度优先的epoch-99 fp32三件套
cp encoder-epoch-99-avg-1-chunk-16-left-64.onnx ../models/encoder.onnx
cp decoder-epoch-99-avg-1-chunk-16-left-64.onnx ../models/decoder.onnx
cp joiner-epoch-99-avg-1-chunk-16-left-64.onnx ../models/joiner.onnx
# 复制配套文件
cp tokens.txt ../models/tokens.txt
# keywords_raw.txt 如果模型包里附带,可自行保留;runtime 不依赖它
```
**方案二:速度优先**
```bash
cd sherpa-onnx-kws-zipformer-wenetspeech-3.3M-2024-01-01
# 创建模型目录
mkdir -p ../models
# 复制速度优先的epoch-99 int8三件套
cp encoder-epoch-99-avg-1-chunk-16-left-64.int8.onnx ../models/encoder.onnx
cp decoder-epoch-99-avg-1-chunk-16-left-64.onnx ../models/decoder.onnx
cp joiner-epoch-99-avg-1-chunk-16-left-64.int8.onnx ../models/joiner.onnx
# 复制配套文件
cp tokens.txt ../models/tokens.txt
```
**注意事项**:
- **不要混用 fp32 与 int8**:三个模型文件必须保持一致的精度
- **优先选择 epoch-99**:比 epoch-12 训练更充分,精度更高
- **必需文件**`encoder.onnx` + `decoder.onnx` + `joiner.onnx` + `tokens.txt` + `keywords.txt`
### 最终模型文件结构
配置完成后,模型文件应放在 `wakeword_runtime/models/` 目录下,完整路径为 `main/digital-human/wakeword_runtime/models/`
```
wakeword_runtime/models/
├── encoder.onnx # 编码器模型(重命名后)
├── decoder.onnx # 解码器模型(重命名后)
├── joiner.onnx # 连接器模型(重命名后)
├── tokens.txt # 拼音 Token 映射表(228行版本)
├── keywords.txt # 关键词配置文件(首次启动自动生成)
└── keywords_raw.txt # 可选,runtime 不依赖
```
## 启动方式
`main/digital-human` 目录执行:
```bash
pip install -r wakeword_runtime/requirements.txt
python start.py
```
启动后默认地址:
- 页面地址:`http://127.0.0.1:8006/index.html`
- 事件桥地址:`ws://127.0.0.1:8006/wakeword-ws`
- 健康检查:`http://127.0.0.1:8006/health`
停止方式:
- 在运行终端按 `Ctrl+C`
- 会同时停止静态页面服务、事件桥和唤醒词检测流程
## 配置文件说明
配置文件位于 [main/digital-human/wakeword_runtime/config.json](../main/digital-human/wakeword_runtime/config.json)。
当前主要配置项:
```json
{
"wakeword": {
"enabled": true
},
"model_dir": "models",
"audio": {
"input_device": null,
"sample_rate": 16000,
"channels": 1
},
"detector": {
"num_threads": 4,
"provider": "cpu",
"max_active_paths": 2,
"keywords_score": 1.8,
"keywords_threshold": 0.1,
"num_trailing_blanks": 1,
"cooldown_seconds": 1.5
},
"logging": {
"level": "INFO",
"dir": "logs",
"file": "wakeword-runtime.log"
}
}
```
各字段含义:
| 参数 | 说明 |
| --- | --- |
| `wakeword.enabled` | 是否启用本地唤醒词检测 |
| `model_dir` | 模型和词表所在目录 |
| `audio.input_device` | 麦克风输入设备,默认使用系统默认设备 |
| `audio.sample_rate` | 采样率,默认 `16000` |
| `audio.channels` | 声道数,默认 `1` |
| `detector.num_threads` | 检测器线程数 |
| `detector.provider` | 推理 provider,当前通常为 `cpu` |
| `detector.max_active_paths` | 搜索路径数 |
| `detector.keywords_score` | 关键词增强分数 |
| `detector.keywords_threshold` | 检测阈值 |
| `detector.num_trailing_blanks` | 尾随空白数量 |
| `detector.cooldown_seconds` | 连续触发冷却时间 |
| `logging.level` | 日志等级 |
| `logging.dir` | 日志目录 |
| `logging.file` | 日志文件名 |
## 推荐使用流程
### 首次使用
1. 准备 `models/` 目录下的模型文件和 `tokens.txt`
2. 确认 `models/keywords.txt` 存在
3.`digital-human` 目录运行 `python start.py`
4. 浏览器打开 `http://127.0.0.1:8006/index.html`
5. 进入设置页检查“唤醒词”配置
### 修改唤醒词
1. 打开数字人页面设置
2. 切到“唤醒词”页签
3. 修改启用状态或唤醒词列表
4. 点击“应用唤醒词”
5. 根据提示决定是否立即重启
### 禁用唤醒词
1. 将“启用本地唤醒词”改成禁用
2. 点击“应用唤醒词”
3. 建议立即重启一次
禁用后:
- 页面与事件桥仍然可用
- 唤醒词检测不会继续运行
+66 -13
View File
@@ -1,21 +1,74 @@
# 本地编译docker镜像方法 # 本地编译docker镜像方法
现在本项目已经使用github自动编译docker功能,本文档是提供给有本地编译docker镜像需求的朋友准备的 现在本项目已经使用`github``自动编译docker镜像`功能,如果您拉取的是项目发行的镜像,您没有自己编译镜像需求,那就忽略本文档
1、安装docker 如果您修改了源码,然后想采用`docker`的方式部署运行,可以参照以下步骤操作:
```
## 1、环境准备
安装docker
```bash
sudo apt-get install docker-ce docker-ce-cli containerd.io docker-buildx-plugin docker-compose-plugin sudo apt-get install docker-ce docker-ce-cli containerd.io docker-buildx-plugin docker-compose-plugin
``` ```
2、编译docker镜像
```
#进入项目根目录
# 编译server
docker build -t xiaozhi-esp32-server:server_latest -f ./Dockerfile-server .
# 编译web
docker build -t xiaozhi-esp32-server:web_latest -f ./Dockerfile-web .
# 编译完成后,可以使用docker-compose启动项目 ## 2、编译镜像
# docker-compose.yml你需要修改成自己编译的镜像版本
当你修改好代码后,需要编译新的镜像时,需要按照以下步骤操作:
准备好你的`你的用户名``新的版本号`
- 这个`你的用户名`是你在`docker hub`注册的用户名,例如`xiaozhi`。当然,如果你不需要推送到`docker hub`,你可以自由定义。
- 这个`新的版本号`是你编译的镜像版本,例如`1.2.3`,你可以根据需要自定义或者使用日期格式(例如`20260609`)主要是方便和现在运行的版本号做区分,同时也方便下次回忆你是什么时候构建的,不要和现在你本机运行的版本号相同。
进入`xiaozhi-esp32-server`项目根目录,编译 server 和 web 两个镜像:
```bash
cd 项目根目录
# 编译server镜像
docker build -f Dockerfile-server -t 你的用户名/xiaozhi-esp32-server:新的版本号 .
# 编译web镜像
docker build -f Dockerfile-web -t 你的用户名/xiaozhi-esp32-server-web:新的版本号 .
```
## 3、修改docker-compose配置
```bash
cd main/xiaozhi-server cd main/xiaozhi-server
docker-compose up -d ```
编辑 `docker-compose_all.yml` 文件,将镜像版本替换为你刚才编译的版本:
```yaml
services:
xiaozhi-esp32-server:
image: 你的用户名/xiaozhi-esp32-server:新的版本号 # 修改为你的镜像地址
...
xiaozhi-esp32-server-web:
image: 你的用户名/xiaozhi-esp32-server-web:新的版本号 #修改为你的镜像地址
...
```
## 4、重启服务
```bash
# 停止旧容器
docker compose -f docker-compose_all.yml down
# 启动新容器
docker compose -f docker-compose_all.yml up -d
```
## 5、验证
查看日志确认服务启动正常:
```bash
# 查看server日志
docker logs -f -n 50 xiaozhi-esp32-server
# 查看web日志
docker logs -f -n 50 xiaozhi-esp32-server-web
``` ```
+2 -2
View File
@@ -13,9 +13,9 @@ http://192.168.1.25:8003/xiaozhi/ota/
``` ```
如果显示“OTA接口运行正常,向设备发送的websocket地址是:ws://xxx:8000/xiaozhi/v1/ 如果显示“OTA接口运行正常,向设备发送的websocket地址是:ws://xxx:8000/xiaozhi/v1/
你可以使用项目自带的`test_page.html`测试一下,是否能连上ota页面输出的websocket地址。 你可以启动`digital-human`模块后打开`index.html`测试一下,是否能连上ota页面输出的websocket地址。
如果访问不到,你需要到配置文件`.config.yaml`里修改`server.websocket`的地址,重启后再重新测试,直到`test_page.html`能正常访问。 如果访问不到,你需要到配置文件`.config.yaml`里修改`server.websocket`的地址,重启后再重新测试,直到`index.html`能正常访问。
成功后,请往下进行第2步 成功后,请往下进行第2步
+1 -1
View File
@@ -97,7 +97,7 @@ http://homeassistant.local:8123
使用管理员账号,登录`智控台`。在`智能体管理`,找到你的智能体,再点击`配置角色`。 使用管理员账号,登录`智控台`。在`智能体管理`,找到你的智能体,再点击`配置角色`。
将意图识别设置成`函数调用`或`LLM意图识别`。这时你会看到右侧有一个`编辑功能`。点击`编辑功能`按钮,会弹出`功能管理`的框。 将意图识别设置成`外挂的大模型意图识别`或`大模型自主函数调用`。这时你会看到右侧有一个`编辑功能`。点击`编辑功能`按钮,会弹出`功能管理`的框。
在`功能管理`的框里,你需要勾选`HomeAssistant设备状态查询`和`HomeAssistant设备状态修改`。 在`功能管理`的框里,你需要勾选`HomeAssistant设备状态查询`和`HomeAssistant设备状态修改`。
+51 -22
View File
@@ -1,29 +1,58 @@
# 火山双流式TTS+声音克隆配置教程 # 智控台 火山双流式语音合成+音色克隆配置教程
单模块部署下,使用火山引擎双向流式语音合成服务的同时进行声音克隆,支持WebSocket协议流式调用。
本教程分为4个阶段:准备阶段、配置阶段、克隆阶段、使用阶段。主要是介绍通过智控台配置火山双流式语音合成+音色克隆的过程。
## 第一阶段:准备阶段
超级管理员先预先把火山引擎服务开通好,获取到App Id,Access Token。默认火上引擎会赠送一个音色资源。这个音色资源需要把它复制到本项目里。
如果你想克隆多个音色,需要购买开通多个音色资源。只要把每个音色资源的声音ID(S_xxxxx)复制到本项目。然后分配给系统的账号使用即可。以下是详细步骤:
### 1.开通火山引擎服务 ### 1.开通火山引擎服务
访问 https://console.volcengine.com/speech/app 在应用管理创建应用,勾选语音合成大模型和声音复刻大模型,左边列表点击声音复刻大模型后下滑获得App IdAccess TokenCluster ID以及声音ID(S_xxxxx) 访问 https://console.volcengine.com/speech/app 在应用管理创建应用,勾选语音合成大模型和声音复刻大模型
### 2.克隆音色
克隆音色请参照教程 https://github.com/104gogo/huoshan-voice-copy
准备一段 10-30 秒的音频文件(.wav格式)添加到克隆的项目中,将平台获得的密钥填入```uploadAndStatus.py```和```tts_http_demo.py``` ### 2.获取音色资源ID
访问 https://console.volcengine.com/speech/service/9999 复制三项内容,分别是App IdAccess Token以及声音ID(S_xxxxx)。如图
在uploadAndStatus.py中,将 audio_path=修改成自己的.wav文件名称 ![获取音色资源](images/image-clone-integration-01.png)
```python
train(appid=appid, token=token, audio_path=r".\audios\xiaohe.wav", spk_id=spk_id)
```
运行以下命令生成test_submit.mp3,点击播放试听克隆效果 ## 第二阶段:配置火山引擎服务
```python ### 1.填写火山引擎配置
python uploadAndStatus.py
python tts_http_demo.py
```
回到火山引擎控制台页面,刷新可以看到声音复刻详情的状态是复刻成功。
### 3.填写配置文件
将火山引擎服务申请到的密钥填入.config.yaml的HuoshanDoubleStreamTTS配置文件中
修改 resource_id的参数为``` volc.megatts.default``` 使用超级管理员账号登录智控台,点击顶部【模型配置】,再点击模型配置页面左侧的【语音合成】,搜索找到“火山双流式语音合成”,点击修改,将你火山引擎的`App Id`填入到【应用ID】字段里,将`Access Token`填入到【访问令牌】字段里。然后保存。
(参考官方文档 https://www.volcengine.com/docs/6561/1329505
speaker的参数填入声音ID(S_xxxxx)
启动服务,唤醒小智发出的声音是克隆的音色即成功。 ### 2.将音色资源ID分配给系统账号
使用超级管理员账号登录智控台,点击顶部`参数字典`,在下拉菜单中,点击`系统功能配置`页面。在页面上勾选`音色克隆`,点击保存配置。即可在顶部菜单看到`音色克隆`按钮。
使用超级管理员账号登录智控台,点击顶部【音色克隆】、【音色资源】。
点击新增按钮,在【平台名称】选择“火山双流式语音合成”;
在【音色资源ID】填入你火山引擎的声音资源ID(S_xxxxx),填入后按回车;
在【归属账号】选择你要分配给的系统账号,你可以分配给你自己。然后点击保存
## 第三阶段:克隆阶段
如果登录后,点击顶部【音色克隆】》【音色克隆】,显示【您的账号暂无音色资源请联系管理员分配音色资源】,说明你在第二阶段还没有把音色资源ID分配给这个账号。那就是回到第二阶段,分配音色资源给对应的账号。
如果登录后,点击顶部【音色克隆】》【音色克隆】,能看到对应的音色列表。请继续。
在列表里会看到对应的音色列表。选择其中一个音色资源,点击【上传音频】按钮。上传后,可以试听一下声音或者截取某段声音。确认后点击【上传音频】按钮。
![上传音频](images/image-clone-integration-02.png)
上传音频后,在列表里会看到对应的音色会变成“待复刻”状态。点击【立即复刻】按钮。等1~2秒会返回结果。
如果复刻失败,请将鼠标放到“错误信息”图标上,会显示失败的原因。
如果复刻成功,在列表里会看到对应的音色会变成“训练成功”状态。此时你可以点击【声音名称】栏的修改按钮,修改音色资源的名称,方便后期选择使用。
## 第四阶段:使用阶段
点击顶部【智能体管理】,选择任意一个智能体,点击【配置角色】按钮。
语音合成(TTS)选择“火山双流式语音合成”。在列表里,找到名字带有“克隆音色”的音色资源(如图),选择它,点击保存。
![选择音色](images/image-clone-integration-03.png)
接下来,可以唤醒小智和它对话。
Binary file not shown.

Before

Width:  |  Height:  |  Size: 289 KiB

After

Width:  |  Height:  |  Size: 152 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 122 KiB

After

Width:  |  Height:  |  Size: 118 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 404 KiB

After

Width:  |  Height:  |  Size: 84 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 111 KiB

After

Width:  |  Height:  |  Size: 98 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 325 KiB

After

Width:  |  Height:  |  Size: 143 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 280 KiB

After

Width:  |  Height:  |  Size: 108 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 298 KiB

After

Width:  |  Height:  |  Size: 100 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 210 KiB

After

Width:  |  Height:  |  Size: 260 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 143 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 96 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 129 KiB

+44 -14
View File
@@ -1,11 +1,18 @@
# IndexStreamTTS 使用指南 # IndexStreamTTS 使用指南
## 环境准备 ## 环境准备
### 1. 克隆项目 (这里使用的为VLLM的版本) ### 1. 克隆项目
```bash ```bash
git clone https://github.com/Ksuriuri/index-tts-vllm.git git clone https://github.com/Ksuriuri/index-tts-vllm.git
```
进入解压后的目录
```bash
cd index-tts-vllm cd index-tts-vllm
``` ```
切换到指定版本 (使用VLLM-0.10.2的历史版本)
```bash
git checkout 224e8d5e5c8f66801845c66b30fa765328fd0be3
```
### 2. 创建并激活 conda 环境 ### 2. 创建并激活 conda 环境
```bash ```bash
@@ -13,7 +20,7 @@ conda create -n index-tts-vllm python=3.12
conda activate index-tts-vllm conda activate index-tts-vllm
``` ```
### 3. 安装PyTorch ### 3. 安装PyTorch 需要版本为2.8.0(最新版)
#### 查看显卡最高支持的版本和实际安装的版本 #### 查看显卡最高支持的版本和实际安装的版本
```bash ```bash
nvidia-smi nvidia-smi
@@ -27,12 +34,11 @@ CUDA Version: 12.8
```bash ```bash
Cuda compilation tools, release 12.8, V12.8.89 Cuda compilation tools, release 12.8, V12.8.89
``` ```
#### 那么对应的安装命令 (请注意不要横跨大版本!!!) #### 那么对应的安装命令pytorch默认给的是12.8的驱动版本)
```bash ```bash
pip install torch==2.7.0 torchvision==0.22.0 torchaudio==2.7.0 --index-url https://download.pytorch.org/whl/cu128 pip install torch torchvision
``` ```
优先建议安装 pytorch 2.7.0(对应 vllm 0.9.0),具体安装指令请参考:[pytorch 官网](https://pytorch.org/get-started/locally/]\) 需要 pytorch 版本 2.8.0(对应 vllm 0.10.2),具体安装指令请参考:[pytorch 官网](https://pytorch.org/get-started/locally/)
若显卡不支持,请安装 pytorch 2.5.1(对应 vllm 0.7.3),并将 requirements.txt 中 vllm==0.9.0 修改为 vllm==0.7.3
### 4. 安装依赖 ### 4. 安装依赖
```bash ```bash
@@ -40,6 +46,7 @@ pip install -r requirements.txt
``` ```
### 5. 下载模型权重 ### 5. 下载模型权重
### 方案一:下载官方权重文件后转换
此为官方权重文件,下载到本地任意路径即可,支持 IndexTTS-1.5 的权重 此为官方权重文件,下载到本地任意路径即可,支持 IndexTTS-1.5 的权重
| HuggingFace | ModelScope | | HuggingFace | ModelScope |
|---------------------------------------------------------------|---------------------------------------------------------------------| |---------------------------------------------------------------|---------------------------------------------------------------------|
@@ -47,19 +54,19 @@ pip install -r requirements.txt
| [IndexTTS-1.5](https://huggingface.co/IndexTeam/IndexTTS-1.5) | [IndexTTS-1.5](https://modelscope.cn/models/IndexTeam/IndexTTS-1.5) | | [IndexTTS-1.5](https://huggingface.co/IndexTeam/IndexTTS-1.5) | [IndexTTS-1.5](https://modelscope.cn/models/IndexTeam/IndexTTS-1.5) |
下面以ModelScope的安装方法为例 下面以ModelScope的安装方法为例
### 请注意:git需要安装并初始化启用lfs(如已安装可以跳过) #### 请注意:git需要安装并初始化启用lfs(如已安装可以跳过)
```bash ```bash
sudo apt-get install git-lfs sudo apt-get install git-lfs
git lfs install git lfs install
``` ```
创建模型目录,并拉取模型 创建模型目录,并拉取模型
```bash ```bash
mkkdir model_dir mkdir model_dir
cd model_dir cd model_dir
git clone https://www.modelscope.cn/IndexTeam/IndexTTS-1.5.git git clone https://www.modelscope.cn/IndexTeam/IndexTTS-1.5.git
``` ```
### 5. 模型权重转换 #### 模型权重转换
```bash ```bash
bash convert_hf_format.sh /path/to/your/model_dir bash convert_hf_format.sh /path/to/your/model_dir
``` ```
@@ -72,6 +79,9 @@ bash convert_hf_format.sh model_dir/IndexTTS-1.5
### 6. 更改接口适配一下项目 ### 6. 更改接口适配一下项目
接口返回数据与项目不适配需要调整一下,使其直接返回音频数据 接口返回数据与项目不适配需要调整一下,使其直接返回音频数据
```bash ```bash
vi api_server.py
```
```bash
@app.post("/tts", responses={ @app.post("/tts", responses={
200: {"content": {"application/octet-stream": {}}}, 200: {"content": {"application/octet-stream": {}}},
500: {"content": {"application/json": {}}} 500: {"content": {"application/json": {}}}
@@ -104,7 +114,7 @@ async def tts_api(request: Request):
vi start_api.sh vi start_api.sh
``` ```
### 将下面内容粘贴进去并按:输入wq保存 ### 将下面内容粘贴进去并按:输入wq保存
#### 脚本中的/home/system/indexTTS/index-tts-vllm/model_dir/IndexTTS-1.5 请自行修改为实际路径 #### 脚本中的/home/system/index-tts-vllm/model_dir/IndexTTS-1.5 请自行修改为实际路径
```bash ```bash
# 激活conda环境 # 激活conda环境
conda activate index-tts-vllm conda activate index-tts-vllm
@@ -129,23 +139,43 @@ else
echo "已终止进程 $PID_VLLM" echo "已终止进程 $PID_VLLM"
fi fi
# 查找占用VLLM::EngineCore进程
GPU_PIDS=$(ps aux | grep -E "VLLM|EngineCore" | grep -v grep | awk '{print $2}')
# 检查是否找到进程号
if [ -z "$GPU_PIDS" ]; then
echo "没有找到VLLM相关进程"
else
echo "找到VLLM相关进程,进程号为: $GPU_PIDS"
# 先尝试普通kill,等待2秒
kill $GPU_PIDS
sleep 2
# 检查进程是否还在
if ps -p $GPU_PIDS > /dev/null; then
echo "进程仍在运行,强制终止..."
kill -9 $GPU_PIDS
fi
echo "已终止进程 $GPU_PIDS"
fi
# 创建tmp目录(如果不存在) # 创建tmp目录(如果不存在)
mkdir -p tmp mkdir -p tmp
# 后台运行api_server.py,日志重定向到tmp/server.log # 后台运行api_server.py,日志重定向到tmp/server.log
export VLLM_USE_V1=0 nohup python api_server.py --model_dir /home/system/index-tts-vllm/model_dir/IndexTTS-1.5 --port 11996 > tmp/server.log 2>&1 &
nohup python api_server.py --model_dir /home/system/indexTTS/index-tts-vllm/model_dir/IndexTTS-1.5 --port 11996 > tmp/server.log 2>&1 &
echo "api_server.py 已在后台运行,日志请查看 tmp/server.log" echo "api_server.py 已在后台运行,日志请查看 tmp/server.log"
``` ```
给脚本执行权限并运行脚本 给脚本执行权限并运行脚本
```bash ```bash
chmod +x tmp chmod +x start_api.sh
./start_api.sh ./start_api.sh
``` ```
日志会在tmp/server.log中输出,可以通过以下命令查看日志情况 日志会在tmp/server.log中输出,可以通过以下命令查看日志情况
```bash ```bash
tail -f tmp/server.log tail -f tmp/server.log
``` ```
如果显卡内存足够,可在脚本中添加启动参数 ----gpu_memory_utilization 来调整显存占用比例,默认值为 0.25
## 音色配置 ## 音色配置
index-tts-vllm支持通过配置文件注册自定义音色,支持单音色和混合音色配置。 index-tts-vllm支持通过配置文件注册自定义音色,支持单音色和混合音色配置。
在项目根目录下的assets/speaker.json文件中配置自定义音色 在项目根目录下的assets/speaker.json文件中配置自定义音色
@@ -161,5 +191,5 @@ index-tts-vllm支持通过配置文件注册自定义音色,支持单音色和
] ]
} }
``` ```
### 注意 ### 注意 (配置角色后需重启服务进行音色注册)
添加后需在智控台中添加相应的说话人(单模块则更换相应的voice) 添加后需在智控台中添加相应的说话人(单模块则更换相应的voice)
+267
View File
@@ -0,0 +1,267 @@
# manager-api FastAPI 兼容性矩阵
> 生成依据:`main/manager-api-fastapi/compatibility/java-routes.json`、
> `main/manager-api-fastapi/compatibility/consumer-routes.json`、`route-surface-results.json`、
> `authenticated-route-results.json`、`contract-results.json` 和当前 Java 源码。接口路径均省略
> 共同前缀 `/xiaozhi`。
## 结论与状态口径
Java 基线共有 **154** 条 Spring MVC 路由;FastAPI 已注册 **154/154100%**,并由
`tests/test_java_route_manifest.py` 对源码清单 freshness、数量和 method/path 注册闭合进行检查。
此外实现 3 条仅由仓库消费者使用、Java Controller 中不存在的兼容路由,因此这 3 条不计入
154 条 Java 覆盖率。三端 188 个调用点均能解析到 FastAPI 路由。
矩阵状态必须按下列含义阅读:
- `结构✓`method/path 已注册且清单闭合;它不等于业务行为逐接口实测。
- `请求面差分✓1`:本行已向隔离 Java/FastAPI 各发送一次缺少鉴权或安全非法输入,精确比较
HTTP status、body 与 Content-Type;最终为 **154/154 通过、0 失败、0 跳过**,且不发送成功写请求。
- `认证业务面差分✓1`:本行已使用有效 DB Token、server-secret 或匿名身份,再向隔离
Java/FastAPI 各发送一次安全业务/校验请求,精确比较 HTTP status、body 与 Content-Type
最终为 **154/154 通过、0 失败、0 跳过**,且不主动发送成功写请求。该状态不等于每条路由的
完整成功生命周期均已差分,完整副作用证据仍以 `差分✓N` 为准。
- `领域✓(x,域级)`:该领域有 service/repository/协议自动测试,但不保证本行每条成功与错误路径
都被直接请求。`领域—` 表示除结构测试外没有可归属的域级直接测试证据。
- `差分✓N`:本行除安全请求面外,还参与了成功、主要错误、协议或数据库副作用的深度对照;
括号说明覆盖面。深度结果为 **49/49 checks 通过、0 失败、0 跳过**,覆盖 **21/154** 条路由。
`差分间接✓` 表示 J125 作为下载链路的 URL 生成步骤被间接覆盖;`差分—` 表示没有深度对照,
不能把 154/154 请求面差分误读成 154 条全部成功路径与副作用都已逐接口对照。
- 所有 `Result<T>` 均表示 `{code,msg,data}` envelope;原 Java 为 HTTP 200 的认证、权限、业务和
参数错误由全局兼容层维持 HTTP 200。二进制/OTA 裸响应在“响应类型”列单独标明。
## 三端消费者闭合
| 消费者 | 调用点 | 唯一结构路由 | 方法分布 |
|---|---:|---:|---|
| `manager-web` | 134 | 130 | DELETE 12、GET 59、POST 40、PUT 23 |
| `manager-mobile` | 46 | 40 | DELETE 3、GET 26、POST 12、PUT 5 |
| `xiaozhi-server` | 8 | 8 | GET 2、POST 6 |
| **合计** | **188** | **140** | — |
### 3 条消费者孤儿兼容路由
| Method/path | 来源 | FastAPI 语义 | 鉴权 | 状态 |
|---|---|---|---|---|
| `GET /api/ping` | manager-mobile 环境设置探活 | `{code:0,msg:"success",data:"pong"}` | 匿名 | 实现✓;consumer resolve✓ |
| `PUT /user/configDevice/{device_id}` | manager-web 遗留设备配置调用 | 按现有设备更新契约处理 body | DB Token | 实现✓;consumer resolve✓ |
| `GET /device/address-book/lookup` | xiaozhi-server 管理客户端 | `callerMac/nickname/answer` 地址簿查询/呼叫兼容别名 | server-secret | 实现✓;consumer resolve✓;device 域测试✓ |
`GET /admin/dict/data/type/FIRMWARE_TYPE` 是动态 Java 路由
`GET /admin/dict/data/type/{dictType}` 的一个字面调用,不是第四条孤儿路由。
## Java 基线静态盘点
- Controller24 个、154 条映射。按 Controller 的路由数为:`AdminController`(5)、`AgentChatHistoryController`(4)、`AgentController`(21)、`AgentMcpAccessPointController`(2)、`AgentSnapshotController`(4)、`AgentTemplateController`(6)、`AgentVoicePrintController`(4)、`ConfigController`(3)、`CorrectWordController`(7)、`DeviceController`(13)、`KnowledgeBaseController`(7)、`KnowledgeFilesController`(8)、`LoginController`(8)、`ModelController`(11)、`ModelProviderController`(5)、`OTAController`(3)、`OTAMagController`(9)、`ServerSideManageController`(2)、`SysDictDataController`(6)、`SysDictTypeController`(5)、`SysParamsController`(5)、`TimbreController`(4)、`VoiceCloneController`(6)、`VoiceResourceController`(6)。
- 数据分层:`entity/` 29 个 Java 文件(28 个 `*Entity.java``BaseEntity`)、`dto/` 58 个、
`vo/` 14 个、`dao/` 29 个、`service/` 树 78 个文件(其中
`service/impl/` 38 个)。FastAPI 对应落在 `schemas/``repositories/`
`services/``routers/``integrations/``jobs/`,没有把跨表事务放进路由。
- MyBatis XML20 个,分别是 `mapper/agent/AgentCorrectWordMappingDao.xml``mapper/agent/AgentDao.xml``mapper/agent/AgentPluginMappingMapper.xml``mapper/agent/AgentSnapshotDao.xml``mapper/agent/AgentTagDao.xml``mapper/agent/AgentTagRelationDao.xml``mapper/agent/AgentTemplateMapper.xml``mapper/agent/AiAgentChatHistoryDao.xml``mapper/correctword/CorrectWordItemDao.xml``mapper/device/DeviceAddressBookDao.xml``mapper/device/DeviceDao.xml``mapper/knowledge/KnowledgeBaseDao.xml``mapper/model/ModelConfigDao.xml``mapper/model/ModelProviderDao.xml``mapper/security/SysUserTokenDao.xml``mapper/sys/SysDictDataDao.xml``mapper/sys/SysDictTypeDao.xml``mapper/sys/SysParamsDao.xml``mapper/sys/SysUserDao.xml``mapper/voiceclone/VoiceCloneDao.xml`
- Liquibase`db.changelog-master.yaml` 含 101 个 `changeSet` 引用,目录中恰有
101 个 SQLPython 部署继续执行这 101 个原始 SQL,不改写历史。
- 定时工作:`DocumentStatusSyncTask` 每次完成后延迟 30 秒,扫描 RAGFlow RUNNING 文档并
回写 SUCCESS/FAIL/CANCEL 与统计;当前 Java 源码另有 `AgentSnapshotRedactionRunner`,启动时
执行一次并在滚动部署期每 15 秒补偿脱敏旧快照。FastAPI 将工作移到独立 jobs 进程,并以
Redis 分布式锁/watchdog 防止多 worker 重复执行。
- 外部集成:RAGFlow dataset/document/chunk/retrieval/upload;阿里云短信;火山语音克隆训练与
音频;声纹 HTTPOpenAI-compatible LLM 摘要/标题;MQTT gateway HTTPMCP/管理动作
WebSocketOTA/WS/MQTT 的 HMAC、Base64、时间戳与下载文件存储。自动测试只访问可重复 mock,
未使用真实付费凭证。
## 154 条 Java→FastAPI 逐接口矩阵
副作用缩写:`DB-R/W`=数据库读/写,`Redis-R/W/DEL`=缓存读/写/失效,`文件-R/W`=文件
读取/写入;外部调用均在 service/integration 层。权限为空时表示只需对应鉴权身份。
| # | Method/path | Java Controller.handler | 请求面 | 响应类型 | 鉴权 / 权限 | DB/Redis/文件/外部副作用 | 实现与测试状态 |
|---:|---|---|---|---|---|---|---|
| J001 | `GET /admin/device/all` | `AdminController.pageDevice` | Query:params:Map<String, Object> | envelope <PageData<UserShowDeviceListVO>> | DB Token / `sys:role:superAdmin` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J002 | `POST /admin/dict/data/delete` | `SysDictDataController.delete` | Body:Long[] | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(dict cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J003 | `GET /admin/dict/data/page` | `SysDictDataController.page` | Query:params:Map<String, Object> | envelope <PageData<SysDictDataVO>> | DB Token / `sys:role:superAdmin` | DB-R; Redis-R/W(dict cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J004 | `POST /admin/dict/data/save` | `SysDictDataController.save` | Body:SysDictDataDTO | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(dict cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J005 | `GET /admin/dict/data/type/{dictType}` | `SysDictDataController.getDictDataByType` | Path:dictType | envelope <List<SysDictDataItem>> | DB Token / `sys:role:normal` | DB-R; Redis-R/W(dict cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J006 | `PUT /admin/dict/data/update` | `SysDictDataController.update` | Body:SysDictDataDTO | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(dict cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J007 | `GET /admin/dict/data/{id}` | `SysDictDataController.get` | Path:id | envelope <SysDictDataVO> | DB Token / `sys:role:superAdmin` | DB-R; Redis-R/W(dict cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J008 | `POST /admin/dict/type/delete` | `SysDictTypeController.delete` | Body:Long[] | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(dict cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J009 | `GET /admin/dict/type/page` | `SysDictTypeController.page` | Query:params:Map<String, Object> | envelope <PageData<SysDictTypeVO>> | DB Token / `sys:role:superAdmin` | DB-R; Redis-R/W(dict cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J010 | `POST /admin/dict/type/save` | `SysDictTypeController.save` | Body:SysDictTypeDTO | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(dict cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J011 | `PUT /admin/dict/type/update` | `SysDictTypeController.update` | Body:SysDictTypeDTO | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(dict cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J012 | `GET /admin/dict/type/{id}` | `SysDictTypeController.get` | Path:id | envelope <SysDictTypeVO> | DB Token / `sys:role:superAdmin` | DB-R; Redis-R/W(dict cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J013 | `POST /admin/params` | `SysParamsController.save` | Body:SysParamsDTO | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-W/DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J014 | `PUT /admin/params` | `SysParamsController.update` | Body:SysParamsDTO | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-W; 外部-配置端点探测(按 paramCode) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J015 | `POST /admin/params/delete` | `SysParamsController.delete` | Body:String[] | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-W/DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J016 | `GET /admin/params/page` | `SysParamsController.page` | Query:params:Map<String, Object> | envelope <PageData<SysParamsDTO>> | DB Token / `sys:role:superAdmin` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J017 | `GET /admin/params/{id}` | `SysParamsController.get` | Path:id | envelope <SysParamsDTO> | DB Token / `sys:role:superAdmin` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J018 | `POST /admin/server/emit-action` | `ServerSideManageController.emitServerAction` | Body:EmitSeverActionDTO | envelope <Boolean> | DB Token / `sys:role:superAdmin` | DB/Redis-R(secret/WS); Redis-W(one-shot); 外部-WebSocket | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J019 | `GET /admin/server/server-list` | `ServerSideManageController.getWsServerList` | — | envelope <List<String>> | DB Token / `sys:role:superAdmin` | DB/Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J020 | `GET /admin/users` | `AdminController.pageUser` | Query:params:Map<String, Object> | envelope <PageData<AdminPageUserVO>> | DB Token / `sys:role:superAdmin` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分✓3(权限/序列化/非法分页) |
| J021 | `PUT /admin/users/changeStatus/{status}` | `AdminController.changeStatus` | Path:status; Body:String[] | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W(user/password/status/token) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J022 | `DELETE /admin/users/{id}` | `AdminController.delete` | Path:id | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W(用户/token/device/agent 级联) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J023 | `PUT /admin/users/{id}` | `AdminController.update` | Path:id | envelope <String> | DB Token / `sys:role:superAdmin` | DB-W(user/password/status/token) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(sys,域级);差分— |
| J024 | `POST /agent` | `AgentController.save` | Body:AgentCreateDTO | envelope <String> | DB Token / `sys:role:normal` | DB-W(含快照/映射/标签事务); Redis-DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J025 | `GET /agent/all` | `AgentController.adminAgentList` | Query:params:Map<String, Object> | envelope <PageData<AgentEntity>> | DB Token / `sys:role:superAdmin` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J026 | `POST /agent/audio/{audioId}` | `AgentController.getAudioId` | Path:audioId | envelope <String> | DB Token / `sys:role:normal` | DB-R(audio); Redis-W(one-shot URL) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J027 | `GET /agent/chat-history/download/{uuid}/current` | `AgentChatHistoryController.downloadCurrentSession` | Path:uuid | 流式/二进制 + 原下载 headers | 匿名 / — | DB/Redis-R(one-shot); 文件-R/流式 | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J028 | `GET /agent/chat-history/download/{uuid}/previous` | `AgentChatHistoryController.downloadCurrentSessionWithPrevious` | Path:uuid | 流式/二进制 + 原下载 headers | 匿名 / — | DB/Redis-R(one-shot); 文件-R/流式 | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J029 | `POST /agent/chat-history/getDownloadUrl/{agentId}/{sessionId}` | `AgentChatHistoryController.getDownloadUrl` | Path:agentId,sessionId | envelope <String> | DB Token / — | DB-R(chat/session); Redis-W(download token TTL) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J030 | `POST /agent/chat-history/report` | `AgentChatHistoryController.uploadFile` | Body:AgentChatHistoryReportDTO | envelope <Boolean> | server-secret / — | DB-W(chat/session); server-secret | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J031 | `POST /agent/chat-summary/{sessionId}/save` | `AgentController.generateAndSaveChatSummary` | Path:sessionId | envelope <Void> | server-secret / — | DB-R/W(chat); 外部-OpenAI-compatible LLM | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J032 | `POST /agent/chat-title/{sessionId}/generate` | `AgentController.generateAndSaveChatTitle` | Path:sessionId | envelope <Void> | server-secret / — | DB-R/W(chat); 外部-OpenAI-compatible LLM | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J033 | `GET /agent/list` | `AgentController.getUserAgents` | Query:keyword:String,searchType:String | envelope <List<AgentDTO>> | DB Token / `sys:role:normal` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分✓1 |
| J034 | `GET /agent/mcp/address/{agentId}` | `AgentMcpAccessPointController.getAgentMcpAccessAddress` | Path:agentId | envelope <String> | DB Token / `sys:role:normal` | DB/Redis-R; AES token 生成 | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J035 | `GET /agent/mcp/tools/{agentId}` | `AgentMcpAccessPointController.getAgentMcpToolsList` | Path:agentId | envelope <List<String>> | DB Token / `sys:role:normal` | DB/Redis-R; 外部-WebSocket MCP | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J036 | `GET /agent/play/{uuid}` | `AgentController.playAudio` | Path:uuid | 流式/二进制 + 原下载 headers | 匿名 / — | DB/Redis-R(one-shot); 文件-R/流式 | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J037 | `PUT /agent/saveMemory/{macAddress}` | `AgentController.updateByDeviceId` | Path:macAddress; Body:AgentMemoryDTO | envelope <Void> | DB Token / `sys:role:normal` | DB-W(含快照/映射/标签事务); Redis-DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J038 | `POST /agent/tag` | `AgentController.createTag` | Body:Map<String, String> | envelope <AgentTagEntity> | DB Token / `sys:role:normal` | DB-W(含快照/映射/标签事务); Redis-DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J039 | `GET /agent/tag/list` | `AgentController.getAllTags` | — | envelope <List<AgentTagDTO>> | DB Token / `sys:role:normal` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J040 | `DELETE /agent/tag/{id}` | `AgentController.deleteTag` | Path:id | envelope <Void> | DB Token / `sys:role:normal` | DB-W(含快照/映射/标签事务); Redis-DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J041 | `GET /agent/template` | `AgentController.templateList` | — | envelope <List<AgentTemplateEntity>> | DB Token / `sys:role:normal` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J042 | `POST /agent/template` | `AgentTemplateController.createAgentTemplate` | Body:AgentTemplateEntity | envelope <AgentTemplateEntity> | DB Token / `sys:role:superAdmin` | DB-W(含快照/映射/标签事务); Redis-DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J043 | `PUT /agent/template` | `AgentTemplateController.updateAgentTemplate` | Body:AgentTemplateEntity | envelope <AgentTemplateEntity> | DB Token / `sys:role:superAdmin` | DB-W(含快照/映射/标签事务); Redis-DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J044 | `POST /agent/template/batch-remove` | `AgentTemplateController.batchRemoveAgentTemplates` | Body:List<String> | envelope <String> | DB Token / `sys:role:superAdmin` | DB-W(含快照/映射/标签事务); Redis-DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J045 | `GET /agent/template/page` | `AgentTemplateController.getAgentTemplatesPage` | Query:params:Map<String, Object> | envelope <PageData<AgentTemplateVO>> | DB Token / `sys:role:superAdmin` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J046 | `DELETE /agent/template/{id}` | `AgentTemplateController.deleteAgentTemplate` | Path:id | envelope <String> | DB Token / `sys:role:superAdmin` | DB-W(含快照/映射/标签事务); Redis-DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J047 | `GET /agent/template/{id}` | `AgentTemplateController.getAgentTemplateById` | Path:id | envelope <AgentTemplateVO> | DB Token / `sys:role:superAdmin` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J048 | `POST /agent/voice-print` | `AgentVoicePrintController.save` | Body:AgentVoicePrintSaveDTO | envelope <Void> | DB Token / `sys:role:normal` | DB-W; 外部-voiceprint HTTP | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J049 | `PUT /agent/voice-print` | `AgentVoicePrintController.update` | Body:AgentVoicePrintUpdateDTO | envelope <Void> | DB Token / `sys:role:normal` | DB-W; 外部-voiceprint HTTP | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J050 | `GET /agent/voice-print/list/{id}` | `AgentVoicePrintController.list` | Path:id | envelope <List<AgentVoicePrintVO>> | DB Token / `sys:role:normal` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J051 | `DELETE /agent/voice-print/{id}` | `AgentVoicePrintController.delete` | Path:id | envelope <Void> | DB Token / `sys:role:normal` | DB-W; 外部-voiceprint HTTP | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J052 | `GET /agent/{agentId}/snapshots` | `AgentSnapshotController.page` | Path:agentId; Query:params:AgentSnapshotPageDTO | envelope <PageData<AgentSnapshotVO>> | DB Token / `sys:role:normal` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J053 | `DELETE /agent/{agentId}/snapshots/{snapshotId}` | `AgentSnapshotController.deleteSnapshot` | Path:agentId,snapshotId | envelope <Void> | DB Token / `sys:role:normal` | DB-W(含快照/映射/标签事务); Redis-DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J054 | `GET /agent/{agentId}/snapshots/{snapshotId}` | `AgentSnapshotController.getSnapshot` | Path:agentId,snapshotId | envelope <AgentSnapshotVO> | DB Token / `sys:role:normal` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J055 | `POST /agent/{agentId}/snapshots/{snapshotId}/restore` | `AgentSnapshotController.restore` | Path:agentId,snapshotId; Body:AgentSnapshotRestoreDTO | envelope <Void> | DB Token / `sys:role:normal` | DB-W(含快照/映射/标签事务); Redis-DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J056 | `DELETE /agent/{id}` | `AgentController.delete` | Path:id | envelope <Void> | DB Token / `sys:role:normal` | DB-W(含快照/映射/标签事务); Redis-DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J057 | `GET /agent/{id}` | `AgentController.getAgentById` | Path:id | envelope <AgentInfoVO> | DB Token / `sys:role:normal` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J058 | `PUT /agent/{id}` | `AgentController.update` | Path:id; Body:AgentUpdateDTO | envelope <Void> | DB Token / `sys:role:normal` | DB-W(含快照/映射/标签事务); Redis-DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J059 | `GET /agent/{id}/chat-history/audio` | `AgentController.getContentByAudioId` | Path:id | envelope <String> | DB Token / `sys:role:normal` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J060 | `GET /agent/{id}/chat-history/user` | `AgentController.getRecentlyFiftyByAgentId` | Path:id | envelope <List<AgentChatHistoryUserVO>> | DB Token / `sys:role:normal` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J061 | `GET /agent/{id}/chat-history/{sessionId}` | `AgentController.getAgentChatHistory` | Path:id,sessionId | envelope <List<AgentChatHistoryDTO>> | DB Token / `sys:role:normal` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J062 | `GET /agent/{id}/sessions` | `AgentController.getAgentSessions` | Path:id; Query:params:Map<String, Object> | envelope <PageData<AgentChatSessionDTO>> | DB Token / `sys:role:normal` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J063 | `GET /agent/{id}/tags` | `AgentController.getAgentTags` | Path:id | envelope <List<AgentTagDTO>> | DB Token / `sys:role:normal` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J064 | `PUT /agent/{id}/tags` | `AgentController.saveAgentTags` | Path:id; Body:Map<String, Object> | envelope <Void> | DB Token / `sys:role:normal` | DB-W(含快照/映射/标签事务); Redis-DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(agent,域级);差分— |
| J065 | `POST /config/agent-models` | `ConfigController.getAgentModels` | Body:AgentModelsDTO | envelope <Object> | server-secret / — | DB-R; Redis-R/W(runtime/model/timbre cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(config,域级);差分— |
| J066 | `POST /config/correct-words` | `ConfigController.getCorrectWords` | Body:CorrectWordsDTO | envelope <Object> | server-secret / — | DB-R; Redis-R/W(runtime/model/timbre cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(config,域级);差分— |
| J067 | `POST /config/server-base` | `ConfigController.getConfig` | — | envelope <Object> | server-secret / — | DB-R; Redis-R/W(runtime/model/timbre cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(config,域级);差分✓3(缺失/错误/正确 secret) |
| J068 | `POST /correct-word/file` | `CorrectWordController.createFile` | Body:CorrectWordFileCreateDTO | envelope <CorrectWordFileVO> | DB Token / `sys:role:normal` | DB-W(file/items/mapping 事务) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(correctword,域级);差分✓2(响应/DB |
| J069 | `POST /correct-word/file/batch-delete` | `CorrectWordController.batchDeleteFiles` | Body:List<String> | envelope <Void> | DB Token / `sys:role:normal` | DB-W(file/items/mapping 事务) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(correctword,域级);差分— |
| J070 | `GET /correct-word/file/download/{fileId}` | `CorrectWordController.downloadFile` | Path:fileId | 流式/二进制 + 原下载 headers | DB Token / `sys:role:normal` | DB-R(content); 二进制 | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(correctword,域级);差分✓2(二进制/更新后下载) |
| J071 | `GET /correct-word/file/list` | `CorrectWordController.listFiles` | Query:params:Map<String, Object> | envelope <PageData<CorrectWordFileVO>> | DB Token / `sys:role:normal` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(correctword,域级);差分✓1 |
| J072 | `GET /correct-word/file/select` | `CorrectWordController.listAllFiles` | — | envelope <List<CorrectWordFileVO>> | DB Token / `sys:role:normal` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(correctword,域级);差分— |
| J073 | `DELETE /correct-word/file/{fileId}` | `CorrectWordController.deleteFile` | Path:fileId | envelope <Void> | DB Token / `sys:role:normal` | DB-W(file/items/mapping 事务) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(correctword,域级);差分✓1(级联副作用) |
| J074 | `PUT /correct-word/file/{fileId}` | `CorrectWordController.updateFile` | Path:fileId; Body:CorrectWordFileCreateDTO | envelope <Void> | DB Token / `sys:role:normal` | DB-W(file/items/mapping 事务) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(correctword,域级);差分✓2(响应/DB |
| J075 | `GET /datasets` | `KnowledgeBaseController.getPageList` | Query:name:String,page:Integer,page_size:Integer | envelope <PageData<KnowledgeBaseDTO>> | DB Token / `sys:role:normal` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(knowledge,域级);差分— |
| J076 | `POST /datasets` | `KnowledgeBaseController.save` | Body:KnowledgeBaseDTO | envelope <KnowledgeBaseDTO> | DB Token / `sys:role:normal` | DB-R/W; 外部-RAGFlow HTTP(upload/dataset/document/chunk/retrieval) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(knowledge,域级);差分— |
| J077 | `DELETE /datasets/batch` | `KnowledgeBaseController.deleteBatch` | Query:ids:String | envelope <Void> | DB Token / `sys:role:normal` | DB-R/W; 外部-RAGFlow HTTP(upload/dataset/document/chunk/retrieval) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(knowledge,域级);差分— |
| J078 | `GET /datasets/rag-models` | `KnowledgeBaseController.getRAGModels` | — | envelope <List<ModelConfigEntity>> | DB Token / `sys:role:normal` | DB-R(model config) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(knowledge,域级);差分— |
| J079 | `DELETE /datasets/{dataset_id}` | `KnowledgeBaseController.delete` | Path:dataset_id | envelope <Void> | DB Token / `sys:role:normal` | DB-R/W; 外部-RAGFlow HTTP(upload/dataset/document/chunk/retrieval) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(knowledge,域级);差分— |
| J080 | `GET /datasets/{dataset_id}` | `KnowledgeBaseController.getByDatasetId` | Path:dataset_id | envelope <KnowledgeBaseDTO> | DB Token / `sys:role:normal` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(knowledge,域级);差分— |
| J081 | `PUT /datasets/{dataset_id}` | `KnowledgeBaseController.update` | Path:dataset_id; Body:KnowledgeBaseDTO | envelope <KnowledgeBaseDTO> | DB Token / `sys:role:normal` | DB-R/W; 外部-RAGFlow HTTP(upload/dataset/document/chunk/retrieval) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(knowledge,域级);差分— |
| J082 | `POST /datasets/{dataset_id}/chunks` | `KnowledgeFilesController.parseDocuments` | Path:dataset_id; Body:Map<String, List<String>> | envelope <Void> | DB Token / `sys:role:normal` | DB-R/W; 外部-RAGFlow HTTP(upload/dataset/document/chunk/retrieval) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(knowledge,域级);差分— |
| J083 | `DELETE /datasets/{dataset_id}/documents` | `KnowledgeFilesController.delete` | Path:dataset_id; Body:DocumentDTO.BatchIdReq | envelope <Void> | DB Token / `sys:role:normal` | DB-R/W; 外部-RAGFlow HTTP(upload/dataset/document/chunk/retrieval) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(knowledge,域级);差分— |
| J084 | `GET /datasets/{dataset_id}/documents` | `KnowledgeFilesController.getPageList` | Path:dataset_id; Query:name:String,status:String,page:Integer,page_size:Integer | envelope <PageData<KnowledgeFilesDTO>> | DB Token / `sys:role:normal` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(knowledge,域级);差分— |
| J085 | `POST /datasets/{dataset_id}/documents` | `KnowledgeFilesController.uploadDocument` | Path:dataset_id; Query:name:String,chunkMethod:String,metaFields:String,parserConfig:String; Multipart:file | envelope <KnowledgeFilesDTO> | DB Token / `sys:role:normal` | DB-R/W; 外部-RAGFlow HTTP(upload/dataset/document/chunk/retrieval) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(knowledge,域级);差分— |
| J086 | `GET /datasets/{dataset_id}/documents/status/{status}` | `KnowledgeFilesController.getPageListByStatus` | Path:dataset_id,status; Query:page:Integer,page_size:Integer | envelope <PageData<KnowledgeFilesDTO>> | DB Token / `sys:role:normal` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(knowledge,域级);差分— |
| J087 | `DELETE /datasets/{dataset_id}/documents/{document_id}` | `KnowledgeFilesController.deleteSingle` | Path:dataset_id,document_id | envelope <Void> | DB Token / `sys:role:normal` | DB-R/W; 外部-RAGFlow HTTP(upload/dataset/document/chunk/retrieval) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(knowledge,域级);差分— |
| J088 | `GET /datasets/{dataset_id}/documents/{document_id}/chunks` | `KnowledgeFilesController.listChunks` | Path:dataset_id,document_id; Query:page:Integer,pageSize:Integer,keywords:String,id:String | envelope <ChunkDTO.ListVO> | DB Token / `sys:role:normal` | DB-R/W; 外部-RAGFlow HTTP(upload/dataset/document/chunk/retrieval) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(knowledge,域级);差分— |
| J089 | `POST /datasets/{dataset_id}/retrieval-test` | `KnowledgeFilesController.retrievalTest` | Path:dataset_id; Body:RetrievalDTO.TestReq | envelope <RetrievalDTO.ResultVO> | DB Token / `sys:role:normal` | DB-R/W; 外部-RAGFlow HTTP(upload/dataset/document/chunk/retrieval) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(knowledge,域级);差分— |
| J090 | `PUT /device/address-book/alias` | `DeviceController.updateAlias` | Body:DeviceAddressBookAliasDTO | envelope <Void> | DB Token / `sys:role:normal` | DB-W(device/bind/address-book); Redis-R/W | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分— |
| J091 | `GET /device/address-book/call` | `DeviceController.callByNickname` | Query:callerMac:String,nickname:String,answer:boolean | envelope <Map<String, Object>> | server-secret / — | DB-R; 外部-MQTT gateway HTTP; server-secret | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分— |
| J092 | `PUT /device/address-book/permission` | `DeviceController.updatePermission` | Body:DeviceAddressBookPermissionDTO | envelope <Void> | DB Token / `sys:role:normal` | DB-W(device/bind/address-book); Redis-R/W | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分— |
| J093 | `GET /device/address-book/{macAddress}` | `DeviceController.getAddressBook` | Path:macAddress | envelope <Object> | DB Token / `sys:role:normal` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分— |
| J094 | `GET /device/bind/{agentId}` | `DeviceController.getUserDevices` | Path:agentId | envelope <List<UserShowDeviceListVO>> | DB Token / `sys:role:normal` | DB-R; Redis-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分✓1 |
| J095 | `POST /device/bind/{agentId}` | `DeviceController.forwardToMqttGateway` | Path:agentId; Body:String | envelope <String> | DB Token / `sys:role:normal` | DB/Redis-R; 外部-MQTT gateway HTTP + daily auth | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分— |
| J096 | `POST /device/bind/{agentId}/{deviceCode}` | `DeviceController.bindDevice` | Path:agentId,deviceCode | envelope <Void> | DB Token / `sys:role:normal` | DB-W(device/bind/address-book); Redis-R/W | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分— |
| J097 | `POST /device/manual-add` | `DeviceController.manualAddDevice` | Body:DeviceManualAddDTO | envelope <Void> | DB Token / `sys:role:normal` | DB-W(device/bind/address-book); Redis-R/W | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分— |
| J098 | `POST /device/register` | `DeviceController.registerDevice` | Body:DeviceRegisterDTO | envelope <String> | DB Token / — | DB-W(device/bind/address-book); Redis-R/W | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分— |
| J099 | `POST /device/tools/call/{deviceId}` | `DeviceController.callDeviceTool` | Path:deviceId; Body:DeviceToolsCallReqDTO | envelope <Object> | DB Token / `sys:role:normal` | DB/Redis-R; 外部-MQTT gateway HTTP + daily auth | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分— |
| J100 | `POST /device/tools/list/{deviceId}` | `DeviceController.getDeviceTools` | Path:deviceId | envelope <Object> | DB Token / `sys:role:normal` | DB/Redis-R; 外部-MQTT gateway HTTP + daily auth | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分✓2(响应/外呼格式) |
| J101 | `POST /device/unbind` | `DeviceController.unbindDevice` | Body:DeviceUnBindDTO | envelope <Void> | DB Token / `sys:role:normal` | DB-W(device/bind/address-book); Redis-R/W | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分— |
| J102 | `PUT /device/update/{id}` | `DeviceController.updateDeviceInfo` | Path:id; Body:DeviceUpdateDTO | envelope <Void> | DB Token / `sys:role:normal` | DB-W(device/bind/address-book); Redis-R/W | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分✓3(上下界/UTF-16 长度) |
| J103 | `PUT /models/default/{id}` | `ModelController.setDefaultModel` | Path:id | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(model/config cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分— |
| J104 | `PUT /models/enable/{id}/{status}` | `ModelController.enableModelConfig` | Path:id,status | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(model/config cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分— |
| J105 | `GET /models/list` | `ModelController.getModelConfigList` | Query:modelType:String,modelName:String,page:String,limit:String | envelope <PageData<ModelConfigDTO>> | DB Token / `sys:role:superAdmin` | DB-R; Redis-R/W(model cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分— |
| J106 | `GET /models/llm/names` | `ModelController.getLlmModelCodeList` | Query:modelName:String | envelope <List<LlmModelBasicInfoDTO>> | DB Token / `sys:role:normal` | DB-R; Redis-R/W(model cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分— |
| J107 | `GET /models/names` | `ModelController.getModelNames` | Query:modelType:String,modelName:String | envelope <List<ModelBasicInfoDTO>> | DB Token / `sys:role:normal` | DB-R; Redis-R/W(model cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分— |
| J108 | `GET /models/provider` | `ModelProviderController.getListPage` | Query:modelProviderDTO:ModelProviderDTO,page:String,limit:String | envelope <PageData<ModelProviderDTO>> | DB Token / `sys:role:superAdmin` | DB-R; Redis-R/W(model cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分✓1 |
| J109 | `POST /models/provider` | `ModelProviderController.add` | Body:ModelProviderDTO | envelope <ModelProviderDTO> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(model/config cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分✓1(约束集合) |
| J110 | `PUT /models/provider` | `ModelProviderController.edit` | Body:ModelProviderDTO | envelope <ModelProviderDTO> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(model/config cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分— |
| J111 | `POST /models/provider/delete` | `ModelProviderController.delete` | Body:List<String> | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(model/config cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分— |
| J112 | `GET /models/provider/plugin/names` | `ModelProviderController.getPluginNameList` | — | envelope <List<ModelProviderDTO>> | DB Token / — | DB-R; Redis-R/W(model cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分— |
| J113 | `DELETE /models/{id}` | `ModelController.deleteModelConfig` | Path:id | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(model/config cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分— |
| J114 | `GET /models/{id}` | `ModelController.getModelConfig` | Path:id | envelope <ModelConfigDTO> | DB Token / `sys:role:superAdmin` | DB-R; Redis-R/W(model cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分— |
| J115 | `GET /models/{modelId}/voices` | `ModelController.getVoiceList` | Path:modelId; Query:voiceName:String | envelope <List<VoiceDTO>> | DB Token / `sys:role:normal` | DB-R; Redis-R/W(model cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分— |
| J116 | `GET /models/{modelType}/provideTypes` | `ModelController.getModelProviderList` | Path:modelType | envelope <List<ModelProviderDTO>> | DB Token / `sys:role:superAdmin` | DB-R; Redis-R/W(model cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分— |
| J117 | `POST /models/{modelType}/{provideCode}` | `ModelController.addModelConfig` | Path:modelType,provideCode; Body:ModelConfigBodyDTO | envelope <ModelConfigDTO> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(model/config cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分— |
| J118 | `PUT /models/{modelType}/{provideCode}/{id}` | `ModelController.editModelConfig` | Path:modelType,provideCode,id; Body:ModelConfigBodyDTO | envelope <ModelConfigDTO> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(model/config cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(model,域级);差分— |
| J119 | `GET /ota/` | `OTAController.getOTA` | — | 裸 text/plain | 匿名 / — | — | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分✓1MIME/body |
| J120 | `POST /ota/` | `OTAController.checkOTAVersion` | Header:Device-Id,Client-Id; Body:DeviceReportReqDTO | 裸 application/json | 匿名 / — | DB/Redis-R(设备/固件/配置); HMAC/Base64/时间戳凭证 | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分✓4(必填/格式/凭证/密码学) |
| J121 | `POST /ota/activate` | `OTAController.activateDevice` | Header:Device-Id,Client-Id | 裸 application/json | 匿名 / — | DB-R/W(device activation); Redis-R/W(TTL) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分✓3 |
| J122 | `GET /otaMag` | `OTAMagController.page` | Query:params:Map<String, Object> | envelope <PageData<OtaEntity>> | DB Token / `sys:role:superAdmin` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分— |
| J123 | `POST /otaMag` | `OTAMagController.save` | Body:OtaEntity | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W(OTA metadata) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分✓2(响应/DB |
| J124 | `GET /otaMag/download/{uuid}` | `OTAMagController.downloadFirmware` | Path:uuid | 流式/二进制 + 原下载 headers | 匿名 / — | Redis-R/W(一次性/次数); 文件-R/流式 | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分✓4(次数限制及二进制) |
| J125 | `GET /otaMag/getDownloadUrl/{id}` | `OTAMagController.getDownloadUrl` | Path:id | envelope <String> | DB Token / `sys:role:superAdmin` | DB-R; Redis-W(download token TTL) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分间接✓(供下载链路) |
| J126 | `POST /otaMag/upload` | `OTAMagController.uploadFirmware` | Multipart:file | envelope <String> | DB Token / `sys:role:superAdmin` | 文件-W(MD5/扩展名/大小) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分✓2(上传/扩展名错误) |
| J127 | `POST /otaMag/uploadAssetsBin` | `OTAMagController.uploadAssetsBin` | Multipart:file | envelope <String> | DB Token / `sys:role:normal` | 文件-W(MD5/扩展名/大小) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分— |
| J128 | `DELETE /otaMag/{id}` | `OTAMagController.delete` | Path:id | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W(OTA metadata); 文件-DEL | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分— |
| J129 | `GET /otaMag/{id}` | `OTAMagController.get` | Path:id | envelope <OtaEntity> | DB Token / `sys:role:superAdmin` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分— |
| J130 | `PUT /otaMag/{id}` | `OTAMagController.update` | Path:id; Body:OtaEntity | envelope <?> | DB Token / `sys:role:superAdmin` | DB-W(OTA metadata) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(device,域级);差分— |
| J131 | `GET /ttsVoice` | `TimbreController.page` | Query:params:Map<String, Object> | envelope <PageData<TimbreDetailsVO>> | DB Token / `sys:role:superAdmin` | DB-R; Redis-R/W(timbre cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(timbre,域级);差分— |
| J132 | `POST /ttsVoice` | `TimbreController.save` | Body:TimbreDataDTO | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(timbre/config cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(timbre,域级);差分— |
| J133 | `POST /ttsVoice/delete` | `TimbreController.delete` | Body:String[] | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(timbre/config cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(timbre,域级);差分— |
| J134 | `PUT /ttsVoice/{id}` | `TimbreController.update` | Path:id; Body:TimbreDataDTO | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W; Redis-DEL(timbre/config cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(timbre,域级);差分— |
| J135 | `GET /user/captcha` | `LoginController.captcha` | Query:uuid:String | image/gif 二进制 | 匿名 / — | Redis-W(captcha TTL); GIF | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(security,域级);差分— |
| J136 | `PUT /user/change-password` | `LoginController.changePassword` | Body:PasswordDTO | envelope <?> | DB Token / — | DB-W(user/token); Redis-R/DEL(SMS) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(security,域级);差分— |
| J137 | `GET /user/info` | `LoginController.info` | — | envelope <UserDetail> | DB Token / — | DB-R; Redis-R/W(cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(security,域级);差分✓9(七语言/过期 Token/Long |
| J138 | `POST /user/login` | `LoginController.login` | Body:LoginDTO | envelope <TokenDTO> | 匿名 / — | DB-R/W(token); Redis-R/DEL(captcha) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(security,域级);差分— |
| J139 | `GET /user/pub-config` | `LoginController.pubConfig` | — | envelope <Map<String, Object>> | 匿名 / — | DB-R; Redis-R/W(cache) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(security,域级);差分✓1 |
| J140 | `POST /user/register` | `LoginController.register` | Body:LoginDTO | envelope <Void> | 匿名 / — | DB-W(user/token); Redis-R/DEL(SMS) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(security,域级);差分— |
| J141 | `PUT /user/retrieve-password` | `LoginController.retrievePassword` | Body:RetrievePasswordDTO | envelope <?> | 匿名 / — | DB-W(user/token); Redis-R/DEL(SMS) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(security,域级);差分— |
| J142 | `POST /user/smsVerification` | `LoginController.smsVerification` | Body:SmsVerificationDTO | envelope <Void> | 匿名 / — | Redis-R/W(TTL/频控); 外部-Aliyun SMS | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(security,域级);差分— |
| J143 | `GET /voiceClone` | `VoiceCloneController.page` | Query:params:Map<String, Object> | envelope <PageData<VoiceCloneResponseDTO>> | DB Token / `sys:role:normal` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(voiceclone,域级);差分— |
| J144 | `POST /voiceClone/audio/{id}` | `VoiceCloneController.getAudioId` | Path:id | envelope <String> | DB Token / `sys:role:normal` | DB-R; Redis-W(one-shot URL) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(voiceclone,域级);差分— |
| J145 | `POST /voiceClone/cloneAudio` | `VoiceCloneController.cloneAudio` | Body:Map<String, String> | envelope <String> | DB Token / `sys:role:normal` | DB-R/W(train state); 文件-W; 外部-火山语音克隆 HTTP | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(voiceclone,域级);差分— |
| J146 | `GET /voiceClone/play/{uuid}` | `VoiceCloneController.playVoice` | Path:uuid | 流式/二进制 + 原下载 headers | 匿名 / — | Redis-R/DEL(one-shot); 文件/外部音频-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(voiceclone,域级);差分— |
| J147 | `POST /voiceClone/updateName` | `VoiceCloneController.updateName` | Body:Map<String, String> | envelope <String> | DB Token / `sys:role:normal` | DB-W(train record name) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(voiceclone,域级);差分— |
| J148 | `POST /voiceClone/upload` | `VoiceCloneController.uploadVoice` | Query:id:String; Multipart:voiceFile | envelope <String> | DB Token / `sys:role:normal` | DB-R/W(train state); 文件-W; 外部-火山语音克隆 HTTP | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(voiceclone,域级);差分— |
| J149 | `GET /voiceResource` | `VoiceResourceController.page` | Query:params:Map<String, Object> | envelope <PageData<VoiceCloneResponseDTO>> | DB Token / `sys:role:superAdmin` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(voiceclone,域级);差分— |
| J150 | `POST /voiceResource` | `VoiceResourceController.save` | Body:VoiceCloneDTO | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W(voice resource) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(voiceclone,域级);差分— |
| J151 | `GET /voiceResource/ttsPlatforms` | `VoiceResourceController.getTtsPlatformList` | — | envelope <List<Map<String, Object>>> | DB Token / `sys:role:superAdmin` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(voiceclone,域级);差分— |
| J152 | `GET /voiceResource/user/{userId}` | `VoiceResourceController.getByUserId` | Path:userId | envelope <List<VoiceCloneResponseDTO>> | DB Token / `sys:role:normal` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(voiceclone,域级);差分— |
| J153 | `DELETE /voiceResource/{id}` | `VoiceResourceController.delete` | Path:id | envelope <Void> | DB Token / `sys:role:superAdmin` | DB-W(voice resource) | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(voiceclone,域级);差分— |
| J154 | `GET /voiceResource/{id}` | `VoiceResourceController.get` | Path:id | envelope <VoiceCloneResponseDTO> | DB Token / `sys:role:superAdmin` | DB-R | 结构✓;请求面差分✓1;认证业务面差分✓1;领域✓(voiceclone,域级);差分— |
## 已观测差异与未覆盖面
- 154 条安全请求面差分最终全部一致。首轮曾发现 5 个空 Body 映射差异;修复 FastAPI 对
Spring `HttpMessageNotReadableException` 的 code-500 语义后,重新从零执行才得到 154/154。
- 154 条认证业务面差分最终全部一致;该轮使用有效鉴权与安全业务/校验输入,在不主动成功
写入的前提下逐路由对照。证据是 `authenticated-route-results.json`,渲染器会在结果不是
154/154、存在失败或跳过时硬失败。
- 2026-07-20 的隔离差分报告未在 49 个 checks 中观测到响应/所选 headers/数据库副作用
不一致;证据是 `main/manager-api-fastapi/compatibility/contract-results.json`,不是人工推断。
- Hibernate Validator 的 `ConstraintViolation Set` 首条消息无稳定顺序;模型 provider 必填
用例比较“消息属于 Java 声明约束集合”与相同错误码,而不伪造一个固定顺序。
- OTA 时间戳/token 是动态值,差分先比较归一化结构,再分别校验两端 HMAC/Base64 密码学
有效性;这属于有意的测试归一化,不是声称字节恒等。
- 深度差分未直接命中的 133 条中,J125 是下载链路间接覆盖,另 132 条标为 `差分—`
它们有请求面、认证业务面与所属领域测试,但尚无逐路由成功+主要错误+副作用深度对照,不能据此宣称
每一种业务状态均已逐接口行为等价。
- FastAPI 额外提供上述 3 条消费者兼容路由与 live/ready 健康检查;它们没有 Java
Controller 基线,属于明确、可回退的加法差异。
- Java 把定时任务放在 Spring 进程;FastAPI 使用独立 jobs 进程和 Redis 分布式锁。这是
部署拓扑差异,业务状态和幂等目标保持一致。
- RAGFlow、阿里云短信、火山语音克隆、真实声纹、真实 LLM、真实 MQTT/MCP/WS 均未用
生产凭证联调;自动化只证明 mock 请求格式、超时/错误映射/重试中的已覆盖场景。
## 可复现检查
```bash
cd main/manager-api-fastapi
.venv/bin/python scripts/extract_java_routes.py --output compatibility/java-routes.json
.venv/bin/python scripts/extract_consumer_routes.py > /tmp/consumer-routes.json
.venv/bin/pytest -q tests/test_java_route_manifest.py tests/test_consumer_route_manifest.py tests/test_compatibility_document.py
```
逐接口差分的启动、隔离库、mock 与执行命令见 `docs/manager-api-fastapi-test-report.md`
本文件只陈述已落盘的结果,不把缺少真实密钥的外部联调列为通过。
+290
View File
@@ -0,0 +1,290 @@
# manager-api 到 FastAPI 迁移说明
## 目标与边界
`main/manager-api-fastapi``main/manager-api` 的兼容替代实现。迁移只替换管理 API
进程,不改变 MySQL 表、Liquibase 历史、Redis 业务语义,也不要求 manager-web、
manager-mobile 或 xiaozhi-server 修改现有 URL。Java 实现继续保留,作为行为基线和
回滚实现。
本次迁移不采用双写。灰度期间,每个业务域在任一时刻只有一个写入方;读流量可以按
请求切分,写流量必须按业务域整体切换。
## 架构
```mermaid
flowchart LR
C["Web / Mobile / xiaozhi-server / Device"] --> N["Nginx /xiaozhi"]
N --> A["FastAPI API workers"]
A --> R["Routers"]
R --> S["Services / transaction boundaries"]
S --> P["Repositories / SQLAlchemy async"]
S --> I["External integration clients"]
P --> M[("Existing MySQL schema")]
S --> D[("Redis Java compatibility layer")]
J["Standalone jobs worker"] --> S
L["Original Liquibase changelog"] --> X["Migration runner"]
X --> M
```
代码按职责分为:
- `app/routers`URL、HTTP 方法、Header/Query/Body 绑定和响应形态。
- `app/services`:权限后的业务规则、事务边界、跨表级联和外部调用编排。
- `app/repositories`:现有 MySQL 表上的参数化 SQL、行锁和分页。
- `app/schemas`:兼容现有字段别名及请求模型。
- `app/integrations`LLM、MQTT gateway、MCP、RAGFlow、语音克隆和声纹客户端。
- `app/jobs`:独立的定时任务进程;API worker 不启动定时任务。
- `app/core`:数据库 Token 认证、国际化、Java JSON/Long 兼容、SM2、Redis 编解码、
Snowflake ID 和健康检查。
所有业务路由保留 `/xiaozhi` 前缀。普通 API 返回 `{code,msg,data}`;认证过滤器、
业务异常和请求校验均由兼容处理器转换,不把 FastAPI 默认的 401、403 或 422 直接
暴露给客户端。OTA、播放和文件下载等原本返回裸 JSON、文本或二进制的接口继续保持
其原始响应类型。
## 数据库兼容策略
### Schema 与迁移
原目录 `main/manager-api/src/main/resources/db/changelog/` 仍是唯一 schema source of
truth。不得把既有 changeset 改写成 Alembic,也不得修改已应用 changeset 的 ID、作者或
校验和。
Python 部署必须先运行独立迁移镜像或 `scripts/run-migrations.sh`。迁移 runner 直接打包
原 Java Liquibase 资源,因此与保留的 Java 服务使用相同的 `DATABASECHANGELOG` 历史。
宿主机执行脚本时,`JAVA_RESOURCES_DIR` 必须指向仓库中的
`main/manager-api/src/main/resources`;不要使用只存在于应用镜像内的
`/opt/xiaozhi/java-resources`。本机隔离数据库的完整命令见“配置与启动 / 本地”。
`docker compose``manager-api-fastapi``manager-api-jobs` 都等待
`manager-api-migrate` 成功退出,避免应用在未完成迁移时接流量。
### 事务、锁与 ID
- Service 层在一次事务中完成智能体创建/更新/删除、插件/标签/纠错词映射、设备解绑、
快照恢复及其他多表操作;异常时显式回滚。
- 智能体更新和恢复先对 `ai_agent` 执行 `SELECT ... FOR UPDATE`,序列化同一智能体的
快照版本分配和状态令牌校验。
- 快照版本仍使用 `(agent_id, version_no)` 唯一约束及同一事务内的 `MAX+1` 插入;行锁
防止并发恢复或更新竞争。
- 需要 Long 主键的管理表继续使用与 Java epoch/node/sequence 布局一致的 Snowflake
生成器;原本使用 32 位 UUID 的业务表仍使用无连字符 UUID。
- 不新增数据库外键,也不改表、索引、字符集或 MySQL 类型。
### Java/Python 并存
并存期必须给写请求建立确定的域路由,例如 `agent/*` 全部指向一个实现,不能把同一域
中的增删改请求在 Java 与 Python 之间随机分配。建议域切换顺序为只读配置、系统管理、
模型/音色、设备、智能体、知识库和外部集成。域回切前先停止该域的新写入并等待在途
请求完成。
## Redis 兼容策略
默认继续使用 Java 已有 key 名称和 TTL。`app/core/redis.py` 实现 Spring Data
`RedisSerializer.json()` 使用的 Jackson wire format,包括:
- Map 的 `@class`、List/Set 的 wrapper-array
- Object 槽位中 Long 的 `java.lang.Long` 包装;
- `java.util.Date` 的 epoch 毫秒包装;
- Java DTO/Entity 缓存所需的具体类名和字段类型;
- Hash 缓存写入后的 86400 秒默认 TTL。
该兼容层使 Java 回滚进程可以读取 FastAPI 写入的缓存。应用启动和正常测试不会执行
`FLUSHALL`;隔离测试脚本的 `reset` 只操作其自建 Redis 实例。定时任务使用 Redis
分布式锁和自动续租 watchdog,即使启动多个 jobs 容器,同一个任务也只有一个执行者。
## 安全与协议兼容
- 用户 Token 仍保存在 `sys_user_token`,按数据库过期时间校验;没有替换为 JWT。
- 登录密文继续使用 SM2 C1C3C2,黄金向量由 Java 和 Python 双向解密测试校验。
- `/config/*`、聊天记录上报/摘要/标题及地址簿内部接口继续校验数据库中的
`server.secret` Bearer 值。
- OTA、WebSocket 和 MQTT 保留原 HMAC、Base64、时间戳、Client-Id/Device-Id 及 token
格式。
- 密钥只通过环境变量或原参数表提供;`.env.example` 和部署文档不包含真实凭证。
## 配置与启动
### 本地
`.env.example` 是容器 Compose 模板,其中的 `mysql``redis` 是 Compose 服务名,
`/opt/xiaozhi/java-resources` 是镜像内路径,不能原样复制后用于宿主机进程。下面的流程
显式使用 `127.0.0.1:13316` 上的隔离 MySQL、`127.0.0.1:16379` 上的隔离 Redis,以及
仓库原始 Liquibase/i18n 资源;不会连接或迁移现有开发数据库:
```bash
cd main/manager-api-fastapi
uv sync --locked
./scripts/isolated-env.sh start
LIQUIBASE_URL='jdbc:mysql://127.0.0.1:13316/manager_fastapi_test?useUnicode=true&characterEncoding=UTF-8&serverTimezone=Asia/Shanghai&allowMultiQueries=true' \
LIQUIBASE_USERNAME='xiaozhi_test' \
LIQUIBASE_PASSWORD='isolated-test-only' \
JAVA_RESOURCES_DIR="$PWD/../manager-api/src/main/resources" \
MAVEN_BIN="$PWD/../../.runtime/maven/bin/mvn" \
MAVEN_LOCAL_REPOSITORY="$PWD/../../.runtime/m2" \
JAVA_HOME="$PWD/../../.runtime/jdk" \
./scripts/run-migrations.sh
eval "$(./scripts/isolated-env.sh env)"
export APP_ENVIRONMENT=development
export APP_DATABASE_URL="$TEST_FASTAPI_DATABASE_URL"
export APP_REDIS_URL="$TEST_FASTAPI_REDIS_URL"
export APP_JAVA_RESOURCES_DIR="$PWD/../manager-api/src/main/resources"
export APP_UPLOAD_DIR="$PWD/.test-runtime/local-uploadfile"
mkdir -p "$APP_UPLOAD_DIR"
./scripts/start-api.sh
```
上述 `eval` 会得到明确的 localhost URLFastAPI 测试库和 Redis DB 2)。定时任务必须
作为单独进程启动;新终端需要重复 `eval` 及四个 `APP_*` 路径/URL 导出,不能让 jobs
进程落回 `.env.example` 的 Docker DNS
```bash
cd main/manager-api-fastapi
eval "$(./scripts/isolated-env.sh env)"
export APP_ENVIRONMENT=development
export APP_DATABASE_URL="$TEST_FASTAPI_DATABASE_URL"
export APP_REDIS_URL="$TEST_FASTAPI_REDIS_URL"
export APP_JAVA_RESOURCES_DIR="$PWD/../manager-api/src/main/resources"
export APP_UPLOAD_DIR="$PWD/.test-runtime/local-uploadfile"
./scripts/start-jobs.sh
```
API 默认监听 `0.0.0.0:8002`,兼容根路径为
`http://127.0.0.1:8002/xiaozhi``APP_WORKERS` 可以大于 1;任务不会随 API worker
复制。验证结束后运行 `./scripts/isolated-env.sh stop`。生产环境不得设置
`APP_ALLOW_START_WITHOUT_DEPENDENCIES=true`
仓库统一启动脚本继承当前 shell 的上述环境变量;FastAPI 是默认实现,保留的 Java
实现可直接用于本地回滚:
```bash
cd "$(git rev-parse --show-toplevel)"
scripts/restart-local-services.sh --manager-api fastapi --wait 180
scripts/restart-local-services.sh --manager-api java --wait 180
```
### 容器
下面的 `mysql``redis` 只在容器网络确实提供对应 DNS 名时有效,否则必须替换为该网络
可访问的真实主机名。API 镜像内的 Java 资源路径是 `/opt/xiaozhi/java-resources`,迁移
镜像则把同一仓库资源打包到 `/migration/java-resources`。以下示例中的凭证必须替换,
并应通过部署平台的 secret 注入而不是提交到仓库:
```bash
cd main/manager-api-fastapi
export LIQUIBASE_URL='jdbc:mysql://mysql:3306/xiaozhi_esp32_server?serverTimezone=Asia/Shanghai'
export MYSQL_USER='xiaozhi'
export MYSQL_PASSWORD='replace-me'
export FASTAPI_DATABASE_URL='mysql+asyncmy://xiaozhi:replace-me@mysql:3306/xiaozhi_esp32_server?charset=utf8mb4'
export REDIS_URL='redis://redis:6379/0'
export MANAGER_API_UPSTREAM='manager-api-fastapi:8002'
docker compose build
docker compose up -d
```
`MANAGER_API_UPLOAD_SOURCE` 指向保留 Java 服务的宿主机上传目录,必须在启动前让 Java
运行用户与容器 UID 10001 都具备读写和目录遍历权限;不要盲目 `chown -R` 导致 Java 失去
访问权,应使用部署环境的共享组或 ACL。空 named volume 在 Docker 通常会继承镜像中
`/data/uploads` 的 UID,但并非所有 OCI runtime 都实现相同 copy-up 语义;必须以容器内
UID 10001 做一次写入预检。`/xiaozhi/health/ready` 同时检查上传目录可写性,权限不正确时
返回 HTTP 503 和 `data.uploads=false`,不得绕过该检查接入流量。Apple Container 的一次性
卷初始化实测命令和结果记录在测试报告中。
`MANAGER_API_UPSTREAM` 默认值是 `manager-api-fastapi:8002`。修改变量后必须重建 Nginx
容器才能重新渲染配置;切到 FastAPI 和整服务回滚到 Java 的命令分别为:
```bash
MANAGER_API_UPSTREAM='manager-api-fastapi:8002' \
docker compose up -d --no-deps --force-recreate manager-api-nginx
MANAGER_API_UPSTREAM='<Nginx 容器可访问的 Java 主机名或 IP>:8002' \
docker compose up -d --no-deps --force-recreate manager-api-nginx
```
Java 地址必须能从 Nginx 容器网络解析和访问。内置 Nginx 的这个变量会切换整个
`/xiaozhi`;逐业务域灰度需要在上层网关按路径配置两个 upstream,仍须遵守“同一业务域
只有一个写入方”。
API 镜像(包括 jobs 命令)和迁移镜像都声明 `USER 10001:10001`,以非 root 用户运行。
Compose 还为 API/jobs 设置只读根文件系统和 `/tmp` tmpfs`/data/uploads` 是它们唯一的
持久写目录,并通过 `/app/uploadfile` 符号链接兼容数据库中的 Java 相对路径。迁移容器是
一次性非 root 进程,但 Compose 没有把它标为只读根文件系统。Nginx 镜像没有声明
非 root `USER`,不能将其描述为非 root 镜像;它通过 Compose 的 `read_only: true`
`/var/cache/nginx``/var/run``/tmp` 三个 tmpfs 加固。Nginx 保留 `/xiaozhi/` 路径,
关闭上传请求缓冲并设置 100 MiB 上限。健康检查分为:
- `/xiaozhi/health/live`:进程存活;
- `/xiaozhi/health/ready`MySQL、Redis 和上传目录写权限都可用才返回 HTTP 200,否则
HTTP 503`data.database``data.redis``data.uploads` 可直接定位失败项。
SIGTERM 触发 Uvicorn 优雅关闭;compose 给 API 40 秒清理在途请求和连接池。
### 容器验证边界
本次仓库内的实际容器运行验证使用 Apple Container 1.0.0,覆盖镜像构建、隔离 MySQL
迁移、双 API worker、独立 jobs、Nginx 路由、只读文件系统、上传卷和 SIGTERM 优雅
关闭。`docker-compose.yml` 已通过 YAML 解析和自动化部署断言,但当前验证主机没有执行
`docker compose up`,因此不能把 Apple Container 的运行结果表述为 Docker Compose
端到端通过。镜像摘要、实际命令和运行结果见
`docs/manager-api-fastapi-test-report.md`;在目标 Docker/Compose 环境切流前,仍需按上面
命令执行一次迁移、ready 检查和 Nginx upstream 冒烟测试。
## 灰度切流
1. 备份当前配置并确认 Java 基线健康;不要停止 Java。
2. 对目标数据库执行原 Liquibase runner,确认 changeset 数量和校验和无差异。
3. 启动 FastAPI API 但暂不接写流量,检查 live/ready、日志和外部 mock。Python jobs
只在隔离环境验证后停止,生产中暂不持续运行,避免与 Java 调度器同时写入。
4. 先镜像或回放脱敏的只读请求,比较状态、Body、Header、数据库读结果和缓存读取。
5. 按业务域把只读流量从 1% 提升到 10%、50%、100%,监控错误码、P95、数据库连接、
Redis 命中与外部服务错误。
6. 对一个完整业务域建立维护窗口,停止该域 Java 新写入,等待在途事务结束,然后把该域
写路由切到 FastAPI。记录切换时间和最后写入方。
7. 逐域重复;稳定观察期内保留 Java 镜像、配置和回切路由。
8. 所有域稳定后才把 jobs 所有权切给 Python;Java 的定时任务进程必须同时停用,避免
两套调度器并行。
## 回滚
1. 冻结待回滚业务域的新写入,等待 FastAPI 在途请求和 jobs 当前轮次结束。
2. 停止 Python jobs,确认 Redis 分布式锁已释放;不要清空 Redis。
3. 将该域的 Nginx upstream 切回原 Java `manager-api`,保持 `/xiaozhi` 路径不变。
4. 用 Java 健康检查和代表性读请求确认 Token、缓存、上传文件与数据库数据可读。
5. 恢复 Java 写流量并记录回滚边界;不要让 FastAPI 继续写该域。
6. 若问题来自新 changeset,只能新增一个经评审的 Liquibase 前向修复;不得删除或改写
`DATABASECHANGELOG` 历史。
容器整服务回滚时,先把 `<JAVA_UPSTREAM>` 替换为 Nginx 容器可访问的真实地址,再只
重建代理;`--no-deps` 可避免回滚命令意外重启 FastAPI 或重复执行迁移:
```bash
MANAGER_API_UPSTREAM='<JAVA_UPSTREAM>:8002' \
docker compose up -d --no-deps --force-recreate manager-api-nginx
```
FastAPI 没有改变现有 schema,且 Redis 写入采用 Java 兼容格式,因此正常应用回滚不需要
数据反向迁移。若外部系统已接收不可撤销操作,按对应供应商的业务补偿流程处理,不能用
数据库回滚伪造外部成功或失败。
## 隔离验证
`scripts/isolated-env.sh` 只创建 `manager_java_test``manager_fastapi_test` 两个测试库和
端口 `16379` 上的独立 Redis;其中的测试密码仅用于本机隔离环境。标准流程是:
```bash
cd main/manager-api-fastapi
./scripts/isolated-env.sh start
./scripts/isolated-env.sh reset
./scripts/isolated-env.sh migrate
eval "$(./scripts/isolated-env.sh env)"
.venv/bin/pytest -m integration -q
./scripts/isolated-env.sh stop
```
实际执行结果、差分用例和不能使用真实凭证完成的联调项记录在
`docs/manager-api-fastapi-test-report.md`;逐接口状态记录在
`docs/manager-api-fastapi-compatibility.md`
+582
View File
@@ -0,0 +1,582 @@
# manager-api FastAPI 迁移测试报告
> 执行日期:2026-07-20Asia/Shanghai
>
> 工作目录:`/Users/mie/Desktop/Repo/xiaozhi-esp32-server`
>
> FastAPI 目标:`main/manager-api-fastapi`
> Java 基线:`main/manager-api`
本报告只记录实际执行并有输出或落盘证据的检查。结构路由闭合、154 条未认证/非法请求面
差分、154 条已认证安全业务/校验差分、领域测试和 49 条深度 Java/FastAPI 差分是不同强度
的证据,不互相替代。生产外部服务和真实硬件没有验证的部分,均不会写成通过。
## 1. 结果摘要
| 检查项 | 通过 | 失败/错误 | 跳过 | 结论 |
|---|---:|---:|---:|---|
| Java 基线测试 | 98 | 0 | 0 | 最终复跑 `BUILD SUCCESS`15.602 秒 |
| FastAPI 全量 pytest(最终回归) | 139 | 0 | 0 | 12.75 秒;含上传卷 readiness 与证据脱敏回归 |
| 隔离 MySQL/Redis 集成测试复跑 | 7 | 0 | 0 | 事务、锁、TTL、job 单实例与 watchdog 全绿 |
| Java→FastAPI 未认证/非法请求面差分 | 154 | 0 | 0 | 每条 Java 路由各 1 个无成功写入的缺认证或非法请求,逐项比较 status/body/Content-Type |
| Java→FastAPI 已认证安全业务/校验差分 | 154 | 0 | 0 | 每条 Java 路由各 1 个带正确认证的安全业务或校验请求;runner 有意不执行成功写入 |
| Java→FastAPI 深度差分契约 | 49 | 0 | 0 | 成功、主要错误与数据库副作用;直接覆盖 21/154 条 Java 路由 |
| 简单性能测试 | 480 请求 | 0 请求错误 | 不适用 | 4 场景 × 2 服务 × 60 次计量请求 |
| Java 路由结构清单 | 154/154 | 0 | 0 | FastAPI 注册闭合;结构证据,不等同逐接口行为证据 |
| 三端消费者调用点 | 188/188 | 0 | 0 | Web 134、Mobile 46、xiaozhi-server 8 个调用点均可解析 |
| FastAPI Ruff | 通过 | 0 | 不适用 | `app tests scripts` |
| FastAPI mypy | 70 个源文件 | 0 | 不适用 | strict 配置下通过 |
| FastAPI compileall | 通过 | 0 | 不适用 | `app tests scripts` |
| 锁文件与依赖同步 | 通过 | 0 | 不适用 | locked 环境共 55 packages |
| Python sdist/wheel | 2 个产物 | 0 | 不适用 | 冷环境完整依赖安装后可导入,路由数 163 |
| manager-web | i18n、5 unit、13 snapshot、build 全过 | 0 | 0 | build 有 4 条既有 size/precache warning |
| manager-mobile | type、lint、14 snapshot、mp-weixin build 全过 | 0 | 0 | 使用仓库声明的 pnpm 10.10.0 |
| xiaozhi-server | compileall 通过 | 0 | 不适用 | 除性能脚本外没有自动单测;8 个调用由 consumer 契约检查 |
| 真实付费/生产外部服务 | 0 | 不适用 | 不适用 | 无真实凭证,不声称联调通过 |
| 容器迁移、API、jobs 与 Nginx | 通过 | 0 | 0 | Apple Container 1.0.0 实际 build/runCompose 仅做静态验证,未伪装为 `docker compose up` |
最终可重复执行的全量测试、隔离差分、集成、构建和容器运行验证均为绿色。以下范围限制必须
与绿色测试分开陈述:
1. 全部 154 条 Java 路由均执行了两次差分:一次未认证/非法请求,一次已认证安全业务/校验
请求。第二个 runner 为保护隔离 fixture,有意不执行成功写入;49 个深度 checks 直接命中
21 条路由并覆盖代表性成功、错误和数据库副作用。因此两层全路由差分仍不等同于每条路由
的完整成功写入生命周期和全部错误路径差分。
2. 没有真实凭证、生产网络或硬件的外部集成未被计入通过。
## 2. 验证环境
| 组件 | 实际版本/配置 |
|---|---|
| 主机 | macOS 27.0arm64 |
| Java | Oracle JDK 21.0.11 LTS |
| Maven | 3.9.9,使用仓库 `.runtime/m2` |
| Python | 3.10.20`main/manager-api-fastapi/.venv` |
| MySQL | Community Server 8.0.46;隔离端口 `13316` |
| Redis | 8.8.0;隔离端口 `16379` |
| Node.js | v24.18.0 |
| npm | 11.16.0 |
| manager-mobile pnpm | Corepack 解析的 10.10.0 |
| OCI runtime | Apple Container 1.0.0;本机没有可用 Docker/Podman daemon |
| 容器架构 | linux/arm64 |
| 时区 | Asia/Shanghai |
隔离测试只重置 `manager_java_test``manager_fastapi_test` 两个测试 schema 和端口
`16379` 上的专用 Redis;没有连接、修改或清空开发 MySQL/Redis。Java 差分使用 Redis DB 1
FastAPI 使用 DB 2Java 单元验证显式使用 DB 3。
## 3. 实际执行命令
### 3.1 Java 基线
```bash
cd main/manager-api && \
JAVA_HOME=../../.runtime/jdk \
PATH="../../.runtime/jdk/bin:../../.runtime/maven/bin:$PATH" \
../../.runtime/maven/bin/mvn -o \
-Dmaven.repo.local=../../.runtime/m2 \
-Dspring.datasource.druid.url='jdbc:mysql://127.0.0.1:13316/manager_java_test?useUnicode=true&characterEncoding=UTF-8&serverTimezone=Asia/Shanghai&nullCatalogMeansCurrent=true&allowMultiQueries=true' \
-Dspring.datasource.druid.username=xiaozhi_test \
-Dspring.datasource.druid.password=isolated-test-only \
-Dspring.data.redis.host=127.0.0.1 \
-Dspring.data.redis.port=16379 \
-Dspring.data.redis.database=3 \
-Dspring.data.redis.password= \
-DskipTests=false test
```
最终复跑结果:98 tests0 failures0 errors0 skipped`BUILD SUCCESS`15.602 秒。Surefire
XML 位于 `main/manager-api/target/surefire-reports/`,各 suite 的 tests 合计为 98。
### 3.2 FastAPI 全量测试
```bash
cd main/manager-api-fastapi && \
eval "$(./scripts/isolated-env.sh env)" && \
APP_DATABASE_URL="$TEST_FASTAPI_DATABASE_URL" \
APP_REDIS_URL="$TEST_FASTAPI_REDIS_URL" \
APP_ENVIRONMENT=test \
.venv/bin/pytest -q
```
最终结果:139 passed、0 failed、0 skipped12.75 秒。
### 3.3 隔离集成、两层全路由差分、深度差分和性能测试
```bash
cd main/manager-api-fastapi && ./scripts/run-isolated-contract-tests.sh
```
最终执行结果为 exit 0。脚本实际完成以下阶段:
- 启动并重置隔离 MySQL `13316`、Redis `16379`
- 对 Java 与 FastAPI 两个 schema 分别执行原 Liquibase 101 个 changeSets
- 启动 Java 基线 `18082`、FastAPI `18083`、确定性外部 mock `18084`
- 生成并镜像固定用户、DB Token、Long ID、设备、agent、模型、纠错词和 OTA fixture
- 执行 7 条隔离集成测试;
- 对 154 条 Java 路由各执行一条未认证或非法的安全请求面差分,不产生成功写入;
- 对 154 条 Java 路由各执行一条带正确认证的安全业务或校验差分,仍不产生成功写入;
- 执行 49 条 Java/FastAPI 差分检查并写入 JSON
- 执行 480 次计量性能请求并写入 JSON;
- 对 FastAPI/mock 日志执行 warning、traceback、error 门禁;
- 退出时关闭 Java、FastAPI 和 mock,最终 `18082``18084` 没有监听进程。
最终阶段摘要(两行 154 分别对应未认证/非法和已认证安全业务/校验):
```text
7 passed
{"total": 154, "passed": 154, "failed": 0, "skipped": 0}
{"total": 154, "passed": 154, "failed": 0, "skipped": 0}
{"total": 49, "passed": 49, "failed": 0, "skipped": 0}
{"measurements": 8, "requests_measured": 480, "errors": 0}
Isolated integration, two 154-route surfaces, deep differential, and performance tests passed.
```
机器结果:
- `main/manager-api-fastapi/compatibility/route-surface-results.json`
- 生成时间:`2026-07-20T07:12:28.099864+00:00`
- 154 passed、0 failed、0 skipped
- `main/manager-api-fastapi/compatibility/authenticated-route-results.json`
- 生成时间:`2026-07-20T07:12:30.818073+00:00`
- 154 passed、0 failed、0 skipped
- `main/manager-api-fastapi/compatibility/contract-results.json`
- 生成时间:`2026-07-20T07:12:31.520810+00:00`
- 49 passed、0 failed、0 skipped
- `main/manager-api-fastapi/compatibility/performance-results.json`
- 生成时间:`2026-07-20T07:12:33.158826+00:00`
- 480 requests、0 errors
### 3.4 FastAPI 静态、依赖和构建验证
```bash
cd main/manager-api-fastapi
.venv/bin/ruff check app tests scripts
.venv/bin/mypy app
.venv/bin/python -m compileall -q app tests scripts
uv lock --check && uv sync --locked
uv build --no-cache
```
结果:
- Ruff 通过;
- mypy70 个源文件无问题;
- compileallexit 0
- lock check 与 locked syncexit 0,共 55 packages
- 无缓存构建 sdist 与 wheel 均成功。
构建产物另在临时冷虚拟环境验证。首次执行 `uv pip install --python
<临时环境>/bin/python --no-deps <wheel>` 后直接 import,因刻意没有安装 FastAPI 等运行依赖而
失败;这暴露的是冷 wheel 检查命令不完整,不是把失败隐藏为通过。随后执行带依赖的安装:
```bash
uv pip install --python <临时环境>/bin/python <wheel>
```
共安装 39 个锁定依赖。从仓库外 `/tmp` 导入成功,输出
`xiaozhi-manager-api 163`,证明不是依赖当前工作目录导入源码。
### 3.5 manager-web
```bash
cd main/manager-web && \
npm run check:i18n && \
npm run test:unit && \
npm run test:snapshot && \
npm run build
```
结果:
- i18n6 个 locale,每个 1527 keyskey 结构一致;
- unit5/5
- snapshot13/13
- Vue 生产构建 exit 0hash `71b64d002eb434ee`1069 ms);
- 输出有 4 条既有 bundle size/precache warning,没有将 warning 写成失败,也没有删除或放宽
测试来取得绿色结果;另有 `caniuse-lite` 数据过期 17 个月提示,未擅自更新依赖。
### 3.6 manager-mobile
```bash
cd main/manager-mobile && \
corepack pnpm type-check && \
corepack pnpm lint && \
corepack pnpm test:snapshot && \
corepack pnpm build:mp
```
结果:type-check exit 0、lint exit 0、snapshot 14/14、mp-weixin build exit 0。构建提示
`caniuse-lite` 数据过期 20 个月及 uni-app 有新版本;两项均不影响退出码,未擅自更新依赖。
一次较早的失败尝试直接调用普通 `pnpm`,环境解析到 v11,并因非 TTY 下依赖目录清理提示而
终止;没有把该次尝试写成通过。最终命令显式使用 Corepack,解析到仓库声明的 pnpm 10.10.0。
### 3.7 xiaozhi-server
```bash
cd main/xiaozhi-server && \
../manager-api-fastapi/.venv/bin/python --version && \
PYTHONPYCACHEPREFIX=/tmp/xiaozhi-server-pycache \
../manager-api-fastapi/.venv/bin/python -m compileall -q . && \
rg --files -g '*test*.py' -g '!performance_tester/**'
```
结果:Python 3.10.20compileall exit 0。排除性能测试目录后没有自动单测文件,因此没有虚构
pytest 通过数量;`xiaozhi-server` 的 8 个 manager-api 调用点由 consumer manifest 和 FastAPI
兼容测试验证为可解析。
首次误用不存在的 `../../.runtime/python/bin/python3.10`,结果为 exit 127;改用上面实际存在的
FastAPI Python 3.10.20 后通过。
## 4. 两层全路由差分与深度差分覆盖
### 4.1 154 条未认证/非法安全请求面差分
`tests/compatibility/route_surface_runner.py` 从 Java route manifest 逐条构造不发生成功写入的请求:
133 条 DB Token 路由省略 Token,14 条匿名路由发送安全非法输入,7 条内部路由省略
server-secret。Java 与 FastAPI 精确比较 HTTP status、解析后的 body 和 Content-Type;最终
154 passed、0 failed、0 skipped。它证明每条 Java 路由至少有一个请求路径兼容,不代表每条
路由的成功、全部错误与数据库副作用均已逐项对照。
### 4.2 154 条已认证安全业务/校验差分
`tests/compatibility/authenticated_route_runner.py` 使用与 Java 基线语义一致的 DB Token、
server-secret 或匿名认证方式,对同一份 154 路由清单逐条发送安全业务/校验请求。runner 对
动态管理员密码和 OTA 下载 UUID 先独立验证格式再做最小归一化,并同步会影响响应的固定审计
时间;最终 154 passed、0 failed、0 skipped。为不污染 fixture 或产生不可逆副作用,该 runner
有意选择资源不存在、单一约束失败、幂等空操作等不会成功写入的路径。因此这里的“已认证”
证明请求已经越过认证层并进入业务/校验逻辑,不表示每条写接口都完成了一次成功写入。
### 4.3 49 项深度结果分布
| 类别 | 通过 | 失败 |
|---|---:|---:|
| configuration | 1 | 0 |
| authentication-i18n | 7 | 0 |
| authentication | 1 | 0 |
| authorization | 1 | 0 |
| serialization | 2 | 0 |
| agent | 1 | 0 |
| device | 1 | 0 |
| model | 1 | 0 |
| correct-word | 1 | 0 |
| binary-download | 6 | 0 |
| validation | 5 | 0 |
| server-secret | 3 | 0 |
| ota | 1 | 0 |
| ota-validation | 2 | 0 |
| ota-signing | 2 | 0 |
| activation | 3 | 0 |
| external-mock | 2 | 0 |
| crud | 3 | 0 |
| database-side-effect | 4 | 0 |
| upload | 1 | 0 |
| upload-validation | 1 | 0 |
| **合计** | **49** | **0** |
直接覆盖内容包括:
- 默认语言、`zh-CN``zh-TW``en-US``de-DE``vi-VN``pt-BR` 七种
`Accept-Language` 情形;
- 未登录、DB Token 过期、普通用户访问管理员接口;
- Long ID 字符串、日期、Asia/Shanghai/UTC 兼容、null、别名和分页;
- 缺字段、数字格式、最小/最大边界和 Java UTF-16 长度语义;
- server-secret 缺失、错误和正确三条路径;
- OTA health、缺失/非法 `Device-Id`、激活、WS/MQTT credential
- HMAC-SHA256、URL-safe Base64、MQTT Base64 密码的独立密码学验证;
- 纠错词创建、更新、下载、删除及数据库副作用;
- OTA multipart 上传、扩展名错误、元数据副作用、三次下载与第四次 404;
- 二进制 MIME、`Content-Disposition``Content-Length` 和字节摘要;
- MQTT mock 的请求 body 和按日期生成的 Authorization。
### 4.4 有意的动态值处理
- OTA timestamp、WebSocket token 和生成 UUID 先做最小范围归一化,再分别验证格式与 HMAC;
不是把动态字段全部忽略。
- Hibernate Validator 使用无序 `ConstraintViolation Set`。模型 provider 空 body 用例要求
Java/FastAPI 均返回 HTTP 200、错误码 10034,消息必须属于 Java DTO 声明的五个精确约束,
不强行固定 Java 本身不稳定的首条消息。
- 报告落盘前递归脱敏 `private_key`、server secret、MQTT signature key、Token、password 和
Authorization;比较与密码学验证仍使用未脱敏的内存值。
### 4.5 覆盖边界
Java 清单共有 154 条路由,FastAPI 结构注册为 154/154;另有 3 条消费者兼容路由。全部 154
条均有一次未认证/非法差分和一次已认证安全业务/校验差分;49 个深度 checks 直接命中
21/154 条 Java 路由,并对代表性成功、主要错误和数据库副作用做更完整的生命周期验证。
其余路由虽有两层逐路由差分及相关领域 service/repository/protocol 测试,仍不能宣称其全部
成功写入和错误路径均已逐项深度差分。逐行状态见
`docs/manager-api-fastapi-compatibility.md`
## 5. 隔离数据库、Redis 和 job 测试
7 条集成测试均连接隔离 MySQL/Redis,而不是纯 mock
1. MySQL 事务异常回滚;
2. `SELECT FOR UPDATE` 在并发写入下串行化;
3. Redis key TTL 到期且不删除无关 key
4. Redis 分布式锁只允许一个并发 job 执行;
5. watchdog 在任务超过原始 lease 后续租,仍保持单实例;
6. 实际 knowledge job 函数在 Redis 锁下只执行一次;
7. Java hash 兼容写入的默认 TTL 为 86400 秒。
差分 CRUD 另检查 Java/FastAPI 各自隔离 schema 的行级副作用;没有实施双写,也没有操作开发
数据库。最终 FastAPI 与 external mock 日志没有 warning/error/traceback。两层全路由差分会
让 Java 基线按其既有全局异常处理记录 36 条预期 ERROR,严格门禁将其精确分为 8 类:缺 body
13 条、缺 `Device-Id` 5 条、对象/数组反序列化 8 条、缺 query 4 条、非 multipart 3 条、
`callerMac` 空值 1 条、空消息 1 条及 `null` 消息 1 条。未分类 Java ERROR 为 0;任一类别数量
变化或出现未分类日志都会使脚本失败。
## 6. 性能对比
参数:每个服务每场景先顺序 warmup 10 次,再以并发 6 计量 60 次;4 个场景、2 个服务共
480 次计量请求。结果来自最终 `performance-results.json`
| 场景 | 服务 | p50 ms | p95 ms | 吞吐 req/s | 错误 |
|---|---|---:|---:|---:|---:|
| representative-read | Java | 7.662 | 16.239 | 601.128 | 0 |
| representative-read | FastAPI | 6.749 | 12.552 | 751.538 | 0 |
| representative-crud-update | Java | 9.331 | 15.302 | 543.013 | 0 |
| representative-crud-update | FastAPI | 11.252 | 20.599 | 476.870 | 0 |
| runtime-configuration | Java | 8.746 | 16.127 | 568.863 | 0 |
| runtime-configuration | FastAPI | 6.488 | 13.722 | 765.126 | 0 |
| ota-check-and-signing | Java | 11.301 | 19.854 | 454.655 | 0 |
| ota-check-and-signing | FastAPI | 16.208 | 20.344 | 354.807 | 0 |
FastAPI/Java 比率:读取 p50 `0.881`、p95 `0.773`、吞吐 `1.250`CRUD p50 `1.206`
p95 `1.346`、吞吐 `0.878`;配置 p50 `0.742`、p95 `0.851`、吞吐 `1.345`OTA p50
`1.434`、p95 `1.025`、吞吐 `0.780`。因此不能笼统声称所有 FastAPI 接口都更快:本轮读取和
配置场景更快,CRUD p50/p95 分别较慢约 20.6%/34.6%OTA p50 较慢约 43.4%、p95 接近、
吞吐低约 22.0%。
这是同机、短时、固定 fixture 的简单对比,用于发现数量级回退,不是容量、长稳、生产网络或
多 worker 极限测试。
## 7. 三端兼容验证
- `manager-web`:134 个调用点、130 条唯一结构路由;i18n、unit、snapshot 与 production build
均通过,现有调用无需修改 URL。
- `manager-mobile`:46 个调用点、40 条唯一结构路由;type-check、lint、snapshot 与微信小程序
build 均通过。
- `xiaozhi-server`:8 个调用点、8 条唯一结构路由;compileall 通过,consumer manifest 确认
全部能解析到 FastAPI。该模块没有可执行的一般单测,不能把 compileall 写成运行时集成通过。
三端合计 188 个调用点、140 条唯一结构路由。此结论证明 path/method 解析闭合;全部 Java
路由另有未认证/非法和已认证安全业务/校验两层差分,参与 49 项深度差分或领域测试的调用拥有
更完整的成功、错误或副作用证据。
## 8. 迁移验证过程中发现并修正的问题
测试没有通过删除、跳过或放宽失败用例获得绿色。差分在实现过程中实际暴露并促成修复的兼容
问题包括:分页 `total` 类型、运行时配置字段查询、设备日期时区、认证与 OTA MIME、缺失
`Device-Id` envelope、provider 校验语义、非法分页消息、运行时配置 key 命名。修复后才生成
当前 49/49 深度报告。
154 路由请求面 runner 首次执行只有 149/154 通过,准确暴露 5 条“整个 JSON body 缺失”差异:
`POST /ota/``POST /user/login``POST /user/register`
`POST /user/retrieve-password``POST /user/smsVerification`。Java 对整个 body 缺失返回
HTTP 200、`code=500` 的通用 envelopeFastAPI 当时返回 `code=10034`。全局校验兼容层随后
只对定位恰为 body 根节点的 missing 错误映射 `code=500`,字段级 missing 仍保持 `10034`
新增回归用例并完整重跑后才得到 154/154。
已认证安全业务/校验 runner 首次执行为 111/154,通过实际差分先后修正了根 body 类型、必填
query/multipart 的 Java envelope、knowledge/device/agent/voice resource 的检查顺序、权限与资源
不存在语义,以及 DTO 单约束消息等差异;后续结果依次为 149/154、152/154,最终才达到
154/154。对 Hibernate Validator 无序约束的请求,runner 改用只触发一个约束的定向 payload,
没有跳过路由、忽略响应字段或放宽比较。该 runner 始终保持“已认证但不成功写入”的安全边界。
另有测试基础设施、运行时和构建问题被明确记录:
- Pydantic 2.13 与 FastAPI 0.116 的 TypeAdapter alias 路径产生
`UnsupportedFieldAttributeWarning`;依赖锁定到 Pydantic 2.11.7/core 2.33.2 后,实际 OTA
body alias 验证与最终日志门禁均无 warning。
- 第一版日志门禁把 Logback 初始化文本 `ERROR_FILE` 误判为运行时 ERROR,虽然 7、49、480
阶段均绿,脚本仍按门禁返回 1。正则收紧为带完整日期的 Java 应用 ERROR 后,重新从头执行
完整流程并获得 exit 0;没有直接忽略门禁失败。
- fixture 的 MySQL `VALUES()` upsert 产生 8.0 弃用 warning;改为 row alias 语法并重新执行后
无该 warning。
- 已认证差分报告最初虽未含真实凭证,但 `paramCode=server.secret` / `paramValue` 结构仍写入了
隔离 fixture 值;递归脱敏器补充键值对识别及回归测试后,再次完整执行差分。最终四份 JSON
`contract-server-secret`、测试 Token 和测试数据库密码的扫描均为 0 命中。
- Python 3.10 下 fixed-delay jobs 等待超时抛出 `asyncio.TimeoutError`;原捕获路径导致 worker
首轮后退出。worker 改为捕获该异常并增加 Python 3.10 回归测试;实际 jobs 容器随后观察到
knowledge job 跨 30 秒重复运行,snapshot redaction 多轮执行,SIGTERM 后干净退出。
- API 镜像构建初期遇到 uv/pip registry 传输失败;Dockerfile 固定 uv 版本并增加 timeout、retry
与缓存后完成构建。迁移镜像的 Maven Central 并发下载两次卡住,改为串行 resolver、超时与
retry 后成功。Nginx 初版配置在镜像 build 期校验失败,改用 template + `envsubst` 并在 build
内执行 `nginx -t` 后通过。
- Apple Container 自定义网络没有提供本次验证所需的容器名 DNS,host publish 和单文件挂载也
与 Docker 行为不同;验证改用容器 IP、显式 TCP bridge、named volume 和运行时 upstream 模板。它们
是测试 runtime 限制,不被记作应用通过或失败,也没有据此声称 Docker Compose 已实际启动。
## 9. 外部服务与真实联调状态
所有自动化外部调用只访问本地确定性 mock/fixture,不访问真实付费服务。
| 外部能力 | 自动化证据 | 真实联调状态 |
|---|---|---|
| RAGFlow dataset/document/chunk/retrieval/upload | 请求 JSON/query/header、30 秒 timeout、强 DTO、Long/null、错误映射与补偿路径测试 | 无真实 RAGFlow 凭证/实例,未联调 |
| 阿里云短信 | 配置、错误 envelope 与业务路径测试 | 无真实 AccessKey,不发送短信,未联调 |
| 火山语音克隆/音频 | multipart/JSON、状态及错误映射 mock | 无真实付费凭证,未联调 |
| 声纹 HTTP | Java multipart 形状与错误映射 mock | 无真实声纹服务,未联调 |
| OpenAI-compatible LLM | 请求格式、thinking policy、摘要/标题相关 mock | 无真实模型 key,不访问付费模型,未联调 |
| MQTT gateway HTTP | 差分验证 body、按日期 Authorization 和 401 retry 语义 | 无真实 MQTT broker/gateway,未联调 |
| MCP/管理 WebSocket | token、URL、path/scheme/form 兼容测试 | 无真实远端 MCP/WS,未联调 |
| OTA/WS/MQTT credential | 本地 HMAC/Base64/时间戳和下载行为实测 | 无 ESP32 真机和生产 broker,不属于硬件联调 |
因此,本报告只证明 mock 下已覆盖的请求格式、超时、错误映射、重试和本地密码学行为;不能把
任何一项写成供应商或生产环境端到端通过。
## 10. 已知行为/部署差异
- Java 的 Hibernate Validator 首条约束消息顺序不稳定;FastAPI 保持相同 envelope、错误码和
声明消息集合,而不是伪造固定顺序。
- Java 在 Spring 进程内运行定时任务;FastAPI 把 jobs 分离为独立进程,并用 Redis 锁和
watchdog 防止多 worker 重复执行。集成测试验证单实例和续租语义,但部署拓扑有意不同。
- FastAPI 增加 3 条消费者兼容路由和 live/ready health endpoints;它们没有 Java Controller
基线,属于明确的加法差异。
- 49 项已执行深度差分中没有观测到响应、所选 header 或数据库副作用差异;这句话只适用于
报告中的 49 项,不外推为全部 154 条路由均完成了成功写入和全部错误路径生命周期验证。
## 11. 实际容器与 Nginx 验证
### 11.1 Runtime、镜像与 Compose 口径
本机没有可用的 Docker/Podman daemon,实际 OCI build/run 使用 Apple Container 1.0.0 的
linux/arm64 VM,并显式使用隔离 app/log/install root
```bash
CLI=/Users/mie/.cache/xiaozhi-migration-tools/container-1.0.0-prefix/bin/container
ROOT=/Users/mie/.cache/xiaozhi-migration-tools/container-1.0.0-prefix
"$CLI" system start \
--app-root "$ROOT/runtime-data" \
--install-root "$ROOT" \
--log-root "$ROOT/runtime-logs" \
--disable-kernel-install
"$CLI" builder start
"$CLI" build --tag xiaozhi/manager-api-fastapi:0.1.0 \
--file main/manager-api-fastapi/Dockerfile .
"$CLI" build --tag xiaozhi/manager-api-migrate:fastapi-0.1.0 \
--file main/manager-api-fastapi/Dockerfile.migrations .
"$CLI" build --tag xiaozhi/manager-api-nginx:fastapi-0.1.0 \
--file main/manager-api-fastapi/Dockerfile.nginx .
```
三张镜像均实际构建并运行。迁移镜像 OCI index 为
`sha256:613faace4314b03392e65b64d9b4a9ba7a694cdd751c1a45009824d55f0647f7`,其 arm64
manifest 为 `sha256:6a10850841370d033a3b521fbb1100cb64b5cc6837fac35d00c4257343c0f2f9`
Nginx 镜像 OCI index 为
`sha256:2e6a188ad6d38b62fa4e77329a629ada00c4e773da9ded73c5f3289e40da477a`,其 arm64
manifest 为 `sha256:ff653bc2d11d4a3b1640747626055d6551fb33324500fb2e65b9333142da8526`
API 镜像在上传目录 readiness 最后一处源码变更后重新 build;最终 OCI index 为
`sha256:04ae1a98307b7369368b9665c6caf9f0911c8b2a967f5a91f23c6dde7c7baa16`,其 arm64
manifest 为 `sha256:c3267d307c9898975372539121f118549bfe51012c6c9bfdc3e84f99f3e56214`
config 为 `sha256:7c0b13757da041c0d118d14342e92c5310e4fb2140f029e385e97de9fe21d8cc`
manifest size 为 84,551,252 bytes,镜像配置创建时间为 `2026-07-20T07:04:45Z`
`docker-compose.yml` 已由 `tests/test_deployment_artifacts.py` 静态验证 migration dependency、
read-only root、tmpfs、upload volume、healthcheck、graceful timeout 与可切换 upstreamNginx
镜像 build 内也实际执行 `nginx -t`。由于本机没有 Docker Compose runtime,本报告明确只把
Compose 记为静态通过,不声称执行过 `docker compose up`
### 11.2 Liquibase migration
迁移镜像以 UID 10001 一次性运行,只读取原 Java resources 内的 Liquibase 历史。目标为隔离
schema `manager_container_test`;最终容器回归中再次运行并报告 101 个 changeSets 均
up-to-date。随后实查 `DATABASECHANGELOG` 为 101 条、业务及 Liquibase 表合计 30 张,
`DATABASECHANGELOGLOCK.LOCKED=0`,证明历史完整且锁已释放。没有连接、修改或清空开发数据库。
### 11.3 API、jobs、health、卷与优雅关闭
Apple Container VM 访问 host-only MySQL/Redis 时使用仓库内 TCP bridge,而不是暴露开发服务:
```bash
cd main/manager-api-fastapi
.venv/bin/python -m tests.compatibility.tcp_proxy \
--listen-port 13317 --target-host 127.0.0.1 --target-port 13316
.venv/bin/python -m tests.compatibility.tcp_proxy \
--listen-port 16380 --target-host 127.0.0.1 --target-port 16379
```
API 容器使用 `APP_WORKERS=2`、隔离 schema、Redis DB 4、read-only root、`/tmp` tmpfs 和
named upload volume 启动。实测结果:
- 容器内 UID 为 10001,最终层没有 `/bin/uv``/usr/bin/gcc`,应用路由数为 163
- 日志确认 2 个 Uvicorn worker(容器内 PID 3、4);`/xiaozhi/health/live` 为 HTTP 200
- `Accept-Language: en-US` 的未认证业务请求保持 HTTP 200、英文 `{code:401,...}`
`POST /xiaozhi/user/login` 整个 JSON body 缺失保持 Java 的 HTTP 200、`code=500`
- read-only root 生效。Apple Container 新建空 named volume 首次以其默认 root ownership 挂载,
新增 readiness 检查准确返回 HTTP 503、`database=true``redis=true``uploads=false`,没有让
无法上传的实例接流量;该失败没有伪装为通过。随后用一次性 root 容器仅对卷执行 `chown`
ownership 初始化,ready 变为 HTTP 200 且 `uploads=true`UID 10001 的 API 成功写入,重启后
文件 SHA256 `1ad4cb4f879aa1ddf43a14e1a84cc5dbf8f65e91295165e126b8b08be3cd9a50` 保持不变;
- 发送 SIGTERM 后 worker 完成 lifespan shutdown 并以 exit 0 退出,无 traceback/error。
同一 API 镜像另以 `python -m app.jobs.worker`、read-only root 和 `/tmp` tmpfs 启动。实际等待
超过 31 秒后,knowledge fixed-delay job 执行两次且相隔 30 秒,snapshot redaction 多次执行;
这验证 Python 3.10 timeout 修复与真实调度循环。SIGTERM 后 jobs 也干净退出。API 多 worker
本身不加载 jobs,独立 worker 再由 Redis lock/watchdog 保证单实例。
上述 ownership 初始化是 Apple Container 空 named volume 的实测处理;本机没有 Docker
Compose runtime,因此 Docker Compose 的 named-volume copy-up 行为没有实际验证,不能用
Apple Container 的结果代替。
### 11.4 Nginx 切流与 Java 回滚
Nginx 镜像以 read-only root 和 `/var/cache/nginx``/var/run``/tmp` 三个 tmpfs 运行;其
entrypoint 将 `MANAGER_API_UPSTREAM` 注入模板后 `exec nginx`。Apple Container 自定义网络在
本次环境没有容器名 DNS,因此实测使用 runtime 分配的 API/Java 容器 IP,语义与生产 hostname
upstream 相同。FastAPI upstream 下实际验证:
- `/xiaozhi/health/ready` 为 HTTP 200
- `/xiaozhi` 精确返回 308 到 `/xiaozhi/`
- `Accept-Language: en-US` 的未认证 envelope 由 Nginx 转发后与直连 FastAPI 一致,整个 JSON
body 缺失也保持 HTTP 200、`code=500`
- Nginx、API 均在 SIGTERM 下以 exit 0 干净退出。
随后仅替换 `MANAGER_API_UPSTREAM` 指向保留的 Java 容器并重建 Nginx 运行实例;`/xiaozhi/ota/`
回滚探针的 response body 与直连 Java 按字节完全一致。此步骤证明回滚不需要删除 Java 服务、
改数据库或双写,只需切换 upstream。Nginx 基础镜像未声明非 root USER,因此这里不虚构其
non-root 属性;实际硬化证据是 read-only root、最小 tmpfs 和无持久写入。应用与迁移镜像则
均以 UID 10001 运行。
## 12. 证据文件
- 逐接口矩阵:`docs/manager-api-fastapi-compatibility.md`
- 迁移、切流与回滚说明:`docs/manager-api-fastapi-migration.md`
- Java 路由清单:`main/manager-api-fastapi/compatibility/java-routes.json`
- 三端调用清单:`main/manager-api-fastapi/compatibility/consumer-routes.json`
- 154 路由未认证/非法请求面机器报告:
`main/manager-api-fastapi/compatibility/route-surface-results.json`
- 154 路由已认证安全业务/校验机器报告:
`main/manager-api-fastapi/compatibility/authenticated-route-results.json`
- 深度差分机器报告:`main/manager-api-fastapi/compatibility/contract-results.json`
- 性能机器报告:`main/manager-api-fastapi/compatibility/performance-results.json`
- 一键隔离脚本:`main/manager-api-fastapi/scripts/run-isolated-contract-tests.sh`
- 未认证/非法请求面 runner
`main/manager-api-fastapi/tests/compatibility/route_surface_runner.py`
- 已认证安全业务/校验 runner:
`main/manager-api-fastapi/tests/compatibility/authenticated_route_runner.py`
- 深度差分 runner`main/manager-api-fastapi/tests/compatibility/differential_runner.py`
- 外部 mock`main/manager-api-fastapi/tests/compatibility/external_mock.py`
- 集成测试:`main/manager-api-fastapi/tests/integration/test_isolated_runtime.py`
- 容器静态断言:`main/manager-api-fastapi/tests/test_deployment_artifacts.py`
- 容器网络 bridge`main/manager-api-fastapi/tests/compatibility/tcp_proxy.py`
- API/migration/Nginx 构建定义:`main/manager-api-fastapi/Dockerfile`
`main/manager-api-fastapi/Dockerfile.migrations``main/manager-api-fastapi/Dockerfile.nginx`
- Nginx runtime 配置:`main/manager-api-fastapi/deploy/nginx.conf`
`main/manager-api-fastapi/deploy/nginx-entrypoint.sh`
- Java Surefire`main/manager-api/target/surefire-reports/`
## 13. 当前结论
Java 98、FastAPI 全量 139、隔离集成 7、未认证/非法请求面 154/154、已认证安全业务/校验
154/154、深度差分 49/49、性能 480/0,以及 Web/Mobile 构建、xiaozhi-server compileall 和
实际容器/Nginx 验证均按上述命令完成;各测试集合均为 0 failed、0 errors、0 skipped。原 Java
服务和 Liquibase 历史均未删除。
本地可安全执行的兼容、集成、构建、消费者和容器验证已经通过。每条 Java 路由虽已有两次
全覆盖差分,但已认证 runner 有意不执行成功写入,所以不能将其表述为 154 条全部成功、错误
和副作用生命周期均已深度验证;真实 RAGFlow、短信、语音克隆、声纹、模型、MQTT/MCP/WS
及 ESP32 硬件因没有真实凭证或设备而未联调,也没有在本报告中描述为已通过。
+2 -1
View File
@@ -71,6 +71,7 @@ docker logs -f mcp-endpoint-server
请你保留好上面两个`接口地址`,下一步要用到。 请你保留好上面两个`接口地址`,下一步要用到。
# 2、全模块部署时,怎么配置MCP接入点 # 2、全模块部署时,怎么配置MCP接入点
首先,你要开启MCP接入点功能。在智控台,点击顶部`参数字典`,在下拉菜单中,点击`系统功能配置`页面。在页面上勾选`MCP接入点`,点击`保存配置`。在`角色配置`页面,点击`编辑功能`按钮,即可看到`mcp接入点`功能。
如果你是全模块部署,使用管理员账号,登录智控台,点击顶部`参数字典`,选择`参数管理`功能。 如果你是全模块部署,使用管理员账号,登录智控台,点击顶部`参数字典`,选择`参数管理`功能。
@@ -117,7 +118,7 @@ mcp_endpoint: ws://192.168.1.25:8004/mcp_endpoint/mcp/?token=def
250705[__main__]-INFO-mcp接入点是 ws://192.168.1.25:8004/mcp_endpoint/mcp/?token=abc 250705[__main__]-INFO-mcp接入点是 ws://192.168.1.25:8004/mcp_endpoint/mcp/?token=abc
250705[__main__]-INFO-Websocket地址是 ws://192.168.1.25:8000/xiaozhi/v1/ 250705[__main__]-INFO-Websocket地址是 ws://192.168.1.25:8000/xiaozhi/v1/
250705[__main__]-INFO-=======上面的地址是websocket协议地址,请勿用浏览器访问======= 250705[__main__]-INFO-=======上面的地址是websocket协议地址,请勿用浏览器访问=======
250705[__main__]-INFO-如想测试websocket请用谷歌浏览器打开test目录下的test_page.html 250705[__main__]-INFO-如想测试websocket请启动digital-human模块,打开浏览器交互测试
250705[__main__]-INFO-============================================================= 250705[__main__]-INFO-=============================================================
``` ```
+1 -1
View File
@@ -23,7 +23,7 @@
250705[__main__]-INFO-mcp接入点是 ws://192.168.1.25:8004/mcp_endpoint/mcp/?token=abc 250705[__main__]-INFO-mcp接入点是 ws://192.168.1.25:8004/mcp_endpoint/mcp/?token=abc
250705[__main__]-INFO-Websocket地址是 ws://192.168.1.25:8000/xiaozhi/v1/ 250705[__main__]-INFO-Websocket地址是 ws://192.168.1.25:8000/xiaozhi/v1/
250705[__main__]-INFO-=======上面的地址是websocket协议地址,请勿用浏览器访问======= 250705[__main__]-INFO-=======上面的地址是websocket协议地址,请勿用浏览器访问=======
250705[__main__]-INFO-如想测试websocket请用谷歌浏览器打开test目录下的test_page.html 250705[__main__]-INFO-如想测试websocket请启动digital-human模块,打开浏览器交互测试
250705[__main__]-INFO-============================================================= 250705[__main__]-INFO-=============================================================
``` ```
+40
View File
@@ -0,0 +1,40 @@
# MCP 方法如何获取设备信息
本教程将指导你如何使用MCP方法获取设备信息。
第一步:自定义你的`agent-base-prompt.txt`文件
把xiaozhi-server目录的`agent-base-prompt.txt`文件内容复制到你的`data`目录下,并重命名为`.agent-base-prompt.txt`
第二步:修改`data/.agent-base-prompt.txt`文件,找到`<context>`标签,在标签内容中添加以下代码内容:
```
- **设备ID** {{device_id}}
```
添加完成后,你的`data/.agent-base-prompt.txt`文件的`<context>`标签内容大致如下:
```
<context>
【重要!以下信息已实时提供,无需调用工具查询,请直接使用:】
- **设备ID** {{device_id}}
- **当前时间:** {{current_time}}
- **今天日期:** {{today_date}} ({{today_weekday}})
- **今天农历:** {{lunar_date}}
- **用户所在城市:** {{local_address}}
- **当地未来7天天气:** {{weather_info}}
</context>
```
第三步:修改`data/.config.yaml`文件,找到`agent-base-prompt`配置,修改前内容如下:
```
prompt_template: agent-base-prompt.txt
```
修改成
```
prompt_template: data/.agent-base-prompt.txt
```
第四步:重启你的xiaozhi-server服务。
第五步:在你的mcp方法增加名称为`device_id`,类型为`string`,描述为`设备ID`的参数。
第六步:重新唤醒小智,让他调用mcp方法,查看你的mcp方法是否可以获取`设备ID`
+2 -2
View File
@@ -58,7 +58,7 @@ docker restart xiaozhi-esp32-server
2025-06-01 **** - 视觉分析接口是 http://192.168.4.7:8003/mcp/vision/explain 2025-06-01 **** - 视觉分析接口是 http://192.168.4.7:8003/mcp/vision/explain
2025-06-01 **** - Websocket地址是 ws://192.168.4.7:8000/xiaozhi/v1/ 2025-06-01 **** - Websocket地址是 ws://192.168.4.7:8000/xiaozhi/v1/
2025-06-01 **** - =======上面的地址是websocket协议地址,请勿用浏览器访问======= 2025-06-01 **** - =======上面的地址是websocket协议地址,请勿用浏览器访问=======
2025-06-01 **** - 如想测试websocket请用谷歌浏览器打开test目录下的test_page.html 2025-06-01 **** - 如想测试websocket请启动digital-human模块,打开浏览器交互测试
2025-06-01 **** - ============================================================= 2025-06-01 **** - =============================================================
``` ```
@@ -133,7 +133,7 @@ docker restart xiaozhi-esp32-server
2025-06-01 **** - 视觉分析接口是 http://192.168.4.7:8003/mcp/vision/explain 2025-06-01 **** - 视觉分析接口是 http://192.168.4.7:8003/mcp/vision/explain
2025-06-01 **** - Websocket地址是 ws://192.168.4.7:8000/xiaozhi/v1/ 2025-06-01 **** - Websocket地址是 ws://192.168.4.7:8000/xiaozhi/v1/
2025-06-01 **** - =======上面的地址是websocket协议地址,请勿用浏览器访问======= 2025-06-01 **** - =======上面的地址是websocket协议地址,请勿用浏览器访问=======
2025-06-01 **** - 如想测试websocket请用谷歌浏览器打开test目录下的test_page.html 2025-06-01 **** - 如想测试websocket请启动digital-human模块,打开浏览器交互测试
2025-06-01 **** - ============================================================= 2025-06-01 **** - =============================================================
``` ```
+62 -11
View File
@@ -1,24 +1,34 @@
# MQTT 网关部署教程 # MQTT 网关部署教程
`xiaozhi-esp32-server`项目,可结合虾哥开源的[xiaozhi-mqtt-gateway](https://github.com/78/xiaozhi-mqtt-gateway) 项目进行简单改造,即可实现小智硬件MQTT+UDP连接。 `xiaozhi-esp32-server`项目,可结合虾哥开源的[xiaozhi-mqtt-gateway](https://github.com/78/xiaozhi-mqtt-gateway) 项目进行简单改造,即可实现小智硬件MQTT+UDP连接。
本教程分为三部分,你可以根据你是全模块部署还是单模块部署,选择对应的部分接入MQTT网关:
- 第一部分:部署MQTT网关
- 第二部分:全模块运行实现小智硬件MQTT+UDP连接
- 第三部分:单模块运行xiaozhi-server实现小智硬件MQTT+UDP连接
## 准备工作 ## 准备阶段
查看你智控台首页底部的版本号,确认你的智控台版本是否是`0.7.7`及以上版本。如果不是,需要升级智控台。
准备好你的`xiaozhi-server``mqtt-websocket`连接地址。在你原来的`websocket地址`基础上,添加`?from=mqtt_gateway`字符,就可以得到`mqtt-websocket`连接地址 准备好你的`xiaozhi-server``mqtt-websocket`连接地址。在你原来的`websocket地址`基础上,添加`?from=mqtt_gateway`字符,就可以得到`mqtt-websocket`连接地址
1、如果你是源码部署,你的`mqtt-websocket`地址是: 1、如果你是源码部署,你的`mqtt-websocket`地址是:
``` ```
ws://127.0.0.1:8000/xiaozhi/v1?from=mqtt_gateway ws://127.0.0.1:8000/xiaozhi/v1/?from=mqtt_gateway
``` ```
2、如果你是docker部署,你的`mqtt-websocket`地址是 2、如果你是docker部署,你的`mqtt-websocket`地址是
``` ```
ws://你宿主机局域网IP:8000/xiaozhi/v1?from=mqtt_gateway ws://你宿主机局域网IP:8000/xiaozhi/v1/?from=mqtt_gateway
``` ```
## 第一步 部署MQTT网关 ## 重要提示
如果你是服务器部署,需要确保服务器`1883``8884``8007`端口都对外开放。`8884`选择的协议类型是`UDP`,其他是`TCP`
如果你是服务器部署,需要确保服务器`1883``8884``8007`端口都对外开放。`8884`选择的协议类型是`UDP`,其他是`TCP`
如果你是服务器部署,需要确保服务器`1883``8884``8007`端口都对外开放。`8884`选择的协议类型是`UDP`,其他是`TCP`
## 第一部分:部署MQTT网关
1. 克隆[改造后的xiaozhi-mqtt-gateway项目](https://github.com/xinnan-tech/xiaozhi-mqtt-gateway.git) 1. 克隆[改造后的xiaozhi-mqtt-gateway项目](https://github.com/xinnan-tech/xiaozhi-mqtt-gateway.git)
```bash ```bash
@@ -43,7 +53,7 @@ cp config/mqtt.json.example config/mqtt.json
{ {
"production": { "production": {
"chat_servers": [ "chat_servers": [
"ws://127.0.0.1:8000/xiaozhi/v1?from=mqtt_gateway" "ws://127.0.0.1:8000/xiaozhi/v1/?from=mqtt_gateway"
] ]
}, },
"debug": false, "debug": false,
@@ -66,6 +76,7 @@ MQTT_PORT=1883 # MQTT服务器端口
UDP_PORT=8884 # UDP服务器端口 UDP_PORT=8884 # UDP服务器端口
API_PORT=8007 # 管理API端口 API_PORT=8007 # 管理API端口
MQTT_SIGNATURE_KEY=test # MQTT签名密钥 MQTT_SIGNATURE_KEY=test # MQTT签名密钥
SERVER_SECRET=Te1st12134 # 服务器密钥,请保持和智控台(server.secret)一致或者和xiaozhi-server里(server.auth_key)保持一致
``` ```
请注意`PUBLIC_IP`配置,确保其与实际公网IP一致,如果有域名就填域名。 请注意`PUBLIC_IP`配置,确保其与实际公网IP一致,如果有域名就填域名。
@@ -75,6 +86,13 @@ MQTT_SIGNATURE_KEY=test # MQTT签名密钥
- 注意不要用简单的密码,比如`123456`、`test`等。 - 注意不要用简单的密码,比如`123456`、`test`等。
- 注意不要用简单的密码,比如`123456`、`test`等。 - 注意不要用简单的密码,比如`123456`、`test`等。
`SERVER_SECRET` 是用生成websocket连接的认证信息。
1、如果你是全模块部署,且你的智控台的参数管理里`server.auth.enabled`设置成了`true`,那么,`SERVER_SECRET`需要和智控台(`server.secret`)保持一致。
2、如果你是单模块部署,且你在配置文件里把`server.auth.enabled`设置成了`true`,那么,`SERVER_SECRET`需要和配置文件里(`server.auth_key`)保持一致。
6. 启动MQTT网关 6. 启动MQTT网关
``` ```
# 启动服务 # 启动服务
@@ -95,7 +113,9 @@ pm2 logs xz-mqtt
pm2 restart xz-mqtt pm2 restart xz-mqtt
``` ```
## 第二步 配置智控台 ## 第二部分:全模块运行实现小智硬件MQTT+UDP连接
查看你智控台首页底部的版本号,确认你的智控台版本是否是`0.7.7`及以上版本。如果不是,需要升级智控台。
1. 在智控台顶部,点击`参数管理`,搜索`server.mqtt_gateway`,点击编辑,填入你在`.env`文件中设置的`PUBLIC_IP`+`:`+`MQTT_PORT`。类似这样 1. 在智控台顶部,点击`参数管理`,搜索`server.mqtt_gateway`,点击编辑,填入你在`.env`文件中设置的`PUBLIC_IP`+`:`+`MQTT_PORT`。类似这样
``` ```
@@ -107,7 +127,7 @@ pm2 restart xz-mqtt
``` ```
192.168.0.7:8884 192.168.0.7:8884
``` ```
4. 在智控台顶部,点击`参数管理`,搜索`server.mqtt_manager_api`,点击编辑,填入你在`.env`文件中设置的`PUBLIC_IP`+`:`+`UDP_PORT`。类似这样 4. 在智控台顶部,点击`参数管理`,搜索`server.mqtt_manager_api`,点击编辑,填入你在`.env`文件中设置的`PUBLIC_IP`+`:`+`API_PORT`。类似这样
``` ```
192.168.0.7:8007 192.168.0.7:8007
``` ```
@@ -127,7 +147,38 @@ curl 'http://localhost:8002/xiaozhi/ota/' \
{"server_time":{"timestamp":1757567894012,"timeZone":"Asia/Shanghai","timezone_offset":480},"activation":{"code":"460609","message":"http://xiaozhi.server.com\n460609","challenge":"11:22:33:44:55:66"},"firmware":{"version":"1.0.1","url":"http://xiaozhi.server.com:8002/xiaozhi/otaMag/download/NOT_ACTIVATED_FIRMWARE_THIS_IS_A_INVALID_URL"},"websocket":{"url":"ws://192.168.4.23:8000/xiaozhi/v1/"},"mqtt":{"endpoint":"192.168.0.7:1883","client_id":"GID_default@@@11_22_33_44_55_66@@@7b94d69a-9808-4c59-9c9b-704333b38aff","username":"eyJpcCI6IjA6MDowOjA6MDowOjA6MSJ9","password":"Y8XP9xcUhVIN9OmbCHT9ETBiYNE3l3Z07Wk46wV9PE8=","publish_topic":"device-server","subscribe_topic":"devices/p2p/11_22_33_44_55_66"}} {"server_time":{"timestamp":1757567894012,"timeZone":"Asia/Shanghai","timezone_offset":480},"activation":{"code":"460609","message":"http://xiaozhi.server.com\n460609","challenge":"11:22:33:44:55:66"},"firmware":{"version":"1.0.1","url":"http://xiaozhi.server.com:8002/xiaozhi/otaMag/download/NOT_ACTIVATED_FIRMWARE_THIS_IS_A_INVALID_URL"},"websocket":{"url":"ws://192.168.4.23:8000/xiaozhi/v1/"},"mqtt":{"endpoint":"192.168.0.7:1883","client_id":"GID_default@@@11_22_33_44_55_66@@@7b94d69a-9808-4c59-9c9b-704333b38aff","username":"eyJpcCI6IjA6MDowOjA6MDowOjA6MSJ9","password":"Y8XP9xcUhVIN9OmbCHT9ETBiYNE3l3Z07Wk46wV9PE8=","publish_topic":"device-server","subscribe_topic":"devices/p2p/11_22_33_44_55_66"}}
``` ```
## 第三步 重启小智设备 由于MQTT信息是需要靠OTA地址下发的,因此只有你保证能正常连接服务器的OTA地址,重启唤醒即可。
唤醒后留意mqtt-gateway的日志,确认是否有连接成功的日志。
```
pm2 logs xz-mqtt
```
## 第三部分:单模块运行xiaozhi-server实现小智硬件MQTT+UDP连接
打开你的`data/.config.yaml`文件,在`server`下找到`mqtt_gateway`填入你在`.env`文件中设置的`PUBLIC_IP`+`:`+`MQTT_PORT`。类似这样
```
192.168.0.7:1883
```
在`server`下找到`mqtt_signature_key`填入你在`.env`文件中设置的`MQTT_SIGNATURE_KEY`。
在`server`下找到`udp_gateway`填入你在`.env`文件中设置的`PUBLIC_IP`+`:`+`UDP_PORT`。类似这样
```
192.168.0.7:8884
```
上面的配置完成后,你可以使用curl命令,验证你的ota地址是否会下发mqtt配置,把下面的`http://localhost:8002/xiaozhi/ota/`改成你的ota地址
```
curl 'http://localhost:8002/xiaozhi/ota/' \
-H 'Device-Id: 11:22:33:44:55:66' \
--data-raw $'{\n "application": {\n "version": "1.0.1",\n "elf_sha256": "1"\n },\n "board": {\n "mac": "11:22:33:44:55:66"\n }\n}'
```
如果返回的内容包含`mqtt`相关的配置,说明配置成功。类似这样
```
{"server_time":{"timestamp":1758781561083,"timeZone":"GMT+08:00","timezone_offset":480},"activation":{"code":"527111","message":"http://xiaozhi.server.com\n527111","challenge":"11:22:33:44:55:66"},"firmware":{"version":"1.0.1","url":"http://xiaozhi.server.com:8002/xiaozhi/otaMag/download/NOT_ACTIVATED_FIRMWARE_THIS_IS_A_INVALID_URL"},"websocket":{"url":"ws://192.168.1.15:8000/xiaozhi/v1/"},"mqtt":{"endpoint":"192.168.1.15:1883","client_id":"GID_default@@@11_22_33_44_55_66@@@11_22_33_44_55_66","username":"eyJpcCI6IjE5Mi4xNjguMS4xNSJ9","password":"fjAYs49zTJecWqJ3jBt+kqxVn/x7vkXRAc85ak/va7Y=","publish_topic":"device-server","subscribe_topic":"devices/p2p/11_22_33_44_55_66"}}
```
由于MQTT信息是需要靠OTA地址下发的,因此只有你保证能正常连接服务器的OTA地址,重启唤醒即可。 由于MQTT信息是需要靠OTA地址下发的,因此只有你保证能正常连接服务器的OTA地址,重启唤醒即可。
唤醒后留意mqtt-gateway的日志,确认是否有连接成功的日志。 唤醒后留意mqtt-gateway的日志,确认是否有连接成功的日志。
+142
View File
@@ -0,0 +1,142 @@
# 单模块部署固件OTA自动升级配置指南
本教程将指导你如何在**单模块部署**场景下配置固件OTA自动升级功能,实现设备固件的自动更新。
如果你已经使用**全模块部署**,请忽略本教程。
## 功能介绍
在单模块部署中,xiaozhi-server内置了OTA固件管理功能,可以自动检测设备版本并下发升级固件。系统会根据设备型号和当前版本,自动匹配并推送最新的固件版本。
## 前提条件
- 你已经成功进行**单模块部署**并运行xiaozhi-server
- 设备能够正常连接到服务器
## 第一步 准备固件文件
### 1. 创建固件存放目录
固件文件需要放在`data/bin/`目录下。如果该目录不存在,请手动创建:
```bash
mkdir -p data/bin
```
### 2. 固件文件命名规则
固件文件必须遵循以下命名格式:
```
{设备型号}_{版本号}.bin
```
**命名规则说明:**
- `设备型号`:设备的型号名称,例如 `lichuang-dev``bread-compact-wifi`
- `版本号`:固件版本号,必须以数字开头,支持数字、字母、点号、下划线和短横线,例如 `1.6.6``2.0.0`
- 文件扩展名必须是 `.bin`
**命名示例:**
```
bread-compact-wifi_1.6.6.bin
lichuang-dev_2.0.0.bin
```
### 3. 放置固件文件
将准备好的固件文件(.bin文件)复制到`data/bin/`目录下:
重要的事情说三遍:升级的bin文件是`xiaozhi.bin`,不是全量固件文件`merged-binary.bin`!
重要的事情说三遍:升级的bin文件是`xiaozhi.bin`,不是全量固件文件`merged-binary.bin`!
重要的事情说三遍:升级的bin文件是`xiaozhi.bin`,不是全量固件文件`merged-binary.bin`!
```bash
cp xiaozhi.bin data/bin/设备型号_版本号.bin
```
例如:
```bash
cp xiaozhi.bin data/bin/bread-compact-wifi_1.6.6.bin
```
## 第二步 配置公网访问地址(仅公网部署需要)
**注意:此步骤仅适用于单模块公网部署的场景。**
如果你的xiaozhi-server是公网部署(使用公网IP或域名),**必须**配置`server.vision_explain`参数,因为OTA固件下载地址会使用该配置的域名和端口。
如果你是局域网部署,可以跳过此步骤。
### 为什么要配置这个参数?
在单模块部署中,系统生成固件下载地址时,会使用`vision_explain`配置的域名和端口作为基础地址。如果不配置或配置错误,设备将无法访问固件下载地址。
### 配置方法
打开`data/.config.yaml`文件,找到`server`配置段,设置`vision_explain`参数:
```yaml
server:
vision_explain: http://你的域名或IP:端口号/mcp/vision/explain
```
**配置示例:**
局域网部署(默认):
```yaml
server:
vision_explain: http://192.168.1.100:8003/mcp/vision/explain
```
公网域名部署:
```yaml
server:
vision_explain: http://yourdomain.com:8003/mcp/vision/explain
```
### 注意事项
- 域名或IP必须是设备能够访问的地址
- 如果使用Docker部署,不能使用Docker内部地址(如127.0.0.1或localhost
- 如果你使用了nginx反向代理,请填写对外的地址和端口号,不是本项目运行的端口号
## 常见问题
### 1. 设备收不到固件更新
**可能原因和解决方法:**
- 检查固件文件命名是否符合规则:`{型号}_{版本号}.bin`
- 检查固件文件是否正确放置在`data/bin/`目录
- 检查设备型号是否与固件文件名中的型号匹配
- 检查固件版本号是否高于设备当前版本
- 查看服务器日志,确认OTA请求是否正常处理
### 2. 设备报告下载地址无法访问
**可能原因和解决方法:**
- 检查`server.vision_explain`配置的域名或IP是否正确
- 确认端口号配置正确(默认8003
- 如果是公网部署,确保设备能够访问该公网地址
- 如果是Docker部署,确保不是使用了内部地址(127.0.0.1)
- 检查防火墙是否开放了对应端口
- 如果你使用了nginx反向代理,请填写对外的地址和端口号,不是本项目运行的端口号
### 3. 如何确认设备当前版本
查看OTA请求日志,日志中会显示设备上报的版本号:
```
[ota_handler] - 设备 AA:BB:CC:DD:EE:FF 固件已是最新: 1.6.6
```
### 4. 固件文件放置后没有生效
系统有30秒的缓存时间(默认),可以:
- 等待30秒后再让设备发起OTA请求
- 重启xiaozhi-server服务
- 调整`firmware_cache_ttl`配置为更短的时间
+17 -7
View File
@@ -1,5 +1,9 @@
# PaddleSpeechTTS集成xiaozhi服务 # PaddleSpeechTTS集成xiaozhi服务
## 重点说明
- 优点:本地离线部署、速度快
- 缺点:截止2025年9月25日,默认的模型是中文模型,不支持英文转语音。如果含英文会发不出声音,如需同时支持中英文需要自己训练。
## 一、基础环境要求 ## 一、基础环境要求
操作系统:Windows / Linux / WSL 2 操作系统:Windows / Linux / WSL 2
@@ -16,15 +20,21 @@ git clone https://github.com/PaddlePaddle/PaddleSpeech.git
``` ```
### 2.建立虚拟环境 ### 2.建立虚拟环境
```bash ```bash
#请根据Paddle官方支持的python版本建立环境 ```https://www.paddlepaddle.org.cn/install```
conda create -n paddle_env python=3.10 -y conda create -n paddle_env python=3.10 -y
conda activate paddle_env conda activate paddle_env
``` ```
### 3.进入paddlespeech目录 ### 3.安装paddle
因CPU架构、GPU架构不同,请根据Paddle官方支持的python版本建立环境
```
https://www.paddlepaddle.org.cn/install
```
### 4.进入paddlespeech目录
```bash ```bash
cd PaddleSpeech cd PaddleSpeech
``` ```
### 4.安装paddlespeech ### 5.安装paddlespeech
```bash ```bash
pip install pytest-runner -i https://pypi.tuna.tsinghua.edu.cn/simple pip install pytest-runner -i https://pypi.tuna.tsinghua.edu.cn/simple
@@ -32,17 +42,17 @@ pip install pytest-runner -i https://pypi.tuna.tsinghua.edu.cn/simple
pip install paddlepaddle -i https://mirror.baidu.com/pypi/simple pip install paddlepaddle -i https://mirror.baidu.com/pypi/simple
pip install paddlespeech -i https://pypi.tuna.tsinghua.edu.cn/simple pip install paddlespeech -i https://pypi.tuna.tsinghua.edu.cn/simple
``` ```
### 5.使用命令自动下载语音模型 ### 6.使用命令自动下载语音模型
```bash ```bash
paddlespeech tts --input "你好,这是一次测试" paddlespeech tts --input "你好,这是一次测试"
``` ```
此步骤会自动下载模型缓存至本地 .paddlespeech/models 目录 此步骤会自动下载模型缓存至本地 .paddlespeech/models 目录
### 6.修改tts_online_application.yaml配置 ### 7.修改tts_online_application.yaml配置
参考目录 ```"PaddleSpeech\demos\streaming_tts_server\conf\tts_online_application.yaml"``` 参考目录 ```"PaddleSpeech\demos\streaming_tts_server\conf\tts_online_application.yaml"```
选择```tts_online_application.yaml```文件用编辑器打开,设置```protocol```为```websocket``` 选择```tts_online_application.yaml```文件用编辑器打开,设置```protocol```为```websocket```
### 7.启动服务 ### 8.启动服务
```yaml ```yaml
paddlespeech_server start --config_file ./demos/streaming_tts_server/conf/tts_online_application.yaml paddlespeech_server start --config_file ./demos/streaming_tts_server/conf/tts_online_application.yaml
#官方默认启动命令: #官方默认启动命令:
@@ -81,7 +91,7 @@ TTS:
```py ```py
python app.py python app.py
``` ```
打开test目录下的test_page.html,测试连接和发送消息时paddlespeech端是否有输出日志 启动`main/digital-human`下的`python start.py`后,打开`http://127.0.0.1:8006/index.html`,测试连接和发送消息时paddlespeech端是否有输出日志
输出日志参考: 输出日志参考:
``` ```
+345
View File
@@ -0,0 +1,345 @@
# PowerMem 记忆组件集成指南
## 简介
[PowerMem](https://www.powermem.ai/) 是由 OceanBase 开源的 Agent 记忆组件,通过本地 LLM 进行记忆总结和智能检索,为 AI 代理提供高效的记忆管理功能。
费用说明:PowerMem 本身开源免费,实际费用取决于您选择的 LLM 和数据库:
- 使用 SQLite + 免费 LLM(如智谱 glm-4-flash= **完全免费**
- 使用云端 LLM 或云端数据库 = 按对应服务收费
> 💡 **最佳性能提示**PowerMem 配合 OceanBase 使用可实现最大性能释放,SQLite 仅建议在资源不足的情况下使用。
- **GitHub**: https://github.com/oceanbase/powermem
- **官网**: https://www.powermem.ai/
- **使用示例**: https://github.com/oceanbase/powermem/tree/main/examples
## 功能特性
- **本地总结**:通过 LLM 在本地进行记忆总结和提取
- **用户画像**:通过 `UserMemory` 自动提取用户信息(姓名、职业、兴趣等),持续更新用户画像
- **智能遗忘**:基于艾宾浩斯遗忘曲线,自动"遗忘"过时噪声信息
- **多种存储后端**:支持 OceanBase(推荐,最佳性能)、SeekDB(推荐,AI应用存储一体)、PostgreSQL、SQLite(轻量备选)
- **多种 LLM 支持**:通义千问、智谱(glm-4-flash 免费)、OpenAI 等
- **智能检索**:基于向量搜索的语义检索能力
- **私有部署**:完全支持本地私有化部署
- **异步操作**:高效的异步记忆管理
## 安装
PowerMem 已添加到项目依赖中,如果需要手动安装:
```bash
pip install powermem
```
## 配置说明
### 基础配置
`config.yaml` 中配置 PowerMem
```yaml
selected_module:
Memory: powermem
Memory:
powermem:
type: powermem
# 是否启用用户画像功能
# 用户画像支持: oceanbase、seekdb、sqlite (powermem 0.3.0+)
enable_user_profile: true
# ========== LLM 配置 ==========
llm:
provider: openai # 可选: qwen, openai, zhipu 等
config:
api_key: 你的LLM API密钥
model: qwen-plus
# openai_base_url: https://api.openai.com/v1 # 可选,自定义服务地址
# ========== Embedding 配置 ==========
embedder:
provider: openai # 可选: qwen, openai 等
config:
api_key: 你的嵌入模型API密钥
model: text-embedding-v4
openai_base_url: https://dashscope.aliyuncs.com/compatible-mode/v1
# embedding_dims: 1024 # 向量维度,非1536时需配置
# ========== Database 配置 ==========
vector_store:
provider: sqlite # 可选: oceanbase(推荐), seekdb(推荐), postgres, sqlite(轻量)
config: {} # SQLite 无需额外配置
```
### 配置参数详解
#### LLM 配置
| 参数 | 说明 | 可选值 |
|------|------|--------|
| `llm.provider` | LLM 提供商 | `qwen`, `openai`, `zhipu` 等 |
| `llm.config.api_key` | API 密钥 | - |
| `llm.config.model` | 模型名称 | 根据提供商选择 |
| `llm.config.openai_base_url` | 自定义服务地址(可选) | - |
#### Embedding 配置
| 参数 | 说明 | 可选值 |
|------|------|--------|
| `embedder.provider` | 嵌入模型提供商 | `qwen`, `openai` 等 |
| `embedder.config.api_key` | API 密钥 | - |
| `embedder.config.model` | 模型名称 | 根据提供商选择 |
| `embedder.config.openai_base_url` | 自定义服务地址(可选) | - |
#### Database 配置
| 参数 | 说明 | 可选值 |
|------|------|--------|
| `vector_store.provider` | 存储后端类型 | `oceanbase`(推荐), `seekdb`(推荐), `postgres`, `sqlite`(轻量) |
| `vector_store.config` | 数据库连接配置 | 根据 provider 设置 |
### 记忆模式说明
PowerMem 支持两种记忆模式:
| 模式 | 配置 | 功能 | 存储要求 |
|------|------|------|----------|
| **普通记忆** | `enable_user_profile: false` | 对话记忆存储与检索 | 支持所有数据库 |
| **用户画像** | `enable_user_profile: true` | 记忆 + 自动提取用户画像 | oceanbase、seekdb、sqlite |
> 📌 **版本说明**PowerMem 0.3.0+ 版本,用户画像功能支持 OceanBase、SeekDB、SQLite 三种存储后端。
### 使用通义千问(推荐)
1. 访问 [阿里云百炼平台](https://bailian.console.aliyun.com/) 注册账号
2. 在 [API Key 管理](https://bailian.console.aliyun.com/?apiKey=1#/api-key) 页面获取 API 密钥
3. 配置如下:
```yaml
Memory:
powermem:
type: powermem
enable_user_profile: true
llm:
provider: qwen
config:
api_key: sk-xxxxxxxxxxxxxxxx
model: qwen-plus
embedder:
provider: openai
config:
api_key: sk-xxxxxxxxxxxxxxxx
model: text-embedding-v4
openai_base_url: https://dashscope.aliyuncs.com/compatible-mode/v1
vector_store:
provider: sqlite
config: {}
```
### 使用智谱免费 LLM(完全免费方案)
智谱提供免费的 glm-4-flash 模型,配合 SQLite 可实现完全免费使用:
1. 访问 [智谱AI开放平台](https://bigmodel.cn/) 注册账号
2. 在 [API Keys](https://bigmodel.cn/usercenter/proj-mgmt/apikeys) 页面获取 API 密钥
3. 配置如下:
```yaml
Memory:
powermem:
type: powermem
enable_user_profile: true
llm:
provider: openai # 使用 openai 兼容模式
config:
api_key: xxxxxxxxxxxxxxxx.xxxxxxxxxxxxxxxx
model: glm-4-flash
openai_base_url: https://open.bigmodel.cn/api/paas/v4/
embedder:
provider: openai
config:
api_key: xxxxxxxxxxxxxxxx.xxxxxxxxxxxxxxxx
model: embedding-3
openai_base_url: https://open.bigmodel.cn/api/paas/v4/
vector_store:
provider: sqlite
config: {}
```
### 使用 OpenAI
```yaml
Memory:
powermem:
type: powermem
enable_user_profile: true
llm:
provider: openai
config:
api_key: sk-xxxxxxxxxxxxxxxx
model: gpt-4o-mini
openai_base_url: https://api.openai.com/v1
embedder:
provider: openai
config:
api_key: sk-xxxxxxxxxxxxxxxx
model: text-embedding-3-small
openai_base_url: https://api.openai.com/v1
vector_store:
provider: sqlite
config: {}
```
### 使用 OceanBase(最佳性能方案)
OceanBase 是 PowerMem 的最佳搭档,可实现最大性能释放:
1. 部署 OceanBase 数据库(支持开源本地部署或使用云服务)
- 开源部署:https://github.com/oceanbase/oceanbase
- 云服务:https://www.oceanbase.com/
2. 配置如下:
```yaml
Memory:
powermem:
type: powermem
enable_user_profile: true
llm:
provider: qwen
config:
api_key: sk-xxxxxxxxxxxxxxxx
model: qwen-plus
embedder:
provider: openai
config:
api_key: sk-xxxxxxxxxxxxxxxx
model: text-embedding-v4
openai_base_url: https://dashscope.aliyuncs.com/compatible-mode/v1
vector_store:
provider: oceanbase
config:
host: 127.0.0.1
port: 2881
user: root@test
password: your_password
db_name: powermem
collection_name: memories # 默认值
embedding_model_dims: 1536 # 嵌入向量维度,必需参数
```
## 设备记忆隔离
PowerMem 会自动使用设备 ID`device_id`)作为 `user_id` 进行记忆隔离。这意味着:
- 每个设备拥有独立的记忆空间
- 不同设备之间的记忆完全隔离
- 同一设备的多次对话可以共享记忆上下文
## 用户画像(UserMemory
PowerMem 提供 `UserMemory` 类,可自动从对话中提取用户画像信息。
> 📌 **版本说明**PowerMem 0.3.0+ 版本,用户画像功能支持 OceanBase、SeekDB、SQLite 三种存储后端。
### 启用用户画像
在配置中设置 `enable_user_profile: true` 即可启用:
```yaml
Memory:
powermem:
type: powermem
enable_user_profile: true # 启用用户画像
llm:
provider: qwen
config:
api_key: sk-xxxxxxxxxxxxxxxx
model: qwen-plus
embedder:
provider: openai
config:
api_key: sk-xxxxxxxxxxxxxxxx
model: text-embedding-v4
openai_base_url: https://dashscope.aliyuncs.com/compatible-mode/v1
vector_store:
provider: sqlite # 用户画像支持: oceanbase、seekdb、sqlite
config: {}
```
### 用户画像能力
| 能力 | 说明 |
|------|------|
| **信息提取** | 自动从对话中提取姓名、年龄、职业、兴趣等 |
| **持续更新** | 随着对话进行,不断完善用户画像 |
| **画像检索** | 将用户画像与记忆搜索结合,提升检索相关性 |
| **智能遗忘** | 基于艾宾浩斯遗忘曲线,淡化过时信息 |
### 工作原理
启用用户画像后,小智在查询记忆时会自动返回:
1. **用户画像**:用户的基本信息、兴趣爱好等
2. **相关记忆**:与当前对话相关的历史记忆
> ✅ **版本说明**PowerMem 0.3.0+ 版本,用户画像功能支持 OceanBase、SeekDB、SQLite 三种存储后端。
## 与其他记忆组件的对比
| 特性 | PowerMem | mem0ai | mem_local_short |
|------|----------|--------|-----------------|
| 工作方式 | 本地总结 | 云端接口 | 本地总结 |
| 存储位置 | 本地/云端DB | 云端 | 本地YAML |
| 费用 | 取决于LLM和DB | 1000次/月免费 | 完全免费 |
| 智能检索 | ✅ 向量搜索 | ✅ 向量搜索 | ❌ 全量返回 |
| 用户画像 | ✅ UserMemory | ❌ | ❌ |
| 智能遗忘 | ✅ 遗忘曲线 | ❌ | ❌ |
| 私有部署 | ✅ 支持 | ❌ 仅云端 | ✅ 支持 |
| 数据库支持 | OceanBase(推荐)/SeekDB/PostgreSQL/SQLite | - | YAML 文件 |
## 常见问题
### 1. API 密钥错误
如果出现 `API key is required` 错误,请检查:
- `llm_api_key``embedding_api_key` 是否正确填写
- API 密钥是否有效
### 2. 模型不存在
如果出现模型不存在的错误,请确认:
- `llm_model``embedding_model` 名称是否正确
- 对应的模型服务是否已开通
### 3. 连接超时
如果出现连接超时,可以尝试:
- 检查网络连接
- 如果使用代理,配置 `llm_base_url``embedding_base_url`
## 测试验证
可以在虚拟环境中测试 PowerMem 是否正常工作:
```bash
# 激活虚拟环境
source .venv/bin/activate
# 测试 PowerMem 导入
python -c "from powermem import AsyncMemory; print('PowerMem 导入成功')"
# 测试 UserMemory 导入(用户画像功能)
python -c "from powermem import UserMemory; print('UserMemory 导入成功')"
```
## 更多资源
- [PowerMem 官方文档](https://www.powermem.ai/)
- [PowerMem GitHub 仓库](https://github.com/oceanbase/powermem)
- [PowerMem 使用示例](https://github.com/oceanbase/powermem/tree/main/examples)
- [OceanBase 官网](https://www.oceanbase.com/)
- [OceanBase GitHub](https://github.com/oceanbase/oceanbase)
- [SeekDB GitHub](https://github.com/oceanbase/seekdb)AI原生搜索数据库)
- [阿里云百炼平台](https://bailian.console.aliyun.com/)
+269
View File
@@ -0,0 +1,269 @@
# ragflow 集成指南
本教程主要是是两部分
- 一、如何部署ragflow
- 二、如何在智控台配置ragflow接口
如果您对ragflow很熟悉,且已经部署了ragflow,可直接跳过第一部分,直接进入第二部分。但是如果你希望有人指导你部署ragflow,让它能够和`xiaozhi-esp32-server`共同使用`mysql``redis`基础服务,以减少资源成本,你需要从第一部分开始。
# 第一部分 如何部署ragflow
## 第一步, 确认mysql、redis是否可用
ragflow需要依赖`mysql`数据库。如果你之前已经部署`智控台`,说明你已经安装了`mysql`。你可以共用它。
你可以你试一下在宿主机使用`telnet`命令,看看能不能正常访问`mysql``3306`端口。
``` shell
telnet 127.0.0.1 3306
telnet 127.0.0.1 6379
```
如果能访问到`3306`端口和`6379`端口,请忽略以下的内容,直接进入第二步。
如果不能访问,你需要回忆一下,你的`mysql`是怎么安装的。
如果你的mysql是通过自己使用安装包安装的,说明你的`mysql`做了网络隔离。你可能先解决访问`mysql`的`3306`端口这个问题。
如果你`mysql`是通过本项目的`docker-compose_all.yml`安装的。你需要找一下你当时创建数据库的`docker-compose_all.yml`文件,修改以下的内容
修改前
``` yaml
xiaozhi-esp32-server-db:
...
networks:
- default
expose:
- "3306:3306"
xiaozhi-esp32-server-redis:
...
expose:
- 6379
```
修改后
``` yaml
xiaozhi-esp32-server-db:
...
networks:
- default
ports:
- "3306:3306"
xiaozhi-esp32-server-redis:
...
ports:
- "6379:6379"
```
注意是将`xiaozhi-esp32-server-db`和`xiaozhi-esp32-server-redis`下面的`expose`改成`ports`。改完后,需要重新启动。以下是重启mysql的命令:
``` shell
# 进入你docker-compose_all.yml所在的文件夹,例如我的是xiaozhi-server
cd xiaozhi-server
docker compose -f docker-compose_all.yml down
docker compose -f docker-compose.yml up -d
```
启动完后,在宿主机再使用`telnet`命令,看看能不能正常访问`mysql`的`3306`端口。
``` shell
telnet 127.0.0.1 3306
telnet 127.0.0.1 6379
```
正常来说这样就可以访问的了。
## 第二步, 创建数据库和表
如果你的宿主机,能正常访问mysql数据库,那就在mysql上创建一个名字为`rag_flow`的数据库和`rag_flow`用户,密码为`infini_rag_flow`。
``` sql
-- 创建数据库
CREATE DATABASE IF NOT EXISTS rag_flow CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci;
-- 创建用户并授权
CREATE USER IF NOT EXISTS 'rag_flow'@'%' IDENTIFIED BY 'infini_rag_flow';
GRANT ALL PRIVILEGES ON rag_flow.* TO 'rag_flow'@'%';
-- 刷新权限
FLUSH PRIVILEGES;
```
## 第三步, 下载ragflow项目
你需要在你电脑找一个文件夹,用来存放ragflow项目。例如我在`/home/system/xiaozhi`文件夹。
你可以使用`git`命令,将ragflow项目下载到这个文件夹,本教程使用的是`v0.22.0`版本进行安装部署。
```
git clone https://ghfast.top/https://github.com/infiniflow/ragflow.git
cd ragflow
git checkout v0.22.0
```
下载完后,进入`docker`文件夹。
``` shell
cd docker
```
修改`ragflow/docker`文件夹下的`docker-compose.yml`文件,将`ragflow-cpu`和`ragflow-gpu`服务的`depends_on`配置去掉,用于解除`ragflow-cpu`服务对`mysql`的依赖。
这是修改前:
``` yaml
ragflow-cpu:
depends_on:
mysql:
condition: service_healthy
profiles:
- cpu
...
ragflow-gpu:
depends_on:
mysql:
condition: service_healthy
profiles:
- gpu
```
这是修改后:
``` yaml
ragflow-cpu:
profiles:
- cpu
...
ragflow-gpu:
profiles:
- gpu
```
接着,修改`ragflow/docker`文件夹下的`docker-compose-base.yml`文件,去掉`mysql`和`redis`的配置。
例如,删除前:
``` yaml
services:
minio:
image: quay.io/minio/minio:RELEASE.2025-06-13T11-33-47Z
...
mysql:
image: mysql:8.0
...
redis:
image: redis:6.2-alpine
...
```
删除后
``` yaml
services:
minio:
image: quay.io/minio/minio:RELEASE.2025-06-13T11-33-47Z
...
```
## 第四步,修改环境变量配置
编辑`ragflow/docker`文件夹下的`.env`文件,找到以下配置,逐个搜索,逐个修改!逐个搜索,逐个修改!
下面对于`.env`文件的修改,60%的人会忽略`MYSQL_USER`配置导致ragflow启动不成功,因此,需要强调三次:
强调第一次:如果你的`.env`文件如果没有`MYSQL_USER`配置,请在配置文件增加这项!
强调第二次:如果你的`.env`文件如果没有`MYSQL_USER`配置,请在配置文件增加这项!
强调第三次:如果你的`.env`文件如果没有`MYSQL_USER`配置,请在配置文件增加这项!
``` env
# 端口设置
SVR_WEB_HTTP_PORT=8008 # HTTP端口
SVR_WEB_HTTPS_PORT=8009 # HTTPS端口
# MySQL配置 - 修改为您本地MySQL的信息
MYSQL_HOST=host.docker.internal # 使用host.docker.internal让容器访问主机服务
MYSQL_PORT=3306 # 本地MySQL端口
MYSQL_USER=rag_flow # 上面创建的用户名,如果没有这项就增加这一项
MYSQL_PASSWORD=infini_rag_flow # 上面设置的密码
MYSQL_DBNAME=rag_flow # 数据库名称
# Redis配置 - 修改为您本地Redis的信息
REDIS_HOST=host.docker.internal # 使用host.docker.internal让容器访问主机服务
REDIS_PORT=6379 # 本地Redis端口
REDIS_PASSWORD= # 如果你的Redis没有设置密码,就按这样子填写,否则填写密码
```
注意,如果你的Redis没有设置密码,还要修改`ragflow/docker`文件夹下`service_conf.yaml.template`,将`infini_rag_flow`替换成空字符串。
修改前
``` shell
redis:
db: 1
password: '${REDIS_PASSWORD:-infini_rag_flow}'
host: '${REDIS_HOST:-redis}:6379'
```
修改后
``` shell
redis:
db: 1
password: '${REDIS_PASSWORD:-}'
host: '${REDIS_HOST:-redis}:6379'
```
## 第五步,启动ragflow服务
执行命令:
``` shell
docker-compose -f docker-compose.yml up -d
```
执行成功后,你可以使用`docker logs -n 20 -f docker-ragflow-cpu-1`命令,查看`docker-ragflow-cpu-1`服务的日志。
如果日志中没有报错,说明ragflow服务启动成功。
# 第五步,注册账号
你可以在浏览器中访问`http://127.0.0.1:8008`,点击`Sign Up`,注册一个账号。
注册成功后,你可以点击`Sign In`,登录到ragflow服务。如果你想关闭ragflow服务的注册服务,不想让其他人注册账号,你可以在`ragflow/docker`文件夹下的`.env`文件中,将`REGISTER_ENABLED`配置项设置为`0`。
``` dotenv
REGISTER_ENABLED=0
```
修改后,重启启动ragflow服务。
``` shell
docker-compose -f docker-compose.yml down
docker-compose -f docker-compose.yml up -d
```
# 第六步,配置ragflow服务的模型
你可以在浏览器中访问`http://127.0.0.1:8008`,点击`Sign In`,登录到ragflow服务。点击页面右上角的`头像`,进入设置页面。
首先,在左侧导航栏中,点击`模型供应商`,进入到模型配置页面。在右侧的`可选模型`搜索框下,选择`LLM`,在列表选择你使用的模型供应商,点击`添加`,输入你的密钥;
然后,选择`TEXT EMBEDDING`,在列表选择你使用的模型供应商,点击`添加`,输入你的密钥。
最后,刷新一下页面,分别点击`设置默认模型`列表的LLM和Embedding,选择你使用的模型即可。请确认你的密钥开通了相应的服务,比如我是用的Embedding模型是xxx供应商的,需要去这个供应商官网查看这个模型是否需要购买资源包才能使用。
# 第二部分 配置ragflow服务
# 第一步 登录ragflow服务
你可以在浏览器中访问`http://127.0.0.1:8008`,点击`Sign In`,登录到ragflow服务。
然后点击右上角的`头像`,进入设置页面。在左侧导航栏中,点击`API`功能,然后点击"API Key"按钮。出现一个弹框,
在弹框中,点击"Create new Key"按钮,生成一个API Key。复制这个`API Key`,你稍后会用到。
# 第二步 配置到智控台
确保你的智控台版本是`0.8.7`或以上。使用超级管理员账号登录到智控台。
首先,你要先开启知识库功能。在顶部导航栏中,点击`参数字典`,在下拉菜单中,点击`系统功能配置`页面。在页面上勾选`知识库`,点击`保存配置`。即可在导航栏看到`知识库`功能。
在顶部导航栏中,点击`模型配置`,在左侧导航栏中,点击`知识库`。在列表中找到`RAG_RAGFlow`,点击`编辑`按钮。
在`服务地址`中,填写`http://你的ragflow服务的局域网IP:8008`,例如我的ragflow服务的局域网IP是`192.168.1.100`,那么我就填写`http://192.168.1.100:8008`。
在`API密钥`中,填写之前复制的`API Key`。
最后点击保存按钮。
# 第二步 创建一个知识库
使用超级管理员账号登录到智控台。在顶部导航栏中,点击`知识库`,在列表左下脚,点击`新增`按钮。填写一个知识库的名字和描述。点击保存。
为了提高大模型对知识库的理解和召回能力,建议在创建知识库时,填写一个有意义的名字和描述。例如,如果你要创建一个关于`公司介绍`的知识库,那么知识库的名字可以是`公司介绍`,描述可以是`关于公司的相关信息例如公司基本信息、服务项目、联系电话、地址等。`。
保存后,你可以在知识库列表中看到这个知识库。点击刚才创建的知识库的`查看`按钮,进入知识库详情页面。
在知识库详情页面中,左下角点击`新增`按钮,可以上传文档到知识库。
上传后,你可以在知识库详情页面中,看到上传的文档。此时可以点击文档的`解析`按钮,解析文档。
解析完成后,你可以查看解析后的切片信息。你可以在知识库详情页面中,点击`召回测试`按钮,可以测试知识库的召回/检索功能。
# 第三步 让小智使用ragflow知识库
登录到智控台。在顶部导航栏中,点击`智能体`,找到你要配置的智能体,点击`配置角色`按钮。
在意图识别左侧,点击`编辑功能`按钮,弹出一个弹框。在弹框中选择你要添加的知识库。保存即可。
+361
View File
@@ -0,0 +1,361 @@
[![Banners](../images/banner1.png)](https://github.com/xinnan-tech/xiaozhi-esp32-server)
<h1 align="center">Xiaozhi Backend-Service xiaozhi-esp32-server</h1>
<p align="center">
Dieses Projekt basiert auf der Theorie und Technologie der Mensch-Maschine-symbiotischen Intelligenz zur Entwicklung intelligenter Terminal-Hardware- und Software-Systeme<br/>und bietet Backend-Dienste für das Open-Source-Hardware-Projekt
<a href="https://github.com/78/xiaozhi-esp32">xiaozhi-esp32</a><br/>
Implementiert mit Python, Java und Vue gemäß dem <a href="https://ccnphfhqs21z.feishu.cn/wiki/M0XiwldO9iJwHikpXD5cEx71nKh">Xiaozhi-Kommunikationsprotokoll</a><br/>
Unterstützt MQTT+UDP-Protokoll, Websocket-Protokoll, MCP-Endpunkte und Stimmabdruckerkennung
</p>
<p align="center">
<a href="../FAQ.md">Häufige Fragen</a>
· <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/issues">Probleme melden</a>
· <a href="../../README.md#%E9%83%A8%E7%BD%B2%E6%96%87%E6%A1%A3">Deployment-Dokumentation</a>
· <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/releases">Release-Hinweise</a>
</p>
<p align="center">
<a href="../../README.md"><img alt="简体中文版自述文件" src="https://img.shields.io/badge/简体中文-DFE0E5"></a>
<a href="./README_en.md"><img alt="README in English" src="https://img.shields.io/badge/English-DFE0E5"></a>
<a href="./README_vi.md"><img alt="Tiếng Việt" src="https://img.shields.io/badge/Tiếng Việt-DFE0E5"></a>
<a href="./README_de.md"><img alt="Deutsch" src="https://img.shields.io/badge/Deutsch-DBEDFA"></a>
<a href="./README_pt_BR.md"><img alt="Português (Brasil)" src="https://img.shields.io/badge/Português (Brasil)-DFE0E5"></a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/releases">
<img alt="GitHub Contributors" src="https://img.shields.io/github/v/release/xinnan-tech/xiaozhi-esp32-server?logo=docker" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/blob/main/LICENSE">
<img alt="GitHub pull requests" src="https://img.shields.io/badge/license-MIT-white?labelColor=black" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server">
<img alt="stars" src="https://img.shields.io/github/stars/xinnan-tech/xiaozhi-esp32-server?color=ffcb47&labelColor=black" />
</a>
</p>
<p align="center">
Geleitet vom Team von Professor Siyuan Liu (South China University of Technology)
</br>
刘思源教授团队主导研发(华南理工大学)
</br>
<img src="../images/hnlg.jpg" alt="South China University of Technology" width="50%">
</p>
---
## Zielgruppe 👥
Dieses Projekt erfordert ESP32-Hardware-Geräte zum Betrieb. Wenn Sie ESP32-bezogene Hardware erworben haben, erfolgreich eine Verbindung zu Brother Xias bereitgestelltem Backend-Service hergestellt haben und Ihren eigenen `xiaozhi-esp32`-Backend-Service unabhängig aufbauen möchten, dann ist dieses Projekt perfekt für Sie.
Möchten Sie die Nutzungseffekte sehen? Klicken Sie auf die Videos unten 🎥
<table>
<tr>
<td>
<a href="https://www.bilibili.com/video/BV1FMFyejExX" target="_blank">
<picture>
<img alt="响应速度感受" src="docs/images/demo9.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1vchQzaEse" target="_blank">
<picture>
<img alt="速度优化秘诀" src="docs/images/demo6.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1WEcxzFEAT" target="_blank">
<picture>
<img alt="小智数字人 支持语音唤醒" src="docs/images/demo8.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1CKVz6UEuB" target="_blank">
<picture>
<img alt="设备呼叫设备,打电话" src="docs/images/demo0.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1C1tCzUEZh" target="_blank">
<picture>
<img alt="复杂医疗场景" src="docs/images/demo1.png" /></picture>
</a>
</td>
</tr>
<tr>
<td>
<a href="https://www.bilibili.com/video/BV1VC96Y5EMH" target="_blank">
<picture>
<img alt="播放音乐查询天气播报新闻" src="docs/images/demo7.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV12J7WzBEaH" target="_blank">
<picture>
<img alt="实时打断" src="docs/images/demo10.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1Co76z7EvK" target="_blank">
<picture>
<img alt="拍照识物品" src="docs/images/demo12.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1pNXWYGEx1" target="_blank">
<picture>
<img alt="控制家电开关" src="docs/images/demo5.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1TJ7WzzEo6" target="_blank">
<picture>
<img alt="多指令任务" src="docs/images/demo11.png" /></picture>
</a>
</td>
</tr>
<tr>
<td>
<a href="https://www.bilibili.com/video/BV1ZQKUzYExM" target="_blank">
<picture>
<img alt="MCP接入点" src="docs/images/demo13.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1zUW5zJEkq" target="_blank">
<picture>
<img alt="MQTT指令下发" src="docs/images/demo4.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1Exu3zqEDe" target="_blank">
<picture>
<img alt="声纹识别" src="docs/images/demo14.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1CDKWemEU6" target="_blank">
<picture>
<img alt="自定义音色" src="docs/images/demo2.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV12yA2egEaC" target="_blank">
<picture>
<img alt="使用粤语交流" src="docs/images/demo3.png" /></picture>
</a>
</td>
</tr>
</table>
---
## Warnungen ⚠️
1. Dieses Projekt ist Open-Source-Software. Diese Software hat keine kommerzielle Partnerschaft mit Drittanbieter-API-Dienstleistern (einschließlich, aber nicht beschränkt auf Spracherkennung, große Modelle, Sprachsynthese und andere Plattformen), mit denen sie sich verbindet, und bietet keine Garantie für deren Servicequalität oder finanzielle Sicherheit. Es wird empfohlen, dass Benutzer Dienstleister mit entsprechenden Geschäftslizenzen bevorzugen und deren Servicevereinbarungen und Datenschutzrichtlinien sorgfältig lesen. Diese Software hostet keine Kontoschlüssel, nimmt nicht an Geldströmen teil und trägt nicht das Risiko von Verlusten bei Guthaben-Aufladungen.
2. Die Funktionalität dieses Projekts ist nicht vollständig und hat keine Netzwerksicherheitsbewertung bestanden. Bitte verwenden Sie es nicht in Produktionsumgebungen. Wenn Sie dieses Projekt zu Lernzwecken in einer öffentlichen Netzwerkumgebung bereitstellen, stellen Sie bitte sicher, dass notwendige Schutzmaßnahmen vorhanden sind.
---
## Deployment-Dokumentation
![Banners](../images/banner2.png)
Dieses Projekt bietet zwei Bereitstellungsmethoden. Bitte wählen Sie basierend auf Ihren spezifischen Anforderungen:
#### 🚀 Auswahl der Bereitstellungsmethode
| Bereitstellungsmethode | Funktionen | Anwendungsszenarien | Deployment-Dokumente | Konfigurationsanforderungen | Video-Tutorials |
|---------|------|---------|---------|---------|---------|
| **Vereinfachte Installation** | Intelligenter Dialog, Einzel-Agenten-Verwaltung | Umgebungen mit geringer Konfiguration, Daten in Konfigurationsdateien gespeichert, keine Datenbank erforderlich | [①Docker-Version](../Deployment.md#%E6%96%B9%E5%BC%8F%E4%B8%80docker%E5%8F%AA%E8%BF%90%E8%A1%8Cserver) / [②Quellcode-Deployment](../Deployment.md#%E6%96%B9%E5%BC%8F%E4%BA%8C%E6%9C%AC%E5%9C%B0%E6%BA%90%E7%A0%81%E5%8F%AA%E8%BF%90%E8%A1%8Cserver)| 2 Kerne 4GB bei Verwendung von `FunASR`, 2 Kerne 2GB bei allen APIs | - |
| **Vollständige Modulinstallation** | Intelligenter Dialog, Mehrbenutzerverwaltung, Mehr-Agenten-Verwaltung, Intelligente Steuerkonsole-Bedienung | Vollständige Funktionserfahrung, Daten in Datenbank gespeichert |[①Docker-Version](../Deployment_all.md#%E6%96%B9%E5%BC%8F%E4%B8%80docker%E8%BF%90%E8%A1%8C%E5%85%A8%E6%A8%A1%E5%9D%97) / [②Quellcode-Deployment](../Deployment_all.md#%E6%96%B9%E5%BC%8F%E4%BA%8C%E6%9C%AC%E5%9C%B0%E6%BA%90%E7%A0%81%E8%BF%90%E8%A1%8C%E5%85%A8%E6%A8%A1%E5%9D%97) / [③Quellcode-Deployment Auto-Update-Tutorial](../dev-ops-integration.md) | 4 Kerne 8GB bei Verwendung von `FunASR`, 2 Kerne 4GB bei allen APIs| [Video-Tutorial für lokalen Quellcode-Start](https://www.bilibili.com/video/BV1wBJhz4Ewe) |
Häufige Fragen und entsprechende Tutorials finden Sie unter [diesem Link](../FAQ.md)
> 💡 Hinweis: Unten ist eine Testplattform, die mit dem neuesten Code bereitgestellt wurde. Sie können bei Bedarf brennen und testen. Gleichzeitige Benutzer: 6, Daten werden täglich gelöscht.
```
Adresse der intelligenten Steuerkonsole: https://2662r3426b.vicp.fun
Adresse der intelligenten Steuerkonsole (H5): https://2662r3426b.vicp.fun/h5/index.html
Service-Testtool: https://2662r3426b.vicp.fun/test/
OTA-Schnittstellenadresse: https://2662r3426b.vicp.fun/xiaozhi/ota/
Websocket-Schnittstellenadresse: wss://2662r3426b.vicp.fun/xiaozhi/v1/
```
#### 🚩 Konfigurationsbeschreibung und Empfehlungen
> [!Note]
> Dieses Projekt bietet zwei Konfigurationsschemata:
>
> 1. `Einstiegslevel Kostenlose Einstellungen`: Geeignet für den persönlichen und privaten Gebrauch, alle Komponenten verwenden kostenlose Lösungen, keine zusätzliche Zahlung erforderlich.
>
> 2. `Streaming-Konfiguration`: Geeignet für Demonstrationen, Schulungen, Szenarien mit mehr als 2 gleichzeitigen Benutzern usw. Verwendet Streaming-Verarbeitungstechnologie für schnellere Reaktionsgeschwindigkeit und bessere Erfahrung.
>
> Ab Version `0.5.2` unterstützt das Projekt Streaming-Konfiguration. Im Vergleich zu früheren Versionen ist die Reaktionsgeschwindigkeit um ca. `2,5 Sekunden` verbessert, was die Benutzererfahrung erheblich verbessert.
| Modulname | Einstiegslevel Kostenlose Einstellungen | Streaming-Konfiguration |
|:---:|:---:|:---:|
| ASR (Spracherkennung) | FunASR (Lokal) | 👍XunfeiStreamASR (Xunfei-Streaming) |
| LLM (Großes Modell) | glm-4-flash (Zhipu) | 👍qwen-flash (Alibaba Bailian) |
| VLLM (Vision Large Model) | glm-4v-flash (Zhipu) | 👍qwen3.5-flash (Alibaba Bailian) |
| TTS (Sprachsynthese) | EdgeTTS (Microsoft) | 👍HuoshanDoubleStreamTTS (Volcano-Streaming) |
| Intent (Absichtserkennung) | function_call (Funktionsaufruf) | function_call (Funktionsaufruf) |
| Memory (Gedächtnisfunktion) | mem_local_short (Lokales Kurzzeitgedächtnis) | mem_local_short (Lokales Kurzzeitgedächtnis) |
Wenn Sie sich um die Latenz jeder Komponente kümmern, lesen Sie bitte den [Xiaozhi-Komponenten-Leistungstestbericht](https://github.com/xinnan-tech/xiaozhi-performance-research). Sie können gemäß den Testmethoden im Bericht in Ihrer Umgebung tatsächlich testen.
#### 🔧 Testwerkzeuge
Dieses Projekt bietet die folgenden Testwerkzeuge, um Ihnen bei der Überprüfung des Systems und der Auswahl geeigneter Modelle zu helfen:
| Werkzeugname | Standort | Verwendungsmethode | Funktionsbeschreibung |
|:---:|:---|:---:|:---:|
| Audio-Interaktionstesttool | main》digital-human》index.html | In `main/digital-human` `python start.py` ausführen und dann `http://127.0.0.1:8006/index.html` öffnen | Testet Audio-Wiedergabe- und Empfangsfunktionen, überprüft, ob die Python-seitige Audioverarbeitung normal ist |
| Modell-Reaktionstesttool | main》xiaozhi-server》performance_tester.py | Ausführen `python performance_tester.py` | Testet die Reaktionsgeschwindigkeit von ASR (Spracherkennung), LLM (großes Modell), VLLM (Vision-Modell), TTS (Sprachsynthese) drei Kernmodulen |
> 💡 Hinweis: Beim Testen der Modellgeschwindigkeit werden nur Modelle mit konfigurierten Schlüsseln getestet.
---
## Funktionsliste ✨
### Implementiert ✅
![请参考-全模块安装架构图](../images/deploy2.png)
| Funktionsmodul | Beschreibung |
|:---:|:---|
| Kernarchitektur | Basierend auf [MQTT+UDP-Gateway](https://github.com/xinnan-tech/xiaozhi-esp32-server/blob/main/docs/mqtt-gateway-integration.md), WebSocket und HTTP-Servern, bietet vollständiges Konsolenverwaltungs- und Authentifizierungssystem |
| Sprachinteraktion | Unterstützt Streaming-ASR (Spracherkennung), Streaming-TTS (Sprachsynthese), VAD (Sprachaktivitätserkennung), unterstützt mehrsprachige Erkennung und Sprachverarbeitung |
| Stimmabdruckerkennung | Unterstützt Mehrbenutzer-Stimmabdruckregistrierung, -verwaltung und -erkennung, verarbeitet parallel mit ASR, Echtzeit-Sprecheridentitätserkennung und Weitergabe an LLM für personalisierte Antworten |
| Intelligenter Dialog | Unterstützt mehrere LLM (große Sprachmodelle), implementiert intelligenten Dialog |
| Visuelle Wahrnehmung | Unterstützt mehrere VLLM (Vision Large Models), implementiert multimodale Interaktion |
| Absichtserkennung | Unterstützt LLM-Absichtserkennung, Function Call-Funktionsaufruf, bietet plugin-basierten Absichtsverarbeitungsmechanismus |
| Gedächtnissystem | Unterstützt lokales Kurzzeitgedächtnis, mem0ai-Schnittstellengedächtnis, PowerMem intelligentes Gedächtnis, mit Gedächtniszusammenfassungsfunktion |
| Wissensdatenbank | Unterstützt RAGFlow-Wissensdatenbank, ermöglicht großem Modell die Bewertung, ob Wissensdatenbank benötigt wird, bevor geantwortet wird |
| Werkzeugaufruf | Unterstützt Client-IOT-Protokoll, Client-MCP-Protokoll, Server-MCP-Protokoll, MCP-Endpunktprotokoll, benutzerdefinierte Werkzeugfunktionen |
| Befehlsübermittlung | Basierend auf MQTT-Protokoll, unterstützt die Übermittlung von MCP-Befehlen von der intelligenten Steuerkonsole an ESP32-Geräte |
| Verwaltungs-Backend | Bietet Web-Verwaltungsoberfläche, unterstützt Benutzerverwaltung, Systemkonfiguration und Geräteverwaltung; Oberfläche unterstützt vereinfachtes Chinesisch, traditionelles Chinesisch und englische Anzeige |
| Testwerkzeuge | Bietet Leistungstestwerkzeuge, Vision-Modell-Testwerkzeuge und Audio-Interaktionstestwerkzeuge |
| Deployment-Unterstützung | Unterstützt Docker-Deployment und lokales Deployment, bietet vollständige Konfigurationsdateiverwaltung |
| Plugin-System | Unterstützt funktionale Plugin-Erweiterungen, benutzerdefinierte Plugin-Entwicklung und Plugin-Hot-Loading |
### In Entwicklung 🚧
Um über spezifische Entwicklungsplanfortschritte zu erfahren, [klicken Sie hier](https://github.com/users/xinnan-tech/projects/3). Häufige Fragen und entsprechende Tutorials finden Sie unter [diesem Link](../FAQ.md)
Wenn Sie ein Softwareentwickler sind, finden Sie hier einen [Offenen Brief an Entwickler](../contributor_open_letter.md). Willkommen beim Beitritt!
---
## Produktökosystem 👬
Xiaozhi ist ein Ökosystem. Wenn Sie dieses Produkt verwenden, können Sie sich auch andere [hervorragende Projekte](https://github.com/78/xiaozhi-esp32/blob/main/README_zh.md#%E7%9B%B8%E5%85%B3%E5%BC%80%E6%BA%90%E9%A1%B9%E7%9B%AE) in diesem Ökosystem ansehen
---
## Liste der von diesem Projekt unterstützten Plattformen/Komponenten 📋
### LLM-Sprachmodelle
| Verwendungsmethode | Unterstützte Plattformen | Kostenlose Plattformen |
|:---:|:---:|:---:|
| OpenAI-Schnittstellenaufrufe | Alibaba Bailian, Volcano Engine Doubao, DeepSeek, Zhipu ChatGLM, Gemini | Zhipu ChatGLM, Gemini |
| Ollama-Schnittstellenaufrufe | Ollama | - |
| Dify-Schnittstellenaufrufe | Dify | - |
| FastGPT-Schnittstellenaufrufe | FastGPT | - |
| Coze-Schnittstellenaufrufe | Coze | - |
| Xinference-Schnittstellenaufrufe | Xinference | - |
| HomeAssistant-Schnittstellenaufrufe | HomeAssistant | - |
Tatsächlich kann jedes LLM, das OpenAI-Schnittstellenaufrufe unterstützt, integriert und verwendet werden.
---
### VLLM-Vision-Modelle
| Verwendungsmethode | Unterstützte Plattformen | Kostenlose Plattformen |
|:---:|:---:|:---:|
| OpenAI-Schnittstellenaufrufe | Alibaba Bailian, Zhipu ChatGLMVLLM | Zhipu ChatGLMVLLM |
Tatsächlich kann jedes VLLM, das OpenAI-Schnittstellenaufrufe unterstützt, integriert und verwendet werden.
---
### TTS-Sprachsynthese
| Verwendungsmethode | Unterstützte Plattformen | Kostenlose Plattformen |
|:---:|:---:|:---:|
| Schnittstellenaufrufe | EdgeTTS, Volcano Engine Doubao TTS, Tencent Cloud, Alibaba Cloud TTS, AliYun Stream TTS, CosyVoiceSiliconflow, TTS302AI, CozeCnTTS, GizwitsTTS, ACGNTTS, OpenAITTS, Lingxi Streaming TTS, MinimaxTTS, Volcano Dual-Stream TTS | Lingxi Streaming TTS, EdgeTTS, CosyVoiceSiliconflow (teilweise) |
| Lokale Dienste | FishSpeech, GPT_SOVITS_V2, GPT_SOVITS_V3, Index-TTS, PaddleSpeech | Index-TTS, PaddleSpeech, FishSpeech, GPT_SOVITS_V2, GPT_SOVITS_V3 |
---
### VAD-Sprachaktivitätserkennung
| Typ | Plattformname | Verwendungsmethode | Preismodell | Hinweise |
|:---:|:---------:|:----:|:----:|:--:|
| VAD | SileroVAD | Lokale Verwendung | Kostenlos | |
---
### ASR-Spracherkennung
| Verwendungsmethode | Unterstützte Plattformen | Kostenlose Plattformen |
|:---:|:---:|:---:|
| Lokale Verwendung | FunASR, SherpaASR | FunASR, SherpaASR |
| Schnittstellenaufrufe | DoubaoASR, Doubao Streaming ASR, FunASRServer, TencentASR, AliyunASR, Aliyun Streaming ASR, Baidu ASR, OpenAI ASR | FunASRServer |
---
### Voiceprint-Stimmabdruckerkennung
| Verwendungsmethode | Unterstützte Plattformen | Kostenlose Plattformen |
|:---:|:---:|:---:|
| Lokale Verwendung | 3D-Speaker | 3D-Speaker |
---
### Memory-Gedächtnisspeicher
| Typ | Plattformname | Verwendungsmethode | Preismodell | Hinweise |
|:------:|:---------------:|:----:|:---------:|:--:|
| Memory | mem0ai | Schnittstellenaufrufe | 1000 Mal/Monat Kontingent | |
| Memory | [powermem](../powermem-integration.md) | Lokale Zusammenfassung | Abhängig von LLM und DB | OceanBase Open Source, unterstützt intelligente Abfrage |
| Memory | mem_local_short | Lokale Zusammenfassung | Kostenlos | |
| Memory | nomem | Kein Gedächtnismodus | Kostenlos | |
---
### Intent-Absichtserkennung
| Typ | Plattformname | Verwendungsmethode | Preismodell | Hinweise |
|:------:|:-------------:|:----:|:-------:|:---------------------:|
| Intent | intent_llm | Schnittstellenaufrufe | Basierend auf LLM-Preisen | Erkennt Absicht durch große Modelle, starke Allgemeingültigkeit |
| Intent | function_call | Schnittstellenaufrufe | Basierend auf LLM-Preisen | Vervollständigt Absicht durch Funktionsaufruf großer Modelle, schnelle Geschwindigkeit, guter Effekt |
| Intent | nointent | Kein Absichtsmodus | Kostenlos | Führt keine Absichtserkennung durch, gibt direkt Dialogergebnis zurück |
---
### Rag Retrieval Augmented Generation
| Typ | Plattformname | Verwendungsmethode | Preismodell | Hinweise |
|:------:|:-------------:|:----:|:-------:|:---------------------:|
| Rag | ragflow | Schnittstellenaufrufe | Gebühren basierend auf Token-Verbrauch für Segmentierung und Tokenisierung | Nutzt RAGFlow's Retrieval-Augmented-Generation-Funktion für präzisere Dialogantworten |
---
## Danksagungen 🙏
| Logo | Projekt/Unternehmen | Beschreibung |
|:---:|:---:|:---|
| <img src="../images/logo_bailing.png" width="160"> | [Bailing Voice Dialogue Robot](https://github.com/wwbin2017/bailing) | Dieses Projekt wurde von [Bailing Voice Dialogue Robot](https://github.com/wwbin2017/bailing) inspiriert und auf dessen Basis implementiert |
| <img src="../images/logo_tenclass.png" width="160"> | [Tenclass](https://www.tenclass.com/) | Dank an [Tenclass](https://www.tenclass.com/) für die Formulierung von Standardkommunikationsprotokollen, Multi-Geräte-Kompatibilitätslösungen und High-Concurrency-Szenario-Praxisdemonstrationen für das Xiaozhi-Ökosystem; für die Bereitstellung vollständiger technischer Dokumentationsunterstützung für dieses Projekt |
| <img src="../images/logo_xuanfeng.png" width="160"> | [Xuanfeng Technology](https://github.com/Eric0308) | Dank an [Xuanfeng Technology](https://github.com/Eric0308) für den Beitrag des Funktionsaufruf-Frameworks, des MCP-Kommunikationsprotokolls und der plugin-basierten Aufrufmechanismus-Implementierungscode. Durch standardisiertes Befehlsplanungssystem und dynamische Erweiterungsfähigkeiten wird die Interaktionseffizienz und funktionale Erweiterbarkeit von Frontend-Geräten (IoT) erheblich verbessert |
| <img src="../images/logo_junsen.png" width="160"> | [huangjunsen](https://github.com/huangjunsen0406) | Dank an [huangjunsen](https://github.com/huangjunsen0406) für den Beitrag des `Smart Control Console Mobile`-Moduls, das eine effiziente Steuerung und Echtzeit-Interaktion über mobile Geräte ermöglicht und die Betriebsbequemlichkeit und Verwaltungseffizienz des Systems in mobilen Szenarien erheblich verbessert |
| <img src="../images/logo_huiyuan.png" width="160"> | [Huiyuan Design](http://ui.kwd988.net/) | Dank an [Huiyuan Design](http://ui.kwd988.net/) für die Bereitstellung professioneller visueller Lösungen für dieses Projekt, unter Verwendung ihrer Design-Praxiserfahrung im Dienst von über tausend Unternehmen, um die Produktbenutzererfahrung dieses Projekts zu stärken |
| <img src="../images/logo_qinren.png" width="160"> | [Xi'an Qinren Information Technology](https://www.029app.com/) | Dank an [Xi'an Qinren Information Technology](https://www.029app.com/) für die Vertiefung des visuellen Systems dieses Projekts und die Sicherstellung der Konsistenz und Erweiterbarkeit des Gesamtdesignstils in Multi-Szenario-Anwendungen |
| <img src="../images/logo_contributors.png" width="160"> | [Code-Mitwirkende](https://github.com/xinnan-tech/xiaozhi-esp32-server/graphs/contributors) | Dank an [alle Code-Mitwirkenden](https://github.com/xinnan-tech/xiaozhi-esp32-server/graphs/contributors), Ihre Bemühungen haben das Projekt robuster und leistungsfähiger gemacht. |
<a href="https://star-history.com/#xinnan-tech/xiaozhi-esp32-server&Date">
<picture>
<source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=xinnan-tech/xiaozhi-esp32-server&type=Date&theme=dark" />
<source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=xinnan-tech/xiaozhi-esp32-server&type=Date" />
<img alt="Star History Chart" src="https://api.star-history.com/svg?repos=xinnan-tech/xiaozhi-esp32-server&type=Date" />
</picture>
</a>
+361
View File
@@ -0,0 +1,361 @@
[![Banners](../images/banner1.png)](https://github.com/xinnan-tech/xiaozhi-esp32-server)
<h1 align="center">Xiaozhi Backend Service xiaozhi-esp32-server</h1>
<p align="center">
This project is based on human-machine symbiotic intelligence theory and technology to develop intelligent terminal hardware and software systems<br/>providing backend services for the open-source intelligent hardware project
<a href="https://github.com/78/xiaozhi-esp32">xiaozhi-esp32</a><br/>
Implemented using Python, Java, and Vue according to the <a href="https://ccnphfhqs21z.feishu.cn/wiki/M0XiwldO9iJwHikpXD5cEx71nKh">Xiaozhi Communication Protocol</a><br/>
Support for MQTT+UDP protocol, Websocket protocol, MCP access point, voiceprint recognition, and knowledge base
</p>
<p align="center">
<a href="../FAQ.md">FAQ</a>
· <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/issues">Report Issues</a>
· <a href="../../README.md#%E9%83%A8%E7%BD%B2%E6%96%87%E6%A1%A3">Deployment Docs</a>
· <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/releases">Release Notes</a>
</p>
<p align="center">
<a href="../../README.md"><img alt="简体中文版自述文件" src="https://img.shields.io/badge/简体中文-DFE0E5"></a>
<a href="./README_en.md"><img alt="README in English" src="https://img.shields.io/badge/English-DBEDFA"></a>
<a href="./README_vi.md"><img alt="Tiếng Việt" src="https://img.shields.io/badge/Tiếng Việt-DFE0E5"></a>
<a href="./README_de.md"><img alt="Deutsch" src="https://img.shields.io/badge/Deutsch-DFE0E5"></a>
<a href="./README_pt_BR.md"><img alt="Português (Brasil)" src="https://img.shields.io/badge/Português (Brasil)-DFE0E5"></a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/releases">
<img alt="GitHub Contributors" src="https://img.shields.io/github/v/release/xinnan-tech/xiaozhi-esp32-server?logo=docker" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/blob/main/LICENSE">
<img alt="GitHub pull requests" src="https://img.shields.io/badge/license-MIT-white?labelColor=black" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server">
<img alt="stars" src="https://img.shields.io/github/stars/xinnan-tech/xiaozhi-esp32-server?color=ffcb47&labelColor=black" />
</a>
</p>
<p align="center">
Spearheaded by Professor Siyuan Liu's Team (South China University of Technology)
</br>
刘思源教授团队主导研发(华南理工大学)
</br>
<img src="../images/hnlg.jpg" alt="South China University of Technology" width="50%">
</p>
---
## Target Users 👥
This project requires ESP32 hardware devices to work. If you have purchased ESP32-related hardware, successfully connected to Brother Xia's deployed backend service, and want to build your own `xiaozhi-esp32` backend service independently, then this project is perfect for you.
Want to see the usage effects? Click the videos below 🎥
<table>
<tr>
<td>
<a href="https://www.bilibili.com/video/BV1FMFyejExX" target="_blank">
<picture>
<img alt="响应速度感受" src="docs/images/demo9.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1vchQzaEse" target="_blank">
<picture>
<img alt="速度优化秘诀" src="docs/images/demo6.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1WEcxzFEAT" target="_blank">
<picture>
<img alt="小智数字人 支持语音唤醒" src="docs/images/demo8.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1CKVz6UEuB" target="_blank">
<picture>
<img alt="设备呼叫设备,打电话" src="docs/images/demo0.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1C1tCzUEZh" target="_blank">
<picture>
<img alt="复杂医疗场景" src="docs/images/demo1.png" /></picture>
</a>
</td>
</tr>
<tr>
<td>
<a href="https://www.bilibili.com/video/BV1VC96Y5EMH" target="_blank">
<picture>
<img alt="播放音乐查询天气播报新闻" src="docs/images/demo7.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV12J7WzBEaH" target="_blank">
<picture>
<img alt="实时打断" src="docs/images/demo10.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1Co76z7EvK" target="_blank">
<picture>
<img alt="拍照识物品" src="docs/images/demo12.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1pNXWYGEx1" target="_blank">
<picture>
<img alt="控制家电开关" src="docs/images/demo5.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1TJ7WzzEo6" target="_blank">
<picture>
<img alt="多指令任务" src="docs/images/demo11.png" /></picture>
</a>
</td>
</tr>
<tr>
<td>
<a href="https://www.bilibili.com/video/BV1ZQKUzYExM" target="_blank">
<picture>
<img alt="MCP接入点" src="docs/images/demo13.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1zUW5zJEkq" target="_blank">
<picture>
<img alt="MQTT指令下发" src="docs/images/demo4.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1Exu3zqEDe" target="_blank">
<picture>
<img alt="声纹识别" src="docs/images/demo14.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1CDKWemEU6" target="_blank">
<picture>
<img alt="自定义音色" src="docs/images/demo2.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV12yA2egEaC" target="_blank">
<picture>
<img alt="使用粤语交流" src="docs/images/demo3.png" /></picture>
</a>
</td>
</tr>
</table>
---
## Warnings ⚠️
1. This project is open-source software. This software has no commercial partnership with any third-party API service providers (including but not limited to speech recognition, large models, speech synthesis, and other platforms) that it interfaces with, and does not provide any form of guarantee for their service quality or financial security. It is recommended that users prioritize service providers with relevant business licenses and carefully read their service agreements and privacy policies. This software does not host any account keys, does not participate in fund flows, and does not bear the risk of recharge fund losses.
2. The functionality of this project is not complete and has not passed network security assessment. Please do not use it in production environments. If you deploy this project for learning purposes in a public network environment, please ensure necessary protection measures are in place.
---
## Deployment Documentation
![Banners](../images/banner2.png)
This project provides two deployment methods. Please choose based on your specific needs:
#### 🚀 Deployment Method Selection
| Deployment Method | Features | Applicable Scenarios | Deployment Docs | Configuration Requirements | Video Tutorials |
|---------|------|---------|---------|---------|---------|
| **Simplified Installation** | Intelligent dialogue, single agent management | Low-configuration environments, data stored in config files, no database required | [①Docker Version](../Deployment.md#%E6%96%B9%E5%BC%8F%E4%B8%80docker%E5%8F%AA%E8%BF%90%E8%A1%8Cserver) / [②Source Code Deployment](../Deployment.md#%E6%96%B9%E5%BC%8F%E4%BA%8C%E6%9C%AC%E5%9C%B0%E6%BA%90%E7%A0%81%E5%8F%AA%E8%BF%90%E8%A1%8Cserver)| 2 cores 4GB if using `FunASR`, 2 cores 2GB if all APIs | - |
| **Full Module Installation** | Intelligent dialogue, multi-user management, multi-agent management, intelligent console interface operation | Complete functionality experience, data stored in database |[①Docker Version](../Deployment_all.md#%E6%96%B9%E5%BC%8F%E4%B8%80docker%E8%BF%90%E8%A1%8C%E5%85%A8%E6%A8%A1%E5%9D%97) / [②Source Code Deployment](../Deployment_all.md#%E6%96%B9%E5%BC%8F%E4%BA%8C%E6%9C%AC%E5%9C%B0%E6%BA%90%E7%A0%81%E8%BF%90%E8%A1%8C%E5%85%A8%E6%A8%A1%E5%9D%97) / [③Source Code Deployment Auto-Update Tutorial](../dev-ops-integration.md) | 4 cores 8GB if using `FunASR`, 2 cores 4GB if all APIs| [Local Source Code Startup Video Tutorial](https://www.bilibili.com/video/BV1wBJhz4Ewe) |
For frequently asked questions and related tutorials, please refer to [this link](../FAQ.md)
> 💡 Note: Below is a test platform deployed with the latest code. You can burn and test if needed. Concurrent users: 6, data will be cleared daily.
```
Intelligent Control Console Address: https://2662r3426b.vicp.fun
Intelligent Control Console Address (H5): https://2662r3426b.vicp.fun/h5/index.html
Service Test Tool: https://2662r3426b.vicp.fun/test/
OTA Interface Address: https://2662r3426b.vicp.fun/xiaozhi/ota/
Websocket Interface Address: wss://2662r3426b.vicp.fun/xiaozhi/v1/
```
#### 🚩 Configuration Description and Recommendations
> [!Note]
> This project provides two configuration schemes:
>
> 1. `Entry Level Free Settings`: Suitable for personal and home use, all components use free solutions, no additional payment required.
>
> 2. `Streaming Configuration`: Suitable for demonstrations, training, scenarios with more than 2 concurrent users, etc. Uses streaming processing technology for faster response speed and better experience.
>
> Starting from version `0.5.2`, the project supports streaming configuration. Compared to earlier versions, response speed is improved by approximately `2.5 seconds`, significantly improving user experience.
| Module Name | Entry Level Free Settings | Streaming Configuration |
|:---:|:---:|:---:|
| ASR(Speech Recognition) | FunASR(Local) | 👍XunfeiStreamASR(Xunfei Streaming) |
| LLM(Large Model) | glm-4-flash(Zhipu) | 👍qwen-flash(Alibaba Bailian) |
| VLLM(Vision Large Model) | glm-4v-flash(Zhipu) | 👍qwen3.5-flash(Alibaba Bailian) |
| TTS(Speech Synthesis) | EdgeTTS(Microsoft) | 👍HuoshanDoubleStreamTTS(Volcano Streaming) |
| Intent(Intent Recognition) | function_call(Function calling) | function_call(Function calling) |
| Memory(Memory function) | mem_local_short(Local short-term memory) | mem_local_short(Local short-term memory) |
If you are concerned about the latency of each component, please refer to the [Xiaozhi Component Performance Test Report](https://github.com/xinnan-tech/xiaozhi-performance-research), and test in your own environment following the test methods in the report.
#### 🔧 Testing Tools
This project provides the following testing tools to help you verify the system and choose suitable models:
| Tool Name | Location | Usage Method | Function Description |
|:---:|:---|:---:|:---:|
| Audio Interaction Test Tool | main》digital-human》index.html | Run `python start.py` in `main/digital-human`, then open `http://127.0.0.1:8006/index.html` | Tests audio playback and reception functions, verifies if Python-side audio processing is normal |
| Model Response Test Tool | main》xiaozhi-server》performance_tester.py | Execute `python performance_tester.py` | Tests response speed of three core modules: ASR(speech recognition), LLM(large model), VLLM(vision model), TTS(speech synthesis) |
> 💡 Note: When testing model speed, only models with configured keys will be tested.
---
## Feature List ✨
### Implemented ✅
![请参考-全模块安装架构图](../images/deploy2.png)
| Feature Module | Description |
|:---:|:---|
| Core Architecture | Based on [MQTT+UDP gateway](https://github.com/xinnan-tech/xiaozhi-esp32-server/blob/main/docs/mqtt-gateway-integration.md), WebSocket and HTTP servers, provides complete console management and authentication system |
| Voice Interaction | Supports streaming ASR(speech recognition), streaming TTS(speech synthesis), VAD(voice activity detection), supports multi-language recognition and voice processing |
| Voiceprint Recognition | Supports multi-user voiceprint registration, management, and recognition, processes in parallel with ASR, real-time speaker identity recognition and passes to LLM for personalized responses |
| Intelligent Dialogue | Supports multiple LLM(large language models), implements intelligent dialogue |
| Visual Perception | Supports multiple VLLM(vision large models), implements multimodal interaction |
| Intent Recognition | Supports LLM intent recognition, Function Call function calling, provides plugin-based intent processing mechanism |
| Memory System | Supports local short-term memory, mem0ai interface memory, PowerMem intelligent memory, with memory summarization functionality |
| Knowledge Base | Supports RAGFlow knowledge base, enabling LLM to judge whether to schedule the knowledge base after receiving the user's question, and then answer the question |
| Tool Calling | Supports client IOT protocol, client MCP protocol, server MCP protocol, MCP endpoint protocol, custom tool functions |
| Command Delivery | Supports MCP command delivery to ESP32 devices via MQTT protocol from Smart Console |
| Management Backend | Provides Web management interface, supports user management, system configuration and device management; Supports Simplified Chinese, Traditional Chinese and English display |
| Testing Tools | Provides performance testing tools, vision model testing tools, and audio interaction testing tools |
| Deployment Support | Supports Docker deployment and local deployment, provides complete configuration file management |
| Plugin System | Supports functional plugin extensions, custom plugin development, and plugin hot-loading |
### Under Development 🚧
To learn about specific development plan progress, [click here](https://github.com/users/xinnan-tech/projects/3). For frequently asked questions and related tutorials, please refer to [this link](../FAQ.md)
If you are a software developer, here is an [Open Letter to Developers](../contributor_open_letter.md). Welcome to join!
---
## Product Ecosystem 👬
Xiaozhi is an ecosystem. When using this product, you can also check out other [excellent projects](https://github.com/78/xiaozhi-esp32/blob/main/README_zh.md#%E7%9B%B8%E5%85%B3%E5%BC%80%E6%BA%90%E9%A1%B9%E7%9B%AE) in this ecosystem
---
## Supported Platforms/Components List 📋
### LLM Language Models
| Usage Method | Supported Platforms | Free Platforms |
|:---:|:---:|:---:|
| OpenAI interface calls | Alibaba Bailian, Volcano Engine, DeepSeek, Zhipu, Gemini, iFLYTEK | Zhipu, Gemini |
| Ollama interface calls | Ollama | - |
| Dify interface calls | Dify | - |
| FastGPT interface calls | FastGPT | - |
| Coze interface calls | Coze | - |
| Xinference interface calls | Xinference | - |
| HomeAssistant interface calls | HomeAssistant | - |
In fact, any LLM that supports OpenAI interface calls can be integrated and used.
---
### VLLM Vision Models
| Usage Method | Supported Platforms | Free Platforms |
|:---:|:---:|:---:|
| OpenAI interface calls | Alibaba Bailian, Zhipu ChatGLMVLLM | Zhipu ChatGLMVLLM |
In fact, any VLLM that supports OpenAI interface calls can be integrated and used.
---
### TTS Speech Synthesis
| Usage Method | Supported Platforms | Free Platforms |
|:---:|:---:|:---:|
| Interface calls | EdgeTTS, iFLYTEK, Volcano Engine, Tencent Cloud, Alibaba Cloud and Bailian, CosyVoiceSiliconflow, TTS302AI, CozeCnTTS, GizwitsTTS, ACGNTTS, OpenAITTS, Lingxi Streaming TTS, MinimaxTTS | Lingxi Streaming TTS, EdgeTTS, CosyVoiceSiliconflow(partial) |
| Local services | FishSpeech, GPT_SOVITS_V2, GPT_SOVITS_V3, Index-TTS, PaddleSpeech | Index-TTS, PaddleSpeech, FishSpeech, GPT_SOVITS_V2, GPT_SOVITS_V3 |
---
### VAD Voice Activity Detection
| Type | Platform Name | Usage Method | Pricing Model | Notes |
|:---:|:---------:|:----:|:----:|:--:|
| VAD | SileroVAD | Local use | Free | |
---
### ASR Speech Recognition
| Usage Method | Supported Platforms | Free Platforms |
|:---:|:---:|:---:|
| Local use | FunASR, SherpaASR | FunASR, SherpaASR |
| Interface calls | FunASRServer, Volcano Engine, iFLYTEK, Tencent Cloud, Alibaba Cloud, Baidu Cloud, OpenAI ASR | FunASRServer |
---
### Voiceprint Recognition
| Usage Method | Supported Platforms | Free Platforms |
|:---:|:---:|:---:|
| Local use | 3D-Speaker | 3D-Speaker |
---
### Memory Storage
| Type | Platform Name | Usage Method | Pricing Model | Notes |
|:------:|:---------------:|:----:|:---------:|:--:|
| Memory | mem0ai | Interface calls | 1000 times/month quota | |
| Memory | [powermem](../powermem-integration.md) | Local summarization | Depends on LLM and DB | OceanBase open source, supports intelligent retrieval |
| Memory | mem_local_short | Local summarization | Free | |
| Memory | nomem | No memory mode | Free | |
---
### Intent Recognition
| Type | Platform Name | Usage Method | Pricing Model | Notes |
|:------:|:-------------:|:----:|:-------:|:---------------------:|
| Intent | intent_llm | Interface calls | Based on LLM pricing | Recognizes intent through large models, strong generalization |
| Intent | function_call | Interface calls | Based on LLM pricing | Completes intent through large model function calling, fast speed, good effect |
| Intent | nointent | No intent mode | Free | Does not perform intent recognition, directly returns dialogue result |
---
### Rag Retrieval-Augmented Generation
| Type | Platform Name | Usage Method | Pricing Model | Notes |
|:------:|:-------------:|:----:|:-------:|:---------------------:|
| Rag | ragflow | Interface calls | Charged based on tokens consumed for slicing and word segmentation | Utilizes RagFlow's retrieval-augmented generation feature to provide more accurate dialog responses |
---
## Acknowledgments 🙏
| Logo | Project/Company | Description |
|:---:|:---:|:---|
| <img src="../images/logo_bailing.png" width="160"> | [Bailing Voice Dialogue Robot](https://github.com/wwbin2017/bailing) | This project is inspired by [Bailing Voice Dialogue Robot](https://github.com/wwbin2017/bailing) and implemented on its basis |
| <img src="../images/logo_tenclass.png" width="160"> | [Tenclass](https://www.tenclass.com/) | Thanks to [Tenclass](https://www.tenclass.com/) for formulating standard communication protocols, multi-device compatibility solutions, and high-concurrency scenario practice demonstrations for the Xiaozhi ecosystem; providing full-link technical documentation support for this project |
| <img src="../images/logo_xuanfeng.png" width="160"> | [Xuanfeng Technology](https://github.com/Eric0308) | Thanks to [Xuanfeng Technology](https://github.com/Eric0308) for contributing function calling framework, MCP communication protocol, and plugin-based calling mechanism implementation code. Through standardized instruction scheduling system and dynamic expansion capabilities, it significantly improves the interaction efficiency and functional extensibility of frontend devices (IoT) |
| <img src="../images/logo_junsen.png" width="160"> | [huangjunsen](https://github.com/huangjunsen0406) | Thanks to [huangjunsen](https://github.com/huangjunsen0406) for contributing the `Smart Control Console Mobile` module, which enables efficient control and real-time interaction across mobile devices, significantly enhancing the system's operational convenience and management efficiency in mobile scenarios. |
| <img src="../images/logo_huiyuan.png" width="160"> | [Huiyuan Design](http://ui.kwd988.net/) | Thanks to [Huiyuan Design](http://ui.kwd988.net/) for providing professional visual solutions for this project, using their design practical experience serving over a thousand enterprises to empower this project's product user experience |
| <img src="../images/logo_qinren.png" width="160"> | [Xi'an Qinren Information Technology](https://www.029app.com/) | Thanks to [Xi'an Qinren Information Technology](https://www.029app.com/) for deepening this project's visual system, ensuring consistency and extensibility of overall design style in multi-scenario applications |
| <img src="../images/logo_contributors.png" width="160"> | [Code Contributors](https://github.com/xinnan-tech/xiaozhi-esp32-server/graphs/contributors) | Thanks to [all code contributors](https://github.com/xinnan-tech/xiaozhi-esp32-server/graphs/contributors), your efforts have made the project more robust and powerful. |
<a href="https://star-history.com/#xinnan-tech/xiaozhi-esp32-server&Date">
<picture>
<source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=xinnan-tech/xiaozhi-esp32-server&type=Date&theme=dark" />
<source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=xinnan-tech/xiaozhi-esp32-server&type=Date" />
<img alt="Star History Chart" src="https://api.star-history.com/svg?repos=xinnan-tech/xiaozhi-esp32-server&type=Date" />
</picture>
</a>
+361
View File
@@ -0,0 +1,361 @@
[![Banners](../images/banner1.png)](https://github.com/xinnan-tech/xiaozhi-esp32-server)
<h1 align="center">Serviço Backend Xiaozhi xiaozhi-esp32-server</h1>
<p align="center">
Este projeto é baseado na teoria e tecnologia de inteligência simbiótica humano-máquina para desenvolver sistemas inteligentes de hardware e software para terminais<br/>fornecendo serviços de backend para o projeto de hardware inteligente de código aberto
<a href="https://github.com/78/xiaozhi-esp32">xiaozhi-esp32</a><br/>
Implementado usando Python, Java e Vue de acordo com o <a href="https://ccnphfhqs21z.feishu.cn/wiki/M0XiwldO9iJwHikpXD5cEx71nKh">Protocolo de Comunicação Xiaozhi</a><br/>
Suporte ao protocolo MQTT+UDP, protocolo WebSocket, ponto de acesso MCP, reconhecimento de impressão vocal e base de conhecimento
</p>
<p align="center">
<a href="../FAQ.md">Perguntas Frequentes</a>
· <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/issues">Reportar Problemas</a>
· <a href="../../README.md#%E9%83%A8%E7%BD%B2%E6%96%87%E6%A1%A3">Documentação de Implantação</a>
· <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/releases">Notas de Lançamento</a>
</p>
<p align="center">
<a href="../../README.md"><img alt="简体中文版自述文件" src="https://img.shields.io/badge/简体中文-DFE0E5"></a>
<a href="./README_en.md"><img alt="README in English" src="https://img.shields.io/badge/English-DFE0E5"></a>
<a href="./README_vi.md"><img alt="Tiếng Việt" src="https://img.shields.io/badge/Tiếng Việt-DFE0E5"></a>
<a href="./README_de.md"><img alt="Deutsch" src="https://img.shields.io/badge/Deutsch-DFE0E5"></a>
<a href="./README_pt_BR.md"><img alt="Português (Brasil)" src="https://img.shields.io/badge/Português (Brasil)-DBEDFA"></a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/releases">
<img alt="GitHub Contributors" src="https://img.shields.io/github/v/release/xinnan-tech/xiaozhi-esp32-server?logo=docker" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/blob/main/LICENSE">
<img alt="GitHub pull requests" src="https://img.shields.io/badge/license-MIT-white?labelColor=black" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server">
<img alt="stars" src="https://img.shields.io/github/stars/xinnan-tech/xiaozhi-esp32-server?color=ffcb47&labelColor=black" />
</a>
</p>
<p align="center">
Liderado pela Equipe do Professor Siyuan Liu (Universidade de Tecnologia do Sul da China)
</br>
刘思源教授团队主导研发(华南理工大学)
</br>
<img src="../images/hnlg.jpg" alt="Universidade de Tecnologia do Sul da China (华南理工大学)" width="50%">
</p>
---
## Público-Alvo 👥
Este projeto requer dispositivos de hardware ESP32 para funcionar. Se você adquiriu hardware relacionado ao ESP32, conectou-se com sucesso ao serviço backend implantado pelo Brother Xia e deseja construir seu próprio serviço backend `xiaozhi-esp32` de forma independente, então este projeto é perfeito para você.
Quer ver os efeitos de uso? Clique nos vídeos abaixo 🎥
<table>
<tr>
<td>
<a href="https://www.bilibili.com/video/BV1FMFyejExX" target="_blank">
<picture>
<img alt="响应速度感受" src="docs/images/demo9.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1vchQzaEse" target="_blank">
<picture>
<img alt="速度优化秘诀" src="docs/images/demo6.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1WEcxzFEAT" target="_blank">
<picture>
<img alt="小智数字人 支持语音唤醒" src="docs/images/demo8.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1CKVz6UEuB" target="_blank">
<picture>
<img alt="设备呼叫设备,打电话" src="docs/images/demo0.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1C1tCzUEZh" target="_blank">
<picture>
<img alt="复杂医疗场景" src="docs/images/demo1.png" /></picture>
</a>
</td>
</tr>
<tr>
<td>
<a href="https://www.bilibili.com/video/BV1VC96Y5EMH" target="_blank">
<picture>
<img alt="播放音乐查询天气播报新闻" src="docs/images/demo7.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV12J7WzBEaH" target="_blank">
<picture>
<img alt="实时打断" src="docs/images/demo10.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1Co76z7EvK" target="_blank">
<picture>
<img alt="拍照识物品" src="docs/images/demo12.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1pNXWYGEx1" target="_blank">
<picture>
<img alt="控制家电开关" src="docs/images/demo5.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1TJ7WzzEo6" target="_blank">
<picture>
<img alt="多指令任务" src="docs/images/demo11.png" /></picture>
</a>
</td>
</tr>
<tr>
<td>
<a href="https://www.bilibili.com/video/BV1ZQKUzYExM" target="_blank">
<picture>
<img alt="MCP接入点" src="docs/images/demo13.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1zUW5zJEkq" target="_blank">
<picture>
<img alt="MQTT指令下发" src="docs/images/demo4.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1Exu3zqEDe" target="_blank">
<picture>
<img alt="声纹识别" src="docs/images/demo14.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1CDKWemEU6" target="_blank">
<picture>
<img alt="自定义音色" src="docs/images/demo2.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV12yA2egEaC" target="_blank">
<picture>
<img alt="使用粤语交流" src="docs/images/demo3.png" /></picture>
</a>
</td>
</tr>
</table>
---
## Avisos ⚠️
1. Este projeto é um software de código aberto. Este software não possui parceria comercial com nenhum provedor de serviços de API de terceiros (incluindo, mas não se limitando a reconhecimento de fala, modelos de linguagem, síntese de voz e outras plataformas) com os quais se conecta, e não fornece nenhuma forma de garantia quanto à qualidade de serviço ou segurança financeira desses provedores. Recomenda-se que os usuários priorizem provedores de serviço com licenças comerciais relevantes e leiam cuidadosamente seus termos de serviço e políticas de privacidade. Este software não armazena nenhuma chave de conta, não participa de fluxos de fundos e não assume o risco de perda de fundos recarregados.
2. A funcionalidade deste projeto não está completa e não passou por avaliação de segurança de rede. Por favor, não o utilize em ambientes de produção. Se você implantar este projeto para fins de aprendizado em um ambiente de rede pública, certifique-se de que as medidas de proteção necessárias estejam em vigor.
---
## Documentação de Implantação
![Banners](../images/banner2.png)
Este projeto oferece dois métodos de implantação. Por favor, escolha de acordo com suas necessidades específicas:
#### 🚀 Seleção do Método de Implantação
| Método de Implantação | Funcionalidades | Cenários Aplicáveis | Documentação de Implantação | Requisitos de Configuração | Tutoriais em Vídeo |
|---------|------|---------|---------|---------|---------|
| **Instalação Simplificada** | Diálogo inteligente, gerenciamento de agente único | Ambientes de baixa configuração, dados armazenados em arquivos de configuração, sem necessidade de banco de dados | [①Versão Docker](../Deployment.md#%E6%96%B9%E5%BC%8F%E4%B8%80docker%E5%8F%AA%E8%BF%90%E8%A1%8Cserver) / [②Implantação via Código-Fonte](../Deployment.md#%E6%96%B9%E5%BC%8F%E4%BA%8C%E6%9C%AC%E5%9C%B0%E6%BA%90%E7%A0%81%E5%8F%AA%E8%BF%90%E8%A1%8Cserver)| 2 núcleos 4GB se usar `FunASR`, 2 núcleos 2GB se todas APIs | - |
| **Instalação de Módulo Completo** | Diálogo inteligente, gerenciamento multiusuário, gerenciamento de múltiplos agentes, operação de interface do console inteligente | Experiência com funcionalidade completa, dados armazenados em banco de dados |[①Versão Docker](../Deployment_all.md#%E6%96%B9%E5%BC%8F%E4%B8%80docker%E8%BF%90%E8%A1%8C%E5%85%A8%E6%A8%A1%E5%9D%97) / [②Implantação via Código-Fonte](../Deployment_all.md#%E6%96%B9%E5%BC%8F%E4%BA%8C%E6%9C%AC%E5%9C%B0%E6%BA%90%E7%A0%81%E8%BF%90%E8%A1%8C%E5%85%A8%E6%A8%A1%E5%9D%97) / [③Tutorial de Atualização Automática via Código-Fonte](../dev-ops-integration.md) | 4 núcleos 8GB se usar `FunASR`, 2 núcleos 4GB se todas APIs| [Tutorial em Vídeo de Inicialização via Código-Fonte Local](https://www.bilibili.com/video/BV1wBJhz4Ewe) |
Perguntas frequentes e tutoriais relacionados podem ser consultados [neste link](../FAQ.md)
> 💡 Nota: Abaixo está uma plataforma de teste implantada com o código mais recente. Você pode gravar e testar se necessário. Usuários simultâneos: 6, os dados serão limpos diariamente.
```
Endereço do Console de Controle Inteligente: https://2662r3426b.vicp.fun
Endereço do Console de Controle Inteligente (H5): https://2662r3426b.vicp.fun/h5/index.html
Ferramenta de Teste de Serviço: https://2662r3426b.vicp.fun/test/
Endereço da Interface OTA: https://2662r3426b.vicp.fun/xiaozhi/ota/
Endereço da Interface WebSocket: wss://2662r3426b.vicp.fun/xiaozhi/v1/
```
#### 🚩 Descrição e Recomendações de Configuração
> [!Note]
> Este projeto oferece dois esquemas de configuração:
>
> 1. `Configurações Gratuitas Nível Básico`: Adequado para uso pessoal e doméstico, todos os componentes utilizam soluções gratuitas, sem necessidade de pagamento adicional.
>
> 2. `Configuração de Streaming`: Adequado para demonstrações, treinamentos, cenários com mais de 2 usuários simultâneos, etc. Utiliza tecnologia de processamento em streaming para velocidade de resposta mais rápida e melhor experiência.
>
> A partir da versão `0.5.2`, o projeto suporta configuração de streaming. Em comparação com versões anteriores, a velocidade de resposta é melhorada em aproximadamente `2,5 segundos`, melhorando significativamente a experiência do usuário.
| Nome do Módulo | Configurações Gratuitas Nível Básico | Configuração de Streaming |
|:---:|:---:|:---:|
| ASR(Reconhecimento de Fala) | FunASR(Local) | 👍XunfeiStreamASR(Xunfei Streaming) |
| LLM(Modelo de Linguagem) | glm-4-flash(Zhipu) | 👍qwen-flash(Alibaba Bailian) |
| VLLM(Modelo de Visão) | glm-4v-flash(Zhipu) | 👍qwen3.5-flash(Alibaba Bailian) |
| TTS(Síntese de Voz) | EdgeTTS(Microsoft) | 👍HuoshanDoubleStreamTTS(Volcano Streaming) |
| Intent(Reconhecimento de Intenção) | function_call(Chamada de função) | function_call(Chamada de função) |
| Memory(Função de Memória) | mem_local_short(Memória local de curto prazo) | mem_local_short(Memória local de curto prazo) |
Se você está preocupado com o tempo de resposta de cada componente, consulte o [Relatório de Teste de Desempenho dos Componentes Xiaozhi](https://github.com/xinnan-tech/xiaozhi-performance-research), e teste em seu próprio ambiente seguindo os métodos de teste do relatório.
#### 🔧 Ferramentas de Teste
Este projeto fornece as seguintes ferramentas de teste para ajudá-lo a verificar o sistema e escolher modelos adequados:
| Nome da Ferramenta | Localização | Método de Uso | Descrição da Função |
|:---:|:---|:---:|:---:|
| Ferramenta de Teste de Interação por Áudio | main》digital-human》index.html | Execute `python start.py` em `main/digital-human` e depois abra `http://127.0.0.1:8006/index.html` | Testa as funções de reprodução e recepção de áudio, verifica se o processamento de áudio no lado Python está normal |
| Ferramenta de Teste de Resposta de Modelo | main》xiaozhi-server》performance_tester.py | Execute `python performance_tester.py` | Testa a velocidade de resposta dos três módulos principais: ASR(reconhecimento de fala), LLM(modelo de linguagem), VLLM(modelo de visão), TTS(síntese de voz) |
> 💡 Nota: Ao testar a velocidade dos modelos, apenas os modelos com chaves configuradas serão testados.
---
## Lista de Funcionalidades ✨
### Implementado ✅
![请参考-全模块安装架构图](../images/deploy2.png)
| Módulo de Funcionalidade | Descrição |
|:---:|:---|
| Arquitetura Principal | Baseado em [gateway MQTT+UDP](https://github.com/xinnan-tech/xiaozhi-esp32-server/blob/main/docs/mqtt-gateway-integration.md), servidores WebSocket e HTTP, fornece sistema completo de gerenciamento de console e autenticação |
| Interação por Voz | Suporta ASR em streaming (reconhecimento de fala), TTS em streaming (síntese de voz), VAD (detecção de atividade vocal), suporta reconhecimento multilíngue e processamento de voz |
| Reconhecimento de Impressão Vocal | Suporta registro, gerenciamento e reconhecimento de impressão vocal de múltiplos usuários, processa em paralelo com o ASR, reconhecimento de identidade do falante em tempo real e repassa ao LLM para respostas personalizadas |
| Diálogo Inteligente | Suporta múltiplos LLM (modelos de linguagem de grande porte), implementa diálogo inteligente |
| Percepção Visual | Suporta múltiplos VLLM (modelos de visão de grande porte), implementa interação multimodal |
| Reconhecimento de Intenção | Suporta reconhecimento de intenção por LLM, Function Call (chamada de função), fornece mecanismo de processamento de intenção baseado em plugins |
| Sistema de Memória | Suporta memória local de curto prazo, memória via interface mem0ai, memória inteligente PowerMem, com funcionalidade de resumo de memória |
| Base de Conhecimento | Suporta base de conhecimento RAGFlow, permitindo que o LLM julgue se deve acionar a base de conhecimento após receber a pergunta do usuário, e então responda à pergunta |
| Chamada de Ferramentas | Suporta protocolo IOT do cliente, protocolo MCP do cliente, protocolo MCP do servidor, protocolo de endpoint MCP, funções de ferramentas personalizadas |
| Envio de Comandos | Suporta envio de comandos MCP para dispositivos ESP32 via protocolo MQTT a partir do Console Inteligente |
| Backend de Gerenciamento | Fornece interface de gerenciamento Web, suporta gerenciamento de usuários, configuração do sistema e gerenciamento de dispositivos; Suporta exibição em Chinês Simplificado, Chinês Tradicional e Inglês |
| Ferramentas de Teste | Fornece ferramentas de teste de desempenho, ferramentas de teste de modelo de visão e ferramentas de teste de interação por áudio |
| Suporte à Implantação | Suporta implantação via Docker e implantação local, fornece gerenciamento completo de arquivos de configuração |
| Sistema de Plugins | Suporta extensões de plugins funcionais, desenvolvimento de plugins personalizados e carregamento dinâmico de plugins |
### Em Desenvolvimento 🚧
Para conhecer o progresso específico do plano de desenvolvimento, [clique aqui](https://github.com/users/xinnan-tech/projects/3). Perguntas frequentes e tutoriais relacionados podem ser consultados [neste link](../FAQ.md)
Se você é um desenvolvedor de software, aqui está uma [Carta Aberta aos Desenvolvedores](../contributor_open_letter.md). Seja bem-vindo a participar!
---
## Ecossistema do Produto 👬
Xiaozhi é um ecossistema. Ao utilizar este produto, você também pode conferir outros [projetos excelentes](https://github.com/78/xiaozhi-esp32/blob/main/README_zh.md#%E7%9B%B8%E5%85%B3%E5%BC%80%E6%BA%90%E9%A1%B9%E7%9B%AE) neste ecossistema
---
## Lista de Plataformas/Componentes Suportados 📋
### LLM Modelos de Linguagem
| Método de Uso | Plataformas Suportadas | Plataformas Gratuitas |
|:---:|:---:|:---:|
| Chamadas via interface OpenAI | Alibaba Bailian, Volcano Engine, DeepSeek, Zhipu, Gemini, iFLYTEK | Zhipu, Gemini |
| Chamadas via interface Ollama | Ollama | - |
| Chamadas via interface Dify | Dify | - |
| Chamadas via interface FastGPT | FastGPT | - |
| Chamadas via interface Coze | Coze | - |
| Chamadas via interface Xinference | Xinference | - |
| Chamadas via interface HomeAssistant | HomeAssistant | - |
Na verdade, qualquer LLM que suporte chamadas via interface openai pode ser integrado e utilizado.
---
### VLLM Modelos de Visão
| Método de Uso | Plataformas Suportadas | Plataformas Gratuitas |
|:---:|:---:|:---:|
| Chamadas via interface OpenAI | Alibaba Bailian, Zhipu ChatGLMVLLM | Zhipu ChatGLMVLLM |
Na verdade, qualquer VLLM que suporte chamadas via interface OpenAI pode ser integrado e utilizado.
---
### TTS Síntese de Voz
| Método de Uso | Plataformas Suportadas | Plataformas Gratuitas |
|:---:|:---:|:---:|
| Chamadas via interface | EdgeTTS, iFLYTEK, Volcano Engine, Tencent Cloud, Alibaba Cloud e Bailian, CosyVoiceSiliconflow, TTS302AI, CozeCnTTS, GizwitsTTS, ACGNTTS, OpenAITTS, Lingxi Streaming TTS, MinimaxTTS | Lingxi Streaming TTS, EdgeTTS, CosyVoiceSiliconflow(parcial) |
| Serviços locais | FishSpeech, GPT_SOVITS_V2, GPT_SOVITS_V3, Index-TTS, PaddleSpeech | Index-TTS, PaddleSpeech, FishSpeech, GPT_SOVITS_V2, GPT_SOVITS_V3 |
---
### VAD Detecção de Atividade Vocal
| Tipo | Nome da Plataforma | Método de Uso | Modelo de Preço | Observações |
|:---:|:---------:|:----:|:----:|:--:|
| VAD | SileroVAD | Uso local | Gratuito | |
---
### ASR Reconhecimento de Fala
| Método de Uso | Plataformas Suportadas | Plataformas Gratuitas |
|:---:|:---:|:---:|
| Uso local | FunASR, SherpaASR | FunASR, SherpaASR |
| Chamadas via interface | FunASRServer, Volcano Engine, iFLYTEK, Tencent Cloud, Alibaba Cloud, Baidu Cloud, OpenAI ASR | FunASRServer |
---
### Reconhecimento de Impressão Vocal
| Método de Uso | Plataformas Suportadas | Plataformas Gratuitas |
|:---:|:---:|:---:|
| Uso local | 3D-Speaker | 3D-Speaker |
---
### Armazenamento de Memória
| Tipo | Nome da Plataforma | Método de Uso | Modelo de Preço | Observações |
|:------:|:---------------:|:----:|:---------:|:--:|
| Memória | mem0ai | Chamadas via interface | Cota de 1000 vezes/mês | |
| Memória | [powermem](../powermem-integration.md) | Resumo local | Depende do LLM e BD | OceanBase de código aberto, suporta busca inteligente |
| Memória | mem_local_short | Resumo local | Gratuito | |
| Memória | nomem | Modo sem memória | Gratuito | |
---
### Reconhecimento de Intenção
| Tipo | Nome da Plataforma | Método de Uso | Modelo de Preço | Observações |
|:------:|:-------------:|:----:|:-------:|:---------------------:|
| Intenção | intent_llm | Chamadas via interface | Baseado no preço do LLM | Reconhece intenção através de modelos de linguagem, forte generalização |
| Intenção | function_call | Chamadas via interface | Baseado no preço do LLM | Completa a intenção através de chamada de função do modelo de linguagem, velocidade rápida, bom resultado |
| Intenção | nointent | Modo sem intenção | Gratuito | Não realiza reconhecimento de intenção, retorna diretamente o resultado do diálogo |
---
### RAG Geração Aumentada por Recuperação
| Tipo | Nome da Plataforma | Método de Uso | Modelo de Preço | Observações |
|:------:|:-------------:|:----:|:-------:|:---------------------:|
| RAG | ragflow | Chamadas via interface | Cobrado com base nos tokens consumidos para fatiamento e segmentação de palavras | Utiliza o recurso de geração aumentada por recuperação do RagFlow para fornecer respostas de diálogo mais precisas |
---
## Agradecimentos 🙏
| Logo | Projeto/Empresa | Descrição |
|:---:|:---:|:---|
| <img src="../images/logo_bailing.png" width="160"> | [Robô de Diálogo por Voz Bailing](https://github.com/wwbin2017/bailing) | Este projeto foi inspirado pelo [Robô de Diálogo por Voz Bailing](https://github.com/wwbin2017/bailing) e implementado com base nele |
| <img src="../images/logo_tenclass.png" width="160"> | [Tenclass](https://www.tenclass.com/) | Agradecimentos à [Tenclass](https://www.tenclass.com/) por formular protocolos de comunicação padrão, soluções de compatibilidade multidispositivo e demonstrações práticas de cenários de alta concorrência para o ecossistema Xiaozhi; fornecendo suporte completo de documentação técnica para este projeto |
| <img src="../images/logo_xuanfeng.png" width="160"> | [Xuanfeng Technology (玄凤科技)](https://github.com/Eric0308) | Agradecimentos à [Xuanfeng Technology](https://github.com/Eric0308) por contribuir com o framework de chamada de função, protocolo de comunicação MCP e implementação do mecanismo de chamada baseado em plugins. Através de um sistema padronizado de agendamento de instruções e capacidades de expansão dinâmica, melhora significativamente a eficiência de interação e extensibilidade funcional dos dispositivos de frontend (IoT) |
| <img src="../images/logo_junsen.png" width="160"> | [huangjunsen](https://github.com/huangjunsen0406) | Agradecimentos a [huangjunsen](https://github.com/huangjunsen0406) por contribuir com o módulo `Console de Controle Inteligente Mobile`, que permite controle eficiente e interação em tempo real em dispositivos móveis, melhorando significativamente a conveniência operacional e a eficiência de gerenciamento do sistema em cenários móveis. |
| <img src="../images/logo_huiyuan.png" width="160"> | [Huiyuan Design (汇远设计)](http://ui.kwd988.net/) | Agradecimentos à [Huiyuan Design](http://ui.kwd988.net/) por fornecer soluções visuais profissionais para este projeto, utilizando sua experiência prática de design atendendo mais de mil empresas para potencializar a experiência do usuário deste produto |
| <img src="../images/logo_qinren.png" width="160"> | [Xi'an Qinren Information Technology (西安勤人信息科技)](https://www.029app.com/) | Agradecimentos à [Xi'an Qinren Information Technology](https://www.029app.com/) por aprofundar o sistema visual deste projeto, garantindo consistência e extensibilidade do estilo de design geral em aplicações de múltiplos cenários |
| <img src="../images/logo_contributors.png" width="160"> | [Contribuidores de Código](https://github.com/xinnan-tech/xiaozhi-esp32-server/graphs/contributors) | Agradecimentos a [todos os contribuidores de código](https://github.com/xinnan-tech/xiaozhi-esp32-server/graphs/contributors), seus esforços tornaram o projeto mais robusto e poderoso. |
<a href="https://star-history.com/#xinnan-tech/xiaozhi-esp32-server&Date">
<picture>
<source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=xinnan-tech/xiaozhi-esp32-server&type=Date&theme=dark" />
<source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=xinnan-tech/xiaozhi-esp32-server&type=Date" />
<img alt="Star History Chart" src="https://api.star-history.com/svg?repos=xinnan-tech/xiaozhi-esp32-server&type=Date" />
</picture>
</a>
+362
View File
@@ -0,0 +1,362 @@
[![Banners](../images/banner1.png)](https://github.com/xinnan-tech/xiaozhi-esp32-server)
<h1 align="center">Dịch vụ Backend Xiaozhi xiaozhi-esp32-server</h1>
<p align="center">
Dự án này dựa trên lý thuyết và công nghệ trí tuệ cộng sinh người-máy để phát triển hệ thống phần mềm và phần cứng thiết bị đầu cuối thông minh<br/>Cung cấp dịch vụ backend cho dự án phần cứng thông minh mã nguồn mở
<a href="https://github.com/78/xiaozhi-esp32">xiaozhi-esp32</a><br/>
Được triển khai bằng Python, Java, Vue theo <a href="https://ccnphfhqs21z.feishu.cn/wiki/M0XiwldO9iJwHikpXD5cEx71nKh">giao thức truyền thông Xiaozhi</a><br/>
Hỗ trợ giao thức MQTT+UDP, giao thức Websocket, điểm truy cập MCP, nhận dạng giọng nói và kho tri thức
</p>
<p align="center">
<a href="../FAQ.md">Câu hỏi thường gặp</a>
· <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/issues">Báo cáo vấn đề</a>
· <a href="../../README.md#%E9%83%A8%E7%BD%B2%E6%96%87%E6%A1%A3">Tài liệu triển khai</a>
· <a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/releases">Nhật ký cập nhật</a>
</p>
<p align="center">
<a href="../../README.md"><img alt="简体中文版自述文件" src="https://img.shields.io/badge/简体中文-DFE0E5"></a>
<a href="./README_en.md"><img alt="README in English" src="https://img.shields.io/badge/English-DFE0E5"></a>
<a href="./README_vi.md"><img alt="Tiếng Việt" src="https://img.shields.io/badge/Tiếng Việt-DBEDFA"></a>
<a href="./README_de.md"><img alt="Deutsch" src="https://img.shields.io/badge/Deutsch-DFE0E5"></a>
<a href="./README_pt_BR.md"><img alt="Português (Brasil)" src="https://img.shields.io/badge/Português (Brasil)-DFE0E5"></a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/releases">
<img alt="GitHub Contributors" src="https://img.shields.io/github/v/release/xinnan-tech/xiaozhi-esp32-server?logo=docker" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server/blob/main/LICENSE">
<img alt="GitHub pull requests" src="https://img.shields.io/badge/license-MIT-white?labelColor=black" />
</a>
<a href="https://github.com/xinnan-tech/xiaozhi-esp32-server">
<img alt="stars" src="https://img.shields.io/github/stars/xinnan-tech/xiaozhi-esp32-server?color=ffcb47&labelColor=black" />
</a>
</p>
<p align="center">
Spearheaded by Professor Siyuan Liu's Team (South China University of Technology)
</br>
Được dẫn dắt bởi nhóm Giáo sư Lưu Tư Nguyên (Đại học Bách khoa Nam Trung Quốc)
</br>
<img src="../images/hnlg.jpg" alt="华南理工大学" width="50%">
</p>
---
## Người dùng phù hợp 👥
Dự án này cần được sử dụng cùng với thiết bị phần cứng ESP32. Nếu bạn đã mua phần cứng liên quan đến ESP32, đã thành công kết nối với dịch vụ backend do anh Xia triển khai, và muốn xây dựng dịch vụ backend `xiaozhi-esp32` riêng của mình, thì dự án này rất phù hợp với bạn.
Muốn xem hiệu quả sử dụng? Hãy xem video 🎥
<table>
<tr>
<td>
<a href="https://www.bilibili.com/video/BV1FMFyejExX" target="_blank">
<picture>
<img alt="响应速度感受" src="docs/images/demo9.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1vchQzaEse" target="_blank">
<picture>
<img alt="速度优化秘诀" src="docs/images/demo6.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1WEcxzFEAT" target="_blank">
<picture>
<img alt="小智数字人 支持语音唤醒" src="docs/images/demo8.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1CKVz6UEuB" target="_blank">
<picture>
<img alt="设备呼叫设备,打电话" src="docs/images/demo0.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1C1tCzUEZh" target="_blank">
<picture>
<img alt="复杂医疗场景" src="docs/images/demo1.png" /></picture>
</a>
</td>
</tr>
<tr>
<td>
<a href="https://www.bilibili.com/video/BV1VC96Y5EMH" target="_blank">
<picture>
<img alt="播放音乐查询天气播报新闻" src="docs/images/demo7.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV12J7WzBEaH" target="_blank">
<picture>
<img alt="实时打断" src="docs/images/demo10.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1Co76z7EvK" target="_blank">
<picture>
<img alt="拍照识物品" src="docs/images/demo12.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1pNXWYGEx1" target="_blank">
<picture>
<img alt="控制家电开关" src="docs/images/demo5.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1TJ7WzzEo6" target="_blank">
<picture>
<img alt="多指令任务" src="docs/images/demo11.png" /></picture>
</a>
</td>
</tr>
<tr>
<td>
<a href="https://www.bilibili.com/video/BV1ZQKUzYExM" target="_blank">
<picture>
<img alt="MCP接入点" src="docs/images/demo13.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1zUW5zJEkq" target="_blank">
<picture>
<img alt="MQTT指令下发" src="docs/images/demo4.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1Exu3zqEDe" target="_blank">
<picture>
<img alt="声纹识别" src="docs/images/demo14.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV1CDKWemEU6" target="_blank">
<picture>
<img alt="自定义音色" src="docs/images/demo2.png" /></picture>
</a>
</td>
<td>
<a href="https://www.bilibili.com/video/BV12yA2egEaC" target="_blank">
<picture>
<img alt="使用粤语交流" src="docs/images/demo3.png" /></picture>
</a>
</td>
</tr>
</table>
---
## Cảnh báo ⚠️
1. Dự án này là phần mềm mã nguồn mở, phần mềm này không có quan hệ hợp tác thương mại với bất kỳ nhà cung cấp dịch vụ API bên thứ ba nào (bao gồm nhưng không giới hạn ở các nền tảng nhận dạng giọng nói, mô hình lớn, tổng hợp giọng nói, v.v.), và không đảm bảo chất lượng dịch vụ cũng như an toàn tài chính của họ.
Khuyến nghị người dùng ưu tiên lựa chọn nhà cung cấp dịch vụ có giấy phép kinh doanh liên quan và đọc kỹ thỏa thuận dịch vụ và chính sách bảo mật của họ. Phần mềm này không lưu trữ bất kỳ khóa tài khoản nào, không tham gia vào luồng tiền và không chịu rủi ro mất tiền nạp.
2. Chức năng của dự án này chưa hoàn thiện và chưa qua đánh giá bảo mật mạng, vui lòng không sử dụng trong môi trường sản xuất. Nếu bạn triển khai dự án này trong môi trường mạng công cộng để học tập, vui lòng thực hiện các biện pháp bảo vệ cần thiết.
---
## Tài liệu triển khai
![Banners](../images/banner2.png)
Dự án này cung cấp hai phương pháp triển khai, vui lòng chọn theo nhu cầu cụ thể của bạn:
#### 🚀 Lựa chọn phương pháp triển khai
| Phương pháp triển khai | Đặc điểm | Tình huống áp dụng | Tài liệu triển khai | Yêu cầu cấu hình | Video hướng dẫn |
|---------|------|---------|---------|---------|---------|
| **Cài đặt tối giản** | Đối thoại thông minh, quản lý đơn tác nhân | Môi trường cấu hình thấp, dữ liệu lưu trong tệp cấu hình, không cần cơ sở dữ liệu | [①Phiên bản Docker](../Deployment.md#%E6%96%B9%E5%BC%8F%E4%B8%80docker%E5%8F%AA%E8%BF%90%E8%A1%8Cserver) / [②Triển khai mã nguồn](../Deployment.md#%E6%96%B9%E5%BC%8F%E4%BA%8C%E6%9C%AC%E5%9C%B0%E6%BA%90%E7%A0%81%E5%8F%AA%E8%BF%90%E8%A1%8Cserver)| 2 nhân 4GB nếu dùng `FunASR`, 2 nhân 2GB nếu toàn API | - |
| **Cài đặt toàn bộ module** | Đối thoại thông minh, quản lý đa người dùng, quản lý đa tác nhân, bảng điều khiển thông minh | Trải nghiệm đầy đủ tính năng, dữ liệu lưu trong cơ sở dữ liệu |[①Phiên bản Docker](../Deployment_all.md#%E6%96%B9%E5%BC%8F%E4%B8%80docker%E8%BF%90%E8%A1%8C%E5%85%A8%E6%A8%A1%E5%9D%97) / [②Triển khai mã nguồn](../Deployment_all.md#%E6%96%B9%E5%BC%8F%E4%BA%8C%E6%9C%AC%E5%9C%B0%E6%BA%90%E7%A0%81%E8%BF%90%E8%A1%8C%E5%85%A8%E6%A8%A1%E5%9D%97) / [③Hướng dẫn tự động cập nhật triển khai mã nguồn](../dev-ops-integration.md) | 4 nhân 8GB nếu dùng `FunASR`, 2 nhân 4GB nếu toàn API| [Video hướng dẫn khởi động mã nguồn cục bộ](https://www.bilibili.com/video/BV1wBJhz4Ewe) |
Câu hỏi thường gặp và hướng dẫn liên quan, vui lòng tham khảo [liên kết này](../FAQ.md)
> 💡 Gợi ý: Dưới đây là nền tảng thử nghiệm được triển khai theo mã mới nhất, có thể flash để thử nghiệm nếu cần, đồng thời là 6, dữ liệu sẽ được xóa mỗi ngày,
```
Địa chỉ bảng điều khiển thông minh: https://2662r3426b.vicp.fun
Bảng điều khiển thông minh (phiên bản h5): https://2662r3426b.vicp.fun/h5/index.html
Công cụ kiểm tra dịch vụ: https://2662r3426b.vicp.fun/test/
Địa chỉ giao diện OTA: https://2662r3426b.vicp.fun/xiaozhi/ota/
Địa chỉ giao diện Websocket: wss://2662r3426b.vicp.fun/xiaozhi/v1/
```
#### 🚩 Mô tả và khuyến nghị cấu hình
> [!Note]
> Dự án này cung cấp hai phương án cấu hình:
>
> 1. Cấu hình `Miễn phí hoàn toàn cho người mới`: Phù hợp với sử dụng gia đình cá nhân, tất cả các thành phần đều sử dụng phương án miễn phí, không cần thanh toán thêm.
>
> 2. `Cấu hình streaming`: Phù hợp với demo, đào tạo, hơn 2 đồng thời, v.v., sử dụng công nghệ xử lý streaming, tốc độ phản hồi nhanh hơn, trải nghiệm tốt hơn.
>
> Từ phiên bản `0.5.2`, dự án hỗ trợ cấu hình streaming, so với phiên bản đầu, tốc độ phản hồi cải thiện khoảng `2.5 giây`, cải thiện đáng kể trải nghiệm người dùng.
| Tên module | Cài đặt miễn phí cho người mới | Cấu hình streaming |
|:---:|:---:|:---:|
| ASR(Nhận dạng giọng nói) | FunASR(Local) | 👍XunfeiStreamASR(Xunfei Streaming) |
| LLM(Mô hình lớn) | glm-4-flash(Zhipu) | 👍qwen-flash(Alibaba Bailian) |
| VLLM(Mô hình lớn thị giác) | glm-4v-flash(Zhipu) | 👍qwen3.5-flash(Alibaba Bailian) |
| TTS(Tổng hợp giọng nói) | EdgeTTS(Microsoft) | 👍HuoshanDoubleStreamTTS(Volcano Streaming) |
| Intent(Nhận dạng ý định) | function_call(Gọi hàm) | function_call(Gọi hàm) |
| Memory(Chức năng bộ nhớ) | mem_local_short(Bộ nhớ ngắn hạn cục bộ) | mem_local_short(Bộ nhớ ngắn hạn cục bộ) |
Nếu bạn quan tâm đến thời gian của từng thành phần, vui lòng xem [Báo cáo kiểm tra hiệu suất các thành phần Xiaozhi](https://github.com/xinnan-tech/xiaozhi-performance-research), có thể kiểm tra thực tế trong môi trường của bạn theo phương pháp kiểm tra trong báo cáo.
#### 🔧 Công cụ kiểm tra
Dự án này cung cấp các công cụ kiểm tra sau để giúp bạn xác minh hệ thống và chọn mô hình phù hợp:
| Tên công cụ | Vị trí | Phương pháp sử dụng | Mô tả chức năng |
|:---:|:---|:---:|:---:|
| Công cụ kiểm tra tương tác âm thanh | main》digital-human》index.html | Chạy `python start.py` trong `main/digital-human`, sau đó mở `http://127.0.0.1:8006/index.html` | Kiểm tra chức năng phát và nhận âm thanh, xác minh xử lý âm thanh phía Python có bình thường không |
| Công cụ kiểm tra phản hồi mô hình | main》xiaozhi-server》performance_tester.py | Thực hiện `python performance_tester.py` | Kiểm tra tốc độ phản hồi của ba module cốt lõi ASR(Nhận dạng giọng nói), LLM(Mô hình lớn), VLLM(Mô hình thị giác), TTS(Tổng hợp giọng nói) |
> 💡 Gợi ý: Khi kiểm tra tốc độ mô hình, chỉ kiểm tra các mô hình đã cấu hình khóa.
---
## Danh sách tính năng ✨
### Đã thực hiện ✅
![请参考-全模块安装架构图](../images/deploy2.png)
| Module chức năng | Mô tả |
|:---:|:---|
| Kiến trúc cốt lõi | Dựa trên [cổng MQTT+UDP](https://github.com/xinnan-tech/xiaozhi-esp32-server/blob/main/docs/mqtt-gateway-integration.md), WebSocket, máy chủ HTTP, cung cấp hệ thống quản lý bảng điều khiển và xác thực hoàn chỉnh |
| Tương tác giọng nói | Hỗ trợ ASR streaming(Nhận dạng giọng nói), TTS streaming(Tổng hợp giọng nói), VAD(Phát hiện hoạt động giọng nói), hỗ trợ nhận dạng đa ngôn ngữ và xử lý giọng nói |
| Nhận dạng vân giọng | Hỗ trợ đăng ký, quản lý và nhận dạng vân giọng đa người dùng, xử lý song song với ASR, nhận dạng danh tính người nói theo thời gian thực và truyền cho LLM để phản hồi cá nhân hóa |
| Đối thoại thông minh | Hỗ trợ nhiều LLM(Mô hình ngôn ngữ lớn), thực hiện đối thoại thông minh |
| Cảm nhận thị giác | Hỗ trợ nhiều VLLM(Mô hình lớn thị giác), thực hiện tương tác đa phương thức |
| Nhận dạng ý định | Hỗ trợ nhận dạng ý định mô hình lớn gắn ngoài, gọi hàm tự chủ mô hình lớn, cung cấp cơ chế xử lý ý định dạng plugin |
| Hệ thống bộ nhớ | Hỗ trợ bộ nhớ ngắn hạn cục bộ, bộ nhớ giao diện mem0ai, bộ nhớ thông minh PowerMem, có chức năng tóm tắt bộ nhớ |
| Kho tri thức | Hỗ trợ kho tri thức RAGFlow, cho phép mô hình lớn đánh giá cần gọi kho tri thức trước khi trả lời |
| Gọi công cụ | Hỗ trợ giao thức IOT phía client, giao thức MCP phía client, giao thức MCP phía server, giao thức điểm truy cập MCP, hàm công cụ tùy chỉnh |
| Gửi lệnh | Dựa vào giao thức MQTT, hỗ trợ gửi lệnh MCP từ bảng điều khiển thông minh xuống thiết bị ESP32 |
| Backend quản lý | Cung cấp giao diện quản lý Web, hỗ trợ quản lý người dùng, cấu hình hệ thống và quản lý thiết bị; giao diện hỗ trợ hiển thị tiếng Trung giản thể, tiếng Trung phồn thể, tiếng Anh |
| Công cụ kiểm tra | Cung cấp công cụ kiểm tra hiệu suất, công cụ kiểm tra mô hình thị giác và công cụ kiểm tra tương tác âm thanh |
| Hỗ trợ triển khai | Hỗ trợ triển khai Docker và triển khai cục bộ, cung cấp quản lý tệp cấu hình hoàn chỉnh |
| Hệ thống plugin | Hỗ trợ mở rộng plugin chức năng, phát triển plugin tùy chỉnh và hot loading plugin |
### Đang phát triển 🚧
Muốn hiểu tiến độ kế hoạch phát triển cụ thể, [vui lòng nhấp vào đây](https://github.com/users/xinnan-tech/projects/3). Câu hỏi thường gặp và hướng dẫn liên quan, vui lòng tham khảo [liên kết này](../FAQ.md)
Nếu bạn là một nhà phát triển phần mềm, đây có một [Lá thư mở gửi các nhà phát triển](../contributor_open_letter.md), chào mừng tham gia!
---
## Hệ sinh thái sản phẩm 👬
Xiaozhi là một hệ sinh thái, khi bạn sử dụng sản phẩm này, bạn cũng có thể xem các [dự án xuất sắc](https://github.com/78/xiaozhi-esp32/blob/main/README_zh.md#%E7%9B%B8%E5%85%B3%E5%BC%80%E6%BA%90%E9%A1%B9%E7%9B%AE) khác trong hệ sinh thái này
---
## Danh sách nền tảng/thành phần được dự án này hỗ trợ 📋
### LLM Mô hình ngôn ngữ
| Phương pháp sử dụng | Nền tảng hỗ trợ | Nền tảng miễn phí |
|:---:|:---:|:---:|
| Gọi giao diện openai | Alibaba Bailian, Volcano Engine, DeepSeek, Zhipu, Gemini, iFlytek | Zhipu, Gemini |
| Gọi giao diện ollama | Ollama | - |
| Gọi giao diện dify | Dify | - |
| Gọi giao diện fastgpt | Fastgpt | - |
| Gọi giao diện coze | Coze | - |
| Gọi giao diện xinference | Xinference | - |
| Gọi giao diện homeassistant | HomeAssistant | - |
Trên thực tế, bất kỳ LLM nào hỗ trợ gọi giao diện openai đều có thể truy cập sử dụng.
---
### VLLM Mô hình thị giác
| Phương pháp sử dụng | Nền tảng hỗ trợ | Nền tảng miễn phí |
|:---:|:---:|:---:|
| Gọi giao diện openai | Alibaba Bailian, Zhipu ChatGLMVLLM | Zhipu ChatGLMVLLM |
Trên thực tế, bất kỳ VLLM nào hỗ trợ gọi giao diện openai đều có thể truy cập sử dụng.
---
### TTS Tổng hợp giọng nói
| Phương pháp sử dụng | Nền tảng hỗ trợ | Nền tảng miễn phí |
|:---:|:---:|:---:|
| Gọi giao diện | EdgeTTS, iFlytek, Volcano Engine, Tencent Cloud, Alibaba Cloud và Bailian, CosyVoiceSiliconflow, TTS302AI, CozeCnTTS, GizwitsTTS, ACGNTTS, OpenAITTS, Lingxi streaming TTS, MinimaxTTS | Lingxi streaming TTS, EdgeTTS, CosyVoiceSiliconflow(một phần) |
| Dịch vụ cục bộ | FishSpeech, GPT_SOVITS_V2, GPT_SOVITS_V3, Index-TTS, PaddleSpeech | Index-TTS, PaddleSpeech, FishSpeech, GPT_SOVITS_V2, GPT_SOVITS_V3 |
---
### VAD Phát hiện hoạt động giọng nói
| Loại | Tên nền tảng | Phương pháp sử dụng | Mô hình thu phí | Ghi chú |
|:---:|:---------:|:----:|:----:|:--:|
| VAD | SileroVAD | Sử dụng cục bộ | Miễn phí | |
---
### ASR Nhận dạng giọng nói
| Phương pháp sử dụng | Nền tảng hỗ trợ | Nền tảng miễn phí |
|:---:|:---:|:---:|
| Sử dụng cục bộ | FunASR, SherpaASR | FunASR, SherpaASR |
| Gọi giao diện | FunASRServer, Volcano Engine, iFlytek, Tencent Cloud, Alibaba Cloud, Baidu Cloud, OpenAI ASR | FunASRServer |
---
### Voiceprint Nhận dạng vân giọng
| Phương pháp sử dụng | Nền tảng hỗ trợ | Nền tảng miễn phí |
|:---:|:---:|:---:|
| Sử dụng cục bộ | 3D-Speaker | 3D-Speaker |
---
### Memory Lưu trữ bộ nhớ
| Loại | Tên nền tảng | Phương pháp sử dụng | Mô hình thu phí | Ghi chú |
|:------:|:---------------:|:----:|:---------:|:--:|
| Memory | mem0ai | Gọi giao diện | Hạn mức 1000 lần/tháng | |
| Memory | [powermem](../powermem-integration.md) | Tóm tắt cục bộ | Phụ thuộc vào LLM và DB | OceanBase mã nguồn mở, hỗ trợ tìm kiếm thông minh |
| Memory | mem_local_short | Tóm tắt cục bộ | Miễn phí | |
| Memory | nomem | Chế độ không có bộ nhớ | Miễn phí | |
---
### Intent Nhận dạng ý định
| Loại | Tên nền tảng | Phương pháp sử dụng | Mô hình thu phí | Ghi chú |
|:------:|:-------------:|:----:|:-------:|:---------------------:|
| Intent | intent_llm | Gọi giao diện | Thu phí theo LLM | Nhận dạng ý định qua mô hình lớn, tính tổng quát mạnh |
| Intent | function_call | Gọi giao diện | Thu phí theo LLM | Hoàn thành ý định qua gọi hàm mô hình lớn, tốc độ nhanh, hiệu quả tốt |
| Intent | nointent | Chế độ không có ý định | Miễn phí | Không thực hiện nhận dạng ý định, trả về trực tiếp kết quả đối thoại |
---
### Rag Tăng cường truy xuất thông tin
| Loại | Tên nền tảng | Phương pháp sử dụng | Mô hình thu phí | Ghi chú |
|:------:|:-------------:|:----:|:-------:|:---------------------:|
| Rag | ragflow | Gọi giao diện | Thu phí theo token tiêu tốn của phân đoạn, phân từ | Sử dụng chức năng tăng cường truy xuất của RagFlow, cung cấp phản hồi đối thoại chính xác hơn |
---
## Lời cảm ơn 🙏
| Logo | Dự án/Công ty | Mô tả |
|:---:|:---:|:---|
| <img src="../images/logo_bailing.png" width="160"> | [Robot đối thoại giọng nói Bailing](https://github.com/wwbin2017/bailing) | Dự án này được lấy cảm hứng từ [Robot đối thoại giọng nói Bailing](https://github.com/wwbin2017/bailing) và được triển khai trên cơ sở đó |
| <img src="../images/logo_tenclass.png" width="160"> | [Shifang Ronghai](https://www.tenclass.com/) | Cảm ơn [Shifang Ronghai](https://www.tenclass.com/) đã xây dựng giao thức truyền thông tiêu chuẩn, phương án tương thích đa thiết bị và mô phạm thực hành tình huống đồng thời cao cho hệ sinh thái Xiaozhi; cung cấp tài liệu hỗ trợ kỹ thuật toàn diện cho dự án này |
| <img src="../images/logo_xuanfeng.png" width="160"> | [Xuanfeng Technology](https://github.com/Eric0308) | Cảm ơn [Xuanfeng Technology](https://github.com/Eric0308) đã đóng góp khung gọi hàm, giao thức truyền thông MCP và mã triển khai cơ chế gọi dạng plugin, thông qua hệ thống điều phối lệnh tiêu chuẩn hóa và khả năng mở rộng động, đã cải thiện đáng kể hiệu suất tương tác và khả năng mở rộng chức năng của thiết bị front-end(IoT) |
| <img src="../images/logo_junsen.png" width="160"> | [huangjunsen](https://github.com/huangjunsen0406) | Cảm ơn [huangjunsen](https://github.com/huangjunsen0406) đã đóng góp module `Bảng điều khiển thông minh di động`, thực hiện điều khiển hiệu quả và tương tác thời gian thực trên thiết bị di động đa nền tảng, cải thiện đáng kể sự tiện lợi vận hành và hiệu quả quản lý của hệ thống trong tình huống di động |
| <img src="../images/logo_huiyuan.png" width="160"> | [Huiyuan Design](http://ui.kwd988.net/) | Cảm ơn [Huiyuan Design](http://ui.kwd988.net/) đã cung cấp giải pháp thị giác chuyên nghiệp cho dự án này, sử dụng kinh nghiệm thực tế thiết kế phục vụ hơn nghìn doanh nghiệp, trao quyền cho trải nghiệm người dùng sản phẩm của dự án này |
| <img src="../images/logo_qinren.png" width="160"> | [Xi'an Qinren Information Technology](https://www.029app.com/) | Cảm ơn [Xi'an Qinren Information Technology](https://www.029app.com/) đã làm sâu sắc hệ thống thị giác của dự án này, đảm bảo tính nhất quán và khả năng mở rộng của phong cách thiết kế tổng thể trong ứng dụng đa tình huống |
| <img src="../images/logo_contributors.png" width="160"> | [Người đóng góp mã](https://github.com/xinnan-tech/xiaozhi-esp32-server/graphs/contributors) | Cảm ơn [tất cả người đóng góp mã](https://github.com/xinnan-tech/xiaozhi-esp32-server/graphs/contributors), sự cống hiến của bạn khiến dự án mạnh mẽ và vững chắc hơn. |
<a href="https://star-history.com/#xinnan-tech/xiaozhi-esp32-server&Date">
<picture>
<source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=xinnan-tech/xiaozhi-esp32-server&type=Date&theme=dark" />
<source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=xinnan-tech/xiaozhi-esp32-server&type=Date" />
<img alt="Star History Chart" src="https://api.star-history.com/svg?repos=xinnan-tech/xiaozhi-esp32-server&type=Date" />
</picture>
</a>
+88
View File
@@ -0,0 +1,88 @@
# 统一 FastAPI 平台开发计划
状态:规划基线
更新时间:2026-07-20
## 1. 项目目标
`manager-api-fastapi``manager-web``xiaozhi-server` 组织成同一套可版本化、可测试、
可部署和可回滚的平台,同时保持当前用户可见功能、设备协议和已存数据可用。
这里的“统一”指同一代码库、同一发布版本和同一公网入口,不要求所有职责运行在同一个
Uvicorn worker。生产环境保留 API、实时连接和定时任务的独立进程边界,轻量环境可以使用
单 worker 的 `all` 运行模式。
## 2. 完成定义
进入真实环境验收前,仓库必须同时满足:
1. 管理端页面及其现有操作均由统一发行物提供。
2. 管理 REST API、设备 OTA、WebSocket、Vision、MCP、IoT 和 MQTT 桥接协议均有自动化契约测试。
3. 虚拟设备可以完成连接、绑定、对话、中断、工具调用、断线和重连的端到端流程。
4. 所有 Provider 均完成接口适配和 Fake/契约验证;无需真实凭证的路径全部自动通过。
5. MySQL、Redis、文件、配置热更新、jobs、升级和回滚均经过隔离集成测试。
6. 统一 Docker/Compose 发行物可启动,健康检查、日志、指标和优雅停机可验证。
7. 所有尚未验证的风险只依赖真实设备、真实外部服务或目标生产环境,并已列入交接清单。
达到以上条件后,项目状态为 `RC / 等待真实环境验收`,不能提前表述为生产验收完成。
## 3. 范围
包含:
- `manager-api-fastapi` 已有管理 API、数据库、Redis、文件和 jobs 能力。
- `manager-web` 的构建、静态交付、PWA 和浏览器业务流程。
- `xiaozhi-server` 的设备 WebSocket、音频会话、Provider、Vision、MCP、IoT、插件和网关桥接能力。
- 本地轻量配置模式与数据库控制台模式,但共享一套实现,不保留重复路由。
- CI、镜像、部署、可观测性、迁移、回滚和真实测试交接。
不包含:
- 改变现有业务规则或重新设计管理端产品体验。
- 在没有凭证时声称云 Provider 已真实联调。
- 把外部 MQTT/UDP gateway、RAGFlow 或云模型服务复制进本仓库。
- 在没有真实 ESP32 和目标环境时批准生产上线。
## 4. 不可破坏的兼容边界
- 设备 OTA URL、请求头、响应字段及激活行为。
- `/xiaozhi/v1/` 的 WebSocket 握手、文本消息、Opus 帧和 MQTT 桥接帧。
- Device-ID、Client-ID、服务密钥和 HMAC Token 语义。
- 已有数据库数据、用户权限、智能体配置、模型标识及文件资产。
- 当前 Web、Mobile 和设备仍在使用的公开 API;内部自调用不属于兼容边界。
- 已启用 Provider 的配置字段和可观察结果。
## 5. 工作原则
1. 先记录现状,再替换实现:没有特征测试的行为不得直接重写。
2. 以可工作的纵向切片提交 PR,避免一次性迁移整个实时服务。
3. 共享代码,隔离运行职责:API、realtime、jobs 可以使用同一镜像但独立启动。
4. 新旧实现可并行对照且可回滚,删除旧入口必须是最后阶段。
5. Provider 按能力族批量适配,默认不为每个供应商启动独立 Agent。
6. 每个 PR 只有一个明确负责人和一个独立审阅角色。
7. 自动化可以判断的事项不等待人工;真实环境不可替代的事项不得用 Mock 冒充通过。
## 6. 文档导航
- [目标架构](architecture.md)
- [阶段路线图与 PR 队列](roadmap.md)
- [多岗位 Agent 协作模型](agent-operating-model.md)
- [质量门禁](quality-gates.md)
- [架构与交付决策](decisions.md)
- [风险登记册](risk-register.md)
- [真实环境测试交接](real-environment-handoff.md)
## 7. 当前基线
- FastAPI 管理 API 已形成 154 条 Java 路由兼容基线和现有测试报告,但逐路由直接深度差分
目前只覆盖 21/154;其余成功路径和副作用证据必须在 M5 前补齐。
- manager-web 的 i18n、unit、snapshot 和生产构建已有绿色记录。
- xiaozhi-server 可通过 Python 语法编译,但缺少正式自动化单测和设备协议回归套件。
- FastAPI、部署文档和脚本已作为集成分支基线提交;后续重构只通过目标为该分支的 PR 进入。
- 真实 RAGFlow、短信、语音、MQTT/MCP 和 ESP32 尚未验收。
项目状态以 [roadmap.md](roadmap.md) 的阶段门禁为唯一进度来源。
开发期间以 `refactor/unified-fastapi-platform` 为长期集成分支。所有规划和实现 PR 均合入该分支,
不会在纯 Agent 阶段向 `main` 创建 PR。
@@ -0,0 +1,118 @@
# 多岗位 Agent 协作模型
## 1. 岗位与责任
| 岗位 | 主要责任 | 不负责 |
| --- | --- | --- |
| PM / 集成负责人 | 范围、优先级、依赖、PR 队列、风险、最终门禁 | 代替专项 Agent 编写所有业务实现 |
| 架构负责人 | 接口边界、ADR、跨模块依赖、并发和状态模型 | 未经测试直接大规模搬迁代码 |
| Realtime 工程 Agent | ASGI WebSocket、DeviceSession、音频队列和生命周期 | 修改管理业务规则 |
| Backend 工程 Agent | 应用服务、配置、数据库、Redis、bootstrap | 修改设备协议 |
| Provider 工程 Agent | 按能力族适配已有 Provider | 改变供应商可见配置语义 |
| Frontend/交付 Agent | manager-web 构建、静态托管和路由 | 重做页面产品设计 |
| Platform/SRE Agent | CI、镜像、Compose、健康、指标、升级回滚 | 在缺少证据时批准上线 |
| QA Agent | 特征测试、差分、虚拟设备、E2E、性能和故障注入 | 为使测试变绿而放宽契约 |
| 独立 Reviewer | 安全、兼容、并发、资源和证据审阅 | 与实现者共享同一结论来源 |
## 2. 工作包准入
分配给 Agent 前,工作包必须明确:
- 目标和不在范围内的事项。
- 允许修改的目录和禁止修改的公共接口。
- 上游依赖及其版本/提交。
- 必须新增或保持的测试。
- 完成命令和预期证据。
- 需要升级给 PM 的决策点。
工作包模板:
```text
目标:
用户可见结果:
允许修改:
禁止修改:
输入契约:
输出契约:
必须运行:
完成条件:
已知风险:
```
## 3. 资源控制
1. 默认使用一个实现 Agent;只有文件所有权互斥且确有关键路径收益时才并行。
2. 同时活动的实现 Agent 不超过三个,另保留一个集成/审阅角色。
3. 不为单个 Provider 启动一个 Agent;按能力族分组,出现特有失败时再拆分。
4. 不让两个 Agent 重复做全仓审计。审阅 Agent 只读取 PR diff、契约和相关文件。
5. 大型上下文按目录和接口裁剪;交接必须写入文档或 PR,不依赖聊天记忆。
6. 长耗时集成、容器、性能测试只在相关代码变化、里程碑收口或夜间任务运行。
7. 发现范围外缺陷时记录后移,不顺手扩张当前 PR。
## 4. 文件所有权与冲突避免
活动 PR 必须声明临时所有权:
| 范围 | 默认岗位 |
| --- | --- |
| `app/realtime/**`、协议测试 | Realtime |
| `app/application/**``app/domain/**` | Architecture/Backend |
| `app/providers/**` | 对应 Provider 能力族 |
| `app/api/**`、现有 routers/services | Backend |
| `main/manager-web/**`、gateway 静态配置 | Frontend/交付 |
| Docker、Compose、CI | Platform/SRE |
| `tests/e2e/**`、虚拟设备、差分工具 | QA |
| 本目录路线图、风险和交接文档 | PM/集成 |
需要跨所有权修改时,先由当前所有者提供最小接口,再继续实现;禁止在合并阶段才发现公共接口冲突。
## 5. 分支与 PR 规则
- 长期集成分支:`refactor/unified-fastapi-platform`;纯 Agent 阶段的开发 PR 均以它为目标。
- 工作分支命名:`platform/<milestone>-<short-topic>`
- 分支名、Commit 和 PR 标题遵守仓库禁用词要求。
- Commit 使用结果导向格式,例如 `test: add device protocol characterization`
- 一个 PR 只交付一个计划工作包或一个可独立回滚的纵向切片。
- PR 描述必须列出阶段、依赖、契约影响、验证命令、风险和回滚方法。
- 实现者不得作为唯一 Reviewer;高风险协议、鉴权、迁移和并发 PR 必须有专项审阅。
- 合并前同步最新目标分支并重新运行必需门禁。
- 不允许未解释的 skip、仅本机通过、手工修改测试数据或把 Mock 结果描述成真实联调。
- 未经项目所有者明确批准,不创建以 `main` 为目标的 PR。
可选 PR 模板位于 `.github/PULL_REQUEST_TEMPLATE/unified-platform.md`
## 6. PR 生命周期
```text
Ready work package
-> implementation + local tests
-> draft PR + evidence
-> independent review
-> required CI gates
-> PM dependency/risk check
-> merge
-> roadmap evidence link
```
发生以下任一情况必须暂停并升级:
- 需要改变设备公开协议或现有数据语义。
- 需要真实凭证、真实设备或目标环境才能判断实现方向。
- 同一测试在旧实现上也失败,且无法确定基线行为。
- 依赖冲突要求删除现有 Provider。
- PR 需要跨越两个以上尚未完成的里程碑。
## 7. Agent 交接格式
```text
完成内容:
变更文件:
契约变化:无 / 具体说明
执行过的命令与结果:
未执行项目及原因:
遗留风险:
下一工作包可依赖的接口:
建议 Reviewer 重点:
```
没有上述交接或可重复证据的工作不进入集成分支。
@@ -0,0 +1,141 @@
# 目标架构
## 1. 架构决策
统一平台采用“模块化单体代码库、按职责运行”的结构:
```text
Public HTTP/WSS
|
v
Gateway / static web
| | |
| | +--> /mcp/vision/explain
| +--------------------> /xiaozhi/v1/ -> realtime
+-------------------------------> /xiaozhi/* -> api
api ----------- application services -------- database / redis / object files
realtime ------ application services -------- provider adapters
jobs ---------- application services -------- scheduled work
```
同一发布版本提供四个角色:
| 角色 | 职责 | 扩缩容方式 |
| --- | --- | --- |
| `gateway` | manager-web、TLS、静态缓存、HTTP/WS 路由 | 无状态横向扩展 |
| `api` | 管理 REST、OTA、文件和内部服务端 API | 多 worker/多副本 |
| `realtime` | 设备 WebSocket、音频会话、Vision、Provider 和工具 | 按连接与模型容量扩展 |
| `jobs` | 定时同步、清理和异步补偿 | Redis 租约保证单任务所有权 |
开发环境可以用一个命令启动全部角色;生产环境不得要求 realtime 与 API 共享 worker。
## 2. 建议代码边界
```text
main/manager-api-fastapi/
├── app/
│ ├── api/ # HTTP 路由和请求适配
│ ├── application/ # 用例服务;HTTP、WS、jobs 共用
│ ├── domain/ # 稳定业务对象和接口
│ ├── realtime/ # ASGI WebSocket、会话和协议
│ ├── providers/ # VAD/ASR/LLM/VLLM/TTS/Memory/Intent/Tools
│ ├── infrastructure/ # MySQL、Redis、文件、外部客户端
│ └── jobs/
├── web/ # manager-web 构建集成或其产物约定
├── tests/
│ ├── unit/
│ ├── contract/
│ ├── integration/
│ ├── protocol/
│ └── e2e/
└── deploy/
```
迁移期间允许现有目录继续存在;目录调整必须跟随可运行的纵向切片,不做只有移动文件的超大 PR。
## 3. 关键内部接口
### 3.1 应用服务
Realtime 不再通过 HTTP 调用同一平台。以下能力通过应用服务接口复用:
- 获取全局配置和设备专属 Agent 配置。
- 设备绑定、激活、在线状态和通讯录查找。
- 聊天记录、音频、摘要、标题和工具调用上报。
- OTA、文件和参数读取。
HTTP 只是这些服务的一个适配器。服务方法不接收 FastAPI `Request`,也不返回 HTTP Response。
### 3.2 实时会话
每个连接由一个 `DeviceSession` 拥有,至少包含:
- 握手元数据与设备鉴权。
- 有界输入/输出音频队列。
- 文本协议路由。
- VAD/ASR/LLM/TTS/工具调用任务。
- 取消、超时、断线保存和资源关闭。
使用 ASGI WebSocket 抽象,不让 Provider 依赖 Starlette 或 `websockets.ServerConnection`
### 3.3 Provider
Provider 通过能力协议注册,配置标识保持现状。同步 SDK 必须通过受限线程池或专用执行器调用,
禁止在事件循环中直接进行阻塞网络或长时间 CPU 工作。
Provider 按以下能力族迁移:
1. VAD + ASR。
2. LLM + VLLM + Memory + Intent。
3. TTS。
4. Tools + MCP + IoT + 插件。
本地 Torch/FunASR/Sherpa 等能力作为可选依赖组和镜像 profile,基础 API 镜像不强制加载模型。
## 4. 配置与控制面
- 数据库是完整模式下的配置事实来源。
- Redis 保存有版本号的缓存,并通过 Pub/Sub 广播配置失效和控制事件。
- worker 原子替换共享配置;现有会话可完成当前轮次,新会话使用新版本。
- `server.secret`、SM2 密钥和其他必需系统参数由并发安全的 bootstrap 初始化。
- 进程重启交给容器编排或服务管理器,业务代码不自行 fork、spawn 或 `os._exit()`
- 轻量模式使用文件配置适配器,但进入相同应用服务,不复制 OTA 或会话实现。
## 5. 路由与兼容策略
| 公共入口 | 所有者 | 迁移策略 |
| --- | --- | --- |
| `/` | manager-web | 构建产物由 gateway 托管,保留 PWA scope |
| `/xiaozhi/*` | api | 保留当前调用方契约 |
| `/xiaozhi/ota/` | api | 完整模式只保留数据库驱动实现 |
| `/xiaozhi/v1/` | realtime | 保持设备协议,gateway 支持 Upgrade |
| `/mcp/vision/explain` | realtime | 独立设备鉴权域,不继承管理用户鉴权 |
迁移期可以继续监听 8000/8002/8003 作为兼容别名;最终公网地址由 gateway 统一,OTA 返回值和
系统参数必须在切换前验证。
## 6. 部署档位
### Lite
- 单 worker。
- 文件或数据库配置。
- 可将 API、realtime 和 jobs 放在同一进程用于本地体验。
- 不作为生产容量结论的依据。
### Production
- gateway、API、realtime、jobs 独立进程或容器。
- 共享 MySQL、Redis 和明确的持久化卷。
- realtime 按模型内存和连接数单独扩容。
- 发布时先摘流,等待连接排空,再终止旧实例。
## 7. 架构完成门禁
- 公开契约清单有可执行测试。
- API 与 realtime 不通过环回 HTTP 互调。
- 多 realtime worker 的配置更新能够广播到全部实例。
- realtime 停止时不产生孤儿线程、遗留任务或自行启动的新进程。
- Lite 与 Production 使用相同业务实现。
- 旧 xiaozhi-server 入口只有在新实现通过全部自动门禁后才能删除。
@@ -0,0 +1,82 @@
# 架构与交付决策
本文件记录跨 PR 的稳定决策。普通实现 PR 不得顺手改变已接受决策;需要变更时,先提交新的
决策记录并说明替代关系。
## D001:统一代码库,按职责运行
- 状态:Accepted
- 决策:gateway、API、realtime 和 jobs 使用同一发布版本,但生产环境独立运行。
- 原因:REST、长连接、模型内存和定时任务的扩缩容及故障边界不同。
- 后果:允许 Lite 单 worker `all` 模式;生产容量不得由 Lite 推断。
- 重新评估:有证据证明单进程能够满足模型内存、多 worker 配置一致性和故障隔离。
## D002:只冻结外部行为
- 状态:Accepted
- 决策:设备、调用方、数据和已启用 Provider 的可观察行为保持;内部类、库、线程模型、目录和
自调用方式可以重构。
- 原因:目标是保留现有功能,不保留实现偶然性。
- 后果:manager-web 与后端可以协同调整内部接口,但设备固件契约必须保持。
- 重新评估:产品所有者批准公开行为变更并提供迁移方案。
## D003:特征测试先于 realtime 重构
- 状态:Accepted
- 决策:M1 的黄金报文、虚拟设备和 Fake Provider 是 ASGI 会话重构的前置条件。
- 原因:xiaozhi-server 当前没有正式协议回归套件。
- 后果:不能以“代码更整洁”为由跳过旧行为记录。
- 重新评估:无。
## D004:应用服务替代内部 HTTP 自调用
- 状态:Accepted
- 决策:完整模式下 API 与 realtime 共享应用服务端口,不通过环回 HTTP 调用自己。
- 原因:减少密钥复制、网络重试和同进程/多进程语义差异。
- 后果:HTTP router 只负责协议适配,长连接不持有数据库 Session。
- 重新评估:跨语言或跨安全域部署成为明确需求。
## D005:Redis 承担配置版本事件,不承担事实来源
- 状态:Accepted
- 决策:数据库是完整配置事实来源;Redis 缓存有版本的快照并广播失效/控制事件。
- 原因:支持多 realtime worker 一致更新和 Java 回滚期兼容。
- 后果:配置事件必须可观测,worker 必须暴露当前版本。
- 重新评估:引入独立且受运维支持的消息系统。
## D006Provider 按能力族迁移
- 状态:Accepted
- 决策:按 VAD/ASR、LLM/VLLM/Memory/Intent、TTS、Tools/MCP/IoT 四组迁移。
- 原因:限制 Agent 数量、统一契约并避免每个供应商重复搭建测试框架。
- 后果:只有出现供应商特有阻塞时才拆分单独工作包。
- 重新评估:某 Provider 需要独立进程或不可兼容的系统依赖。
## D007:集成分支承载全部开发 PR
- 状态:Accepted
- 决策:纯 Agent 阶段所有 PR 目标为 `refactor/unified-fastapi-platform`
- 原因:项目所有者要求开发 PR 不进入主分支,并需要统一的长期集成点。
- 后果:PM 负责持续同步上游并在每个里程碑重跑消费者与协议清单。
- 重新评估:仅由项目所有者明确批准。
## D008:真实环境是独立验收门
- 状态:Accepted
- 决策:ESP32、真实 Provider、MQTT/RAGFlow 和目标网络只能在 M8 判定。
- 原因:Mock 和模拟不能证明硬件、供应商及生产网络事实。
- 后果:M7 可以形成 RC,但不能发布正式版本或描述为生产通过。
- 重新评估:所需真实资产已经安全接入自动化环境。
## 新决策模板
```text
ID / 日期 / 状态
背景:
决策:
备选方案:
后果:
验证方式:
重新评估触发条件:
Owner / 关联 PR
```
@@ -0,0 +1,117 @@
# 质量门禁
## 1. 原则
自动测试负责证明“实现符合可观察的现有行为”;真实测试负责证明“设备、供应商和生产网络确实
按预期工作”。两类证据不得互相替代。
## 2. 测试层次
| 层次 | 目标 | 典型内容 | 运行时机 |
| --- | --- | --- | --- |
| 静态 | 快速发现格式、类型、依赖问题 | Ruff、Mypy、前端 lint/type、配置校验 | 每次提交 |
| 单元 | 验证纯逻辑与状态转换 | Token、协议解析、队列、配置版本、Provider DTO | 每次提交 |
| 特征/契约 | 冻结现有可观察行为 | 旧服务黄金报文、API envelope、设备帧、Provider 请求 | 每个 PR |
| 隔离集成 | 验证真实本地依赖 | MySQL、Redis、文件、jobs、Pub/Sub、事务 | 每个相关 PR |
| 模拟 E2E | 验证用户和设备流程 | manager-web、虚拟设备、Fake Provider、统一 gateway | 每个里程碑 |
| 差分 | 比较旧/新实现 | REST、OTA、WS 会话和副作用 | M1 起持续运行 |
| 非功能 | 容量和恢复 | 多连接、背压、断线、摘流、资源泄漏、安全 | M6/M7 |
| 真实环境 | 外部事实 | ESP32、云 Provider、MQTT/RAGFlow、目标部署 | M8 |
## 3. CI 分层
### Tier A:快速门禁
目标是在常规 PR 中快速反馈,建议控制在 10 分钟内:
- FastAPI Ruff、Mypy、unit/contract pytest。
- manager-web i18n、unit、snapshot。
- xiaozhi-server/迁移代码语法与导入检查。
- 路由、消费者和生成文档一致性。
- Secret 扫描和部署配置静态检查。
从 P01 建立覆盖率基线后,变更行覆盖率不得低于 85%,协议解析和会话状态机的分支覆盖率
不得低于 90%,全局覆盖率不得下降。覆盖率用于发现遗漏,不允许为了数字制造无行为价值的测试。
### Tier BPR 集成门禁
- 隔离 MySQL/Redis。
- manager-web 生产构建及浏览器关键路径。
- 虚拟设备完整对话流程。
- Fake Provider 覆盖成功、超时、断流、取消和错误映射。
- API/realtime/jobs 镜像构建与 Compose readiness。
- 修改范围对应的旧/新差分。
### Tier C:里程碑门禁
- 全量旧/新差分。
- 多 realtime worker 配置广播。
- 并发连接和 60 分钟模拟 soak。
- 优雅摘流、进程终止、Redis/MySQL 短暂故障恢复。
- 依赖漏洞、鉴权边界和恶意输入测试。
- 支持架构的镜像构建或至少可重复的构建证明。
Tier C 不在每个小 PR 重复运行,以节约资源;相关核心代码变化或阶段收口时必须运行。
性能采用相同 runner 上的新旧实现相对比较。初始自动门槛为:
- Fake Provider 下业务错误率为 0。
- WebSocket 握手和控制消息 p95 回退不超过 15%。
- 可完成会话吞吐回退不超过 10%。
- 每个 realtime PR 运行 20 个虚拟连接、5 分钟;RC 运行 100 个虚拟连接、60 分钟。
- 测试结束后 RSS、线程、async task、文件描述符和临时文件不得持续线性增长。
这些指标只约束模拟回归,不代表真机容量结论;如固定测试机不足,PR 必须记录实际档位,不能
静默降低门槛。
## 4. 必须建立的协议场景
虚拟设备测试至少覆盖:
1. OTA 获取 WebSocket/MQTT 信息和未绑定激活码。
2. WebSocket header 与 query 参数两种握手。
3. hello、listen、abort、ping、iot、mcp、server 控制消息。
4. Opus 音频输入、ASR 文本、LLM 流、TTS 文本及音频输出。
5. 唤醒、连续对话、主动中断、无语音超时。
6. 设备绑定前后的行为。
7. Provider 超时、断流、无内容、限流和取消。
8. 客户端断线、服务摘流和重连。
9. MQTT gateway 16 字节桥接帧的编码与解码。
10. 配置更新只影响约定的当前或后续会话。
黄金数据必须脱敏、可提交且带版本说明。不能从一次偶然运行直接认定为规范;旧代码、文档和至少
一个调用方必须交叉确认。
## 5. PR 合并门禁
所有 PR
- 计划工作包和文件所有权明确。
- 新行为有测试,重构行为有特征/差分证据。
- 相关 Tier A 全绿。
- 无未解释 skip、xfail、warning 激增或生成文件漂移。
- 文档和配置随行为同步。
- 回滚方法明确。
高风险 PR 额外要求:
- 设备协议:协议 Reviewer + 虚拟设备差分。
- 鉴权/密钥:Security Reviewer + 失败路径和密钥泄露检查。
- 数据迁移:全新库、已有库、重复执行和回滚测试。
- 并发/任务:取消、超时、资源关闭和故障注入。
- Provider 公共接口:所有能力族契约测试通过。
## 6. RC 自动化退出条件
只有满足以下全部条件,才能转入真实测试:
- 所有计划 PR P00-P12 已合并或明确取消并记录理由。
- 公开兼容边界全部关联到自动测试。
- 虚拟设备和 manager-web E2E 全绿。
- 所有 Provider 适配器通过 Fake/契约矩阵。
- 没有 P0/P1 内部缺陷;P2 风险有接受或后续方案。
- 统一发行物在干净环境完成安装、升级和回滚。
- 模拟负载下没有无界内存、线程、任务或文件增长。
- 未验证项逐条映射到真实环境交接用例。
满足这些条件代表“纯 Agent 阶段完成”,不代表真实 Provider 或硬件已经通过。
@@ -0,0 +1,90 @@
# 真实环境测试交接
## 1. 何时交接
只有 [质量门禁](quality-gates.md) 的 RC 自动化退出条件全部满足,才进入本清单。以下情况不属于
真实环境阻塞,必须由开发阶段解决:
- 本地依赖无法安装或镜像无法构建。
- 虚拟设备协议不一致。
- Fake Provider 成功/错误/取消契约失败。
- 数据库、Redis、文件权限或配置广播失败。
- 缺少测试脚本、日志字段或复现步骤。
## 2. 外部资产
| 资产 | 最低要求 | 提供方 |
| --- | --- | --- |
| ESP32 | 至少一台生产使用型号;记录固件版本 | Hardware QA |
| 网络 | 局域网与公网 WSS;可控制弱网/断网 | Ops/QA |
| 模型凭证 | 实际计划启用的 ASR、LLM、TTS,必要时 VLLM | Service owner |
| MQTT/UDP | 受支持版本的 gateway、broker 和端口 | Integration/Ops |
| RAGFlow | 受支持版本、测试数据集和访问凭证 | Knowledge owner |
| 可选服务 | 短信、声纹、语音克隆、MCP、Memory | 对应服务 owner |
| 目标主机 | 计划上线的 CPU/GPU/架构、Docker/Compose 或编排平台 | Ops |
| 域名证书 | HTTPS/WSS 域名、证书和反向代理权限 | Ops |
凭证不得写入 Issue、PR、测试产物或仓库;通过部署环境的 Secret 管理提供。
## 3. 必测场景
### 设备与音频
- 首次启动、OTA、激活和绑定。
- WebSocket 鉴权和长连接保持。
- 唤醒、单轮/连续对话、打断、静音超时和重连。
- 真实麦克风、扬声器、AEC、Opus 帧节奏和中文/非中文语音。
- 弱网、高延迟、短时断网和服务滚动发布。
### Provider
- 每个计划启用 Provider 的成功、流式、空结果、超时、限流和凭证错误。
- ASR 音频格式、TTS 音频参数、LLM 工具调用及取消行为。
- 供应商控制台侧请求量、错误和费用符合预期。
### 外部集成
- MQTT/UDP 上下行音频与控制消息。
- RAGFlow 上传、解析、检索、删除和故障补偿。
- Vision 图片上传、鉴权、大小/格式限制和真实结果。
- 按实际启用范围验证短信、声纹、语音克隆、MCP 和通讯录呼叫。
### 部署
- 全新安装、已有数据升级、滚动发布和回滚。
- 非 root 文件权限、模型卷、上传卷和日志采集。
- API、realtime、jobs 独立健康与告警。
- 目标容量下的连接数、CPU/GPU、内存、带宽和响应时间。
## 4. 证据格式
每条真实用例记录:
```text
Case ID
日期/测试人:
硬件与固件:
服务版本/Commit
外部服务及版本:
网络与部署环境:
步骤:
预期:
实际:
日志/指标/录屏位置:
结果:Pass / Fail / Blocked
缺陷链接:
```
敏感字段在上传前脱敏。失败必须能关联到服务器会话 ID、设备 ID 的脱敏标识和时间窗口。
## 5. 上线判定
生产批准至少要求:
- 选定 ESP32/固件矩阵通过。
- 实际启用的 Provider 和外部集成通过,不要求未启用供应商全部真实联调。
- 目标环境升级与回滚演练通过。
- 没有 P0/P1 缺陷。
- 监控、告警、值守和回滚负责人明确。
未提供的外部能力应标记为“未验收/不可启用”,不得以自动化 Mock 结果改为 Pass。
@@ -0,0 +1,27 @@
# 风险登记册
| ID | 风险 | 概率/影响 | 缓解措施 | 触发升级条件 | Owner |
| --- | --- | --- | --- | --- | --- |
| R01 | xiaozhi-server 缺少正式协议测试,重构造成隐性漂移 | 高/高 | M1 先建立黄金报文、虚拟设备和旧/新差分 | 无法从旧代码和调用方确定行为 | QA + Realtime |
| R02 | REST API worker 与实时模型共享进程导致内存和延迟失控 | 高/高 | 同代码库、独立运行角色;Lite 仅单 worker | 生产方案要求单进程多 worker | Architecture |
| R03 | `websockets`、PyYAML 及重模型依赖冲突 | 高/高 | 统一锁文件、Provider extras、镜像 profile | 必须删除现有 Provider 才能解析 | Platform + Provider |
| R04 | `server.secret`/SM2 首次初始化或轮换不一致 | 高/高 | 并发安全 bootstrap、单一事实来源、广播测试 | 任一 worker 使用不同密钥 | Backend + Security |
| R05 | 配置热更新只到达部分 realtime worker | 中/高 | Redis 版本事件、ack/指标、全 worker 集成测试 | 版本长时间不一致 | Backend + SRE |
| R06 | 线程、任务或 SDK 阻塞事件循环 | 高/高 | DeviceSession 所有权、有界执行器、soak/lag 指标 | 负载下 API/WS 延迟无界增长 | Realtime + QA |
| R07 | manager-web Service Worker 缓存旧 API/资源 | 中/中 | 保留 scope、版本化资源、升级 E2E | 新旧前端混用造成不可恢复错误 | Frontend |
| R08 | OTA 双实现或公网 URL/端口切换破坏旧固件 | 中/高 | 完整/轻量 profile、兼容别名、OTA 差分 | 需要升级固件才能连接 | Architecture + QA |
| R09 | 本地模型文件、FFmpeg、libopus 和写目录未正确打包 | 高/中 | 显式卷、非 root 权限、profile 构建测试 | 干净容器无法启动选定能力 | Platform |
| R10 | Mock 掩盖真实供应商流式和错误行为 | 高/高 | Fake 只证明契约;M8 真实矩阵单独验收 | 需要凭证才能决定公共设计 | QA + External QA |
| R11 | MQTT/UDP gateway、RAGFlow 等外部项目版本漂移 | 中/高 | 固定支持版本、录制契约、真实交接清单 | 外部接口文档与实际不一致 | Integration |
| R12 | 上游 `main` 持续变化导致长分支难以合并 | 高/中 | 小 PR、每次合并后变基、每阶段重跑调用方清单 | 冲突改变公共契约 | PM/集成 |
| R13 | PR 并行过多造成重复实现和冲突 | 中/中 | 最多三个实现工作包、文件所有权、唯一负责人 | 两个 PR 修改同一协议核心 | PM/集成 |
| R14 | 现有未跟踪 FastAPI 基线没有进入远端 | 高/高 | P00 优先提交并建立 PR | 后续 Agent 无稳定基线 | PM/集成 |
## 风险处理规则
- P0:立即停止受影响工作包,PM 和架构负责人决策。
- P1:不得合并相关 PR,必须有修复或明确的外部阻塞证据。
- P2:可以带风险进入后续阶段,但必须有 Owner、验证计划和截止阶段。
- 外部阻塞:只有满足 [真实环境测试交接](real-environment-handoff.md) 的定义才能标记,不能把内部未完成项转嫁给真实测试。
每个里程碑收口时重新评估概率、影响和 Owner;关闭风险必须附测试、PR 或决策记录链接。
+90
View File
@@ -0,0 +1,90 @@
# 阶段路线图与 PR 队列
## 1. 里程碑
| 阶段 | 目标 | 主要交付物 | 负责人角色 | 依赖 | 退出门禁 |
| --- | --- | --- | --- | --- | --- |
| M0 基线 | 建立可审阅的 FastAPI 与计划基线 | 当前实现、兼容报告、计划文档、首个 PR | PM/集成 | 无 | 基线测试可重复,变更范围清楚 |
| M1 可执行契约 | 把“现有功能”转换成测试资产 | 协议清单、黄金报文、虚拟设备、Fake Provider | QA + Realtime | M0 | 旧服务完整模拟会话可重复 |
| M2 统一发行入口 | 纳入 manager-web 与统一网关 | 前端构建、静态托管、路由、统一 Compose | Platform + Frontend | M0 | 页面/API/WS 路由冒烟通过 |
| M3 共享核心 | 消除配置和业务能力的内部 HTTP 自调用 | bootstrap、应用服务接口、配置版本与广播 | Backend + Architecture | M1 | API/旧 realtime 均可使用共享接口 |
| M4 ASGI 实时内核 | 建立不依赖真实 Provider 的新会话运行时 | WebSocket 适配、DeviceSession、背压、取消和鉴权 | Realtime | M1、M3 | Fake 全会话与旧协议一致 |
| M5 Provider 迁移 | 接回全部现有能力 | 四个 Provider 能力族适配、契约测试 | Provider 专项 | M4 | 所有 Provider 可构造且契约通过 |
| M6 平台硬化 | 达到可部署候选质量 | 多进程、优雅摘流、指标、安全、负载和故障测试 | Platform + QA | M2、M4、M5 | 自动质量门禁全部绿色 |
| M7 切换准备 | 删除自动化范围内的未知项 | 新旧差分、升级/回滚演练、旧入口弃用方案 | PM + Reviewer | M6 | 只剩真实环境清单中的阻塞项 |
| M8 真实验收 | 外部人员/环境介入 | ESP32、真实 Provider、MQTT/RAGFlow、目标环境报告 | External QA/Ops | M7 | 不属于纯 Agent 阶段 |
## 2. 推荐 PR 队列
PR 必须按可独立验证的纵向能力拆分。编号是计划标识,不是 GitHub 实际编号。
| 计划 PR | 内容 | 建议目标分支 | 可并行关系 |
| --- | --- | --- | --- |
| P00 | 开发计划、Agent 协作和质量门禁 | `refactor/unified-fastapi-platform` | FastAPI 基线直接推送后创建 |
| P01 | CI 快速门禁和测试目录重组 | `refactor/unified-fastapi-platform` | 与 P02 并行 |
| P02 | manager-web 构建、gateway 路由和统一 Compose | 同上 | 与 P01 并行 |
| P03 | 设备协议清单、黄金报文和虚拟设备 | 同上 | 与 P02 并行 |
| P04 | 系统密钥 bootstrap、共享配置接口和版本事件 | P01 + P03 | 阻塞 P06 |
| P05 | 聊天/设备配置等内部 HTTP 调用改为应用服务端口 | P04 | 可按用例拆成两个 PR |
| P06 | ASGI WebSocket 握手、鉴权和协议适配层 | P03 + P04 | 与 P05 后半段并行 |
| P07 | DeviceSession 生命周期、队列、取消和 Fake 对话链路 | P06 | 阻塞 Provider 迁移 |
| P08-A | VAD/ASR Provider 适配 | P07 | 与 P08-B/C 并行 |
| P08-B | LLM/VLLM/Memory/Intent 适配 | P07 | 与 P08-A/C 并行 |
| P08-C | TTS Provider 适配 | P07 | 与 P08-A/B 并行 |
| P09 | Tools/MCP/IoT/插件及 Vision | P08-B | 可与 P08 收尾并行 |
| P10 | 配置广播、管理控制、优雅摘流与健康指标 | P07 + P08 | 与 P09 并行 |
| P11 | 全栈模拟 E2E、新旧差分和负载/故障测试 | P02 + P09 + P10 | 集成收口 |
| P12 | 发行、升级、回滚、弃用和真实测试交接 | P11 | 纯 Agent 最后一个 PR |
所有开发 PR 默认以长期集成分支 `refactor/unified-fastapi-platform` 为目标。存在未合并依赖时使用
堆叠 PR;依赖合并后及时变基到最新集成分支。纯 Agent 阶段不向 `main` 创建 PR;真实环境验收
完成后,也只有在项目所有者明确批准时才讨论主分支集成。不得让多个 PR 同时修改同一协议核心文件。
## 3. 建议容量分配
以下比例用于分配 Agent 和审阅资源,不是完成度承诺:
| 工作域 | 参考占比 | 原因 |
| --- | ---: | --- |
| manager-web 与 gateway | 15% | 构建已可用,主要缺静态交付和浏览器 E2E |
| manager-api 深度兼容与共享服务 | 25% | 路由齐全,但成功写入、副作用和 bootstrap 仍需补强 |
| realtime、会话与 Provider | 45% | 缺少协议测试,且存在线程、模型和长连接重构 |
| CI、部署、安全与交接 | 15% | 需要从零建立 PR 门禁和统一发行证据 |
容量应随风险登记和测试证据调整,不按代码行数机械分配。
## 4. 并行执行窗口
为控制资源和冲突,同一时间最多开放三个实现工作包:
1. 一个核心依赖链工作包,例如 P04/P06/P07。
2. 一个交付或前端工作包,例如 P02/P10。
3. 一个测试/独立审阅工作包,例如 P03/P11。
Provider 阶段允许三个能力族并行,但每个能力族只分配一个实现 Agent;共享接口由架构负责人
预先冻结,任何接口变更先更新 ADR 和契约测试。
## 5. 阶段状态规则
每个阶段只有四种状态:
- `Not started`:依赖未完成。
- `Ready`:依赖完成且工作包说明已批准。
- `In progress`:已有唯一负责人和活动 PR。
- `Done`PR 合并且退出门禁有证据链接。
不得使用“代码写完”代替 `Done`。失败、跳过、未执行和缺少外部环境必须分别记录。
## 6. 每周/每轮进度摘要
```text
当前阶段:
已合并 PR
活动 PR(负责人 / 门禁):
本轮新增证据:
阻塞项(内部 / 真实环境):
风险变化:
下一轮最多三个工作包:
```
路线图由 PM/集成负责人维护;实现 Agent 只更新自己 PR 的证据和工作包状态。
+3 -1
View File
@@ -28,7 +28,7 @@ telnet 127.0.0.1 3306
如果不能访问,你需要回忆一下,你的`mysql`是怎么安装的。 如果不能访问,你需要回忆一下,你的`mysql`是怎么安装的。
如果你的mysql是通过自己使用安装包安装的,说明你的`mysql`做了网络隔离。你可能先解访问`mysql``3306`端口这个问题。 如果你的mysql是通过自己使用安装包安装的,说明你的`mysql`做了网络隔离。你可能先解访问`mysql``3306`端口这个问题。
如果你`mysql`是通过本项目的`docker-compose_all.yml`安装的。你需要找一下你当时创建数据库的`docker-compose_all.yml`文件,修改以下的内容 如果你`mysql`是通过本项目的`docker-compose_all.yml`安装的。你需要找一下你当时创建数据库的`docker-compose_all.yml`文件,修改以下的内容
@@ -164,6 +164,8 @@ http://192.168.1.25:8005/voiceprint/health?key=abcd
# 2、全模块部署时,怎么配置声纹识别 # 2、全模块部署时,怎么配置声纹识别
## 第一步 配置接口 ## 第一步 配置接口
首先,你要开启声纹识别功能。在智控台,点击顶部`参数字典`,在下拉菜单中,点击`系统功能配置`页面。在页面上勾选`声纹识别`,点击`保存配置`。即可在新建智能体的卡片上看到`声纹识别`按钮。
如果你是全模块部署,使用管理员账号,登录智控台,点击顶部`参数字典`,选择`参数管理`功能。 如果你是全模块部署,使用管理员账号,登录智控台,点击顶部`参数字典`,选择`参数管理`功能。
然后搜索参数`server.voice_print`,此时,它的值应该是`null`值。 然后搜索参数`server.voice_print`,此时,它的值应该是`null`值。
+71
View File
@@ -0,0 +1,71 @@
# 联网搜索插件使用指南
## 功能简介
联网搜索插件 `web_search` 支持在对话过程中实时联网搜索信息并返回结果。插件支持两个搜索源:秘塔(Metaso)和Tavily,用户可根据需要选择其中一个。
## API Key申请指南
目前我们适配了`秘塔搜索``Tavily搜索`
- Tavily搜索:每个月1000次免费额度。
- 秘塔搜索:拥有较为优质的国内数据源。
## API Key申请指南
### 方式一:使用秘塔搜索
- 访问 [秘塔搜索API](https://metaso.cn/search-api/api-keys),注册并登录账号
- 在API密钥管理页面,点击"创建新的Key"
- 复制生成的API Key(以 `mk-` 为前缀),这是配置所需的关键信息
### 方式二:使用Tavily搜索
- 访问 [Tavily控制台](https://app.tavily.com/home),注册并登录账号
- 在控制台中创建API Key
- 复制生成的API Key(以 `tvly-` 为前缀),这是配置所需的关键信息
## 配置方式
### 方式1. 使用智控台部署(推荐)
- 登录智控台
- 进入"配置角色"页面,选择要配置的智能体
- 点击"编辑功能"按钮,在右侧参数配置区域找到"联网搜索"插件
- 勾选"联网搜索"
- 填入搜索源(`metaso``tavily`),并将对应的`API Key`填入配置项
- 保存配置,再保存智能体配置
### 方式2. 单模块xiaozhi-server部署
`data/.config.yaml` 中配置:
- 将搜索源填入 `provider`,可选值为 `metaso``tavily`
- 将申请到的API Key填入 `api_key`
```yaml
plugins:
web_search:
provider: "metaso"
api_key: "你的API Key"
```
如需自定义返回结果数量和工具描述,可额外配置 `max_results``description`
```yaml
plugins:
web_search:
provider: "metaso"
description: "联网搜索工具。当用户明确需要联网搜索问题时使用此工具。"
max_results: 5
api_key: "你的API Key"
```
同时在 `functions` 列表中确保已启用 `web_search`
```yaml
plugins:
functions:
- web_search
```
配置完成后重启服务即可生效。
+52 -1
View File
@@ -9,6 +9,7 @@
* [3.2. `manager-api` (管理后端 - Java Spring Boot实现)](#32-manager-api-管理后端---java-spring-boot实现) * [3.2. `manager-api` (管理后端 - Java Spring Boot实现)](#32-manager-api-管理后端---java-spring-boot实现)
* [3.3. `manager-web` (Web管理前端 - Vue.js实现)](#33-manager-web-web管理前端---vuejs实现) * [3.3. `manager-web` (Web管理前端 - Vue.js实现)](#33-manager-web-web管理前端---vuejs实现)
* [3.4. `manager-mobile` (移动管理端 - uni-app+Vue3实现)](#34-manager-mobile-移动管理端---uni-appvue3实现) * [3.4. `manager-mobile` (移动管理端 - uni-app+Vue3实现)](#34-manager-mobile-移动管理端---uni-appvue3实现)
* [3.5. `digital-human` (数字人测试模块 - Python+Web实现)](#35-digital-human-数字人测试模块---pythonweb实现)
4. [数据流与交互机制](#4-数据流与交互机制) 4. [数据流与交互机制](#4-数据流与交互机制)
5. [核心功能概要](#5-核心功能概要) 5. [核心功能概要](#5-核心功能概要)
6. [部署与配置概述](#6-部署与配置概述) 6. [部署与配置概述](#6-部署与配置概述)
@@ -67,6 +68,13 @@
* 基于alova + @alova/adapter-uniapp实现网络请求,与manager-api无缝集成。 * 基于alova + @alova/adapter-uniapp实现网络请求,与manager-api无缝集成。
* 使用pinia进行状态管理,确保数据一致性。 * 使用pinia进行状态管理,确保数据一致性。
6. **`digital-human` (数字人测试模块 - Python+Web实现):**
这是一个独立的数字人测试模块,提供本地测试页面、前端交互资源、唤醒词运行时与事件桥能力,用于联调整个数字人交互链路。其主要能力包括:
* 提供浏览器端的数字人测试页面,用于验证音频播放、接收和交互流程。
* 集成本地唤醒词运行时,支持基于Sherpa-ONNX的关键词检测。
* 通过事件桥将页面状态与本地运行时打通,便于调试唤醒词触发与交互效果。
* 作为独立模块与`xiaozhi-server``manager-web``manager-api`并列,便于单独开发和部署。
**高层交互流程概述:** **高层交互流程概述:**
* **语音交互主线:** **ESP32设备**捕捉到用户语音后,通过**WebSocket**将音频数据实时传输给**`xiaozhi-server`**。`xiaozhi-server`完成一系列AI处理(VAD、ASR、LLM交互、TTS)后,再通过WebSocket将合成的语音回复发送回ESP32设备进行播放。所有与语音直接相关的实时交互均在此链路完成。 * **语音交互主线:** **ESP32设备**捕捉到用户语音后,通过**WebSocket**将音频数据实时传输给**`xiaozhi-server`**。`xiaozhi-server`完成一系列AI处理(VAD、ASR、LLM交互、TTS)后,再通过WebSocket将合成的语音回复发送回ESP32设备进行播放。所有与语音直接相关的实时交互均在此链路完成。
@@ -80,7 +88,8 @@ xiaozhi-esp32-server
├─ xiaozhi-server 8000 端口 Python语言开发 负责与esp32通信 ├─ xiaozhi-server 8000 端口 Python语言开发 负责与esp32通信
├─ manager-web 8001 端口 Node.js+Vue开发 负责提供控制台的web界面 ├─ manager-web 8001 端口 Node.js+Vue开发 负责提供控制台的web界面
├─ manager-api 8002 端口 Java语言开发 负责提供控制台的api ├─ manager-api 8002 端口 Java语言开发 负责提供控制台的api
─ manager-mobile 跨平台移动应用 uni-app+Vue3开发 负责提供移动版智控台管理 ─ manager-mobile 跨平台移动应用 uni-app+Vue3开发 负责提供移动版智控台管理
└─ digital-human 数字人测试模块 Python+Web实现 负责本地测试页面、唤醒词运行时与事件桥
``` ```
--- ---
@@ -366,6 +375,48 @@ xiaozhi-esp32-server
--- ---
### 3.5. `digital-human` (数字人测试模块 - Python+Web实现)
`digital-human` 组件是一个独立的数字人测试模块,负责提供本地测试页面、页面资源、唤醒词运行时和事件桥能力。它主要用于联调整个数字人交互链路,帮助开发者验证页面交互、音频能力和本地唤醒词流程。
* **核心目标:**
* 提供独立的本地测试页面,用于验证数字人交互效果。
* 提供本地唤醒词运行时,支持关键词检测与事件上报。
* 为前端页面和本地运行时提供事件桥,打通测试链路。
* 以独立模块形式存在,便于单独开发、调试和部署。
* **核心技术栈:**
* **Python 3:** 作为测试运行时的主要语言,负责启动本地服务、管理唤醒词运行时和事件桥。
* **原生HTML/CSS/JavaScript:** 用于构建数字人测试页面与交互逻辑。
* **Sherpa-ONNX:** 用于本地唤醒词检测。
* **ThreadingHTTPServer / WebSocket桥接:** 用于承载测试页面、健康检查和本地事件通信。
* **关键实现细节:**
1. **模块入口与页面资源:**
* `start.py` 是模块启动入口,负责初始化测试运行时。
* `index.html` 是数字人测试页面入口。
* `js/``css/``images/``resources/` 目录提供页面脚本、样式和资源文件。
2. **本地测试运行时:**
* `wakeword_runtime/` 目录承载本地唤醒词运行时实现。
* 运行时负责配置加载、日志初始化、音频采集、关键词检测和服务生命周期管理。
* 模块默认通过本地HTTP服务暴露页面地址、事件桥地址和健康检查接口。
3. **唤醒词检测链路:**
* 基于Sherpa-ONNX关键词检测模型,支持本地唤醒词触发。
* 模型文件和关键词配置位于 `wakeword_runtime/models/` 及相应配置文件中。
* 页面侧可结合事件桥观察唤醒词服务状态和触发结果。
4. **调试与联调定位:**
* 启动后可通过浏览器访问本地页面,验证音频播放、接收和交互逻辑。
* 通过事件桥可以将唤醒词状态同步到页面,便于排查本地链路问题。
* 详细的模型下载、运行时配置和使用说明已整理在 `docs/digital-human-wakeword.md` 中。
`digital-human` 的存在使得数字人相关能力可以脱离主服务独立验证,降低了页面调试、唤醒词联调和本地环境搭建的复杂度。
---
## 4. 数据流与交互机制 ## 4. 数据流与交互机制
`xiaozhi-esp32-server` 系统通过各组件间定义清晰的数据流和交互协议来协同工作。主要的通信方式依赖于针对实时交互优化的WebSocket协议和适用于客户端-服务器请求的RESTful API。 `xiaozhi-esp32-server` 系统通过各组件间定义清晰的数据流和交互协议来协同工作。主要的通信方式依赖于针对实时交互优化的WebSocket协议和适用于客户端-服务器请求的RESTful API。
+58 -1
View File
@@ -9,6 +9,7 @@
* [3.2. `manager-api` (Management Backend - Java Spring Boot Implementation)](#32-manager-api-management-backend---java-spring-boot-implementation) * [3.2. `manager-api` (Management Backend - Java Spring Boot Implementation)](#32-manager-api-management-backend---java-spring-boot-implementation)
* [3.3. `manager-web` (Web Management Frontend - Vue.js Implementation)](#33-manager-web-web-management-frontend---vuejs-implementation) * [3.3. `manager-web` (Web Management Frontend - Vue.js Implementation)](#33-manager-web-web-management-frontend---vuejs-implementation)
* [3.4. `manager-mobile` (Mobile Management Console - uni-app Implementation)](#34-manager-mobile-mobile-management-console---uniapp-implementation) * [3.4. `manager-mobile` (Mobile Management Console - uni-app Implementation)](#34-manager-mobile-mobile-management-console---uniapp-implementation)
* [3.5. `digital-human` (Digital Human Test Module - Python+Web Implementation)](#35-digital-human-digital-human-test-module---pythonweb-implementation)
4. [Data Flow and Interaction Mechanisms](#4-data-flow-and-interaction-mechanisms) 4. [Data Flow and Interaction Mechanisms](#4-data-flow-and-interaction-mechanisms)
5. [Key Features Summary](#5-key-features-summary) 5. [Key Features Summary](#5-key-features-summary)
6. [Deployment and Configuration Overview](#6-deployment-and-configuration-overview) 6. [Deployment and Configuration Overview](#6-deployment-and-configuration-overview)
@@ -60,6 +61,21 @@ The `xiaozhi-esp32-server` system adopts a **distributed, multi-component collab
* (Potential functionality) Monitoring system operation status, viewing logs, troubleshooting, etc. * (Potential functionality) Monitoring system operation status, viewing logs, troubleshooting, etc.
* Comprehensive interaction with all backend management functions provided by `manager-api`. * Comprehensive interaction with all backend management functions provided by `manager-api`.
5. **`manager-mobile` (Mobile Management Console - uni-app Implementation):**
This is a cross-platform mobile management application based on uni-app v3 + Vue 3 + Vite, supporting App (Android & iOS) and WeChat Mini Program. Its main capabilities include:
* Providing a convenient management interface on mobile devices, similar in functionality to manager-web but optimized for mobile platforms.
* Supporting core functions such as user login, device management, and AI service configuration.
* Cross-platform adaptation, allowing a single codebase to run on iOS, Android, and WeChat Mini Programs.
* Integrating with `manager-api` through alova + `@alova/adapter-uniapp`.
* Using pinia for state management to ensure data consistency.
6. **`digital-human` (Digital Human Test Module - Python+Web Implementation):**
This is an independent digital human test module that provides a local test page, frontend interaction resources, a wake word runtime, and an event bridge for end-to-end debugging of the digital human interaction flow. Its main capabilities include:
* Providing a browser-based digital human test page for validating audio playback, audio reception, and interaction flow.
* Integrating a local wake word runtime with keyword detection based on Sherpa-ONNX.
* Bridging page state and the local runtime through an event bridge to simplify wake word debugging.
* Existing as a standalone module alongside `xiaozhi-server`, `manager-web`, and `manager-api`, making independent development and deployment easier.
**High-Level Interaction Flow Overview:** **High-Level Interaction Flow Overview:**
* **Voice Interaction Main Line:** After the **ESP32 device** captures user voice, it transmits audio data in real-time to **`xiaozhi-server`** through **WebSocket**. After `xiaozhi-server` completes a series of AI processing (VAD, ASR, LLM interaction, TTS), it sends the synthesized voice response back to the ESP32 device for playback through WebSocket. All real-time interactions directly related to voice are completed in this link. * **Voice Interaction Main Line:** After the **ESP32 device** captures user voice, it transmits audio data in real-time to **`xiaozhi-server`** through **WebSocket**. After `xiaozhi-server` completes a series of AI processing (VAD, ASR, LLM interaction, TTS), it sends the synthesized voice response back to the ESP32 device for playback through WebSocket. All real-time interactions directly related to voice are completed in this link.
@@ -73,7 +89,8 @@ xiaozhi-esp32-server
├─ xiaozhi-server Port 8000 Python development Responsible for ESP32 communication ├─ xiaozhi-server Port 8000 Python development Responsible for ESP32 communication
├─ manager-web Port 8001 Node.js+Vue development Responsible for providing web interface for console ├─ manager-web Port 8001 Node.js+Vue development Responsible for providing web interface for console
├─ manager-api Port 8002 Java development Responsible for providing console API ├─ manager-api Port 8002 Java development Responsible for providing console API
─ manager-mobile Cross-platform mobile application uni-app+Vue3 development Responsible for providing mobile console management ─ manager-mobile Cross-platform mobile application uni-app+Vue3 development Responsible for providing mobile console management
└─ digital-human Digital human test module Python+Web implementation Responsible for local test page, wake word runtime, and event bridge
``` ```
--- ---
@@ -257,6 +274,46 @@ The `manager-mobile` component is a cross-platform mobile management application
`manager-mobile` provides users with a fully functional, smooth mobile management tool through the application of these technologies, allowing administrators to perform system management and configuration anytime, anywhere. `manager-mobile` provides users with a fully functional, smooth mobile management tool through the application of these technologies, allowing administrators to perform system management and configuration anytime, anywhere.
### 3.5. `digital-human` (Digital Human Test Module - Python+Web Implementation)
The `digital-human` component is an independent digital human test module responsible for providing a local test page, page resources, a wake word runtime, and an event bridge. It is primarily used to debug the complete digital human interaction flow and to validate frontend interaction, audio capability, and local wake word behavior.
* **Core Objectives:**
* Provide an independent local test page for validating digital human interaction effects.
* Provide a local wake word runtime for keyword detection and event reporting.
* Provide an event bridge between the frontend page and the local runtime to connect the test flow.
* Exist as a standalone module for easier independent development, debugging, and deployment.
* **Core Technologies:**
* **Python 3:** Used for the local test runtime, service startup, wake word lifecycle management, and event bridge coordination.
* **Native HTML/CSS/JavaScript:** Used to build the digital human test page and interaction logic.
* **Sherpa-ONNX:** Used for local wake word keyword detection.
* **ThreadingHTTPServer / WebSocket bridge:** Used to serve the local page, health check endpoint, and event communication.
* **Key Implementation Aspects:**
1. **Module Entry and Page Resources:**
* `start.py` is the module entry point responsible for starting the local test runtime.
* `index.html` is the entry page for digital human testing.
* The `js/`, `css/`, `images/`, and `resources/` directories provide frontend scripts, styles, and static resources.
2. **Local Test Runtime:**
* The `wakeword_runtime/` directory contains the local wake word runtime implementation.
* The runtime is responsible for configuration loading, logging initialization, audio capture, keyword detection, and lifecycle management.
* The module exposes a local page URL, event bridge URL, and health check endpoint through a local HTTP service.
3. **Wake Word Detection Flow:**
* Uses Sherpa-ONNX keyword detection models to support local wake word triggering.
* Model files and keyword configurations are stored in `wakeword_runtime/models/` and related configuration files.
* The page can observe runtime state and trigger results through the event bridge.
4. **Debugging and Integration Positioning:**
* After startup, developers can open the local page in a browser to validate playback, reception, and interaction logic.
* The event bridge synchronizes wake word state to the page, which helps with local debugging.
* Detailed model download, runtime configuration, and usage instructions are documented in `docs/digital-human-wakeword.md`.
The `digital-human` module allows digital human related capabilities to be validated independently from the main service, reducing the complexity of page debugging, wake word integration, and local environment setup.
--- ---
## 4. Data Flow and Interaction Mechanisms ## 4. Data Flow and Interaction Mechanisms
+46
View File
@@ -0,0 +1,46 @@
本文档是开发类文档,如需部署小智服务端,[点击这里查看部署教程](../../README.md#%E9%83%A8%E7%BD%B2%E6%96%87%E6%A1%A3)
如需查看一体机数字人落地部署、Kiosk 全屏启动和系统环境配置,[点击这里查看一体机部署指南](../../docs/all-in-one-digital-human-setup.md)
如需查看唤醒词模型下载、运行时配置和详细使用说明,[点击这里查看唤醒词专题文档](../../docs/digital-human-wakeword.md)
# 项目介绍
digital-human 是独立的数字人测试模块,负责提供本地测试页面、前端交互资源、唤醒词运行时和事件桥能力,用于联调整个数字人交互链路。
# 快速启动
安装依赖:
```bash
pip install -r wakeword_runtime/requirements.txt
```
启动模块:
```bash
python start.py
```
# 访问地址
启动后可访问:
- 页面地址:http://127.0.0.1:8006/index.html
- 事件桥地址:ws://127.0.0.1:8006/wakeword-ws
- 健康检查:http://127.0.0.1:8006/health
# 目录说明
- `start.py`:模块启动入口
- `index.html`:数字人测试页面入口
- `wakeword_runtime`:本地唤醒词运行时与配置目录
- `js``css`:页面前端脚本与样式
- `images``resources`:页面资源文件
# 相关文档
- 一体机部署指南:适用于 x86 设备整机落地部署、Kiosk 展示和开机自启动配置
[../../docs/all-in-one-digital-human-setup.md](../../docs/all-in-one-digital-human-setup.md)
- 唤醒词专题文档:适用于唤醒词模型下载、运行时配置和本地调试说明
[../../docs/digital-human-wakeword.md](../../docs/digital-human-wakeword.md)
Binary file not shown.

After

Width:  |  Height:  |  Size: 185 B

File diff suppressed because it is too large Load Diff
Binary file not shown.

After

Width:  |  Height:  |  Size: 4.2 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 136 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 321 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 290 KiB

+416
View File
@@ -0,0 +1,416 @@
<!DOCTYPE html>
<html lang="zh-CN">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>小智数字人页面</title>
<link rel="stylesheet" href="css/index.css?v=0206">
<script>
// 检测是否使用file://协议打开
if (window.location.protocol === 'file:') {
document.addEventListener('DOMContentLoaded', function () {
// 创建背景模糊遮罩
const overlayDiv = document.createElement('div');
overlayDiv.className = 'file-protocol-overlay';
document.body.appendChild(overlayDiv);
// 创建警告框
const warningDiv = document.createElement('div');
warningDiv.id = 'fileProtocolWarning';
warningDiv.innerHTML = `
<h2>⚠️ 警告:请使用HTTP服务器打开此页面</h2>
<p>您当前使用的是本地文件方式打开页面(file://协议),这可能导致页面功能异常。</p>
<p>您可以使用nginx映射启动小智数字人页面,也可以请按照以下步骤启动测试运行时:</p>
<ol>
<li>打开命令行终端</li>
<li>命令行进入到 digital-human 目录</li>
<li>执行以下命令启动小智数字人页面运行时:</li>
</ol>
<pre>python start.py</pre>
<p>然后在浏览器中访问:<strong>http://localhost:8006/index.html</strong></p>
`;
document.body.appendChild(warningDiv);
});
}
</script>
</head>
<body>
<!-- 背景容器 -->
<div class="background-container" id="backgroundContainer">
<div class="background-overlay"></div>
</div>
<!-- 主容器 -->
<div class="container">
<!-- 摄像头显示区域(可拖动) -->
<div class="camera-container" id="cameraContainer">
<video id="cameraVideo" autoplay playsinline muted></video>
<div class="camera-switch-mask" id="cameraSwitchMask">
<div class="camera-switch-container">
<!-- 摄像头切换按钮 -->
<div class="camera-switch" id="cameraSwitch">
<svg class="btn-icon" viewBox="0 0 1024 1024" version="1.1" xmlns="http://www.w3.org/2000/svg" p-id="7319" width="32" height="32"><path d="M722.1 530.3c0-44.9-13.8-86.5-37.2-120.9-0.3-0.6-0.6-1.2-0.9-1.7-2.9-4.2-6.1-8.1-9.2-12-0.4-0.5-0.7-1-1.1-1.5-21.4-26.3-48.3-46.7-78.7-59.9-0.8-0.4-1.7-0.8-2.5-1.2-4.9-2-9.8-3.8-14.9-5.5-1.8-0.6-3.6-1.3-5.4-1.8-4.4-1.3-8.8-2.4-13.3-3.5-2.5-0.6-4.9-1.2-7.4-1.7-1.2-0.2-2.4-0.6-3.6-0.8-3.3-0.6-6.7-0.8-10-1.3-2.3-0.3-4.6-0.7-6.9-0.9-5.6-0.5-11.2-0.8-16.7-0.9-1 0-2-0.1-3-0.1-0.2 0-0.3 0.1-0.5 0.1-42.8 0-84.6 13.1-120.4 38.6-11.5 8.1-14.2 24.1-6.2 35.7 8 11.6 23.8 14.4 35.2 6.3 27.7-19.7 60-29.6 93.1-29.3 4.8 0 9.5 0.3 14.1 0.7 1.4 0.2 2.8 0.4 4.3 0.5 3.8 0.5 7.6 1 11.3 1.8 1.6 0.3 3.3 0.8 4.9 1.1 3.7 0.8 7.3 1.7 10.9 2.8 1.1 0.4 2.2 0.8 3.4 1.2 4.1 1.3 8.1 2.8 12 4.5 0.4 0.2 0.8 0.4 1.2 0.5 23.6 10.3 44.3 26.1 60.4 45.9 0.1 0.1 0.2 0.3 0.3 0.4 22.7 28 36.4 63.9 36.4 102.9h-42.2L697 632.9l67.5-102.6h-42.4z m-119 133c-27.9 19.8-60.5 29.7-93.9 29.3-4.4-0.1-8.8-0.3-13.1-0.7-1.8-0.2-3.5-0.5-5.3-0.7-3.4-0.5-6.8-0.9-10.1-1.6-2-0.4-4-0.9-6.1-1.4-3.3-0.8-6.5-1.5-9.7-2.5-1.5-0.5-3-1-4.6-1.5-3.7-1.2-7.3-2.5-10.8-4-0.8-0.3-1.6-0.7-2.4-1.1-4.1-1.9-8.2-3.8-12.2-6l-0.6-0.3c-13.4-7.5-25.7-16.8-36.4-27.7-0.2-0.2-0.3-0.4-0.5-0.6-3.3-3.4-6.5-6.9-9.5-10.7-0.6-0.8-1.3-1.6-1.9-2.5-21.9-27.8-35.1-63-35.1-101.2h42.2l-67.5-102.6-67.5 102.6h42.2c0 45 13.9 86.7 37.3 121.1 0.3 0.5 0.5 1 0.8 1.5 2.4 3.5 5.1 6.8 7.7 10.1 1 1.3 1.9 2.6 2.9 3.8 3.9 4.7 7.9 9.2 12.1 13.5 0.4 0.4 0.8 0.9 1.2 1.3 14.1 14.3 30.1 26.4 47.4 36 0.5 0.3 0.9 0.6 1.4 0.8 5 2.7 10.1 5.2 15.3 7.5l3.9 1.8c4.4 1.9 9 3.5 13.6 5.1 2.2 0.7 4.4 1.5 6.6 2.2 4 1.2 8.1 2.2 12.3 3.2 2.8 0.7 5.5 1.4 8.3 1.9 1.2 0.3 2.3 0.6 3.4 0.8 3.9 0.7 7.8 1.1 11.8 1.6l4.2 0.6c7 0.7 14.1 1.2 21.1 1.2 42.9 0 84.7-13.3 120.6-38.8 11.4-8.2 14.2-24.1 6.2-35.8-8-11.5-23.8-14.3-35.3-6.2z m271.3-421.6H763c-26.5 0-82.9-119.7-112.5-119.7H370.6c-29.6 0-85.9 119.7-112.4 119.7H147.5c-46.3 0-83.9 40.2-83.9 89.7v478.8c0 49.6 37.6 89.8 83.9 89.8h726.9c46.3 0 83.9-40.1 83.9-89.8V331.4c0-49.5-37.6-89.7-83.9-89.7z m28 538.6c0 33.1-25 59.8-55.9 59.8h-671c-30.9 0-55.9-26.8-55.9-59.8V361.4c0-33 25-59.8 55.9-59.8h111.2c25.8 0 82-119.7 111.6-119.7h223.2c29.5 0 85.6 119.7 111.5 119.7h113.5c30.9 0 55.9 26.8 55.9 59.8v418.9z" p-id="7320" fill="#ffffff"></path></svg>
</div>
</div>
</div>
</div>
<!-- 连接状态指示器 - 页面顶部中部 -->
<div class="connection-status-top">
<div class="status-indicator">
<span class="status-dot status-disconnected"></span>
<span id="connectionStatus">离线</span>
</div>
</div>
<!-- 模型加载提示 -->
<div class="model-container">
<div class="model-loading" id="modelLoading">
<div class="loading-char-float">
<div class="main-char">
<span></span><span></span><span></span><span></span><span></span><span></span>
</div>
<div class="shadow-char">模型加载中✨</div>
</div>
</div>
</div>
<!-- Live2D Canvas 容器 -->
<canvas id="live2d-stage"></canvas>
<!-- 聊天消息流(弹幕样式) -->
<div class="chat-container">
<div class="chat-stream" id="chatStream">
<!-- 聊天消息将动态插入这里 -->
</div>
<div class="chat-ipt" id="chatIpt">
<input type="text" id="messageInput" autocomplete="off" placeholder="输入消息,按Enter发送">
</div>
</div>
<!-- 底部控制栏 -->
<div class="control-bar">
<button class="control-btn" id="settingsBtn" title="设置">
<svg class="btn-icon" viewBox="0 0 24 24" fill="currentColor">
<path
d="M12 15.5A3.5 3.5 0 0 1 8.5 12A3.5 3.5 0 0 1 12 8.5A3.5 3.5 0 0 1 15.5 12A3.5 3.5 0 0 1 12 15.5M19.43 12.97C19.47 12.65 19.5 12.33 19.5 12C19.5 11.67 19.47 11.34 19.43 11L21.54 9.37C21.73 9.22 21.78 8.95 21.66 8.73L19.66 5.27C19.54 5.05 19.27 4.96 19.05 5.05L16.56 6.05C16.04 5.66 15.5 5.32 14.87 5.07L14.5 2.42C14.46 2.18 14.25 2 14 2H10C9.75 2 9.54 2.18 9.5 2.42L9.13 5.07C8.5 5.32 7.96 5.66 7.44 6.05L4.95 5.05C4.73 4.96 4.46 5.05 4.34 5.27L2.34 8.73C2.22 8.95 2.27 9.22 2.46 9.37L4.57 11C4.53 11.34 4.5 11.67 4.5 12C4.5 12.33 4.53 12.65 4.57 12.97L2.46 14.63C2.27 14.78 2.22 15.05 2.34 15.27L4.34 18.73C4.46 18.95 4.73 19.03 4.95 18.95L7.44 17.94C7.96 18.34 8.5 18.68 9.13 18.93L9.5 21.58C9.54 21.82 9.75 22 10 22H14C14.25 22 14.46 21.82 14.5 21.58L14.87 18.93C15.5 18.68 16.04 18.34 16.56 17.94L19.05 18.95C19.27 19.03 19.54 18.95 19.66 18.73L21.66 15.27C21.78 15.05 21.73 14.78 21.54 14.63L19.43 12.97Z" />
</svg>
<span class="btn-text">设置</span>
</button>
<button class="control-btn" id="cameraBtn" title="请先连接服务器" disabled>
<svg class="btn-icon" viewBox="0 0 24 24" fill="currentColor">
<path
d="M17,10.5V7A1,1 0 0,0 16,6H4A1,1 0 0,0 3,7V17A1,1 0 0,0 4,18H16A1,1 0 0,0 17,17V13.5L21,17.5V6.5L17,10.5Z" />
</svg>
<span class="btn-text">摄像头</span>
</button>
<button class="control-btn dial-btn" id="dialBtn" title="拨号">
<svg class="btn-icon" viewBox="0 0 24 24" fill="currentColor">
<path
d="M6.62,10.79C8.06,13.62 10.38,15.94 13.21,17.38L15.41,15.18C15.69,14.9 16.08,14.82 16.43,14.93C17.55,15.3 18.75,15.5 20,15.5A1,1 0 0,1 21,16.5V20A1,1 0 0,1 20,21A17,17 0 0,1 3,4A1,1 0 0,1 4,3H7.5A1,1 0 0,1 8.5,4C8.5,5.25 8.7,6.45 9.07,7.57C9.18,7.92 9.1,8.31 8.82,8.59L6.62,10.79Z" />
</svg>
<span class="btn-text">拨号</span>
</button>
<button class="control-btn" id="recordBtn" title="开始录音" disabled>
<svg class="btn-icon" viewBox="0 0 24 24" fill="currentColor">
<path
d="M12,2A3,3 0 0,1 15,5V11A3,3 0 0,1 12,14A3,3 0 0,1 9,11V5A3,3 0 0,1 12,2M19,11C19,14.53 16.39,17.44 13,17.93V21H11V17.93C7.61,17.44 5,14.53 5,11H7A5,5 0 0,0 12,16A5,5 0 0,0 17,11H19Z" />
</svg>
<span class="btn-text">录音</span>
</button>
</div>
</div>
<!-- 设置弹窗 -->
<div class="modal" id="settingsModal">
<div class="modal-content settings-modal">
<div class="modal-header">
<h2>设置</h2>
<button class="close-btn" id="closeSettingsBtn">&times;</button>
</div>
<div class="modal-body">
<div class="settings-tabs">
<button class="tab-btn active" data-tab="device">设备配置</button>
<button class="tab-btn" data-tab="wakeword">唤醒词</button>
<button class="tab-btn" data-tab="mcp">MCP工具</button>
<button class="tab-btn" data-tab="other">数字人皮肤</button>
</div>
<div class="tab-content active" id="deviceTab">
<!-- 设备配置面板 -->
<div class="config-panel">
<div class="control-panel">
<div class="config-row">
<div class="config-item">
<label for="deviceMac">设备MAC:</label>
<input type="text" id="deviceMac" placeholder="device-id">
</div>
</div>
<div class="config-row">
<div class="config-item">
<label for="clientId">客户端ID:</label>
<input type="text" id="clientId" value="web_test_client" placeholder="client-id">
</div>
<div class="config-item">
<label for="deviceName">设备名称:</label>
<input type="text" id="deviceName" value="Web测试设备" maxlength="50"
placeholder="deviceName">
</div>
</div>
<div class="config-row">
<div class="config-item">
<label for="emojiEnabled">情绪表情:</label>
<select id="emojiEnabled" class="model-select">
<option value="true">启用</option>
<option value="false" selected>禁用</option>
</select>
</div>
</div>
</div>
</div>
<!-- 连接信息面板 -->
<div class="connection-controls">
<div class="input-group">
<label for="otaUrl">OTA服务器地址:</label>
<input type="text" id="otaUrl" value="http://127.0.0.1:8002/xiaozhi/ota/"
placeholder="OTA服务器地址,如:http://127.0.0.1:8002/xiaozhi/ota/" />
</div>
<div class="input-group">
<label for="serverUrl">WebSocket服务器地址:</label>
<input type="text" id="serverUrl" value="" readonly disabled
placeholder="填写OTA地址后,点击拨号按钮自动连接" />
</div>
<div class="input-group">
<label for="visionUrl">视觉分析地址:</label>
<input type="text" id="visionUrl" value="" readonly disabled placeholder="成功建立ws连接后自动获取" />
</div>
</div>
</div>
<div class="tab-content" id="wakewordTab">
<div class="config-panel">
<div class="control-panel">
<div class="config-row">
<div class="config-item">
<label for="wakewordEnabled">启用本地唤醒词:</label>
<select id="wakewordEnabled" class="model-select">
<option value="true">启用</option>
<option value="false">禁用</option>
</select>
</div>
</div>
<div class="config-row">
<div class="config-item">
<label for="wakewordWsUrl">唤醒词服务地址:</label>
<input type="text" id="wakewordWsUrl" placeholder="ws://127.0.0.1:8006/wakeword-ws">
</div>
</div>
<div class="config-row">
<div class="config-item">
<label for="wakewordList">唤醒词配置 (一行一个):</label>
<textarea id="wakewordList" placeholder="例如:&#10;小智小智&#10;你好小智"></textarea>
</div>
</div>
<div class="modal-actions">
<button type="button" class="btn-primary" id="applyWakewordBtn">应用唤醒词</button>
</div>
</div>
</div>
</div>
<div class="tab-content" id="mcpTab">
<!-- MCP 工具管理区域 -->
<div class="mcp-tools-container">
<div class="mcp-tools-header">
<h3>MCP 工具管理</h3>
</div>
<div class="mcp-tools-panel" id="mcpToolsPanel">
<div class="mcp-tools-list" id="mcpToolsContainer">
<!-- 工具列表将动态插入这里 -->
</div>
<div class="mcp-actions">
<button class="btn-primary" id="addMcpToolBtn">
➕ 添加新工具
</button>
</div>
</div>
</div>
</div>
<div class="tab-content" id="otherTab">
<div class="other-settings-panel">
<div class="control-panel">
<div class="config-row">
<div class="config-item">
<label>选择模型:</label>
<select id="live2dModelSelect" class="model-select">
<option value="hiyori_pro_zh">hiyori (春日)</option>
<option value="natori_pro_zh">natori (名取)</option>
</select>
</div>
</div>
<div class="config-row">
<div class="config-item">
<label></label>
<button class="background-btn" id="backgroundBtn" title="切换背景">
<svg class="btn-icon" viewBox="0 0 24 24" fill="currentColor">
<path
d="M12,18A6,6 0 0,1 6,12C6,11 6.25,10.03 6.7,9.2L5.24,7.74C4.46,8.97 4,10.43 4,12A8,8 0 0,0 12,20V23L16,19L12,15V18M12,4V1L8,5L12,9V6A6,6 0 0,1 18,12C18,13 17.75,13.97 17.3,14.8L18.76,16.26C19.54,15.03 20,13.57 20,12A8,8 0 0,0 12,4Z" />
</svg>
<span class="btn-text">切换背景</span>
</button>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
<!-- MCP 工具编辑模态框 -->
<div id="mcpToolModal" class="modal">
<div class="modal-content">
<div class="modal-header">
<h2 id="mcpModalTitle">添加工具</h2>
<button class="close-btn" id="closeMcpModalBtn">&times;</button>
</div>
<div class="modal-body">
<div id="mcpErrorContainer"></div>
<form id="mcpToolForm">
<div class="input-group">
<label for="mcpToolName">工具名称 *</label>
<input type="text" id="mcpToolName" placeholder="例如: self.get_device_status" required>
</div>
<div class="input-group">
<label for="mcpToolDescription">工具描述 *</label>
<textarea id="mcpToolDescription" placeholder="详细描述工具的功能和使用场景..." required></textarea>
</div>
<div class="input-group">
<div class="input-group-header">
<label>输入参数</label>
<button type="button" class="properties-btn-primary" id="addMcpPropertyBtn">
➕ 添加参数
</button>
</div>
<div class="properties-container">
<div class="mcp-empty-state" id="mcpEmptyState">
暂无参数,点击上方按钮添加参数
</div>
<div class="mcp-properties-list" id="mcpPropertiesContainer">
</div>
</div>
</div>
<div class="input-group">
<label for="mcpMockResponse">
模拟返回结果 (JSON 格式,可选)
<span style="font-size: 12px; color: #999; font-weight: normal;">- 留空则返回默认成功消息</span>
</label>
<textarea id="mcpMockResponse" placeholder='{"success": true, "data": "执行成功"}'></textarea>
</div>
<div class="modal-actions">
<button type="button" class="btn-secondary" id="cancelMcpBtn">取消</button>
<button type="submit" class="btn-primary">保存</button>
</div>
</form>
</div>
</div>
</div>
<!-- 参数编辑模态框 -->
<div id="mcpPropertyModal" class="modal">
<div class="modal-content property-modal">
<div class="modal-header">
<h2 id="mcpPropertyModalTitle">编辑参数</h2>
<button class="close-btn" id="closeMcpPropertyModalBtn">&times;</button>
</div>
<div class="modal-body">
<form id="mcpPropertyForm">
<input type="hidden" id="mcpPropertyIndex" value="-1">
<div class="input-group">
<label for="mcpPropertyName">参数名称 *</label>
<input type="text" id="mcpPropertyName" placeholder="例如: param_1" required>
</div>
<div class="input-group">
<label for="mcpPropertyType">数据类型 *</label>
<select id="mcpPropertyType" class="model-select" required>
<option value="string">字符串</option>
<option value="integer">整数</option>
<option value="number">数字</option>
<option value="boolean">布尔值</option>
<option value="array">数组</option>
<option value="object">对象</option>
</select>
</div>
<div class="input-group" id="mcpPropertyRangeGroup" style="display: none;">
<div class="config-row">
<div class="config-item">
<label for="mcpPropertyMinimum">最小值</label>
<input type="number" id="mcpPropertyMinimum" placeholder="可选">
</div>
<div class="config-item">
<label for="mcpPropertyMaximum">最大值</label>
<input type="number" id="mcpPropertyMaximum" placeholder="可选">
</div>
</div>
</div>
<div class="input-group">
<label for="mcpPropertyDescription">参数描述</label>
<input type="text" id="mcpPropertyDescription" placeholder="可选">
</div>
<div class="input-group">
<label class="mcp-checkbox-label">
<input type="checkbox" id="mcpPropertyRequired">
必填参数
</label>
</div>
<div class="modal-actions">
<button type="button" class="btn-secondary" id="cancelMcpPropertyBtn">取消</button>
<button type="submit" class="btn-primary">保存</button>
</div>
</form>
</div>
</div>
</div>
<!-- 背景加载 -->
<script src="js/ui/background-load.js?v=0205"></script>
<!-- PIXI.js 2D渲染引擎 -->
<script src="js/live2d/pixi.js?v=0205"></script>
<!-- Live2D Cubism 4.0 SDK -->
<script src="js/live2d/live2dcubismcore.min.js?v=0205"></script>
<script src="js/live2d/cubism4.min.js?v=0205"></script>
<!-- Live2D 管理器 -->
<script src="js/live2d/live2d.js?v=0205"></script>
<!-- Opus解码库 -->
<script src="js/utils/libopus.js?v=0205"></script>
<!-- 主应用入口 -->
<script type="module" src="js/app.js?v=0205"></script>
<!-- 全局错误处理 -->
<script>
// 添加全局错误处理
window.addEventListener('error', function (event) {
console.error('全局错误:', event.error);
});
// 添加未处理的Promise拒绝处理
window.addEventListener('unhandledrejection', function (event) {
console.error('未处理的Promise拒绝:', event.reason);
});
</script>
</body>
</html>
+373
View File
@@ -0,0 +1,373 @@
// 主应用入口
import { checkOpusLoaded, initOpusEncoder } from './core/audio/opus-codec.js?v=0205';
import { getAudioPlayer } from './core/audio/player.js?v=0205';
import { checkMicrophoneAvailability, isHttpNonLocalhost } from './core/audio/recorder.js?v=0205';
import { initMcpTools } from './core/mcp/tools.js?v=0205';
import { startWakewordBridgeListener } from './core/network/wakeword-bridge.js?v=0205';
import { uiController } from './ui/controller.js?v=0205';
import { log } from './utils/logger.js?v=0205';
// 辅助函数:将Base64数据转换为Blob
function dataURItoBlob(dataURI) {
const byteString = atob(dataURI.split(',')[1]);
const mimeString = dataURI.split(',')[0].split(':')[1].split(';')[0];
const ab = new ArrayBuffer(byteString.length);
const ia = new Uint8Array(ab);
for (let i = 0; i < byteString.length; i++) {
ia[i] = byteString.charCodeAt(i);
}
return new Blob([ab], { type: mimeString });
}
// 应用类
class App {
constructor() {
this.uiController = null;
this.audioPlayer = null;
this.live2dManager = null;
this.cameraStream = null;
this.currentFacingMode = 'user';
}
// 初始化应用
async init() {
log('正在初始化应用...', 'info');
// 初始化UI控制器
this.uiController = uiController;
this.uiController.init();
// 检查Opus库
checkOpusLoaded();
// 初始化Opus编码器
initOpusEncoder();
// 初始化音频播放器
this.audioPlayer = getAudioPlayer();
await this.audioPlayer.start();
// 初始化MCP工具
initMcpTools();
// 初始化本地唤醒事件监听
startWakewordBridgeListener();
// 检查麦克风可用性
await this.checkMicrophoneAvailability();
// 检查摄像头可用性
this.checkCameraAvailability();
// 初始化Live2D
await this.initLive2D();
// 初始化摄像头
this.initCamera();
// 关闭加载loading
this.setModelLoadingStatus(false);
log('应用初始化完成', 'success');
}
// 初始化Live2D
async initLive2D() {
try {
// 检查Live2DManager是否已加载
if (typeof window.Live2DManager === 'undefined') {
throw new Error('Live2DManager未加载,请检查脚本引入顺序');
}
this.live2dManager = new window.Live2DManager();
await this.live2dManager.initializeLive2D();
// 更新UI状态
const live2dStatus = document.getElementById('live2dStatus');
if (live2dStatus) {
live2dStatus.textContent = '● 已加载';
live2dStatus.className = 'status loaded';
}
log('Live2D初始化完成', 'success');
} catch (error) {
log(`Live2D初始化失败: ${error.message}`, 'error');
// 更新UI状态
const live2dStatus = document.getElementById('live2dStatus');
if (live2dStatus) {
live2dStatus.textContent = '● 加载失败';
live2dStatus.className = 'status error';
}
}
}
// 设置model加载状态
setModelLoadingStatus(isLoading) {
const modelLoading = document.getElementById('modelLoading');
if (modelLoading) {
modelLoading.style.display = isLoading ? 'flex' : 'none';
}
}
/**
* 检查麦克风可用性
* 在应用初始化时调用,检查麦克风是否可用并更新UI状态
*/
async checkMicrophoneAvailability() {
try {
const isAvailable = await checkMicrophoneAvailability();
const isHttp = isHttpNonLocalhost();
// 保存可用性状态到全局变量
window.microphoneAvailable = isAvailable;
window.isHttpNonLocalhost = isHttp;
// 更新UI
if (this.uiController) {
this.uiController.updateMicrophoneAvailability(isAvailable, isHttp);
}
log(`麦克风可用性检查完成: ${isAvailable ? '可用' : '不可用'}`, isAvailable ? 'success' : 'warning');
} catch (error) {
log(`检查麦克风可用性失败: ${error.message}`, 'error');
// 默认设置为不可用
window.microphoneAvailable = false;
window.isHttpNonLocalhost = isHttpNonLocalhost();
if (this.uiController) {
this.uiController.updateMicrophoneAvailability(false, window.isHttpNonLocalhost);
}
}
}
// 检查摄像头可用性
checkCameraAvailability() {
window.cameraAvailable = true;
log('摄像头可用性检查完成: 默认已绑定验证码', 'success');
}
// 初始化摄像头
async initCamera() {
const cameraContainer = document.getElementById('cameraContainer');
const cameraVideo = document.getElementById('cameraVideo');
const cameraSwitch = document.getElementById('cameraSwitch');
const cameraSwitchMask = document.getElementById('cameraSwitchMask');
const dialBtn = document.getElementById('dialBtn');
if (!cameraContainer || !cameraVideo) {
log('摄像头元素未找到,跳过初始化', 'warning');
return Promise.resolve(false);
}
let isDragging = false;
let currentX, currentY, initialX, initialY;
let xOffset = 0, yOffset = 0;
cameraContainer.addEventListener('mousedown', dragStart);
document.addEventListener('mousemove', drag);
document.addEventListener('mouseup', dragEnd);
cameraContainer.addEventListener('touchstart', dragStart, { passive: false });
document.addEventListener('touchmove', drag, { passive: false });
document.addEventListener('touchend', dragEnd);
function dragStart(e) {
if (e.type === 'touchstart') {
initialX = e.touches[0].clientX - xOffset;
initialY = e.touches[0].clientY - yOffset;
} else {
initialX = e.clientX - xOffset;
initialY = e.clientY - yOffset;
}
isDragging = true;
cameraContainer.classList.add('dragging');
}
function drag(e) {
if (isDragging) {
e.preventDefault();
if (e.type === 'touchmove') {
currentX = e.touches[0].clientX - initialX;
currentY = e.touches[0].clientY - initialY;
} else {
currentX = e.clientX - initialX;
currentY = e.clientY - initialY;
}
xOffset = currentX;
yOffset = currentY;
cameraContainer.style.transform = `translate3d(${currentX}px, ${currentY}px, 0)`;
}
}
function dragEnd() {
initialX = currentX;
initialY = currentY;
isDragging = false;
cameraContainer.classList.remove('dragging');
}
return new Promise((resolve) => {
window.startCamera = async () => {
try {
if (!navigator.mediaDevices || !navigator.mediaDevices.getUserMedia) {
log('浏览器不支持摄像头API', 'warning');
return false;
}
log('正在请求摄像头权限...', 'info');
this.cameraStream = await navigator.mediaDevices.getUserMedia({
video: { width: 180, height: 240, facingMode: this.currentFacingMode },
audio: false
});
cameraVideo.srcObject = this.cameraStream;
const devices = await navigator.mediaDevices.enumerateDevices();
const videoDevices = devices.filter(device => device.kind === 'videoinput');
if (videoDevices.length > 1) {
if (cameraSwitch) cameraSwitch.classList.add('active');
}
cameraContainer.classList.add('active');
// 切换时挂断情况
const hasActive = dialBtn.classList.contains('dial-active');
if (!hasActive) {
cameraContainer.classList.remove('active');
cameraSwitch.classList.remove('active');
window.stopCamera();
}
log('摄像头已启动', 'success');
return true;
} catch (error) {
log(`启动摄像头失败: ${error.name} - ${error.message}`, 'error');
if (error.name === 'NotAllowedError') {
log('摄像头权限被拒绝,请检查浏览器设置', 'warning');
} else if (error.name === 'NotFoundError') {
log('未找到摄像头设备', 'warning');
} else if (error.name === 'NotReadableError') {
log('摄像头已被其他程序占用', 'warning');
}
return false;
}
};
window.stopCamera = () => {
if (this.cameraStream) {
this.cameraStream.getTracks().forEach(track => track.stop());
this.cameraStream = null;
cameraVideo.srcObject = null;
log('摄像头已关闭', 'info');
}
};
window.switchCamera = async() => {
if (window.switchCameraTimer) return;
if (this.cameraStream) {
const currentTransform = window.getComputedStyle(cameraContainer).transform;
const originalTransform = currentTransform === 'none' ? 'translate(0px, 0px)' : currentTransform;
cameraContainer.style.setProperty('--original-transform', originalTransform);
cameraContainer.classList.add('flip');
if (cameraSwitchMask) cameraSwitchMask.style.opacity = 0;
this.currentFacingMode = this.currentFacingMode === 'user' ? 'environment' : 'user';
window.stopCamera();
window.startCamera();
window.switchCameraTimer = setTimeout(() => {
if (this.currentFacingMode === 'user') {
cameraVideo.style.transform = 'scaleX(-1)';
} else {
cameraVideo.style.transform = 'scaleX(1)';
}
window.switchCameraTimer = null;
cameraContainer.classList.remove('flip');
cameraContainer.style.removeProperty('--original-transform');
if (cameraSwitchMask) cameraSwitchMask.style.opacity = 1;
}, 500);
}
};
window.takePhoto = (question = '描述一下看到的物品') => {
return new Promise(async (resolve) => {
const canvas = document.createElement('canvas');
const video = cameraVideo;
if (!video || video.readyState !== video.HAVE_ENOUGH_DATA) {
log('无法拍照:摄像头未就绪', 'warning');
resolve({
success: false,
error: '摄像头未就绪,请确保已连接且摄像头已启动'
});
return;
}
canvas.width = video.videoWidth || 180;
canvas.height = video.videoHeight || 240;
const ctx = canvas.getContext('2d');
ctx.drawImage(video, 0, 0, canvas.width, canvas.height);
const photoData = canvas.toDataURL('image/jpeg', 0.8);
log(`拍照成功,图像数据长度: ${photoData.length}`, 'success');
try {
const xz_tester_vision = localStorage.getItem('xz_tester_vision');
if (xz_tester_vision) {
let visionInfo = null;
try {
visionInfo = JSON.parse(xz_tester_vision);
} catch (err) {
throw new Error(`视觉配置解析失败`);
}
const { url, token } = visionInfo || {};
if (!url || !token) {
throw new Error('视觉分析失败:配置缺少接口地址(url)或令牌(token)');
}
log(`正在发送图片到视觉分析接口: ${url}`, 'info');
const deviceId = document.getElementById('deviceMac')?.value || '';
const clientId = document.getElementById('clientId')?.value || 'web_test_client';
const formData = new FormData();
formData.append('question', question);
formData.append('image', dataURItoBlob(photoData), 'photo.jpg');
const response = await fetch(url, {
method: 'POST',
body: formData,
headers: {
'Device-Id': deviceId,
'Client-Id': clientId,
'Authorization': `Bearer ${token}`
}
});
if (!response.ok) {
throw new Error(`HTTP error! status: ${response.status}`);
}
const analysisResult = await response.json();
log(`视觉分析完成: ${JSON.stringify(analysisResult).substring(0, 200)}...`, 'success');
resolve({
success: true,
message: question,
photo_data: photoData,
photo_width: canvas.width,
photo_height: canvas.height,
vision_analysis: analysisResult
});
} else {
log('未配置视觉分析服务', 'warning');
}
} catch (error) {
log(`视觉分析失败: ${error.message}`, 'error');
resolve({
success: true,
message: question,
photo_data: photoData,
photo_width: canvas.width,
photo_height: canvas.height,
vision_analysis: {
success: false,
error: error.message,
fallback: '无法连接到视觉分析服务'
}
});
}
});
};
log('摄像头初始化完成', 'success');
resolve(true);
});
}
}
// 创建并启动应用
const app = new App();
// 将应用实例暴露到全局,供其他模块访问
window.chatApp = app;
document.addEventListener('DOMContentLoaded', () => {
// 初始化应用
app.init();
});
export default app;
@@ -0,0 +1,90 @@
[
{
"name": "self_camera_take_photo",
"description": "Take a photo using the device's camera. This tool captures the current camera frame and returns the image data. You MUST call this tool when user asks to 'take a photo', 'what do you see', 'describe what you see', or similar requests. After getting the photo data, describe what you see in the image. Parameter 'question' is optional, defaults to 'describe objects seen'.",
"inputSchema": {
"type": "object",
"properties": {
"question": {
"type": "string",
"description": "Question to guide the photo analysis, e.g., 'describe the objects seen'. Can be empty."
}
}
},
"mockResponse": {
"success": true,
"photo_data": "base64_image_data",
"response": "Please describe what you see based on the returned photo_data"
}
},
{
"name": "self.get_device_status",
"description": "Provides the real-time information of the device, including the current status of the audio speaker, screen, battery, network, etc.\nUse this tool for: \n1. Answering questions about current condition (e.g. what is the current volume of the audio speaker?)\n2. As the first step to control the device (e.g. turn up / down the volume of the audio speaker, etc.)",
"inputSchema": {
"type": "object",
"properties": {}
},
"mockResponse": {
"audio_speaker": {
"volume": 50,
"muted": false
},
"screen": {
"brightness": 80,
"theme": "light"
},
"battery": {
"level": 85,
"charging": false
},
"network": {
"connected": true,
"type": "wifi"
}
}
},
{
"name": "self.audio_speaker.set_volume",
"description": "Set the volume of the audio speaker. If the current volume is unknown, you must call `self.get_device_status` tool first and then call this tool.",
"inputSchema": {
"type": "object",
"properties": {
"volume": {
"type": "integer",
"minimum": 0,
"maximum": 100
}
},
"required": [
"volume"
]
},
"mockResponse": {
"success": true,
"volume": "${volume}",
"message": "音量已设置为 ${volume}"
}
},
{
"name": "self.screen.set_brightness",
"description": "Set the brightness of the screen.",
"inputSchema": {
"type": "object",
"properties": {
"brightness": {
"type": "integer",
"minimum": 0,
"maximum": 100
}
},
"required": [
"brightness"
]
},
"mockResponse": {
"success": true,
"brightness": "${brightness}",
"message": "亮度已设置为 ${brightness}"
}
}
]
+127
View File
@@ -0,0 +1,127 @@
// 配置管理模块
// 默认唤醒词列表
export const DEFAULT_WAKE_WORDS = '你好小智\n你好小志\n小爱同学\n你好小鑫\n你好小新\n小美同学\n小龙小龙\n喵喵同学\n小滨小滨\n小冰小冰\n嘿你好呀';
// 生成随机MAC地址
function generateRandomMac() {
const hexDigits = '0123456789ABCDEF';
let mac = '';
for (let i = 0; i < 6; i++) {
if (i > 0) mac += ':';
for (let j = 0; j < 2; j++) {
mac += hexDigits.charAt(Math.floor(Math.random() * 16));
}
}
return mac;
}
// 加载配置
export function loadConfig() {
const deviceMacInput = document.getElementById('deviceMac');
const deviceNameInput = document.getElementById('deviceName');
const clientIdInput = document.getElementById('clientId');
const otaUrlInput = document.getElementById('otaUrl');
const wakewordWsUrlInput = document.getElementById('wakewordWsUrl');
const wakewordEnabledInput = document.getElementById('wakewordEnabled');
const wakewordListInput = document.getElementById('wakewordList');
// 从localStorage加载MAC地址,如果没有则生成新的
let savedMac = localStorage.getItem('xz_tester_deviceMac');
if (!savedMac) {
savedMac = generateRandomMac();
localStorage.setItem('xz_tester_deviceMac', savedMac);
}
deviceMacInput.value = savedMac;
// 从localStorage加载其他配置
const savedDeviceName = localStorage.getItem('xz_tester_deviceName');
if (savedDeviceName) {
deviceNameInput.value = savedDeviceName;
}
const savedClientId = localStorage.getItem('xz_tester_clientId');
if (savedClientId) {
clientIdInput.value = savedClientId;
}
const savedOtaUrl = localStorage.getItem('xz_tester_otaUrl');
if (savedOtaUrl) {
otaUrlInput.value = savedOtaUrl;
}
const savedWakewordWsUrl = localStorage.getItem('xz_tester_wakewordWsUrl');
if (savedWakewordWsUrl !== null && wakewordWsUrlInput) {
wakewordWsUrlInput.value = savedWakewordWsUrl;
}
const savedWakewordEnabled = localStorage.getItem('xz_tester_wakewordEnabled');
if (savedWakewordEnabled !== null && wakewordEnabledInput) {
wakewordEnabledInput.value = savedWakewordEnabled;
}
const savedWakewordList = localStorage.getItem('xz_tester_wakewordList');
if (savedWakewordList !== null && wakewordListInput) {
wakewordListInput.value = savedWakewordList;
} else if (wakewordListInput) {
wakewordListInput.value = DEFAULT_WAKE_WORDS;
}
const emojiEnabledInput = document.getElementById('emojiEnabled');
const savedEmojiEnabled = localStorage.getItem('xz_tester_emojiEnabled');
if (savedEmojiEnabled !== null && emojiEnabledInput) {
emojiEnabledInput.value = savedEmojiEnabled;
}
}
// 保存配置
export function saveConfig() {
const deviceMacInput = document.getElementById('deviceMac');
const deviceNameInput = document.getElementById('deviceName');
const clientIdInput = document.getElementById('clientId');
const wakewordWsUrlInput = document.getElementById('wakewordWsUrl');
const wakewordEnabledInput = document.getElementById('wakewordEnabled');
const wakewordListInput = document.getElementById('wakewordList');
localStorage.setItem('xz_tester_deviceMac', deviceMacInput.value);
localStorage.setItem('xz_tester_deviceName', deviceNameInput.value);
localStorage.setItem('xz_tester_clientId', clientIdInput.value);
const emojiEnabledInput = document.getElementById('emojiEnabled');
if (emojiEnabledInput) {
localStorage.setItem('xz_tester_emojiEnabled', emojiEnabledInput.value);
}
if (wakewordEnabledInput) {
localStorage.setItem('xz_tester_wakewordEnabled', wakewordEnabledInput.value);
}
if (wakewordListInput) {
localStorage.setItem('xz_tester_wakewordList', wakewordListInput.value);
}
if (wakewordWsUrlInput && wakewordWsUrlInput.value.trim()) {
localStorage.setItem('xz_tester_wakewordWsUrl', wakewordWsUrlInput.value.trim());
}
}
// 获取配置值
export function getConfig() {
// 从DOM获取值
const deviceMac = document.getElementById('deviceMac')?.value.trim() || '';
const deviceName = document.getElementById('deviceName')?.value.trim() || '';
const clientId = document.getElementById('clientId')?.value.trim() || '';
const emojiEnabled = document.getElementById('emojiEnabled')?.value !== 'false';
return {
deviceId: deviceMac, // 使用MAC地址作为deviceId
deviceName,
deviceMac,
clientId,
emojiEnabled
};
}
// 保存连接URL
export function saveConnectionUrls() {
const otaUrl = document.getElementById('otaUrl').value.trim();
const wsUrl = document.getElementById('serverUrl').value.trim();
localStorage.setItem('xz_tester_otaUrl', otaUrl);
localStorage.setItem('xz_tester_wsUrl', wsUrl);
}
@@ -1,5 +1,4 @@
import { log } from './utils/logger.js'; import { log } from '../../utils/logger.js?v=0205';
import { updateScriptStatus } from './document.js'
// 检查Opus库是否已加载 // 检查Opus库是否已加载
@@ -15,7 +14,6 @@ export function checkOpusLoaded() {
// 使用Module.instance对象替换全局Module对象 // 使用Module.instance对象替换全局Module对象
window.ModuleInstance = Module.instance; window.ModuleInstance = Module.instance;
log('Opus库加载成功(使用Module.instance', 'success'); log('Opus库加载成功(使用Module.instance', 'success');
updateScriptStatus('Opus库加载成功', 'success');
// 3秒后隐藏状态 // 3秒后隐藏状态
const statusElement = document.getElementById('scriptStatus'); const statusElement = document.getElementById('scriptStatus');
@@ -27,7 +25,6 @@ export function checkOpusLoaded() {
if (typeof Module._opus_decoder_get_size === 'function') { if (typeof Module._opus_decoder_get_size === 'function') {
window.ModuleInstance = Module; window.ModuleInstance = Module;
log('Opus库加载成功(使用全局Module', 'success'); log('Opus库加载成功(使用全局Module', 'success');
updateScriptStatus('Opus库加载成功', 'success');
// 3秒后隐藏状态 // 3秒后隐藏状态
const statusElement = document.getElementById('scriptStatus'); const statusElement = document.getElementById('scriptStatus');
@@ -38,7 +35,6 @@ export function checkOpusLoaded() {
throw new Error('Opus解码函数未找到,可能Module结构不正确'); throw new Error('Opus解码函数未找到,可能Module结构不正确');
} catch (err) { } catch (err) {
log(`Opus库加载失败,请检查libopus.js文件是否存在且正确: ${err.message}`, 'error'); log(`Opus库加载失败,请检查libopus.js文件是否存在且正确: ${err.message}`, 'error');
updateScriptStatus('Opus库加载失败,请检查libopus.js文件是否存在且正确', 'error');
} }
} }
+297
View File
@@ -0,0 +1,297 @@
// 音频播放模块
import BlockingQueue from '../../utils/blocking-queue.js?v=0205';
import { log } from '../../utils/logger.js?v=0205';
import { createStreamingContext } from './stream-context.js?v=0205';
// 音频播放器类
export class AudioPlayer {
constructor() {
// 音频参数
this.SAMPLE_RATE = 16000;
this.CHANNELS = 1;
this.FRAME_SIZE = 960;
this.MIN_AUDIO_DURATION = 0.12;
// 状态
this.audioContext = null;
this.opusDecoder = null;
this.streamingContext = null;
this.queue = new BlockingQueue();
this.isPlaying = false;
}
// 获取或创建AudioContext
getAudioContext() {
if (!this.audioContext) {
this.audioContext = new (window.AudioContext || window.webkitAudioContext)({
sampleRate: this.SAMPLE_RATE,
latencyHint: 'interactive'
});
log('创建音频上下文,采样率: ' + this.SAMPLE_RATE + 'Hz', 'debug');
}
return this.audioContext;
}
// 初始化Opus解码器
async initOpusDecoder() {
if (this.opusDecoder) return this.opusDecoder;
try {
if (typeof window.ModuleInstance === 'undefined') {
if (typeof Module !== 'undefined') {
window.ModuleInstance = Module;
log('使用全局Module作为ModuleInstance', 'info');
} else {
throw new Error('Opus库未加载,ModuleInstance和Module对象都不存在');
}
}
const mod = window.ModuleInstance;
this.opusDecoder = {
channels: this.CHANNELS,
rate: this.SAMPLE_RATE,
frameSize: this.FRAME_SIZE,
module: mod,
decoderPtr: null,
init: function () {
if (this.decoderPtr) return true;
const decoderSize = mod._opus_decoder_get_size(this.channels);
log(`Opus解码器大小: ${decoderSize}字节`, 'debug');
this.decoderPtr = mod._malloc(decoderSize);
if (!this.decoderPtr) {
throw new Error("无法分配解码器内存");
}
const err = mod._opus_decoder_init(
this.decoderPtr,
this.rate,
this.channels
);
if (err < 0) {
this.destroy();
throw new Error(`Opus解码器初始化失败: ${err}`);
}
log("Opus解码器初始化成功", 'success');
return true;
},
decode: function (opusData) {
if (!this.decoderPtr) {
if (!this.init()) {
throw new Error("解码器未初始化且无法初始化");
}
}
try {
const mod = this.module;
const opusPtr = mod._malloc(opusData.length);
mod.HEAPU8.set(opusData, opusPtr);
const pcmPtr = mod._malloc(this.frameSize * 2);
const decodedSamples = mod._opus_decode(
this.decoderPtr,
opusPtr,
opusData.length,
pcmPtr,
this.frameSize,
0
);
if (decodedSamples < 0) {
mod._free(opusPtr);
mod._free(pcmPtr);
throw new Error(`Opus解码失败: ${decodedSamples}`);
}
const decodedData = new Int16Array(decodedSamples);
for (let i = 0; i < decodedSamples; i++) {
decodedData[i] = mod.HEAP16[(pcmPtr >> 1) + i];
}
mod._free(opusPtr);
mod._free(pcmPtr);
return decodedData;
} catch (error) {
log(`Opus解码错误: ${error.message}`, 'error');
return new Int16Array(0);
}
},
destroy: function () {
if (this.decoderPtr) {
this.module._free(this.decoderPtr);
this.decoderPtr = null;
}
}
};
if (!this.opusDecoder.init()) {
throw new Error("Opus解码器初始化失败");
}
return this.opusDecoder;
} catch (error) {
log(`Opus解码器初始化失败: ${error.message}`, 'error');
this.opusDecoder = null;
throw error;
}
}
// 启动音频缓冲
async startAudioBuffering() {
log("开始音频缓冲...", 'info');
this.initOpusDecoder().catch(error => {
log(`预初始化Opus解码器失败: ${error.message}`, 'warning');
});
const timeout = 400;
while (true) {
const packets = await this.queue.dequeue(
6,
timeout,
(count) => {
log(`缓冲超时,当前缓冲包数: ${count},开始播放`, 'info');
}
);
if (packets.length) {
log(`已缓冲 ${packets.length} 个音频包,开始播放`, 'info');
this.streamingContext.pushAudioBuffer(packets);
}
while (true) {
const data = await this.queue.dequeue(99, 30);
if (data.length) {
this.streamingContext.pushAudioBuffer(data);
} else {
break;
}
}
}
}
// 播放已缓冲的音频
async playBufferedAudio() {
try {
this.audioContext = this.getAudioContext();
if (!this.opusDecoder) {
log('初始化Opus解码器...', 'info');
try {
this.opusDecoder = await this.initOpusDecoder();
if (!this.opusDecoder) {
throw new Error('解码器初始化失败');
}
log('Opus解码器初始化成功', 'success');
} catch (error) {
log('Opus解码器初始化失败: ' + error.message, 'error');
this.isPlaying = false;
return;
}
}
if (!this.streamingContext) {
this.streamingContext = createStreamingContext(
this.opusDecoder,
this.audioContext,
this.SAMPLE_RATE,
this.CHANNELS,
this.MIN_AUDIO_DURATION
);
}
this.streamingContext.decodeOpusFrames();
this.streamingContext.startPlaying();
} catch (error) {
log(`播放已缓冲的音频出错: ${error.message}`, 'error');
this.isPlaying = false;
this.streamingContext = null;
}
}
// 添加音频数据到队列
enqueueAudioData(opusData) {
if (opusData.length > 0) {
this.queue.enqueue(opusData);
} else {
log('收到空音频数据帧,可能是结束标志', 'warning');
if (this.isPlaying && this.streamingContext) {
this.streamingContext.endOfStream = true;
}
}
}
// 预加载解码器
async preload() {
log('预加载Opus解码器...', 'info');
try {
await this.initOpusDecoder();
log('Opus解码器预加载成功', 'success');
} catch (error) {
log(`Opus解码器预加载失败: ${error.message},将在需要时重试`, 'warning');
}
}
// 启动播放系统
async start() {
await this.preload();
this.playBufferedAudio();
this.startAudioBuffering();
}
// 获取音频包统计信息
getAudioStats() {
if (!this.streamingContext) {
return {
pendingDecode: 0,
pendingPlay: 0,
totalPending: 0
};
}
const pendingDecode = this.streamingContext.getPendingDecodeCount();
const pendingPlay = this.streamingContext.getPendingPlayCount();
return {
pendingDecode, // 待解码包数
pendingPlay, // 待播放包数
totalPending: pendingDecode + pendingPlay // 总待处理包数
};
}
// 清空所有音频缓冲并停止播放
clearAllAudio() {
log('AudioPlayer: 清空所有音频', 'info');
// 清空接收队列(使用clear方法保持对象引用)
this.queue.clear();
// 清空流上下文的所有缓冲
if (this.streamingContext) {
this.streamingContext.clearAllBuffers();
}
log('AudioPlayer: 音频已清空', 'success');
}
}
// 创建单例
let audioPlayerInstance = null;
export function getAudioPlayer() {
if (!audioPlayerInstance) {
audioPlayerInstance = new AudioPlayer();
}
return audioPlayerInstance;
}
@@ -0,0 +1,384 @@
// Audio recording module
import { log } from '../../utils/logger.js?v=0205';
import { initOpusEncoder } from './opus-codec.js?v=0205';
import { getAudioPlayer } from './player.js?v=0205';
// Audio recorder class
export class AudioRecorder {
constructor() {
this.isRecording = false;
this.audioContext = null;
this.analyser = null;
this.audioProcessor = null;
this.audioProcessorType = null;
this.audioSource = null;
this.opusEncoder = null;
this.pcmDataBuffer = new Int16Array();
this.audioBuffers = [];
this.totalAudioSize = 0;
this.visualizationRequest = null;
this.recordingTimer = null;
this.websocket = null;
// Callback functions
this.onRecordingStart = null;
this.onRecordingStop = null;
this.onVisualizerUpdate = null;
}
// Set WebSocket instance
setWebSocket(ws) {
this.websocket = ws;
}
// Get AudioContext instance
getAudioContext() {
return getAudioPlayer().getAudioContext();
}
// Initialize encoder
initEncoder() {
if (!this.opusEncoder) {
this.opusEncoder = initOpusEncoder();
}
return this.opusEncoder;
}
// PCM processor code
getAudioProcessorCode() {
return `
class AudioRecorderProcessor extends AudioWorkletProcessor {
constructor() {
super();
this.buffers = [];
this.frameSize = 960;
this.buffer = new Int16Array(this.frameSize);
this.bufferIndex = 0;
this.isRecording = false;
this.port.onmessage = (event) => {
if (event.data.command === 'start') {
this.isRecording = true;
this.port.postMessage({ type: 'status', status: 'started' });
} else if (event.data.command === 'stop') {
this.isRecording = false;
if (this.bufferIndex > 0) {
const finalBuffer = this.buffer.slice(0, this.bufferIndex);
this.port.postMessage({ type: 'buffer', buffer: finalBuffer });
this.bufferIndex = 0;
}
this.port.postMessage({ type: 'status', status: 'stopped' });
}
};
}
process(inputs, outputs, parameters) {
if (!this.isRecording) return true;
const input = inputs[0][0];
if (!input) return true;
for (let i = 0; i < input.length; i++) {
if (this.bufferIndex >= this.frameSize) {
this.port.postMessage({ type: 'buffer', buffer: this.buffer.slice(0) });
this.bufferIndex = 0;
}
this.buffer[this.bufferIndex++] = Math.max(-32768, Math.min(32767, Math.floor(input[i] * 32767)));
}
return true;
}
}
registerProcessor('audio-recorder-processor', AudioRecorderProcessor);
`;
}
// Create audio processor
async createAudioProcessor() {
this.audioContext = this.getAudioContext();
try {
if (this.audioContext.audioWorklet) {
const blob = new Blob([this.getAudioProcessorCode()], { type: 'application/javascript' });
const url = URL.createObjectURL(blob);
await this.audioContext.audioWorklet.addModule(url);
URL.revokeObjectURL(url);
const audioProcessor = new AudioWorkletNode(this.audioContext, 'audio-recorder-processor');
audioProcessor.port.onmessage = (event) => {
if (event.data.type === 'buffer') {
this.processPCMBuffer(event.data.buffer);
}
};
log('使用AudioWorklet处理音频', 'success');
const silent = this.audioContext.createGain();
silent.gain.value = 0;
audioProcessor.connect(silent);
silent.connect(this.audioContext.destination);
return { node: audioProcessor, type: 'worklet' };
} else {
log('AudioWorklet不可用,使用ScriptProcessorNode作为后备方案', 'warning');
return this.createScriptProcessor();
}
} catch (error) {
log(`创建音频处理器失败: ${error.message},尝试后备方案`, 'error');
return this.createScriptProcessor();
}
}
// Create ScriptProcessor as fallback
createScriptProcessor() {
try {
const frameSize = 4096;
const scriptProcessor = this.audioContext.createScriptProcessor(frameSize, 1, 1);
scriptProcessor.onaudioprocess = (event) => {
if (!this.isRecording) return;
const input = event.inputBuffer.getChannelData(0);
const buffer = new Int16Array(input.length);
for (let i = 0; i < input.length; i++) {
buffer[i] = Math.max(-32768, Math.min(32767, Math.floor(input[i] * 32767)));
}
this.processPCMBuffer(buffer);
};
const silent = this.audioContext.createGain();
silent.gain.value = 0;
scriptProcessor.connect(silent);
silent.connect(this.audioContext.destination);
log('使用ScriptProcessorNode作为后备方案成功', 'warning');
return { node: scriptProcessor, type: 'processor' };
} catch (fallbackError) {
log(`后备方案也失败: ${fallbackError.message}`, 'error');
return null;
}
}
// Process PCM buffer data
processPCMBuffer(buffer) {
if (!this.isRecording) return;
const newBuffer = new Int16Array(this.pcmDataBuffer.length + buffer.length);
newBuffer.set(this.pcmDataBuffer);
newBuffer.set(buffer, this.pcmDataBuffer.length);
this.pcmDataBuffer = newBuffer;
const samplesPerFrame = 960;
while (this.pcmDataBuffer.length >= samplesPerFrame) {
const frameData = this.pcmDataBuffer.slice(0, samplesPerFrame);
this.pcmDataBuffer = this.pcmDataBuffer.slice(samplesPerFrame);
this.encodeAndSendOpus(frameData);
}
}
// Encode and send Opus data
encodeAndSendOpus(pcmData = null) {
if (!this.opusEncoder) {
log('Opus编码器未初始化', 'error');
return;
}
try {
if (pcmData) {
const opusData = this.opusEncoder.encode(pcmData);
if (opusData && opusData.length > 0) {
this.audioBuffers.push(opusData.buffer);
this.totalAudioSize += opusData.length;
if (this.websocket && this.websocket.readyState === WebSocket.OPEN) {
try {
this.websocket.send(opusData.buffer);
} catch (error) {
log(`WebSocket发送错误: ${error.message}`, 'error');
}
}
} else {
log('Opus编码失败,未返回有效数据', 'error');
}
} else {
if (this.pcmDataBuffer.length > 0) {
const samplesPerFrame = 960;
if (this.pcmDataBuffer.length < samplesPerFrame) {
const paddedBuffer = new Int16Array(samplesPerFrame);
paddedBuffer.set(this.pcmDataBuffer);
this.encodeAndSendOpus(paddedBuffer);
} else {
this.encodeAndSendOpus(this.pcmDataBuffer.slice(0, samplesPerFrame));
}
this.pcmDataBuffer = new Int16Array(0);
}
}
} catch (error) {
log(`Opus编码错误: ${error.message}`, 'error');
}
}
// Start recording
async start() {
if (this.isRecording) return false;
try {
if (!this.initEncoder()) {
log('无法开始录音: Opus编码器初始化失败', 'error');
return false;
}
log('请至少录制1-2秒音频以确保收集足够的数据', 'info');
const stream = await navigator.mediaDevices.getUserMedia({ audio: { echoCancellation: true, noiseSuppression: true, sampleRate: 16000, channelCount: 1 } });
this.audioContext = this.getAudioContext();
if (this.audioContext.state === 'suspended') {
await this.audioContext.resume();
}
const processorResult = await this.createAudioProcessor();
if (!processorResult) {
log('无法创建音频处理器', 'error');
return false;
}
this.audioProcessor = processorResult.node;
this.audioProcessorType = processorResult.type;
this.audioSource = this.audioContext.createMediaStreamSource(stream);
this.analyser = this.audioContext.createAnalyser();
this.analyser.fftSize = 2048;
this.audioSource.connect(this.analyser);
this.audioSource.connect(this.audioProcessor);
this.pcmDataBuffer = new Int16Array();
this.audioBuffers = [];
this.totalAudioSize = 0;
this.isRecording = true;
if (this.audioProcessorType === 'worklet' && this.audioProcessor.port) {
this.audioProcessor.port.postMessage({ command: 'start' });
}
// Send listening start message
if (this.websocket && this.websocket.readyState === WebSocket.OPEN) {
log(`已发送录音开始消息`, 'info');
} else {
log('WebSocket未连接,无法发送开始消息', 'error');
return false;
}
// Start visualization
if (this.onVisualizerUpdate) {
const dataArray = new Uint8Array(this.analyser.frequencyBinCount);
this.startVisualization(dataArray);
}
// Immediately notify recording start, update button state
if (this.onRecordingStart) {
this.onRecordingStart(0);
}
// Start recording timer
let recordingSeconds = 0;
this.recordingTimer = setInterval(() => {
recordingSeconds += 0.1;
if (this.onRecordingStart) {
this.onRecordingStart(recordingSeconds);
}
}, 100);
log('已开始PCM直接录音', 'success');
return true;
} catch (error) {
log(`直接录音启动错误: ${error.message}`, 'error');
this.isRecording = false;
return false;
}
}
// Start visualization
startVisualization(dataArray) {
const draw = () => {
this.visualizationRequest = requestAnimationFrame(() => draw());
if (!this.isRecording) return;
this.analyser.getByteFrequencyData(dataArray);
if (this.onVisualizerUpdate) {
this.onVisualizerUpdate(dataArray);
}
};
draw();
}
// Stop recording
stop() {
if (!this.isRecording) return false;
try {
this.isRecording = false;
if (this.audioProcessor) {
if (this.audioProcessorType === 'worklet' && this.audioProcessor.port) {
this.audioProcessor.port.postMessage({ command: 'stop' });
}
this.audioProcessor.disconnect();
this.audioProcessor = null;
}
if (this.audioSource) {
this.audioSource.disconnect();
this.audioSource = null;
}
if (this.visualizationRequest) {
cancelAnimationFrame(this.visualizationRequest);
this.visualizationRequest = null;
}
if (this.recordingTimer) {
clearInterval(this.recordingTimer);
this.recordingTimer = null;
}
// Encode and send remaining data
this.encodeAndSendOpus();
// Send end signal
if (this.websocket && this.websocket.readyState === WebSocket.OPEN) {
const emptyOpusFrame = new Uint8Array(0);
this.websocket.send(emptyOpusFrame);
log('已发送录音停止信号', 'info');
}
if (this.onRecordingStop) {
this.onRecordingStop();
}
log('已停止PCM直接录音', 'success');
return true;
} catch (error) {
log(`直接录音停止错误: ${error.message}`, 'error');
return false;
}
}
// Get analyser
getAnalyser() {
return this.analyser;
}
}
// Create singleton instance
let audioRecorderInstance = null;
export function getAudioRecorder() {
if (!audioRecorderInstance) {
audioRecorderInstance = new AudioRecorder();
}
return audioRecorderInstance;
}
/**
* Check if microphone is available
* @returns {Promise<boolean>} Returns true if available, false if not available
*/
export async function checkMicrophoneAvailability() {
// Check if browser supports getUserMedia API
if (!navigator.mediaDevices || !navigator.mediaDevices.getUserMedia) {
log('浏览器不支持getUserMedia API', 'warning');
return false;
}
try {
// Try to access microphone
const stream = await navigator.mediaDevices.getUserMedia({ audio: { echoCancellation: true, noiseSuppression: true, sampleRate: 16000, channelCount: 1 } });
// Immediately stop all tracks to release microphone
stream.getTracks().forEach(track => track.stop());
log('麦克风可用性检查成功', 'success');
return true;
} catch (error) {
log(`麦克风不可用: ${error.message}`, 'warning');
return false;
}
}
/**
* Check if it is HTTP non-localhost access
* @returns {boolean} Returns true if it is HTTP non-localhost access
*/
export function isHttpNonLocalhost() {
const protocol = window.location.protocol;
const hostname = window.location.hostname;
// Check if it is HTTP protocol
if (protocol !== 'http:') {
return false;
}
// localhost and 127.0.0.1 can use microphone
if (hostname === 'localhost' || hostname === '127.0.0.1') {
return false;
}
// Private IP addresses can also use microphone (browser allows)
if (hostname.startsWith('192.168.') || hostname.startsWith('10.') || hostname.startsWith('172.')) {
return false;
}
// Other HTTP access is considered non-localhost
return true;
}
@@ -1,5 +1,5 @@
import BlockingQueue from './utils/BlockingQueue.js'; import BlockingQueue from '../../utils/blocking-queue.js?v=0205';
import { log } from './utils/logger.js'; import { log } from '../../utils/logger.js?v=0205';
// 音频流播放上下文类 // 音频流播放上下文类
export class StreamingContext { export class StreamingContext {
@@ -22,6 +22,11 @@ export class StreamingContext {
this.source = null; // 当前音频源 this.source = null; // 当前音频源
this.totalSamples = 0; // 累积的总样本数 this.totalSamples = 0; // 累积的总样本数
this.lastPlayTime = 0; // 上次播放的时间戳 this.lastPlayTime = 0; // 上次播放的时间戳
this.scheduledEndTime = 0; // 已调度音频的结束时间
// 初始化分析器节点(供Live2D使用)
this.analyser = this.audioContext.createAnalyser();
this.analyser.fftSize = 256;
} }
// 缓存音频数组 // 缓存音频数组
@@ -31,29 +36,87 @@ export class StreamingContext {
// 获取需要处理缓存队列,单线程:在audioBufferQueue一直更新的状态下不会出现安全问题 // 获取需要处理缓存队列,单线程:在audioBufferQueue一直更新的状态下不会出现安全问题
async getPendingAudioBufferQueue() { async getPendingAudioBufferQueue() {
// 原子交换 + 清空 // 等待数据到达并获取
[this.pendingAudioBufferQueue, this.audioBufferQueue] = [await this.audioBufferQueue.dequeue(), new BlockingQueue()]; const data = await this.audioBufferQueue.dequeue();
// 赋值给待处理队列
this.pendingAudioBufferQueue = data;
} }
// 获取正在播放已解码的PCM队列,单线程:在activeQueue一直更新的状态下不会出现安全问题 // 获取正在播放已解码的PCM队列,单线程:在activeQueue一直更新的状态下不会出现安全问题
async getQueue(minSamples) { async getQueue(minSamples) {
let TepArray = [];
const num = minSamples - this.queue.length > 0 ? minSamples - this.queue.length : 1; const num = minSamples - this.queue.length > 0 ? minSamples - this.queue.length : 1;
// 原子交换 + 清空
[TepArray, this.activeQueue] = [await this.activeQueue.dequeue(num), new BlockingQueue()]; // 等待数据并获取
this.queue.push(...TepArray); const tempArray = await this.activeQueue.dequeue(num);
this.queue.push(...tempArray);
} }
// 将Int16音频数据转换为Float32音频数据 // 将Int16音频数据转换为Float32音频数据
convertInt16ToFloat32(int16Data) { convertInt16ToFloat32(int16Data) {
const float32Data = new Float32Array(int16Data.length); const float32Data = new Float32Array(int16Data.length);
for (let i = 0; i < int16Data.length; i++) { for (let i = 0; i < int16Data.length; i++) {
// 将[-32768,32767]范围转换为[-1,1] // 将[-32768,32767]范围转换为[-1,1],统一使用32768.0避免不对称失真
float32Data[i] = int16Data[i] / (int16Data[i] < 0 ? 0x8000 : 0x7FFF); float32Data[i] = int16Data[i] / 32768.0;
} }
return float32Data; return float32Data;
} }
// 获取待解码包数
getPendingDecodeCount() {
return this.audioBufferQueue.length + this.pendingAudioBufferQueue.length;
}
// 获取待播放样本数(转换为包数,每包960样本)
getPendingPlayCount() {
// 计算已在队列中的样本
const queuedSamples = this.activeQueue.length + this.queue.length;
// 计算已调度但未播放的样本(在Web Audio缓冲区中)
let scheduledSamples = 0;
if (this.playing && this.scheduledEndTime) {
const currentTime = this.audioContext.currentTime;
const remainingTime = Math.max(0, this.scheduledEndTime - currentTime);
scheduledSamples = Math.floor(remainingTime * this.sampleRate);
}
const totalSamples = queuedSamples + scheduledSamples;
return Math.ceil(totalSamples / 960);
}
// 清空所有音频缓冲
clearAllBuffers() {
log('清空所有音频缓冲', 'info');
// 清空所有队列(使用clear方法保持对象引用)
this.audioBufferQueue.clear();
this.pendingAudioBufferQueue = [];
this.activeQueue.clear();
this.queue = [];
// 停止当前播放的音频源
if (this.source) {
try {
this.source.stop();
this.source.disconnect();
} catch (e) {
// 忽略已经停止的错误
}
this.source = null;
}
// 重置状态
this.playing = false;
this.scheduledEndTime = this.audioContext.currentTime;
this.totalSamples = 0;
log('音频缓冲已清空', 'success');
}
// 获取分析器节点(供Live2D使用)
getAnalyser() {
return this.analyser;
}
// 将Opus数据解码为PCM // 将Opus数据解码为PCM
async decodeOpusFrames() { async decodeOpusFrames() {
if (!this.opusDecoder) { if (!this.opusDecoder) {
@@ -97,18 +160,26 @@ export class StreamingContext {
// 开始播放音频 // 开始播放音频
async startPlaying() { async startPlaying() {
this.scheduledEndTime = this.audioContext.currentTime; // 跟踪已调度音频的结束时间
while (true) { while (true) {
// 如果累积了至少0.3秒的音频,开始播放 // 初始缓冲:等待足够的样本再开始播放
const minSamples = this.sampleRate * this.minAudioDuration * 3; const minSamples = this.sampleRate * this.minAudioDuration * 2;
if (!this.playing && this.queue.length < minSamples) { if (!this.playing && this.queue.length < minSamples) {
await this.getQueue(minSamples); await this.getQueue(minSamples);
} }
this.playing = true; this.playing = true;
while (this.playing && this.queue.length) {
// 创建新的音频缓冲区
const minPlaySamples = Math.min(this.queue.length, this.sampleRate);
const currentSamples = this.queue.splice(0, minPlaySamples);
// 持续播放队列中的音频,每次播放一个小块
while (this.playing && this.queue.length > 0) {
// 每次播放120ms的音频(2个Opus包)
const playDuration = 0.12;
const targetSamples = Math.floor(this.sampleRate * playDuration);
const actualSamples = Math.min(this.queue.length, targetSamples);
if (actualSamples === 0) break;
const currentSamples = this.queue.splice(0, actualSamples);
const audioBuffer = this.audioContext.createBuffer(this.channels, currentSamples.length, this.sampleRate); const audioBuffer = this.audioContext.createBuffer(this.channels, currentSamples.length, this.sampleRate);
audioBuffer.copyToChannel(new Float32Array(currentSamples), 0); audioBuffer.copyToChannel(new Float32Array(currentSamples), 0);
@@ -116,28 +187,29 @@ export class StreamingContext {
this.source = this.audioContext.createBufferSource(); this.source = this.audioContext.createBufferSource();
this.source.buffer = audioBuffer; this.source.buffer = audioBuffer;
// 创建增益节点用于平滑过渡 // 精确调度播放时间
const gainNode = this.audioContext.createGain(); const currentTime = this.audioContext.currentTime;
const startTime = Math.max(this.scheduledEndTime, currentTime);
// 应用淡入淡出效果避免爆音 // 连接到分析器和输出
const fadeDuration = 0.02; // 20毫秒 this.source.connect(this.analyser);
gainNode.gain.setValueAtTime(0, this.audioContext.currentTime); this.source.connect(this.audioContext.destination);
gainNode.gain.linearRampToValueAtTime(1, this.audioContext.currentTime + fadeDuration);
log(`调度播放 ${currentSamples.length} 个样本,约 ${(currentSamples.length / this.sampleRate).toFixed(2)}`, 'debug');
this.source.start(startTime);
// 更新下一个音频块的调度时间
const duration = audioBuffer.duration; const duration = audioBuffer.duration;
if (duration > fadeDuration * 2) { this.scheduledEndTime = startTime + duration;
gainNode.gain.setValueAtTime(1, this.audioContext.currentTime + duration - fadeDuration); this.lastPlayTime = startTime;
gainNode.gain.linearRampToValueAtTime(0, this.audioContext.currentTime + duration);
// 如果队列中数据不足,等待新数据
if (this.queue.length < targetSamples) {
break;
}
} }
// 连接节点并开始播放 // 等待新数据
this.source.connect(gainNode);
gainNode.connect(this.audioContext.destination);
this.lastPlayTime = this.audioContext.currentTime;
log(`开始播放 ${currentSamples.length} 个样本,约 ${(currentSamples.length / this.sampleRate).toFixed(2)}`, 'info');
this.source.start();
}
await this.getQueue(minSamples); await this.getQueue(minSamples);
} }
} }
+525
View File
@@ -0,0 +1,525 @@
import { log } from '../../utils/logger.js?v=0205';
// ==========================================
// MCP 工具管理逻辑
// ==========================================
// 全局变量
let mcpTools = [];
let mcpEditingIndex = null;
let mcpProperties = [];
let websocket = null; // 将从外部设置
/**
* 设置 WebSocket 实例
* @param {WebSocket} ws - WebSocket 连接实例
*/
export function setWebSocket(ws) {
websocket = ws;
}
/**
* 初始化 MCP 工具
*/
export async function initMcpTools() {
// 加载默认工具数据
const defaultMcpTools = await fetch("js/config/default-mcp-tools.json").then(res => res.json());
const savedTools = localStorage.getItem('mcpTools');
if (savedTools) {
try {
const parsedTools = JSON.parse(savedTools);
// 合并默认工具和用户保存的工具,保留用户自定义的工具
const defaultToolNames = new Set(defaultMcpTools.map(t => t.name));
// 添加默认工具中不存在的新工具
parsedTools.forEach(tool => {
if (!defaultToolNames.has(tool.name)) {
defaultMcpTools.push(tool);
}
});
mcpTools = defaultMcpTools;
} catch (e) {
log('加载MCP工具失败,使用默认工具', 'warning');
mcpTools = [...defaultMcpTools];
}
} else {
mcpTools = [...defaultMcpTools];
}
renderMcpTools();
setupMcpEventListeners();
}
/**
* 渲染工具列表
*/
function renderMcpTools() {
const container = document.getElementById('mcpToolsContainer');
const countSpan = document.getElementById('mcpToolsCount');
if (!container) {
return; // Container not found, skip rendering
}
if (countSpan) {
countSpan.textContent = `${mcpTools.length} 个工具`;
}
if (mcpTools.length === 0) {
container.innerHTML = '<div style="text-align: center; padding: 30px; color: #999;">暂无工具,点击下方按钮添加新工具</div>';
return;
}
container.innerHTML = mcpTools.map((tool, index) => {
const paramCount = tool.inputSchema.properties ? Object.keys(tool.inputSchema.properties).length : 0;
const requiredCount = tool.inputSchema.required ? tool.inputSchema.required.length : 0;
const hasMockResponse = tool.mockResponse && Object.keys(tool.mockResponse).length > 0;
return `
<div class="mcp-tool-card">
<div class="mcp-tool-header">
<div class="mcp-tool-name">${tool.name}</div>
<div class="mcp-tool-actions">
<button class="mcp-edit-btn" onclick="window.mcpModule.editMcpTool(${index})">
✏️ 编辑
</button>
<button class="mcp-delete-btn" onclick="window.mcpModule.deleteMcpTool(${index})">
🗑️ 删除
</button>
</div>
</div>
<div class="mcp-tool-description">${tool.description}</div>
<div class="mcp-tool-info">
<div class="mcp-tool-info-row">
<span class="mcp-tool-info-label">参数数量:</span>
<span class="mcp-tool-info-value">${paramCount}${requiredCount > 0 ? `(${requiredCount} 个必填)` : ''}</span>
</div>
<div class="mcp-tool-info-row">
<span class="mcp-tool-info-label">模拟返回:</span>
<span class="mcp-tool-info-value">${hasMockResponse ? '✅ 已配置: ' + JSON.stringify(tool.mockResponse) : '⚪ 使用默认'}</span>
</div>
</div>
</div>
`;
}).join('');
}
/**
* 渲染参数列表
*/
function renderMcpProperties() {
const container = document.getElementById('mcpPropertiesContainer');
const emptyState = document.getElementById('mcpEmptyState');
if (!container) {
return; // Container not found, skip rendering
}
if (mcpProperties.length === 0) {
if (emptyState) {
emptyState.style.display = 'block';
}
container.innerHTML = '';
return;
}
if (emptyState) {
emptyState.style.display = 'none';
}
container.innerHTML = mcpProperties.map((prop, index) => `
<div class="mcp-property-card" onclick="window.mcpModule.editMcpProperty(${index})">
<div class="mcp-property-row-label">
<span class="mcp-property-label">参数名称</span>
<span class="mcp-property-value">${prop.name}${prop.required ? ' <span class="mcp-property-required-badge">[必填]</span>' : ''}</span>
</div>
<div class="mcp-property-row-label">
<span class="mcp-property-label">数据类型</span>
<span class="mcp-property-value">${getTypeLabel(prop.type)}</span>
</div>
<div class="mcp-property-row-label">
<span class="mcp-property-label">描述</span>
<span class="mcp-property-value">${prop.description || '-'}</span>
</div>
<div class="mcp-property-row-action">
<button class="mcp-property-delete-btn" onclick="event.stopPropagation(); window.mcpModule.deleteMcpProperty(${index})">删除</button>
</div>
</div>
`).join('');
}
/**
* 获取数据类型标签
*/
function getTypeLabel(type) {
const typeMap = {
'string': '字符串',
'integer': '整数',
'number': '数字',
'boolean': '布尔值',
'array': '数组',
'object': '对象'
};
return typeMap[type] || type;
}
/**
* 添加参数 - 打开参数编辑模态框
*/
function addMcpProperty() {
openPropertyModal();
}
/**
* 编辑参数 - 打开参数编辑模态框
*/
function editMcpProperty(index) {
openPropertyModal(index);
}
/**
* 打开参数编辑模态框
*/
function openPropertyModal(index = null) {
const form = document.getElementById('mcpPropertyForm');
const title = document.getElementById('mcpPropertyModalTitle');
document.getElementById('mcpPropertyIndex').value = index !== null ? index : -1;
if (index !== null) {
const prop = mcpProperties[index];
title.textContent = '编辑参数';
document.getElementById('mcpPropertyName').value = prop.name;
document.getElementById('mcpPropertyType').value = prop.type || 'string';
document.getElementById('mcpPropertyMinimum').value = prop.minimum !== undefined ? prop.minimum : '';
document.getElementById('mcpPropertyMaximum').value = prop.maximum !== undefined ? prop.maximum : '';
document.getElementById('mcpPropertyDescription').value = prop.description || '';
document.getElementById('mcpPropertyRequired').checked = prop.required || false;
} else {
title.textContent = '添加参数';
form.reset();
document.getElementById('mcpPropertyName').value = `param_${mcpProperties.length + 1}`;
document.getElementById('mcpPropertyType').value = 'string';
document.getElementById('mcpPropertyMinimum').value = '';
document.getElementById('mcpPropertyMaximum').value = '';
document.getElementById('mcpPropertyDescription').value = '';
document.getElementById('mcpPropertyRequired').checked = false;
}
updatePropertyRangeVisibility();
document.getElementById('mcpPropertyModal').style.display = 'flex';
}
/**
* 关闭参数编辑模态框
*/
function closePropertyModal() {
document.getElementById('mcpPropertyModal').style.display = 'none';
}
/**
* 更新数值范围输入框的可见性
*/
function updatePropertyRangeVisibility() {
const type = document.getElementById('mcpPropertyType').value;
const rangeGroup = document.getElementById('mcpPropertyRangeGroup');
if (type === 'integer' || type === 'number') {
rangeGroup.style.display = 'block';
} else {
rangeGroup.style.display = 'none';
}
}
/**
* 处理参数表单提交
*/
function handlePropertySubmit(e) {
e.preventDefault();
const index = parseInt(document.getElementById('mcpPropertyIndex').value);
const name = document.getElementById('mcpPropertyName').value.trim();
const type = document.getElementById('mcpPropertyType').value;
const minimum = document.getElementById('mcpPropertyMinimum').value;
const maximum = document.getElementById('mcpPropertyMaximum').value;
const description = document.getElementById('mcpPropertyDescription').value.trim();
const required = document.getElementById('mcpPropertyRequired').checked;
// 检查名称重复
const isDuplicate = mcpProperties.some((p, i) => i !== index && p.name === name);
if (isDuplicate) {
alert('参数名称已存在,请使用不同的名称');
return;
}
const propData = {
name,
type,
description,
required
};
// 数值类型添加范围限制
if (type === 'integer' || type === 'number') {
if (minimum !== '') {
propData.minimum = parseFloat(minimum);
}
if (maximum !== '') {
propData.maximum = parseFloat(maximum);
}
}
if (index >= 0) {
mcpProperties[index] = propData;
} else {
mcpProperties.push(propData);
}
renderMcpProperties();
closePropertyModal();
}
/**
* 删除参数
*/
function deleteMcpProperty(index) {
mcpProperties.splice(index, 1);
renderMcpProperties();
}
/**
* 设置事件监听
*/
function setupMcpEventListeners() {
const panel = document.getElementById('mcpToolsPanel');
const addBtn = document.getElementById('addMcpToolBtn');
const modal = document.getElementById('mcpToolModal');
const closeBtn = document.getElementById('closeMcpModalBtn');
const cancelBtn = document.getElementById('cancelMcpBtn');
const form = document.getElementById('mcpToolForm');
const addPropertyBtn = document.getElementById('addMcpPropertyBtn');
// 参数编辑模态框相关元素
const propertyModal = document.getElementById('mcpPropertyModal');
const closePropertyBtn = document.getElementById('closeMcpPropertyModalBtn');
const cancelPropertyBtn = document.getElementById('cancelMcpPropertyBtn');
const propertyForm = document.getElementById('mcpPropertyForm');
const propertyTypeSelect = document.getElementById('mcpPropertyType');
// Return early if required elements don't exist (e.g., in test environment)
if (!panel || !addBtn || !modal || !closeBtn || !cancelBtn || !form || !addPropertyBtn) {
return;
}
addBtn.addEventListener('click', () => openMcpModal());
closeBtn.addEventListener('click', closeMcpModal);
cancelBtn.addEventListener('click', closeMcpModal);
addPropertyBtn.addEventListener('click', addMcpProperty);
form.addEventListener('submit', handleMcpSubmit);
// 参数编辑模态框事件
if (propertyModal && closePropertyBtn && cancelPropertyBtn && propertyForm && propertyTypeSelect) {
closePropertyBtn.addEventListener('click', closePropertyModal);
cancelPropertyBtn.addEventListener('click', closePropertyModal);
propertyForm.addEventListener('submit', handlePropertySubmit);
propertyTypeSelect.addEventListener('change', updatePropertyRangeVisibility);
}
}
/**
* 打开模态框
*/
function openMcpModal(index = null) {
const isConnected = websocket && websocket.readyState === WebSocket.OPEN;
if (isConnected) {
alert('WebSocket 已连接,无法编辑工具');
return;
}
mcpEditingIndex = index;
const errorContainer = document.getElementById('mcpErrorContainer');
errorContainer.innerHTML = '';
if (index !== null) {
document.getElementById('mcpModalTitle').textContent = '编辑工具';
const tool = mcpTools[index];
document.getElementById('mcpToolName').value = tool.name;
document.getElementById('mcpToolDescription').value = tool.description;
document.getElementById('mcpMockResponse').value = tool.mockResponse ? JSON.stringify(tool.mockResponse, null, 2) : '';
mcpProperties = [];
const schema = tool.inputSchema;
if (schema.properties) {
Object.keys(schema.properties).forEach(key => {
const prop = schema.properties[key];
mcpProperties.push({
name: key,
type: prop.type || 'string',
minimum: prop.minimum,
maximum: prop.maximum,
description: prop.description || '',
required: schema.required && schema.required.includes(key)
});
});
}
} else {
document.getElementById('mcpModalTitle').textContent = '添加工具';
document.getElementById('mcpToolForm').reset();
mcpProperties = [];
}
renderMcpProperties();
document.getElementById('mcpToolModal').style.display = 'flex';
}
/**
* 关闭模态框
*/
function closeMcpModal() {
document.getElementById('mcpToolModal').style.display = 'none';
mcpEditingIndex = null;
document.getElementById('mcpToolForm').reset();
mcpProperties = [];
document.getElementById('mcpErrorContainer').innerHTML = '';
}
/**
* 处理表单提交
*/
function handleMcpSubmit(e) {
e.preventDefault();
const errorContainer = document.getElementById('mcpErrorContainer');
errorContainer.innerHTML = '';
const name = document.getElementById('mcpToolName').value.trim();
const description = document.getElementById('mcpToolDescription').value.trim();
const mockResponseText = document.getElementById('mcpMockResponse').value.trim();
// 检查名称重复
const isDuplicate = mcpTools.some((tool, index) => tool.name === name && index !== mcpEditingIndex);
if (isDuplicate) {
showMcpError('工具名称已存在,请使用不同的名称');
return;
}
// 解析模拟返回结果
let mockResponse = null;
if (mockResponseText) {
try {
mockResponse = JSON.parse(mockResponseText);
} catch (e) {
showMcpError('模拟返回结果不是有效的 JSON 格式: ' + e.message);
return;
}
}
// 构建 inputSchema
const inputSchema = { type: "object", properties: {}, required: [] };
mcpProperties.forEach(prop => {
const propSchema = { type: prop.type };
if (prop.description) {
propSchema.description = prop.description;
}
if ((prop.type === 'integer' || prop.type === 'number')) {
if (prop.minimum !== undefined && prop.minimum !== '') {
propSchema.minimum = prop.minimum;
}
if (prop.maximum !== undefined && prop.maximum !== '') {
propSchema.maximum = prop.maximum;
}
}
inputSchema.properties[prop.name] = propSchema;
if (prop.required) {
inputSchema.required.push(prop.name);
}
});
if (inputSchema.required.length === 0) {
delete inputSchema.required;
}
const tool = { name, description, inputSchema, mockResponse };
if (mcpEditingIndex !== null) {
mcpTools[mcpEditingIndex] = tool;
log(`已更新工具: ${name}`, 'success');
} else {
mcpTools.push(tool);
log(`已添加工具: ${name}`, 'success');
}
saveMcpTools();
renderMcpTools();
closeMcpModal();
}
/**
* 显示错误
*/
function showMcpError(message) {
const errorContainer = document.getElementById('mcpErrorContainer');
errorContainer.innerHTML = `<div class="mcp-error">${message}</div>`;
}
/**
* 编辑工具
*/
function editMcpTool(index) {
openMcpModal(index);
}
/**
* 删除工具
*/
function deleteMcpTool(index) {
const isConnected = websocket && websocket.readyState === WebSocket.OPEN;
if (isConnected) {
alert('WebSocket 已连接,无法编辑工具');
return;
}
if (confirm(`确定要删除工具 "${mcpTools[index].name}" 吗?`)) {
const toolName = mcpTools[index].name;
mcpTools.splice(index, 1);
saveMcpTools();
renderMcpTools();
log(`已删除工具: ${toolName}`, 'info');
}
}
/**
* 保存工具
*/
function saveMcpTools() {
localStorage.setItem('mcpTools', JSON.stringify(mcpTools));
}
/**
* 获取工具列表
*/
export function getMcpTools() {
return mcpTools.map(tool => ({ name: tool.name, description: tool.description, inputSchema: tool.inputSchema }));
}
/**
* 执行工具调用
*/
export async function executeMcpTool(toolName, toolArgs) {
const tool = mcpTools.find(t => t.name === toolName);
if (!tool) {
log(`未找到工具: ${toolName}`, 'error');
return { success: false, error: `未知工具: ${toolName}` };
}
// 处理拍照工具
if (toolName === 'self_camera_take_photo') {
if (typeof window.takePhoto === 'function') {
const question = toolArgs && toolArgs.question ? toolArgs.question : '描述一下看到的物品';
log(`正在执行拍照: ${question}`, 'info');
const result = await window.takePhoto(question);
return result;
} else {
log('拍照功能不可用', 'warning');
return { success: false, error: '摄像头未启动或不支持拍照功能' };
}
}
// 如果有模拟返回结果,使用它
if (tool.mockResponse) {
// 替换模板变量
let responseStr = JSON.stringify(tool.mockResponse);
// 替换 ${paramName} 格式的变量
if (toolArgs) {
Object.keys(toolArgs).forEach(key => {
const regex = new RegExp(`\\$\\{${key}\\}`, 'g');
responseStr = responseStr.replace(regex, toolArgs[key]);
});
}
try {
const response = JSON.parse(responseStr);
log(`工具 ${toolName} 执行成功,返回模拟结果: ${responseStr}`, 'success');
return response;
} catch (e) {
log(`解析模拟返回结果失败: ${e.message}`, 'error');
return tool.mockResponse;
}
}
// 没有模拟返回结果,返回默认成功消息
log(`工具 ${toolName} 执行成功,返回默认结果`, 'success');
return { success: true, message: `工具 ${toolName} 执行成功`, tool: toolName, arguments: toolArgs };
}
// 暴露全局方法供 HTML 内联事件调用
window.mcpModule = { addMcpProperty, editMcpProperty, deleteMcpProperty, editMcpTool, deleteMcpTool };
@@ -1,24 +1,49 @@
import { log } from './utils/logger.js'; import { log } from '../../utils/logger.js?v=0205';
import { otaStatusStyle } from './document.js'
// WebSocket 连接 // WebSocket 连接
export async function webSocketConnect(otaUrl,wsUrl,config){ export async function webSocketConnect(otaUrl, config) {
if (!validateWsUrl(wsUrl)) {
return; // 直接返回,不再往下执行
}
if (!validateConfig(config)) { if (!validateConfig(config)) {
return; return;
} }
const ok = await sendOTA(otaUrl, config);
if (!ok) return;
// 使用自定义WebSocket实现以添加认证头信息 // 发送OTA请求并获取返回的websocket信息
let connUrl = new URL(wsUrl); const otaResult = await sendOTA(otaUrl, config);
// 添加认证参数 if (!otaResult) {
log('无法从OTA服务器获取信息', 'error');
return;
}
// 从OTA响应中提取websocket信息
const { websocket } = otaResult;
if (!websocket || !websocket.url) {
log('OTA响应中缺少websocket信息', 'error');
return;
}
// 使用OTA返回的websocket URL
let connUrl = new URL(websocket.url);
// 添加token参数(从OTA响应中获取)
if (websocket.token) {
if (websocket.token.startsWith("Bearer ")) {
connUrl.searchParams.append('authorization', websocket.token);
} else {
connUrl.searchParams.append('authorization', 'Bearer ' + websocket.token);
}
}
// 添加认证参数(保持原有逻辑)
connUrl.searchParams.append('device-id', config.deviceId); connUrl.searchParams.append('device-id', config.deviceId);
connUrl.searchParams.append('client-id', config.clientId); connUrl.searchParams.append('client-id', config.clientId);
log(`正在连接: ${connUrl.toString()}`, 'info');
const wsurl = connUrl.toString()
log(`正在连接: ${wsurl}`, 'info');
if (wsurl) {
document.getElementById('serverUrl').value = wsurl;
}
return new WebSocket(connUrl.toString()); return new WebSocket(connUrl.toString());
} }
@@ -36,19 +61,7 @@ function validateConfig(config) {
return true; return true;
} }
// 判断wsUrl路径是否存在错误 // OTA发送请求,验证状态,并返回响应数据
function validateWsUrl(wsUrl){
if (wsUrl === '') return false;
// 检查URL格式
if (!wsUrl.startsWith('ws://') && !wsUrl.startsWith('wss://')) {
log('URL格式错误,必须以ws://或wss://开头', 'error');
return false;
}
return true
}
// OTA发送请求,验证状态
async function sendOTA(otaUrl, config) { async function sendOTA(otaUrl, config) {
try { try {
const res = await fetch(otaUrl, { const res = await fetch(otaUrl, {
@@ -70,7 +83,7 @@ async function sendOTA(otaUrl, config) {
}, },
ota: { label: 'xiaozhi-web-test' }, ota: { label: 'xiaozhi-web-test' },
board: { board: {
type: 'xiaozhi-web-test', type: config.deviceName,
ssid: 'xiaozhi-web-test', ssid: 'xiaozhi-web-test',
rssi: 0, rssi: 0,
channel: 0, channel: 0,
@@ -89,15 +102,8 @@ async function sendOTA(otaUrl, config) {
if (!res.ok) throw new Error(`${res.status} ${res.statusText}`); if (!res.ok) throw new Error(`${res.status} ${res.statusText}`);
const result = await res.json(); const result = await res.json();
otaStatusStyle(true) return result; // 返回完整的响应数据
return true; // 成功
} catch (err) { } catch (err) {
otaStatusStyle(false) return null; // 失败返回null
return false; // 失败
} }
} }
@@ -0,0 +1,225 @@
import { uiController } from '../../ui/controller.js?v=0205';
import { log } from '../../utils/logger.js?v=0205';
let wakewordSocket = null;
let reconnectTimer = null;
let reconnectAttempts = 0;
let shouldReconnect = true;
let wakewordRequestSeq = 0;
let onNextBridgeConnectedCallback = null;
const pendingWakewordRequests = new Map();
export function startWakewordBridgeListener() {
if (wakewordSocket) {
return wakewordSocket;
}
shouldReconnect = true;
log('正在连接本地唤醒事件桥...', 'info');
tryConnect();
return wakewordSocket;
}
function tryConnect() {
const bridgeUrl = buildWakewordBridgeUrl();
try {
wakewordSocket = new WebSocket(bridgeUrl);
wakewordSocket.onopen = () => {
reconnectAttempts = 0;
log(`本地唤醒事件桥已连接: ${bridgeUrl}`, 'success');
// 连接成功后自动保存地址,刷新后仍能记住
localStorage.setItem('xz_tester_wakewordWsUrl', bridgeUrl);
const urlInput = document.getElementById('wakewordWsUrl');
if (urlInput) urlInput.value = bridgeUrl;
};
wakewordSocket.onerror = () => {
log(`本地唤醒事件桥连接失败: ${bridgeUrl}`, 'error');
};
wakewordSocket.onmessage = async (event) => {
try {
const message = parseWakewordBridgeMessage(event.data);
if (message.requestId && pendingWakewordRequests.has(message.requestId)) {
settleWakewordRequest(message);
return;
}
if (message.success === false) {
log(`本地唤醒事件桥返回错误: ${message.error || '未知错误'}`, 'error');
return;
}
if (message.type === 'bridge_connected') {
log('本地唤醒监听已就绪', 'info');
if (onNextBridgeConnectedCallback) {
const cb = onNextBridgeConnectedCallback;
onNextBridgeConnectedCallback = null;
cb(message);
}
return;
}
if (message.type === 'service_ready') {
log('本地唤醒服务已启动', 'info');
return;
}
if (message.type === 'wakeword_config') {
uiController.applyWakewordConfig(message.payload || {});
log('已同步本地唤醒词配置', 'info');
return;
}
if (message.type === 'service_stopping') {
log('本地唤醒服务正在停止', 'warning');
return;
}
if (message.type === 'wake_word_detected') {
const wakeWord = message.payload?.wake_word || '唤醒词';
log(`检测到本地唤醒事件: ${wakeWord}`, 'info');
await uiController.triggerWakewordDial(wakeWord);
}
} catch (error) {
log(`解析本地唤醒事件失败: ${error.message}`, 'error');
}
};
wakewordSocket.onclose = () => {
if (wakewordSocket) {
wakewordSocket = null;
}
rejectAllWakewordRequests('本地唤醒事件桥已断开');
if (!shouldReconnect) {
return;
}
if (reconnectTimer) {
return;
}
reconnectAttempts += 1;
const delay = Math.min(1000 * reconnectAttempts, 5000);
log(`本地唤醒事件桥将在 ${delay}ms 后重连: ${bridgeUrl}`, 'warning');
reconnectTimer = window.setTimeout(() => {
reconnectTimer = null;
tryConnect();
}, delay);
};
return wakewordSocket;
} catch (error) {
log(`启动本地唤醒监听失败: ${error.message}`, 'error');
return null;
}
}
export function stopWakewordBridgeListener() {
shouldReconnect = false;
if (reconnectTimer) {
window.clearTimeout(reconnectTimer);
reconnectTimer = null;
}
if (!wakewordSocket) {
return;
}
wakewordSocket.onclose = null;
wakewordSocket.close();
wakewordSocket = null;
}
export function sendWakewordBridgeMessage(type, payload = {}, requestId = null) {
if (!wakewordSocket || wakewordSocket.readyState !== WebSocket.OPEN) {
log('本地唤醒事件桥未连接,无法发送消息', 'warning');
return false;
}
wakewordSocket.send(JSON.stringify({
type,
requestId,
payload,
}));
return true;
}
export function requestWakewordBridge(type, payload = {}, timeout = 5000) {
const requestId = `wakeword-${Date.now()}-${++wakewordRequestSeq}`;
return new Promise((resolve, reject) => {
const timer = window.setTimeout(() => {
pendingWakewordRequests.delete(requestId);
reject(new Error('本地唤醒服务响应超时'));
}, timeout);
pendingWakewordRequests.set(requestId, { resolve, reject, timer });
if (!sendWakewordBridgeMessage(type, payload, requestId)) {
window.clearTimeout(timer);
pendingWakewordRequests.delete(requestId);
reject(new Error('本地唤醒事件桥未连接'));
}
});
}
export function getWakewordBridgeUrl() {
if (wakewordSocket && wakewordSocket.url) {
return wakewordSocket.url;
}
return buildWakewordBridgeUrl();
}
export function onNextBridgeConnected(callback) {
onNextBridgeConnectedCallback = callback;
}
function buildWakewordBridgeUrl() {
const configured = localStorage.getItem('xz_tester_wakewordWsUrl');
if (configured && configured.trim()) {
return configured.trim();
}
return 'ws://127.0.0.1:8006/wakeword-ws';
}
function parseWakewordBridgeMessage(rawData) {
const message = JSON.parse(rawData);
return {
type: message.type || '',
requestId: message.requestId || null,
success: message.success !== false,
payload: message.payload || {},
error: message.error || null,
};
}
function settleWakewordRequest(message) {
const pendingRequest = pendingWakewordRequests.get(message.requestId);
if (!pendingRequest) {
return;
}
window.clearTimeout(pendingRequest.timer);
pendingWakewordRequests.delete(message.requestId);
if (message.success === false) {
pendingRequest.reject(new Error(message.error || '本地唤醒服务返回失败'));
return;
}
pendingRequest.resolve(message);
}
function rejectAllWakewordRequests(errorMessage) {
pendingWakewordRequests.forEach((pendingRequest) => {
window.clearTimeout(pendingRequest.timer);
pendingRequest.reject(new Error(errorMessage));
});
pendingWakewordRequests.clear();
}
@@ -0,0 +1,591 @@
// WebSocket消息处理模块
import { getConfig, saveConnectionUrls } from '../../config/manager.js?v=0205';
import { uiController } from '../../ui/controller.js?v=0205';
import { log } from '../../utils/logger.js?v=0205';
import { getAudioPlayer } from '../audio/player.js?v=0205';
import { getAudioRecorder } from '../audio/recorder.js?v=0205';
import { executeMcpTool, getMcpTools, setWebSocket as setMcpWebSocket } from '../mcp/tools.js?v=0205';
import { webSocketConnect } from './ota-connector.js?v=0205';
// WebSocket处理器类
export class WebSocketHandler {
constructor() {
this.websocket = null;
this.onConnectionStateChange = null;
this.onRecordButtonStateChange = null;
this.onSessionStateChange = null;
this.onSessionEmotionChange = null;
this.onChatMessage = null; // 新增:聊天消息回调
this.currentSessionId = null;
this.isRemoteSpeaking = false;
}
// 发送hello握手消息
async sendHelloMessage() {
if (!this.websocket || this.websocket.readyState !== WebSocket.OPEN) return false;
try {
const config = getConfig();
const helloMessage = {
type: 'hello',
device_id: config.deviceId,
device_name: config.deviceName,
device_mac: config.deviceMac,
token: config.token,
features: {
mcp: true,
emoji: config.emojiEnabled
}
};
log('发送hello握手消息', 'info');
this.websocket.send(JSON.stringify(helloMessage));
return new Promise(resolve => {
const timeout = setTimeout(() => {
log('等待hello响应超时', 'error');
log('提示: 请尝试点击"测试认证"按钮进行连接排查', 'info');
resolve(false);
}, 5000);
const onMessageHandler = (event) => {
try {
const response = JSON.parse(event.data);
if (response.type === 'hello' && response.session_id) {
log(`服务器握手成功,会话ID: ${response.session_id}`, 'success');
clearTimeout(timeout);
this.websocket.removeEventListener('message', onMessageHandler);
resolve(true);
}
} catch (e) {
// 忽略非JSON消息
}
};
this.websocket.addEventListener('message', onMessageHandler);
});
} catch (error) {
log(`发送hello消息错误: ${error.message}`, 'error');
return false;
}
}
_sendWakeupMessages(sessionId) {
if (!this.websocket || this.websocket.readyState !== WebSocket.OPEN) return;
// listen detect
this.websocket.send(JSON.stringify({
session_id: sessionId,
type: 'listen',
state: 'detect',
text: '嘿,你好呀'
}));
log('发送listen detect消息,唤醒词: 嘿,你好呀', 'info');
// listen start:开始监听
this.websocket.send(JSON.stringify({
session_id: sessionId,
type: 'listen',
state: 'start',
mode: 'auto'
}));
log('发送listen start消息', 'info');
}
// 处理文本消息
handleTextMessage(message) {
if (message.type === 'hello') {
log(`服务器回应:${JSON.stringify(message, null, 2)}`, 'success');
window.cameraAvailable = true;
log('连接成功,摄像头已可用', 'success');
uiController.updateDialButton(true);
this._sendWakeupMessages(message.session_id);
uiController.startAIChatSession();
} else if (message.type === 'tts') {
this.handleTTSMessage(message);
} else if (message.type === 'audio') {
log(`收到音频控制消息: ${JSON.stringify(message)}`, 'info');
} else if (message.type === 'stt') {
log(`识别结果: ${message.text}`, 'info');
// 检查是否需要绑定设备
if (message.text && (message.text.includes('绑定') || message.text.includes('bind'))) {
log('收到设备绑定提示,更新摄像头状态', 'warning');
window.cameraAvailable = false;
// 关闭摄像头
if (typeof window.stopCamera === 'function') {
window.stopCamera();
}
// 更新摄像头按钮状态
const cameraBtn = document.getElementById('cameraBtn');
if (cameraBtn) {
cameraBtn.classList.remove('camera-active');
cameraBtn.querySelector('.btn-text').textContent = '摄像头';
cameraBtn.disabled = true;
cameraBtn.title = '请先绑定验证码';
}
}
// 使用新的聊天消息回调显示STT消息
if (this.onChatMessage && message.text) {
this.onChatMessage(message.text, true);
}
} else if (message.type === 'llm') {
log(`大模型回复: ${message.text}`, 'info');
// 使用新的聊天消息回调显示LLM回复
if (this.onChatMessage && message.text) {
this.onChatMessage(message.text, false);
}
// 如果包含表情,更新sessionStatus表情并触发Live2D动作
if (message.text && /[\u{1F300}-\u{1F9FF}]|[\u{2600}-\u{26FF}]|[\u{2700}-\u{27BF}]/u.test(message.text)) {
// 提取表情符号
const emojiMatch = message.text.match(/[\u{1F300}-\u{1F9FF}]|[\u{2600}-\u{26FF}]|[\u{2700}-\u{27BF}]/u);
if (emojiMatch && this.onSessionEmotionChange) {
this.onSessionEmotionChange(emojiMatch[0]);
}
// 触发Live2D情绪动作
if (message.emotion) {
console.log(`收到情绪消息: emotion=${message.emotion}, text=${message.text}`);
this.triggerLive2DEmotionAction(message.emotion);
}
}
// 只有当文本不仅仅是表情时,才添加到对话中
// 移除文本中的表情后检查是否还有内容
const textWithoutEmoji = message.text ? message.text.replace(/[\u{1F300}-\u{1F9FF}]|[\u{2600}-\u{26FF}]|[\u{2700}-\u{27BF}]/gu, '').trim() : '';
if (textWithoutEmoji && this.onChatMessage) {
this.onChatMessage(message.text, false);
}
} else if (message.type === 'mcp') {
this.handleMCPMessage(message);
} else {
log(`未知消息类型: ${message.type}`, 'info');
if (this.onChatMessage) {
this.onChatMessage(`未知消息类型: ${message.type}\n${JSON.stringify(message, null, 2)}`, false);
}
}
}
// 处理TTS消息
handleTTSMessage(message) {
if (message.state === 'start') {
log('服务器开始发送语音', 'info');
this.currentSessionId = message.session_id;
this.isRemoteSpeaking = true;
if (this.onSessionStateChange) {
this.onSessionStateChange(true);
}
// 启动Live2D说话动画
this.startLive2DTalking();
} else if (message.state === 'sentence_start') {
log(`服务器发送语音段: ${message.text}`, 'info');
this.ttsSentenceCount = (this.ttsSentenceCount || 0) + 1;
if (message.text && this.onChatMessage) {
this.onChatMessage(message.text, false);
}
// 确保动画在句子开始时运行
const live2dManager = window.chatApp?.live2dManager;
if (live2dManager && !live2dManager.isTalking) {
this.startLive2DTalking();
}
} else if (message.state === 'sentence_end') {
log(`语音段结束: ${message.text}`, 'info');
// 句子结束时不清除动画,等待下一个句子或最终停止
} else if (message.state === 'stop') {
log('服务器语音传输结束,清空所有音频缓冲', 'info');
// 清空所有音频缓冲并停止播放
const audioPlayer = getAudioPlayer();
audioPlayer.clearAllAudio();
this.isRemoteSpeaking = false;
if (this.onRecordButtonStateChange) {
this.onRecordButtonStateChange(false);
}
if (this.onSessionStateChange) {
this.onSessionStateChange(false);
}
// 延迟停止Live2D说话动画,确保所有句子都播放完毕
setTimeout(() => {
this.stopLive2DTalking();
this.ttsSentenceCount = 0; // 重置计数器
}, 1000); // 1秒延迟,确保所有句子都完成
}
}
// 启动Live2D说话动画
startLive2DTalking() {
try {
// 获取Live2D管理器实例
const live2dManager = window.chatApp?.live2dManager;
if (live2dManager && live2dManager.live2dModel) {
// 使用音频播放器的分析器节点
live2dManager.startTalking();
log('Live2D说话动画已启动', 'info');
}
} catch (error) {
log(`启动Live2D说话动画失败: ${error.message}`, 'error');
}
}
// 停止Live2D说话动画
stopLive2DTalking() {
try {
const live2dManager = window.chatApp?.live2dManager;
if (live2dManager) {
live2dManager.stopTalking();
log('Live2D说话动画已停止', 'info');
}
} catch (error) {
log(`停止Live2D说话动画失败: ${error.message}`, 'error');
}
}
// 初始化Live2D音频分析器
initializeLive2DAudioAnalyzer() {
try {
const live2dManager = window.chatApp?.live2dManager;
if (live2dManager) {
// 初始化音频分析器(使用音频播放器的上下文)
if (live2dManager.initializeAudioAnalyzer()) {
log('Live2D音频分析器初始化完成,已连接到音频播放器', 'success');
} else {
log('Live2D音频分析器初始化失败,将使用模拟动画', 'warning');
}
}
} catch (error) {
log(`初始化Live2D音频分析器失败: ${error.message}`, 'error');
}
}
// 处理MCP消息
handleMCPMessage(message) {
const payload = message.payload || {};
log(`服务器下发: ${JSON.stringify(message)}`, 'info');
if (payload.method === 'tools/list') {
const tools = getMcpTools();
const replyMessage = JSON.stringify({
"session_id": message.session_id || "",
"type": "mcp",
"payload": {
"jsonrpc": "2.0",
"id": payload.id,
"result": {
"tools": tools
}
}
});
log(`客户端上报: ${replyMessage}`, 'info');
this.websocket.send(replyMessage);
log(`回复MCP工具列表: ${tools.length} 个工具`, 'info');
} else if (payload.method === 'tools/call') {
const toolName = payload.params?.name;
const toolArgs = payload.params?.arguments;
log(`调用工具: ${toolName} 参数: ${JSON.stringify(toolArgs)}`, 'info');
executeMcpTool(toolName, toolArgs).then(result => {
const replyMessage = JSON.stringify({
"session_id": message.session_id || "",
"type": "mcp",
"payload": {
"jsonrpc": "2.0",
"id": payload.id,
"result": {
"content": [
{
"type": "text",
"text": JSON.stringify(result)
}
],
"isError": false
}
}
});
log(`客户端上报: ${replyMessage}`, 'info');
this.websocket.send(replyMessage);
}).catch(error => {
log(`工具执行失败: ${error.message}`, 'error');
const errorReply = JSON.stringify({
"session_id": message.session_id || "",
"type": "mcp",
"payload": {
"jsonrpc": "2.0",
"id": payload.id,
"error": {
"code": -32603,
"message": error.message
}
}
});
this.websocket.send(errorReply);
});
} else if (payload.method === 'initialize') {
log(`收到工具初始化请求: ${JSON.stringify(payload.params)}`, 'info');
// 保存视觉分析接口地址
const visionUrl = document.getElementById('visionUrl');
const visionConfig = payload?.params?.capabilities?.vision;
if (visionConfig && typeof visionConfig === 'object' && visionConfig.url && visionConfig.token) {
const visionConfigStr = JSON.stringify(visionConfig);
localStorage.setItem('xz_tester_vision', visionConfigStr);
if (visionUrl) visionUrl.value = visionConfig.url;
} else {
localStorage.removeItem('xz_tester_vision');
if (visionUrl) visionUrl.value = '';
}
const replyMessage = JSON.stringify({
"session_id": message.session_id || "",
"type": "mcp",
"payload": {
"jsonrpc": "2.0",
"id": payload.id,
"result": {
"protocolVersion": "2024-11-05",
"capabilities": {
"tools": {}
},
"serverInfo": {
"name": "xiaozhi-web-test",
"version": "2.1.0"
}
}
}
});
log(`回复初始化响应`, 'info');
this.websocket.send(replyMessage);
} else {
log(`未知的MCP方法: ${payload.method}`, 'warning');
}
}
// 处理二进制消息
async handleBinaryMessage(data) {
try {
let arrayBuffer;
if (data instanceof ArrayBuffer) {
arrayBuffer = data;
} else if (data instanceof Blob) {
arrayBuffer = await data.arrayBuffer();
log(`收到Blob音频数据,大小: ${arrayBuffer.byteLength}字节`, 'debug');
} else {
log(`收到未知类型的二进制数据: ${typeof data}`, 'warning');
return;
}
const opusData = new Uint8Array(arrayBuffer);
const audioPlayer = getAudioPlayer();
audioPlayer.enqueueAudioData(opusData);
} catch (error) {
log(`处理二进制消息出错: ${error.message}`, 'error');
}
}
// 连接WebSocket服务器
async connect() {
const config = getConfig();
log('正在检查OTA状态...', 'info');
saveConnectionUrls();
try {
const otaUrl = document.getElementById('otaUrl').value.trim();
const ws = await webSocketConnect(otaUrl, config);
if (ws === undefined) {
return false;
}
this.websocket = ws;
// 设置接收二进制数据的类型为ArrayBuffer
this.websocket.binaryType = 'arraybuffer';
// 设置 MCP 模块的 WebSocket 实例
setMcpWebSocket(this.websocket);
// 设置录音器的WebSocket
const audioRecorder = getAudioRecorder();
audioRecorder.setWebSocket(this.websocket);
this.setupEventHandlers();
return true;
} catch (error) {
log(`连接错误: ${error.message}`, 'error');
if (this.onConnectionStateChange) {
this.onConnectionStateChange(false);
}
return false;
}
}
// 设置事件处理器
setupEventHandlers() {
this.websocket.onopen = async () => {
const url = document.getElementById('serverUrl').value;
log(`已连接到服务器: ${url}`, 'success');
if (this.onConnectionStateChange) {
this.onConnectionStateChange(true);
}
// 连接成功后,默认状态为聆听中
this.isRemoteSpeaking = false;
if (this.onSessionStateChange) {
this.onSessionStateChange(false);
}
// 在WebSocket连接成功时初始化Live2D音频分析器
this.initializeLive2DAudioAnalyzer();
await this.sendHelloMessage();
};
this.websocket.onclose = () => {
log('已断开连接', 'info');
if (this.onConnectionStateChange) {
this.onConnectionStateChange(false);
}
const audioRecorder = getAudioRecorder();
audioRecorder.stop();
// 关闭摄像头
if (typeof window.stopCamera === 'function') {
window.stopCamera();
}
// 隐藏摄像头显示区域
const cameraContainer = document.getElementById('cameraContainer');
if (cameraContainer) {
cameraContainer.classList.remove('active');
}
};
this.websocket.onerror = (error) => {
log(`WebSocket错误: ${error.message || '未知错误'}`, 'error');
uiController.addChatMessage(`⚠️ WebSocket错误: ${error.message || '未知错误'}`, false);
if (this.onConnectionStateChange) {
this.onConnectionStateChange(false);
}
};
this.websocket.onmessage = (event) => {
try {
if (typeof event.data === 'string') {
const message = JSON.parse(event.data);
this.handleTextMessage(message);
} else {
this.handleBinaryMessage(event.data);
}
} catch (error) {
log(`WebSocket消息处理错误: ${error.message}`, 'error');
// 不再使用旧的addMessage函数,因为conversationDiv元素不存在
// 错误消息将通过其他方式显示
}
};
}
// 断开连接
disconnect() {
if (!this.websocket) return;
this.websocket.close();
const audioRecorder = getAudioRecorder();
audioRecorder.stop();
// 关闭摄像头
if (typeof window.stopCamera === 'function') {
window.stopCamera();
}
// 隐藏摄像头显示区域
const cameraContainer = document.getElementById('cameraContainer');
if (cameraContainer) {
cameraContainer.classList.remove('active');
}
}
// 发送文本消息
sendTextMessage(text) {
if (text === '' || !this.websocket || this.websocket.readyState !== WebSocket.OPEN) {
return false;
}
try {
// 如果对方正在说话,先发送打断消息
if (this.isRemoteSpeaking && this.currentSessionId) {
const abortMessage = {
session_id: this.currentSessionId,
type: 'abort',
reason: 'wake_word_detected'
};
this.websocket.send(JSON.stringify(abortMessage));
log('发送打断消息', 'info');
}
const listenMessage = {
type: 'listen',
state: 'detect',
text: text
};
this.websocket.send(JSON.stringify(listenMessage));
log(`发送文本消息: ${text}`, 'info');
return true;
} catch (error) {
log(`发送消息错误: ${error.message}`, 'error');
return false;
}
}
/**
* 触发Live2D情绪动作
* @param {string} emotion - 情绪名称
*/
triggerLive2DEmotionAction(emotion) {
try {
const live2dManager = window.chatApp?.live2dManager;
if (live2dManager && typeof live2dManager.triggerEmotionAction === 'function') {
live2dManager.triggerEmotionAction(emotion);
log(`触发Live2D情绪动作: ${emotion}`, 'info');
} else {
log(`无法触发Live2D情绪动作: Live2D管理器未找到或方法不可用`, 'warning');
}
} catch (error) {
log(`触发Live2D情绪动作失败: ${error.message}`, 'error');
}
}
// 获取WebSocket实例
getWebSocket() {
return this.websocket;
}
// 检查是否已连接
isConnected() {
return this.websocket && this.websocket.readyState === WebSocket.OPEN;
}
}
// 创建单例
let wsHandlerInstance = null;
export function getWebSocketHandler() {
if (!wsHandlerInstance) {
wsHandlerInstance = new WebSocketHandler();
}
return wsHandlerInstance;
}
File diff suppressed because one or more lines are too long
+824
View File
@@ -0,0 +1,824 @@
/**
* Live2D 管理器
* 负责 Live2D 模型的初始化、嘴部动画控制等功能
*/
class Live2DManager {
constructor() {
this.live2dApp = null;
this.live2dModel = null;
this.isTalking = false;
this.mouthAnimationId = null;
this.mouthParam = 'ParamMouthOpenY';
this.audioContext = null;
this.analyser = null;
this.dataArray = null;
this.lastEmotionActionTime = null;
this.currentModelName = null;
// 模型特定配置
this.modelConfig = {
'hiyori_pro_zh': {
mouthParam: 'ParamMouthOpenY',
mouthAmplitude: 1.0,
mouthThresholds: { low: 0.3, high: 0.7 },
motionMap: {
'FlickUp': 'FlickUp',
'FlickDown': 'FlickDown',
'Tap': 'Tap',
'Tap@Body': 'Tap@Body',
'Flick': 'Flick',
'Flick@Body': 'Flick@Body'
}
},
'natori_pro_zh': {
mouthParam: 'ParamMouthOpenY',
mouthAmplitude: 1.0,
mouthThresholds: { low: 0.1, high: 0.4 },
mouthFormParam: 'ParamMouthForm',
mouthFormAmplitude: 1.0,
mouthForm2Param: 'ParamMouthForm2',
mouthForm2Amplitude: 0.8,
motionMap: {
'FlickUp': 'FlickUp',
'FlickDown': 'Flick@Body',
'Tap': 'Tap',
'Tap@Body': 'Tap@Head',
'Flick': 'Tap',
'Flick@Body': 'Flick@Body'
}
}
};
// 情绪到动作的映射
this.emotionToActionMap = {
'happy': 'FlickUp', // 开心-向上轻扫动作
'laughing': 'FlickUp', // 大笑-向上轻扫动作
'funny': 'FlickUp', // 搞笑-向上轻扫动作
'sad': 'FlickDown', // 伤心-向下轻扫动作
'crying': 'FlickDown', // 哭泣-向下轻扫动作
'angry': 'Tap@Body', // 生气-身体点击动作
'surprised': 'Tap', // 惊讶-点击动作
'neutral': 'Flick', // 平常-轻扫动作
'default': 'Flick@Body' // 默认-身体轻扫动作
};
// 单/双击判定配置与状态
this._lastClickTime = 0;
this._lastClickPos = { x: 0, y: 0 };
this._singleClickTimer = null;
this._doubleClickMs = 280; // 双击时间阈值(ms)
this._doubleClickDist = 16; // 双击允许的最大位移(px)
// 滑动判定
this._pointerDown = false;
this._downPos = { x: 0, y: 0 };
this._downTime = 0;
this._downArea = 'Body';
this._movedBeyondClick = false;
this._swipeMinDist = 24; // 触发滑动的最小距离
}
/**
* 初始化 Live2D
*/
async initializeLive2D() {
try {
const canvas = document.getElementById('live2d-stage');
// 供内部使用
window.PIXI = PIXI;
this.live2dApp = new PIXI.Application({
view: canvas,
height: window.innerHeight,
width: window.innerWidth,
resolution: window.devicePixelRatio,
autoDensity: true,
antialias: true,
backgroundAlpha: 0,
});
// 加载 Live2D 模型 - 动态检测当前目录,适配不同环境
// 获取当前HTML文件所在的目录路径
const currentPath = window.location.pathname;
const lastSlashIndex = currentPath.lastIndexOf('/');
const basePath = currentPath.substring(0, lastSlashIndex + 1);
// 从 localStorage 读取上次选择的模型,如果没有则使用默认
const savedModelName = localStorage.getItem('live2dModel') || 'hiyori_pro_zh';
const modelFileMap = {
'hiyori_pro_zh': 'hiyori_pro_t11.model3.json',
'natori_pro_zh': 'natori_pro_t06.model3.json'
};
const modelFileName = modelFileMap[savedModelName] || 'hiyori_pro_t11.model3.json';
const modelPath = basePath + 'resources/' + savedModelName + '/runtime/' + modelFileName;
this.live2dModel = await PIXI.live2d.Live2DModel.from(modelPath);
this.live2dApp.stage.addChild(this.live2dModel);
// 保存当前模型名称
this.currentModelName = savedModelName;
// 更新下拉框显示
const modelSelect = document.getElementById('live2dModelSelect');
if (modelSelect) {
modelSelect.value = savedModelName;
}
// 设置模型特定的嘴部参数名
if (this.modelConfig[savedModelName]) {
this.mouthParam = this.modelConfig[savedModelName].mouthParam || 'ParamMouthOpenY';
}
// 设置模型属性
this.live2dModel.scale.set(0.33);
this.live2dModel.x = (window.innerWidth - this.live2dModel.width) * 0.5;
this.live2dModel.y = -50;
// 启用交互并监听点击命中(头部/身体等)
this.live2dModel.interactive = true;
this.live2dModel.on('doublehit', (args) => {
const area = Array.isArray(args) ? args[0] : args;
// 触发双击动作
if (area === 'Body') {
this.motion('Flick@Body');
} else if (area === 'Head' || area === 'Face') {
this.motion('Flick');
}
const app = window.chatApp;
const payload = JSON.stringify({ type: 'live2d', event: 'doublehit', area });
if (app && app.dataChannel && app.dataChannel.readyState === 'open') {
app.dataChannel.send(payload);
}
});
this.live2dModel.on('singlehit', (args) => {
const area = Array.isArray(args) ? args[0] : args;
// 触发单击动作
if (area === 'Body') {
this.motion('Tap@Body');
} else if (area === 'Head' || area === 'Face') {
this.motion('Tap');
}
const app = window.chatApp;
const payload = JSON.stringify({ type: 'live2d', event: 'singlehit', area });
if (app && app.dataChannel && app.dataChannel.readyState === 'open') {
app.dataChannel.send(payload);
}
});
this.live2dModel.on('swipe', (args) => {
const area = Array.isArray(args) ? args[0] : args;
const dir = Array.isArray(args) ? args[1] : undefined;
// 触发滑动动作
if (area === 'Body') {
if (dir === 'up') {
this.motion('FlickUp');
} else if (dir === 'down') {
this.motion('FlickDown');
}
} else if (area === 'Head' || area === 'Face') {
if (dir === 'up') {
this.motion('FlickUp');
} else if (dir === 'down') {
this.motion('FlickDown');
}
}
const app = window.chatApp;
const payload = JSON.stringify({ type: 'live2d', event: 'swipe', area, dir });
if (app && app.dataChannel && app.dataChannel.readyState === 'open') {
app.dataChannel.send(payload);
}
});
// 兜底:自定义"头部/身体"命中区域 + 单/双击/滑动区分
this.live2dModel.on('pointerdown', (event) => {
try {
const global = event.data.global;
const bounds = this.live2dModel.getBounds();
// 仅在点击落在模型可见范围内时判定
if (!bounds || !bounds.contains(global.x, global.y)) return;
const relX = (global.x - bounds.x) / (bounds.width || 1);
const relY = (global.y - bounds.y) / (bounds.height || 1);
let area = '';
// 经验阈值:模型可见矩形的上部 20% 视为"头部"区域
if (relX >= 0.4 && relX <= 0.6) {
if (relY <= 0.15) {
area = 'Head';
} else if (relY <= 0.23) {
area = 'Face';
} else {
area = 'Body';
}
}
if (area === '') {
return;
}
// 记录按下状态用于滑动判定
this._pointerDown = true;
this._downPos = { x: global.x, y: global.y };
this._downTime = performance.now();
this._downArea = area;
this._movedBeyondClick = false;
const now = performance.now();
const dt = now - (this._lastClickTime || 0);
const dx = global.x - (this._lastClickPos?.x || 0);
const dy = global.y - (this._lastClickPos?.y || 0);
const dist = Math.hypot(dx, dy);
// 命中确认:仅当点击在模型上时做单/双击判断
if (this._lastClickTime && dt <= this._doubleClickMs && dist <= this._doubleClickDist) {
// 判定为双击:取消待触发的单击事件
if (this._singleClickTimer) {
clearTimeout(this._singleClickTimer);
this._singleClickTimer = null;
}
if (typeof this.live2dModel.emit === 'function') {
this.live2dModel.emit('doublehit', [area]);
}
this._lastClickTime = 0;
this._pointerDown = false; // 双击完成,重置状态
return;
}
// 可能是单击:记录并延迟确认
this._lastClickTime = now;
this._lastClickPos = { x: global.x, y: global.y };
if (this._singleClickTimer) {
clearTimeout(this._singleClickTimer);
this._singleClickTimer = null;
}
this._singleClickTimer = setTimeout(() => {
// 若在等待期间发生了移动超过阈值,则不再当作单击
if (!this._movedBeyondClick && typeof this.live2dModel.emit === 'function') {
this.live2dModel.emit('singlehit', [area]);
}
this._singleClickTimer = null;
this._lastClickTime = 0;
}, this._doubleClickMs);
} catch (e) {
// 忽略自定义命中判断中的异常,避免影响主流程
}
});
// 指针移动:用于判定是否从"点击"升级为"滑动"
this.live2dModel.on('pointermove', (event) => {
try {
if (!this._pointerDown) return;
const global = event.data.global;
const dx = global.x - this._downPos.x;
const dy = global.y - this._downPos.y;
const dist = Math.hypot(dx, dy);
// 使用 _doubleClickDist 作为点击/滑动的判定阈值
if (dist > this._doubleClickDist) {
this._movedBeyondClick = true;
// 若已超出点击阈值,取消可能的单击触发
if (this._singleClickTimer) {
clearTimeout(this._singleClickTimer);
this._singleClickTimer = null;
}
this._lastClickTime = 0;
}
} catch (e) {
// 忽略移动判定中的异常
}
});
// 指针抬起:确认是否为滑动
const handlePointerUp = (event) => {
try {
if (!this._pointerDown) return;
const global = (event && event.data && event.data.global) ? event.data.global : { x: this._downPos.x, y: this._downPos.y };
const dx = global.x - this._downPos.x;
const dy = global.y - this._downPos.y;
const dist = Math.hypot(dx, dy);
// 滑动:超过滑动最小距离则触发 swipe 事件(携带方向与区域)
if (this._movedBeyondClick && dist >= this._swipeMinDist) {
if (typeof this.live2dModel.emit === 'function') {
const dir = Math.abs(dx) >= Math.abs(dy)
? (dx > 0 ? 'right' : 'left')
: (dy > 0 ? 'down' : 'up');
this.live2dModel.emit('swipe', [this._downArea, dir]);
}
// 终止:不再让单击/双击触发
if (this._singleClickTimer) {
clearTimeout(this._singleClickTimer);
this._singleClickTimer = null;
}
this._lastClickTime = 0;
}
} catch (e) {
// 忽略抬起判定中的异常
}
finally {
this._pointerDown = false;
this._movedBeyondClick = false;
}
};
this.live2dModel.on('pointerup', handlePointerUp);
this.live2dModel.on('pointerupoutside', handlePointerUp);
// 添加窗口大小变化监听器,保持模型在Canvas中间和底部
window.addEventListener('resize', () => {
if (this.live2dModel) {
// 使用窗口实际尺寸重新计算模型位置
this.live2dModel.x = (window.innerWidth - this.live2dModel.width) * 0.5;
this.live2dModel.y = -50;
}
});
} catch (err) {
console.error('加载 Live2D 模型失败:', err);
}
}
/**
* 初始化音频分析器 - 使用音频播放器的分析器节点
*/
initializeAudioAnalyzer() {
try {
// 获取音频播放器实例
const audioPlayer = window.chatApp?.audioPlayer;
if (!audioPlayer) {
console.warn('音频播放器未初始化,无法获取分析器节点');
return false;
}
// 获取音频播放器的音频上下文
this.audioContext = audioPlayer.getAudioContext();
if (!this.audioContext) {
console.warn('无法获取音频播放器的音频上下文');
return false;
}
// 创建分析器节点
this.analyser = this.audioContext.createAnalyser();
this.analyser.fftSize = 256;
this.dataArray = new Uint8Array(this.analyser.frequencyBinCount);
return true;
} catch (error) {
console.error('初始化音频分析器失败:', error);
return false;
}
}
/**
* 连接到音频播放器的输出节点
*/
connectToAudioPlayer() {
try {
// 获取音频播放器的流上下文
const audioPlayer = window.chatApp?.audioPlayer;
if (!audioPlayer || !audioPlayer.streamingContext) {
console.warn('音频播放器或流上下文未初始化');
return false;
}
// 获取音频播放器的流上下文
const streamingContext = audioPlayer.streamingContext;
// 获取分析器节点
const analyser = streamingContext.getAnalyser();
if (!analyser) {
console.warn('音频播放器尚未创建分析器节点,无法连接');
return false;
}
// 使用音频播放器的分析器节点
this.analyser = analyser;
this.dataArray = new Uint8Array(this.analyser.frequencyBinCount);
return true;
} catch (error) {
console.error('连接到音频播放器失败:', error);
return false;
}
}
/**
* 嘴部动画循环
*/
animateMouth() {
if (!this.isTalking) return;
if (!this.live2dModel) return;
const internal = this.live2dModel && this.live2dModel.internalModel;
if (internal && internal.coreModel) {
const coreModel = internal.coreModel;
let mouthOpenY = 0;
let mouthForm = 0;
let mouthForm2 = 0;
let average = 0;
if (this.analyser && this.dataArray) {
this.analyser.getByteFrequencyData(this.dataArray);
average = this.dataArray.reduce((a, b) => a + b) / this.dataArray.length;
const normalizedVolume = average / 255;
// 获取模型特定的阈值
let lowThreshold = 0.3;
let highThreshold = 0.7;
if (this.currentModelName && this.modelConfig[this.currentModelName]) {
lowThreshold = this.modelConfig[this.currentModelName].mouthThresholds?.low || 0.3;
highThreshold = this.modelConfig[this.currentModelName].mouthThresholds?.high || 0.7;
}
// 使用模型特定的阈值进行映射
let minOpenY = 0.1;
if (this.currentModelName && this.modelConfig[this.currentModelName]) {
minOpenY = this.modelConfig[this.currentModelName].mouthMinOpenY || 0.1;
}
if (normalizedVolume < lowThreshold) {
mouthOpenY = minOpenY + Math.pow(normalizedVolume / lowThreshold, 1.5) * (0.4 - minOpenY);
} else if (normalizedVolume < highThreshold) {
mouthOpenY = 0.4 + (normalizedVolume - lowThreshold) / (highThreshold - lowThreshold) * 0.4;
} else {
mouthOpenY = 0.8 + Math.pow((normalizedVolume - highThreshold) / (1 - highThreshold), 1.2) * 0.2;
}
// 应用模型特定的嘴部开合幅度
let amplitudeMultiplier = 1.0;
let maxOpenY = 2.5;
if (this.currentModelName && this.modelConfig[this.currentModelName]) {
amplitudeMultiplier = this.modelConfig[this.currentModelName].mouthAmplitude;
maxOpenY = this.modelConfig[this.currentModelName].maxOpenY || 2.5;
}
mouthOpenY = mouthOpenY * amplitudeMultiplier;
mouthOpenY = Math.min(Math.max(mouthOpenY, 0), maxOpenY);
// 计算嘴型参数(仅对支持嘴型变化的模型)
if (this.currentModelName && this.modelConfig[this.currentModelName]?.mouthFormParam) {
const config = this.modelConfig[this.currentModelName];
const formAmplitude = config.mouthFormAmplitude || 0.5;
const form2Amplitude = config.mouthForm2Amplitude || 0;
// 嘴型随音量变化:
// 低音量:嘴型偏"一"字(负值)
// 高音量:嘴型偏"o"字(正值)
// 音量=0时:嘴型=0(自然状态)
mouthForm = (normalizedVolume - 0.5) * 2 * formAmplitude;
mouthForm = Math.max(-formAmplitude, Math.min(formAmplitude, mouthForm));
// 第二嘴型参数(natori特有)
if (config.mouthForm2Param) {
mouthForm2 = (normalizedVolume - 0.3) * 2 * form2Amplitude;
mouthForm2 = Math.max(-form2Amplitude, Math.min(form2Amplitude, mouthForm2));
}
}
// 调试日志:输出嘴部参数
console.log(`[Live2D] 模型: ${this.currentModelName || 'unknown'}, 音量: ${average?.toFixed(0)}, OpenY: ${mouthOpenY.toFixed(3)}, Form: ${mouthForm.toFixed(3)}, Form2: ${mouthForm2.toFixed(3)}`);
}
// 设置嘴部开合参数
coreModel.setParameterValueById(this.mouthParam, mouthOpenY);
// 设置嘴型参数(仅对支持嘴型变化的模型)
if (this.currentModelName && this.modelConfig[this.currentModelName]?.mouthFormParam) {
const config = this.modelConfig[this.currentModelName];
const formParam = config.mouthFormParam;
coreModel.setParameterValueById(formParam, mouthForm);
// 设置第二嘴型参数(natori特有)
if (config.mouthForm2Param) {
coreModel.setParameterValueById(config.mouthForm2Param, mouthForm2);
}
}
coreModel.update();
}
this.mouthAnimationId = requestAnimationFrame(() => this.animateMouth());
}
/**
* 开始说话动画
*/
startTalking() {
if (this.isTalking || !this.live2dModel) return;
// 确保音频分析器已初始化
if (!this.analyser) {
if (!this.initializeAudioAnalyzer()) {
console.warn('音频分析器初始化失败,将使用模拟动画');
// 即使分析器初始化失败,也启动动画(使用模拟数据)
this.isTalking = true;
this.animateMouth();
return;
}
}
// 连接到音频播放器输出
if (!this.connectToAudioPlayer()) {
console.warn('无法连接到音频播放器输出,将使用模拟动画');
}
this.isTalking = true;
this.animateMouth();
}
/**
* 停止说话动画
*/
stopTalking() {
this.isTalking = false;
if (this.mouthAnimationId) {
cancelAnimationFrame(this.mouthAnimationId);
this.mouthAnimationId = null;
}
// 重置嘴部参数
if (this.live2dModel) {
const internal = this.live2dModel.internalModel;
if (internal && internal.coreModel) {
const coreModel = internal.coreModel;
coreModel.setParameterValueById(this.mouthParam, 0);
coreModel.update();
}
}
}
/**
* 基于情绪触发动作
* @param {string} emotion - 情绪名称
*/
triggerEmotionAction(emotion) {
if (!this.live2dModel) return;
// 添加冷却时间控制,避免过于频繁触发
const now = Date.now();
if (this.lastEmotionActionTime && now - this.lastEmotionActionTime < 5000) { // 5秒冷却时间
return;
}
// 根据情绪获取对应的动作
const action = this.emotionToActionMap[emotion] || this.emotionToActionMap['default'];
// 触发动作并记录时间
this.motion(action);
this.lastEmotionActionTime = now;
}
/**
* 触发模型动作(Motion)
* @param {string} name - 动作分组名称,如 'TapBody'、'FlickUp'、'Idle' 等
*/
motion(name) {
try {
if (!this.live2dModel) return;
// 根据当前模型获取对应的动作名称
let actualMotionName = name;
if (this.currentModelName && this.modelConfig[this.currentModelName]) {
const motionMap = this.modelConfig[this.currentModelName].motionMap;
actualMotionName = motionMap[name] || name;
}
this.live2dModel.motion(actualMotionName);
} catch (error) {
console.error('触发动作失败:', error);
}
}
/**
* 设置模型交互事件
*/
setupModelInteractions() {
if (!this.live2dModel) return;
this.live2dModel.interactive = true;
this.live2dModel.on('doublehit', (args) => {
const area = Array.isArray(args) ? args[0] : args;
if (area === 'Body') {
this.motion('Flick@Body');
} else if (area === 'Head' || area === 'Face') {
this.motion('Flick');
}
const app = window.chatApp;
const payload = JSON.stringify({ type: 'live2d', event: 'doublehit', area });
if (app && app.dataChannel && app.dataChannel.readyState === 'open') {
app.dataChannel.send(payload);
}
});
this.live2dModel.on('singlehit', (args) => {
const area = Array.isArray(args) ? args[0] : args;
if (area === 'Body') {
this.motion('Tap@Body');
} else if (area === 'Head' || area === 'Face') {
this.motion('Tap');
}
const app = window.chatApp;
const payload = JSON.stringify({ type: 'live2d', event: 'singlehit', area });
if (app && app.dataChannel && app.dataChannel.readyState === 'open') {
app.dataChannel.send(payload);
}
});
this.live2dModel.on('swipe', (args) => {
const area = Array.isArray(args) ? args[0] : args;
const dir = Array.isArray(args) ? args[1] : undefined;
if (area === 'Body') {
if (dir === 'up') {
this.motion('FlickUp');
} else if (dir === 'down') {
this.motion('FlickDown');
}
}
const app = window.chatApp;
const payload = JSON.stringify({ type: 'live2d', event: 'swipe', area, dir });
if (app && app.dataChannel && app.dataChannel.readyState === 'open') {
app.dataChannel.send(payload);
}
});
this.live2dModel.on('pointerdown', (event) => {
try {
const global = event.data.global;
const bounds = this.live2dModel.getBounds();
if (!bounds || !bounds.contains(global.x, global.y)) return;
const relX = (global.x - bounds.x) / (bounds.width || 1);
const relY = (global.y - bounds.y) / (bounds.height || 1);
let area = '';
if (relX >= 0.4 && relX <= 0.6) {
if (relY <= 0.15) {
area = 'Head';
} else if (relY >= 0.7) {
area = 'Body';
}
}
if (!area) return;
const now = Date.now();
const dt = now - (this._lastClickTime || 0);
const dx = global.x - (this._lastClickPos?.x || 0);
const dy = global.y - (this._lastClickPos?.y || 0);
const dist = Math.hypot(dx, dy);
if (this._lastClickTime && dt <= this._doubleClickMs && dist <= this._doubleClickDist) {
if (this._singleClickTimer) {
clearTimeout(this._singleClickTimer);
this._singleClickTimer = null;
}
this.live2dModel.emit('doublehit', area);
this._lastClickTime = null;
this._lastClickPos = null;
} else {
this._lastClickTime = now;
this._lastClickPos = { x: global.x, y: global.y };
this._singleClickTimer = setTimeout(() => {
this._singleClickTimer = null;
this.live2dModel.emit('singlehit', area);
}, this._doubleClickMs);
}
} catch (e) {
console.warn('pointerdown 处理出错:', e);
}
});
}
/**
* 清理资源
*/
destroy() {
this.stopTalking();
// 清理音频分析器
if (this.audioContext) {
this.audioContext.close();
this.audioContext = null;
}
this.analyser = null;
this.dataArray = null;
// 清理 Live2D 应用
if (this.live2dApp) {
this.live2dApp.destroy(true);
this.live2dApp = null;
}
this.live2dModel = null;
}
/**
* 切换 Live2D 模型
* @param {string} modelName - 模型目录名称,如 'hiyori_pro_zh'、'natori_pro_zh'
* @returns {Promise<boolean>} - 切换是否成功
*/
async switchModel(modelName) {
try {
// 获取模型文件名映射
const modelFileMap = {
'hiyori_pro_zh': 'hiyori_pro_t11.model3.json',
'natori_pro_zh': 'natori_pro_t06.model3.json',
'chitose': 'chitose.model3.json',
'haru_greeter_pro_jp': 'haru_greeter_t05.model3.json'
};
const modelFileName = modelFileMap[modelName];
if (!modelFileName) {
console.error('未知的模型名称:', modelName);
return false;
}
// 获取基础路径
const currentPath = window.location.pathname;
const lastSlashIndex = currentPath.lastIndexOf('/');
const basePath = currentPath.substring(0, lastSlashIndex + 1);
const modelPath = basePath + 'resources/' + modelName + '/runtime/' + modelFileName;
// 如果已存在模型,先移除
if (this.live2dModel) {
this.live2dApp.stage.removeChild(this.live2dModel);
this.live2dModel.destroy();
this.live2dModel = null;
}
// 显示加载状态
const app = window.chatApp;
if (app) {
app.setModelLoadingStatus(true);
}
// 加载新模型
this.live2dModel = await PIXI.live2d.Live2DModel.from(modelPath);
this.live2dApp.stage.addChild(this.live2dModel);
// 设置模型属性
this.live2dModel.scale.set(0.33);
this.live2dModel.x = (window.innerWidth - this.live2dModel.width) * 0.5;
this.live2dModel.y = -50;
// 重新绑定交互事件
this.setupModelInteractions();
// 隐藏加载状态
if (app) {
app.setModelLoadingStatus(false);
}
// 保存当前模型名称
this.currentModelName = modelName;
// 设置模型特定的嘴部参数名
if (this.modelConfig[modelName]) {
this.mouthParam = this.modelConfig[modelName].mouthParam || 'ParamMouthOpenY';
}
// 保存到 localStorage
localStorage.setItem('live2dModel', modelName);
// 更新下拉框显示
const modelSelect = document.getElementById('live2dModelSelect');
if (modelSelect) {
modelSelect.value = modelName;
}
console.log('模型切换成功:', modelName);
return true;
} catch (error) {
console.error('切换模型失败:', error);
const app = window.chatApp;
if (app) {
app.setModelLoadingStatus(false);
}
return false;
}
}
}
// 导出全局实例
window.Live2DManager = Live2DManager;
File diff suppressed because one or more lines are too long
File diff suppressed because it is too large Load Diff
@@ -0,0 +1,27 @@
// 背景图加载检测
(function() {
const backgroundContainer = document.getElementById('backgroundContainer');
// 提取背景图片URL
let bgImageUrl = window.getComputedStyle(backgroundContainer).backgroundImage;
const urlMatch = bgImageUrl && bgImageUrl.match(/url\(["']?(.*?)["']?\)/);
if (!urlMatch || !urlMatch[1]) {
console.warn('未提取到有效的背景图片URL');
return;
}
bgImageUrl = urlMatch[1];
const bgImage = new Image();
bgImage.onerror = function() {
console.error('背景图片加载失败:', bgImageUrl);
};
// 加载成功显示模型加载
bgImage.onload = function() {
modelLoading.style.display = 'flex';
};
bgImage.src = bgImageUrl;
})();
+929
View File
@@ -0,0 +1,929 @@
// UI controller module
import { loadConfig, saveConfig } from '../config/manager.js?v=0205';
import { getAudioPlayer } from '../core/audio/player.js?v=0205';
import { getAudioRecorder } from '../core/audio/recorder.js?v=0205';
import { requestWakewordBridge, stopWakewordBridgeListener, startWakewordBridgeListener, getWakewordBridgeUrl, onNextBridgeConnected } from '../core/network/wakeword-bridge.js?v=0205';
import { getWebSocketHandler } from '../core/network/websocket.js?v=0205';
import { log } from '../utils/logger.js?v=0205';
// UI controller class
class UIController {
constructor() {
this.isEditing = false;
this.visualizerCanvas = null;
this.visualizerContext = null;
this.audioStatsTimer = null;
this.currentBackgroundIndex = localStorage.getItem('backgroundIndex') ? parseInt(localStorage.getItem('backgroundIndex')) : 0;
this.backgroundImages = ['1.png', '2.png', '3.png'];
this.dialBtnDisabled = false;
this.isConnecting = false;
this.lastWakewordDialTime = 0;
// Bind methods
this.init = this.init.bind(this);
this.initEventListeners = this.initEventListeners.bind(this);
this.updateDialButton = this.updateDialButton.bind(this);
this.addChatMessage = this.addChatMessage.bind(this);
this.switchBackground = this.switchBackground.bind(this);
this.switchLive2DModel = this.switchLive2DModel.bind(this);
this.showModal = this.showModal.bind(this);
this.hideModal = this.hideModal.bind(this);
this.switchTab = this.switchTab.bind(this);
this.applyWakewordConfig = this.applyWakewordConfig.bind(this);
this.handleApplyWakeword = this.handleApplyWakeword.bind(this);
this.triggerWakewordDial = this.triggerWakewordDial.bind(this);
}
// Initialize
init() {
console.log('UIController init started');
this.visualizerCanvas = document.getElementById('audioVisualizer');
if (this.visualizerCanvas) {
this.visualizerContext = this.visualizerCanvas.getContext('2d');
this.initVisualizer();
}
// Check if connect button exists during initialization
const connectBtn = document.getElementById('connectBtn');
console.log('connectBtn during init:', connectBtn);
this.initEventListeners();
this.startAudioStatsMonitor();
loadConfig();
// Register recording callback
const audioRecorder = getAudioRecorder();
audioRecorder.onRecordingStart = (seconds) => {
this.updateRecordButtonState(true, seconds);
};
// Initialize status display
this.updateConnectionUI(false);
// Apply saved background
const backgroundContainer = document.querySelector('.background-container');
if (backgroundContainer) {
backgroundContainer.style.backgroundImage = `url('./images/${this.backgroundImages[this.currentBackgroundIndex]}')`;
}
this.updateDialButton(false);
console.log('UIController init completed');
}
// Initialize visualizer
initVisualizer() {
if (this.visualizerCanvas) {
this.visualizerCanvas.width = this.visualizerCanvas.clientWidth;
this.visualizerCanvas.height = this.visualizerCanvas.clientHeight;
this.visualizerContext.fillStyle = '#fafafa';
this.visualizerContext.fillRect(0, 0, this.visualizerCanvas.width, this.visualizerCanvas.height);
}
}
// Initialize event listeners
initEventListeners() {
// Settings button
const settingsBtn = document.getElementById('settingsBtn');
if (settingsBtn) {
settingsBtn.addEventListener('click', () => {
this.showModal('settingsModal');
});
}
// Background switch button
const backgroundBtn = document.getElementById('backgroundBtn');
if (backgroundBtn) {
backgroundBtn.addEventListener('click', this.switchBackground);
}
// Model select change event
const modelSelect = document.getElementById('live2dModelSelect');
if (modelSelect) {
modelSelect.addEventListener('change', () => {
this.switchLive2DModel();
});
}
// Camera switch button
const cameraSwitch = document.getElementById('cameraSwitch');
const cameraSwitchMask = document.getElementById('cameraSwitchMask');
if (cameraSwitchMask) {
cameraSwitchMask.addEventListener('click', () => {
const isCameraActive = cameraSwitch.classList.contains('active');
if (isCameraActive) {
window.switchCamera();
}
})
}
// Dial button
const dialBtn = document.getElementById('dialBtn');
if (dialBtn) {
dialBtn.addEventListener('click', () => {
dialBtn.disabled = true;
this.dialBtnDisabled = true;
setTimeout(() => {
dialBtn.disabled = false;
this.dialBtnDisabled = false;
}, 3000);
const wsHandler = getWebSocketHandler();
const isConnected = wsHandler.isConnected();
if (isConnected) {
wsHandler.disconnect();
this.updateDialButton(false);
if (cameraSwitch) cameraSwitch.classList.remove('active');
this.addChatMessage('Disconnected, see you next time~😊', false);
} else {
// Check if OTA URL is filled
const otaUrlInput = document.getElementById('otaUrl');
if (!otaUrlInput || !otaUrlInput.value.trim()) {
// If OTA URL is not filled, show settings modal and switch to device tab
this.showModal('settingsModal');
this.switchTab('device');
this.addChatMessage('Please fill in OTA server URL', false);
return;
}
// Start connection process
this.handleConnect();
}
});
}
// Camera button
const cameraBtn = document.getElementById('cameraBtn');
let cameraTimer = null;
if (cameraBtn) {
cameraBtn.addEventListener('click', () => {
if (cameraTimer) {
clearTimeout(cameraTimer);
cameraTimer = null;
}
cameraTimer = setTimeout(() => {
const cameraContainer = document.getElementById('cameraContainer');
if (!cameraContainer) {
log('摄像头容器不存在', 'warning');
return;
}
const isActive = cameraContainer.classList.contains('active');
if (isActive) {
// 关闭摄像头
if (typeof window.stopCamera === 'function') {
if (cameraSwitch) cameraSwitch.classList.remove('active');
window.stopCamera();
}
cameraContainer.classList.remove('active');
cameraBtn.classList.remove('camera-active');
cameraBtn.querySelector('.btn-text').textContent = '摄像头';
log('摄像头已关闭', 'info');
} else {
// 打开摄像头
if (typeof window.startCamera === 'function') {
window.startCamera().then(success => {
if (success) {
cameraBtn.classList.add('camera-active');
cameraBtn.querySelector('.btn-text').textContent = '关闭';
} else {
this.addChatMessage('⚠️ 摄像头启动失败,请检查浏览器权限', false);
}
}).catch(error => {
log(`启动摄像头异常: ${error.message}`, 'error');
});
} else {
log('startCamera函数未定义', 'warning');
}
}
}, 300);
});
}
// Record button
const recordBtn = document.getElementById('recordBtn');
if (recordBtn) {
let recordTimer = null;
recordBtn.addEventListener('click', () => {
if (recordTimer) {
clearTimeout(recordTimer);
recordTimer = null;
}
recordTimer = setTimeout(() => {
const audioRecorder = getAudioRecorder();
if (audioRecorder.isRecording) {
audioRecorder.stop();
// Restore record button to normal state
recordBtn.classList.remove('recording');
recordBtn.querySelector('.btn-text').textContent = '录音';
} else {
// Update button state to recording
recordBtn.classList.add('recording');
recordBtn.querySelector('.btn-text').textContent = '录音中';
// Start recording, update button state after delay
setTimeout(() => {
audioRecorder.start();
}, 100);
}
}, 300);
});
}
// Chat input event listener
const chatIpt = document.getElementById('chatIpt');
if (chatIpt) {
const wsHandler = getWebSocketHandler();
chatIpt.addEventListener('keydown', (e) => {
if (e.key === 'Enter') {
if (e.target.value) {
wsHandler.sendTextMessage(e.target.value);
e.target.value = '';
return;
}
}
});
}
// Close button
const closeButtons = document.querySelectorAll('.close-btn');
closeButtons.forEach(btn => {
btn.addEventListener('click', (e) => {
e.stopPropagation();
const modal = e.target.closest('.modal');
if (modal) {
if (modal.id === 'settingsModal') {
saveConfig();
}
this.hideModal(modal.id);
}
});
});
// Settings tab switch
const tabBtns = document.querySelectorAll('.tab-btn');
tabBtns.forEach(btn => {
btn.addEventListener('click', (e) => {
this.switchTab(e.target.dataset.tab);
});
});
const applyWakewordBtn = document.getElementById('applyWakewordBtn');
if (applyWakewordBtn) {
applyWakewordBtn.addEventListener('click', this.handleApplyWakeword);
}
// 点击模态框背景关闭(仅对特定模态框禁用此功能)
const modals = document.querySelectorAll('.modal');
modals.forEach(modal => {
modal.addEventListener('click', (e) => {
if (e.target === modal) {
// settingsModal、mcpToolModal、mcpPropertyModal 只能通过点击X关闭
const nonClosableModals = ['settingsModal', 'mcpToolModal', 'mcpPropertyModal'];
if (nonClosableModals.includes(modal.id)) {
return; // 禁止点击背景关闭
}
this.hideModal(modal.id);
}
});
});
// Add MCP tool button
const addMCPToolBtn = document.getElementById('addMCPToolBtn');
if (addMCPToolBtn) {
addMCPToolBtn.addEventListener('click', (e) => {
e.stopPropagation();
this.addMCPTool();
});
}
// Connect button and send button are not removed, can be added to dial button later
}
// Update connection status UI
updateConnectionUI(isConnected) {
const connectionStatus = document.getElementById('connectionStatus');
const statusDot = document.querySelector('.status-dot');
if (connectionStatus) {
if (isConnected) {
connectionStatus.textContent = '已连接';
if (statusDot) {
statusDot.className = 'status-dot status-connected';
}
} else {
connectionStatus.textContent = '离线';
if (statusDot) {
statusDot.className = 'status-dot status-disconnected';
}
}
}
}
// Update dial button state
updateDialButton(isConnected) {
const dialBtn = document.getElementById('dialBtn');
const recordBtn = document.getElementById('recordBtn');
const cameraBtn = document.getElementById('cameraBtn');
if (dialBtn) {
if (isConnected) {
dialBtn.classList.add('dial-active');
dialBtn.querySelector('.btn-text').textContent = '挂断';
// Update dial button icon to hang up icon
dialBtn.querySelector('svg').innerHTML = `
<path d="M12,9C10.4,9 9,10.4 9,12C9,13.6 10.4,15 12,15C13.6,15 15,13.6 15,12C15,10.4 13.6,9 12,9M12,17C9.2,17 7,14.8 7,12C7,9.2 9.2,7 12,7C14.8,7 17,9.2 17,12C17,14.8 14.8,17 12,17M12,4.5C7,4.5 2.7,7.6 1,12C2.7,16.4 7,19.5 12,19.5C17,19.5 21.3,16.4 23,12C21.3,7.6 17,4.5 12,4.5Z"/>
`;
} else {
dialBtn.classList.remove('dial-active');
dialBtn.querySelector('.btn-text').textContent = '拨号';
// Restore dial button icon
dialBtn.querySelector('svg').innerHTML = `
<path d="M6.62,10.79C8.06,13.62 10.38,15.94 13.21,17.38L15.41,15.18C15.69,14.9 16.08,14.82 16.43,14.93C17.55,15.3 18.75,15.5 20,15.5A1,1 0 0,1 21,16.5V20A1,1 0 0,1 20,21A17,17 0 0,1 3,4A1,1 0 0,1 4,3H7.5A1,1 0 0,1 8.5,4C8.5,5.25 8.7,6.45 9.07,7.57C9.18,7.92 9.1,8.31 8.82,8.59L6.62,10.79Z"/>
`;
}
}
// Update camera button state - reset to default when disconnected
if (cameraBtn && !isConnected) {
const cameraContainer = document.getElementById('cameraContainer');
if (cameraContainer && cameraContainer.classList.contains('active')) {
cameraContainer.classList.remove('active');
}
cameraBtn.classList.remove('camera-active');
cameraBtn.querySelector('.btn-text').textContent = '摄像头';
cameraBtn.disabled = true;
cameraBtn.title = '请先连接服务器';
// 关闭摄像头
if (typeof window.stopCamera === 'function') {
window.stopCamera();
}
}
// Update camera button state - enable when connected and camera is available
if (cameraBtn && isConnected) {
if (window.cameraAvailable) {
cameraBtn.disabled = false;
cameraBtn.title = '打开/关闭摄像头';
} else {
cameraBtn.disabled = true;
cameraBtn.title = '请先绑定验证码';
}
}
// Update record button state
if (recordBtn) {
const microphoneAvailable = window.microphoneAvailable !== false;
if (isConnected && microphoneAvailable) {
recordBtn.disabled = false;
recordBtn.title = '开始录音';
// Restore record button to normal state
recordBtn.querySelector('.btn-text').textContent = '录音';
recordBtn.classList.remove('recording');
} else {
recordBtn.disabled = true;
if (!microphoneAvailable) {
recordBtn.title = window.isHttpNonLocalhost ? '当前由于是http访问,无法录音,只能用文字交互' : '麦克风不可用';
} else {
recordBtn.title = '请先连接服务器';
}
// Restore record button to normal state
recordBtn.querySelector('.btn-text').textContent = '录音';
recordBtn.classList.remove('recording');
}
}
}
// Update record button state
updateRecordButtonState(isRecording, seconds = 0) {
const recordBtn = document.getElementById('recordBtn');
if (recordBtn) {
if (isRecording) {
recordBtn.querySelector('.btn-text').textContent = `录音中`;
recordBtn.classList.add('recording');
} else {
recordBtn.querySelector('.btn-text').textContent = '录音';
recordBtn.classList.remove('recording');
}
// Only enable button when microphone is available
recordBtn.disabled = window.microphoneAvailable === false;
}
}
/**
* Update microphone availability state
* @param {boolean} isAvailable - Whether microphone is available
* @param {boolean} isHttpNonLocalhost - Whether it is HTTP non-localhost access
*/
updateMicrophoneAvailability(isAvailable, isHttpNonLocalhost) {
const recordBtn = document.getElementById('recordBtn');
if (!recordBtn) return;
if (!isAvailable) {
// Disable record button
recordBtn.disabled = true;
// Update button text and title
recordBtn.querySelector('.btn-text').textContent = '录音';
recordBtn.title = isHttpNonLocalhost ? '当前由于是http访问,无法录音,只能用文字交互' : '麦克风不可用';
} else {
// If connected, enable record button
const wsHandler = getWebSocketHandler();
if (wsHandler && wsHandler.isConnected()) {
recordBtn.disabled = false;
recordBtn.title = '开始录音';
}
}
}
// Add chat message
addChatMessage(content, isUser = false) {
const chatStream = document.getElementById('chatStream');
if (!chatStream) return;
const messageDiv = document.createElement('div');
messageDiv.className = `chat-message ${isUser ? 'user' : 'ai'}`;
messageDiv.innerHTML = `<div class="message-bubble">${content}</div>`;
chatStream.appendChild(messageDiv);
// Scroll to bottom
chatStream.scrollTop = chatStream.scrollHeight;
}
// Switch background
switchBackground() {
this.currentBackgroundIndex = (this.currentBackgroundIndex + 1) % this.backgroundImages.length;
const backgroundContainer = document.querySelector('.background-container');
if (backgroundContainer) {
backgroundContainer.style.backgroundImage = `url('./images/${this.backgroundImages[this.currentBackgroundIndex]}')`;
}
localStorage.setItem('backgroundIndex', this.currentBackgroundIndex);
}
// Switch Live2D model
switchLive2DModel() {
const modelSelect = document.getElementById('live2dModelSelect');
if (!modelSelect) {
console.error('模型选择下拉框不存在');
return;
}
const selectedModel = modelSelect.value;
const app = window.chatApp;
if (app && app.live2dManager) {
app.live2dManager.switchModel(selectedModel)
.then(success => {
if (success) {
this.addChatMessage(`已切换到模型: ${selectedModel}`, false);
} else {
this.addChatMessage('模型切换失败', false);
}
})
.catch(error => {
console.error('模型切换错误:', error);
this.addChatMessage('模型切换出错', false);
});
} else {
this.addChatMessage('Live2D管理器未初始化', false);
}
}
// Show modal
showModal(modalId) {
const modal = document.getElementById(modalId);
if (modal) {
modal.style.display = 'flex';
}
}
// Hide modal
hideModal(modalId) {
const modal = document.getElementById(modalId);
if (modal) {
modal.style.display = 'none';
}
}
// Switch tab
switchTab(tabName) {
// Remove active class from all tabs
const tabBtns = document.querySelectorAll('.tab-btn');
const tabContents = document.querySelectorAll('.tab-content');
tabBtns.forEach(btn => btn.classList.remove('active'));
tabContents.forEach(content => content.classList.remove('active'));
// Activate selected tab
const activeTabBtn = document.querySelector(`[data-tab="${tabName}"]`);
const activeTabContent = document.getElementById(`${tabName}Tab`);
if (activeTabBtn && activeTabContent) {
activeTabBtn.classList.add('active');
activeTabContent.classList.add('active');
}
}
applyWakewordConfig(config = {}) {
const wakewordEnabledInput = document.getElementById('wakewordEnabled');
const wakewordListInput = document.getElementById('wakewordList');
if (!wakewordEnabledInput || !wakewordListInput) {
return;
}
const wakeWords = Array.isArray(config.wakeWords)
? config.wakeWords.filter(item => typeof item === 'string' && item.trim())
: [];
wakewordEnabledInput.value = config.enabled === false ? 'false' : 'true';
wakewordListInput.value = wakeWords.join('\n');
saveConfig();
}
async handleApplyWakeword() {
const wakewordEnabledInput = document.getElementById('wakewordEnabled');
const wakewordListInput = document.getElementById('wakewordList');
const wakewordWsUrlInput = document.getElementById('wakewordWsUrl');
if (!wakewordEnabledInput || !wakewordListInput) {
return;
}
const wakeWords = wakewordListInput.value
.split(/\r?\n/u)
.map(item => item.trim())
.filter(Boolean)
.filter((item, index, items) => items.indexOf(item) === index);
const payload = {
enabled: wakewordEnabledInput.value !== 'false',
wakeWords,
};
if (payload.enabled && payload.wakeWords.length === 0) {
this.addChatMessage('启用唤醒词时,至少需要填写一个唤醒词。', false);
return;
}
const applyWakewordBtn = document.getElementById('applyWakewordBtn');
if (applyWakewordBtn) {
applyWakewordBtn.disabled = true;
applyWakewordBtn.textContent = '应用中...';
}
try {
// 保存地址到 localStorage
if (wakewordWsUrlInput && wakewordWsUrlInput.value.trim()) {
localStorage.setItem('xz_tester_wakewordWsUrl', wakewordWsUrlInput.value.trim());
}
// 比较新地址和当前连接地址
const newWsUrl = localStorage.getItem('xz_tester_wakewordWsUrl');
const currentWsUrl = getWakewordBridgeUrl();
const urlChanged = newWsUrl !== currentWsUrl;
if (urlChanged) {
// 地址变了,先确认
const shouldRestart = window.confirm('地址已变更,是否继续?(将断开旧连接并重新连接)');
if (!shouldRestart) {
// 恢复 localStorage 里的旧地址
localStorage.setItem('xz_tester_wakewordWsUrl', currentWsUrl);
this.addChatMessage('已取消地址变更。', false);
return;
}
// 断开旧连接
stopWakewordBridgeListener();
// 启动新连接(自动用新地址)
startWakewordBridgeListener();
// 等 bridge_connected
await new Promise((resolve, reject) => {
const timeout = setTimeout(() => {
reject(new Error('连接新服务器超时'));
}, 5000);
onNextBridgeConnected(() => {
clearTimeout(timeout);
resolve();
});
});
// 发配置和重启
await requestWakewordBridge('set_wakeword_config', payload);
this.applyWakewordConfig(payload);
await requestWakewordBridge('restart_wakeword_service');
this.addChatMessage('唤醒词配置已保存,唤醒词服务正在重启。', false);
} else {
// 地址没变:直接在当前连接操作
const response = await requestWakewordBridge('set_wakeword_config', payload);
this.applyWakewordConfig(response.payload || payload);
const shouldRestart = window.confirm('唤醒词已保存。是否现在重启唤醒词服务以立即生效?');
if (!shouldRestart) {
this.addChatMessage('唤醒词配置已保存,可稍后手动重启服务后生效。', false);
return;
}
await requestWakewordBridge('restart_wakeword_service');
this.addChatMessage('唤醒词配置已保存,唤醒词服务正在重启。', false);
}
} catch (error) {
this.addChatMessage(`应用唤醒词失败: ${error.message}`, false);
} finally {
if (applyWakewordBtn) {
applyWakewordBtn.disabled = false;
applyWakewordBtn.textContent = '应用唤醒词';
}
}
}
// Start AI chat session after connection
startAIChatSession() {
this.addChatMessage('连接成功,开始聊天吧~😊', false);
// Check microphone availability and show error messages if needed
if (!window.microphoneAvailable) {
if (window.isHttpNonLocalhost) {
this.addChatMessage('⚠️ 当前由于是http访问,无法录音,只能用文字交互', false);
} else {
this.addChatMessage('⚠️ 麦克风不可用,请检查权限设置,只能用文字交互', false);
}
}
// Start recording only if microphone is available
if (window.microphoneAvailable) {
const recordBtn = document.getElementById('recordBtn');
if (recordBtn) {
recordBtn.click();
}
}
// Start camera only if camera is available (bound with verification code)
if (window.cameraAvailable && typeof window.startCamera === 'function') {
window.startCamera().then(success => {
if (success) {
const cameraBtn = document.getElementById('cameraBtn');
if (cameraBtn) {
cameraBtn.classList.add('camera-active');
cameraBtn.querySelector('.btn-text').textContent = '关闭';
}
} else {
this.addChatMessage('⚠️ 摄像头启动失败,可能被浏览器拒绝', false);
}
}).catch(error => {
log(`启动摄像头异常: ${error.message}`, 'error');
});
}
}
// Handle connect button click
async handleConnect() {
const wsHandler = getWebSocketHandler();
if (this.isConnecting || (wsHandler && wsHandler.isConnected())) {
log('连接已存在或正在进行,忽略本次拨号请求', 'info');
return;
}
this.isConnecting = true;
console.log('handleConnect called');
try {
// Switch to device settings tab
this.switchTab('device');
// Wait for DOM update
await new Promise(resolve => setTimeout(resolve, 50));
const otaUrlInput = document.getElementById('otaUrl');
console.log('otaUrl element:', otaUrlInput);
if (!otaUrlInput || !otaUrlInput.value) {
this.addChatMessage('请输入OTA服务器地址', false);
return;
}
const otaUrl = otaUrlInput.value;
console.log('otaUrl value:', otaUrl);
// Update dial button state to connecting
const dialBtn = document.getElementById('dialBtn');
if (dialBtn) {
dialBtn.classList.add('dial-active');
dialBtn.querySelector('.btn-text').textContent = '连接中...';
dialBtn.disabled = true;
}
// Show connecting message
this.addChatMessage('正在连接服务器...', false);
const chatIpt = document.getElementById('chatIpt');
if (chatIpt) {
chatIpt.style.display = 'flex';
}
// Get WebSocket handler instance
// Register connection state callback BEFORE connecting
wsHandler.onConnectionStateChange = (isConnected) => {
this.updateConnectionUI(isConnected);
this.updateDialButton(isConnected);
};
// Register chat message callback BEFORE connecting
wsHandler.onChatMessage = (text, isUser) => {
this.addChatMessage(text, isUser);
};
// Register record button state callback BEFORE connecting
wsHandler.onRecordButtonStateChange = (isRecording) => {
const recordBtn = document.getElementById('recordBtn');
if (recordBtn) {
if (isRecording) {
recordBtn.classList.add('recording');
recordBtn.querySelector('.btn-text').textContent = '录音中';
} else {
recordBtn.classList.remove('recording');
recordBtn.querySelector('.btn-text').textContent = '录音';
}
}
};
const isConnected = await wsHandler.connect();
if (isConnected) {
// Check microphone availability (check again after connection)
const { checkMicrophoneAvailability } = await import('../core/audio/recorder.js?v=0205');
const micAvailable = await checkMicrophoneAvailability();
if (!micAvailable) {
const isHttp = window.isHttpNonLocalhost;
if (isHttp) {
this.addChatMessage('⚠️ 当前由于是http访问,无法录音,只能用文字交互', false);
}
// Update global state
window.microphoneAvailable = false;
}
// Update dial button state
const dialBtn = document.getElementById('dialBtn');
if (dialBtn) {
if (!this.dialBtnDisabled) {
dialBtn.disabled = false;
}
dialBtn.querySelector('.btn-text').textContent = '挂断';
dialBtn.classList.add('dial-active');
}
this.hideModal('settingsModal');
} else {
throw new Error('OTA连接失败');
}
} catch (error) {
console.error('Connection error details:', {
message: error.message,
stack: error.stack,
name: error.name
});
// Show error message
const errorMessage = error.message.includes('Cannot set properties of null')
? '连接失败:请检查设备连接'
: `连接失败: ${error.message}`;
this.addChatMessage(errorMessage, false);
// Restore dial button state
const dialBtn = document.getElementById('dialBtn');
if (dialBtn) {
if (!this.dialBtnDisabled) {
dialBtn.disabled = false;
}
dialBtn.querySelector('.btn-text').textContent = '拨号';
dialBtn.classList.remove('dial-active');
console.log('Dial button state restored successfully');
}
} finally {
this.isConnecting = false;
}
}
async triggerWakewordDial(wakeWord = '唤醒词') {
const wsHandler = getWebSocketHandler();
const now = Date.now();
if (wsHandler && wsHandler.isConnected()) {
log('页面已连接,忽略自动拨号', 'info');
return false;
}
if (this.isConnecting || this.dialBtnDisabled) {
log('页面正在连接中,忽略重复唤醒', 'info');
return false;
}
if (now - this.lastWakewordDialTime < 3000) {
log('唤醒触发过于频繁,忽略本次自动拨号', 'warning');
return false;
}
this.lastWakewordDialTime = now;
this.addChatMessage(`检测到唤醒词“${wakeWord}”,准备连接服务器...`, false);
await this.handleConnect();
return true;
}
// Add MCP tool
addMCPTool() {
const mcpToolsList = document.getElementById('mcpToolsList');
if (!mcpToolsList) return;
const toolId = `mcp-tool-${Date.now()}`;
const toolDiv = document.createElement('div');
toolDiv.className = 'properties-container';
toolDiv.innerHTML = `
<div class="property-item">
<input type="text" placeholder="工具名称" value="新工具">
<input type="text" placeholder="工具描述" value="工具描述">
<button class="remove-property" onclick="uiController.removeMCPTool('${toolId}')">删除</button>
</div>
`;
mcpToolsList.appendChild(toolDiv);
}
// Remove MCP tool
removeMCPTool(toolId) {
const toolElement = document.getElementById(toolId);
if (toolElement) {
toolElement.remove();
}
}
// Update audio statistics display
updateAudioStats() {
const audioPlayer = getAudioPlayer();
if (!audioPlayer) return;
const stats = audioPlayer.getAudioStats();
// Here can add audio statistics UI update logic
}
// Start audio statistics monitor
startAudioStatsMonitor() {
// Update audio statistics every 100ms
this.audioStatsTimer = setInterval(() => {
this.updateAudioStats();
}, 100);
}
// Stop audio statistics monitor
stopAudioStatsMonitor() {
if (this.audioStatsTimer) {
clearInterval(this.audioStatsTimer);
this.audioStatsTimer = null;
}
}
// Draw audio visualizer waveform
drawVisualizer(dataArray) {
if (!this.visualizerContext || !this.visualizerCanvas) return;
this.visualizerContext.fillStyle = '#fafafa';
this.visualizerContext.fillRect(0, 0, this.visualizerCanvas.width, this.visualizerCanvas.height);
const barWidth = (this.visualizerCanvas.width / dataArray.length) * 2.5;
let barHeight;
let x = 0;
for (let i = 0; i < dataArray.length; i++) {
barHeight = dataArray[i] / 2;
// Create gradient color: from purple to blue to green
const gradient = this.visualizerContext.createLinearGradient(0, 0, 0, this.visualizerCanvas.height);
gradient.addColorStop(0, '#8e44ad');
gradient.addColorStop(0.5, '#3498db');
gradient.addColorStop(1, '#1abc9c');
this.visualizerContext.fillStyle = gradient;
this.visualizerContext.fillRect(x, this.visualizerCanvas.height - barHeight, barWidth, barHeight);
x += barWidth + 1;
}
}
// Update session status UI
updateSessionStatus(isSpeaking) {
// Here can add session status UI update logic
// For example: update Live2D model's mouth movement status
}
// Update session emotion
updateSessionEmotion(emoji) {
// Here can add emotion update logic
// For example: display emoji in status indicator
}
}
// Create singleton instance
export const uiController = new UIController();
// Export class for module usage
export { UIController };
@@ -95,4 +95,9 @@ export default class BlockingQueue {
get length() { get length() {
return this.#items.length; return this.#items.length;
} }
/* 清空队列(保持对象引用,不影响等待者) */
clear() {
this.#items.length = 0;
}
} }
@@ -1,6 +1,3 @@
import { getLogContainer } from '../document.js'
const logContainer = getLogContainer();
// 日志记录函数 // 日志记录函数
export function log(message, type = 'info') { export function log(message, type = 'info') {
// 将消息按换行符分割成多行 // 将消息按换行符分割成多行
@@ -8,6 +5,15 @@ export function log(message, type = 'info') {
const now = new Date(); const now = new Date();
// const timestamp = `[${now.toLocaleTimeString()}] `; // const timestamp = `[${now.toLocaleTimeString()}] `;
const timestamp = `[${now.toLocaleTimeString()}.${now.getMilliseconds().toString().padStart(3, '0')}] `; const timestamp = `[${now.toLocaleTimeString()}.${now.getMilliseconds().toString().padStart(3, '0')}] `;
// 检查是否存在日志容器
const logContainer = document.getElementById('logContainer');
if (!logContainer) {
// 如果日志容器不存在,只输出到控制台
console.log(`[${type.toUpperCase()}] ${message}`);
return;
}
// 为每一行创建日志条目 // 为每一行创建日志条目
lines.forEach((line, index) => { lines.forEach((line, index) => {
const logEntry = document.createElement('div'); const logEntry = document.createElement('div');
@@ -0,0 +1,72 @@
============================================================
示例模型
桃濑日和 - PRO
============================================================
该示例时基于Cubism3.0制作的标准模型素材。
可用于学习变形器的构造以及参数的使用方法。
模型的肩部应用了新功能【胶水】。
------------------------------
素材使用许可
------------------------------
 普通用户以及小规模企业在同意授权协议的情况下可用于商业用途。
 中/大规模的企业只能用于非公开的内部试用。
 在使用该素材时,请确认以下的【无偿提供素材使用授权协议】中的“授权类型”、“Live2D原创角色”等的相关内容,
 并必须接受【Live2D Cubism 示例模型的使用授权要求】中的利用条件。
 有关许可证的更多信息,请参阅以下页面。
 https://www.live2d.com/zh-CHS/download/sample-data/
------------------------------
创作者
------------------------------
插画:Kani Biimu
模型:Live2D
------------------------------
素材内容
------------------------------
模型文件(cmo3) ※包含物理模拟的设定
动画文件(can3)
嵌入文件列表(runtime文件夹)
・模型数据(moc3)
・动作数据(motion3.json)
・模型设定文件(model3.json)
・物理模拟设定文件(physics3.json)
・姿势设定文件(pose3.json)
・辅助显示的文件(cdi3.json)
------------------------------
更新记录
------------------------------
【cmo3】
 hiyori_pro_t11
 2023年03月08日 修改了部分模型关键点
hiyori_pro_t10
2021年06月10日 公开
【can3】
 hiyori_pro_t04
 2023年03月08日 修改了部分动画关键帧
hiyori_pro_t03
2021年06月10日 公开
 
Binary file not shown.

After

Width:  |  Height:  |  Size: 1.7 MiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 2.4 MiB

@@ -0,0 +1,533 @@
{
"Version": 3,
"Parameters": [
{
"Id": "ParamAngleX",
"GroupId": "ParamGroupFace",
"Name": "角度 X"
},
{
"Id": "ParamAngleY",
"GroupId": "ParamGroupFace",
"Name": "角度 Y"
},
{
"Id": "ParamAngleZ",
"GroupId": "ParamGroupFace",
"Name": "角度 Z"
},
{
"Id": "ParamCheek",
"GroupId": "ParamGroupFace",
"Name": "脸颊泛红"
},
{
"Id": "ParamEyeLOpen",
"GroupId": "ParamGroupEyes",
"Name": "左眼 开闭"
},
{
"Id": "ParamEyeLSmile",
"GroupId": "ParamGroupEyes",
"Name": "左眼 微笑"
},
{
"Id": "ParamEyeROpen",
"GroupId": "ParamGroupEyes",
"Name": "右眼 开闭"
},
{
"Id": "ParamEyeRSmile",
"GroupId": "ParamGroupEyes",
"Name": "右眼 微笑"
},
{
"Id": "ParamEyeBallX",
"GroupId": "ParamGroupEyeballs",
"Name": "眼珠 X"
},
{
"Id": "ParamEyeBallY",
"GroupId": "ParamGroupEyeballs",
"Name": "眼珠 Y"
},
{
"Id": "ParamBrowLY",
"GroupId": "ParamGroupBrows",
"Name": "左眉 上下"
},
{
"Id": "ParamBrowRY",
"GroupId": "ParamGroupBrows",
"Name": "右眉 上下"
},
{
"Id": "ParamBrowLX",
"GroupId": "ParamGroupBrows",
"Name": "左眉 左右"
},
{
"Id": "ParamBrowRX",
"GroupId": "ParamGroupBrows",
"Name": "右眉 左右"
},
{
"Id": "ParamBrowLAngle",
"GroupId": "ParamGroupBrows",
"Name": "左眉 角度"
},
{
"Id": "ParamBrowRAngle",
"GroupId": "ParamGroupBrows",
"Name": "右眉 角度"
},
{
"Id": "ParamBrowLForm",
"GroupId": "ParamGroupBrows",
"Name": "左眉 变形"
},
{
"Id": "ParamBrowRForm",
"GroupId": "ParamGroupBrows",
"Name": "右眉 变形"
},
{
"Id": "ParamMouthForm",
"GroupId": "ParamGroupMouth",
"Name": "嘴 变形"
},
{
"Id": "ParamMouthOpenY",
"GroupId": "ParamGroupMouth",
"Name": "嘴 开闭"
},
{
"Id": "ParamBodyAngleX",
"GroupId": "ParamGroupBody",
"Name": "身体的旋转 X"
},
{
"Id": "ParamBodyAngleY",
"GroupId": "ParamGroupBody",
"Name": "身体的旋转 Y"
},
{
"Id": "ParamBodyAngleZ",
"GroupId": "ParamGroupBody",
"Name": "身体的旋转 Z"
},
{
"Id": "ParamBreath",
"GroupId": "ParamGroupBody",
"Name": "呼吸"
},
{
"Id": "ParamShoulder",
"GroupId": "ParamGroupBody",
"Name": "肩"
},
{
"Id": "ParamLeg",
"GroupId": "ParamGroupBody",
"Name": "腿"
},
{
"Id": "ParamArmLA",
"GroupId": "ParamGroupArms",
"Name": "左臂 A"
},
{
"Id": "ParamArmRA",
"GroupId": "ParamGroupArms",
"Name": "右臂 A"
},
{
"Id": "ParamArmLB",
"GroupId": "ParamGroupArms",
"Name": "左臂 B"
},
{
"Id": "ParamArmRB",
"GroupId": "ParamGroupArms",
"Name": "右臂 B"
},
{
"Id": "ParamHandLB",
"GroupId": "ParamGroupArms",
"Name": "左手B 旋转"
},
{
"Id": "ParamHandRB",
"GroupId": "ParamGroupArms",
"Name": "右手B 旋转"
},
{
"Id": "ParamHandL",
"GroupId": "ParamGroupArms",
"Name": "左手"
},
{
"Id": "ParamHandR",
"GroupId": "ParamGroupArms",
"Name": "右手"
},
{
"Id": "ParamBustY",
"GroupId": "ParamGroupSway",
"Name": "胸部 摇动"
},
{
"Id": "ParamHairAhoge",
"GroupId": "ParamGroupSway",
"Name": "头发摇动 呆毛"
},
{
"Id": "ParamHairFront",
"GroupId": "ParamGroupSway",
"Name": "头发摇动 前"
},
{
"Id": "ParamHairBack",
"GroupId": "ParamGroupSway",
"Name": "头发摇动 后"
},
{
"Id": "ParamSideupRibbon",
"GroupId": "ParamGroupSway",
"Name": "发饰的摇动"
},
{
"Id": "ParamRibbon",
"GroupId": "ParamGroupSway",
"Name": "蝴蝶结的摇动"
},
{
"Id": "ParamSkirt",
"GroupId": "ParamGroupSway",
"Name": "短裙的摇动"
},
{
"Id": "ParamSkirt2",
"GroupId": "ParamGroupSway",
"Name": "短裙的上卷"
},
{
"Id": "Param_Angle_Rotation_1_ArtMesh62",
"GroupId": "ParamGroup2",
"Name": "[0]辫子左"
},
{
"Id": "Param_Angle_Rotation_2_ArtMesh62",
"GroupId": "ParamGroup2",
"Name": "[1]辫子左"
},
{
"Id": "Param_Angle_Rotation_3_ArtMesh62",
"GroupId": "ParamGroup2",
"Name": "[2]辫子左"
},
{
"Id": "Param_Angle_Rotation_4_ArtMesh62",
"GroupId": "ParamGroup2",
"Name": "[3]辫子左"
},
{
"Id": "Param_Angle_Rotation_5_ArtMesh62",
"GroupId": "ParamGroup2",
"Name": "[4]辫子左"
},
{
"Id": "Param_Angle_Rotation_6_ArtMesh62",
"GroupId": "ParamGroup2",
"Name": "[5]辫子左"
},
{
"Id": "Param_Angle_Rotation_7_ArtMesh62",
"GroupId": "ParamGroup2",
"Name": "[6]辫子左"
},
{
"Id": "Param_Angle_Rotation_1_ArtMesh61",
"GroupId": "ParamGroup",
"Name": "[0]辫子右"
},
{
"Id": "Param_Angle_Rotation_2_ArtMesh61",
"GroupId": "ParamGroup",
"Name": "[1]辫子右"
},
{
"Id": "Param_Angle_Rotation_3_ArtMesh61",
"GroupId": "ParamGroup",
"Name": "[2]辫子右"
},
{
"Id": "Param_Angle_Rotation_4_ArtMesh61",
"GroupId": "ParamGroup",
"Name": "[3]辫子右"
},
{
"Id": "Param_Angle_Rotation_5_ArtMesh61",
"GroupId": "ParamGroup",
"Name": "[4]辫子右"
},
{
"Id": "Param_Angle_Rotation_6_ArtMesh61",
"GroupId": "ParamGroup",
"Name": "[5]辫子右"
},
{
"Id": "Param_Angle_Rotation_7_ArtMesh61",
"GroupId": "ParamGroup",
"Name": "[6]辫子右"
},
{
"Id": "Param_Angle_Rotation_1_ArtMesh55",
"GroupId": "ParamGroup4",
"Name": "[0]侧发左"
},
{
"Id": "Param_Angle_Rotation_2_ArtMesh55",
"GroupId": "ParamGroup4",
"Name": "[1]侧发左"
},
{
"Id": "Param_Angle_Rotation_3_ArtMesh55",
"GroupId": "ParamGroup4",
"Name": "[2]侧发左"
},
{
"Id": "Param_Angle_Rotation_4_ArtMesh55",
"GroupId": "ParamGroup4",
"Name": "[3]侧发左"
},
{
"Id": "Param_Angle_Rotation_5_ArtMesh55",
"GroupId": "ParamGroup4",
"Name": "[4]侧发左"
},
{
"Id": "Param_Angle_Rotation_6_ArtMesh55",
"GroupId": "ParamGroup4",
"Name": "[5]侧发左"
},
{
"Id": "Param_Angle_Rotation_7_ArtMesh55",
"GroupId": "ParamGroup4",
"Name": "[6]侧发左"
},
{
"Id": "Param_Angle_Rotation_1_ArtMesh54",
"GroupId": "ParamGroup3",
"Name": "[0]侧发右"
},
{
"Id": "Param_Angle_Rotation_2_ArtMesh54",
"GroupId": "ParamGroup3",
"Name": "[1]侧发右"
},
{
"Id": "Param_Angle_Rotation_3_ArtMesh54",
"GroupId": "ParamGroup3",
"Name": "[2]侧发右"
},
{
"Id": "Param_Angle_Rotation_4_ArtMesh54",
"GroupId": "ParamGroup3",
"Name": "[3]侧发右"
},
{
"Id": "Param_Angle_Rotation_5_ArtMesh54",
"GroupId": "ParamGroup3",
"Name": "[4]侧发右"
},
{
"Id": "Param_Angle_Rotation_6_ArtMesh54",
"GroupId": "ParamGroup3",
"Name": "[5]侧发右"
},
{
"Id": "Param_Angle_Rotation_7_ArtMesh54",
"GroupId": "ParamGroup3",
"Name": "[6]侧发右"
}
],
"ParameterGroups": [
{
"Id": "ParamGroupFace",
"GroupId": "",
"Name": "脸"
},
{
"Id": "ParamGroupEyes",
"GroupId": "",
"Name": "眼睛"
},
{
"Id": "ParamGroupEyeballs",
"GroupId": "",
"Name": "眼珠"
},
{
"Id": "ParamGroupBrows",
"GroupId": "",
"Name": "眉毛"
},
{
"Id": "ParamGroupMouth",
"GroupId": "",
"Name": "嘴"
},
{
"Id": "ParamGroupBody",
"GroupId": "",
"Name": "身体"
},
{
"Id": "ParamGroupArms",
"GroupId": "",
"Name": "手臂"
},
{
"Id": "ParamGroupSway",
"GroupId": "",
"Name": "摇动"
},
{
"Id": "ParamGroup2",
"GroupId": "",
"Name": "摇动 辫子左"
},
{
"Id": "ParamGroup",
"GroupId": "",
"Name": "摇动 辫子右"
},
{
"Id": "ParamGroup4",
"GroupId": "",
"Name": "摇动 侧发左"
},
{
"Id": "ParamGroup3",
"GroupId": "",
"Name": "摇动 侧发右"
}
],
"Parts": [
{
"Id": "PartCore",
"Name": "Core"
},
{
"Id": "PartCheek",
"Name": "脸颊"
},
{
"Id": "PartBrow",
"Name": "眉毛"
},
{
"Id": "PartEye",
"Name": "眼睛"
},
{
"Id": "PartNose",
"Name": "鼻子"
},
{
"Id": "PartMouth",
"Name": "嘴"
},
{
"Id": "PartFace",
"Name": "脸"
},
{
"Id": "PartEar",
"Name": "耳朵"
},
{
"Id": "PartHairSide",
"Name": "侧发"
},
{
"Id": "PartHairFront",
"Name": "前发"
},
{
"Id": "PartHairBack",
"Name": "后发"
},
{
"Id": "PartNeck",
"Name": "脖子"
},
{
"Id": "PartArmA",
"Name": "手臂A"
},
{
"Id": "PartArmB",
"Name": "手臂B"
},
{
"Id": "PartBody",
"Name": "身体"
},
{
"Id": "PartBackground",
"Name": "背景"
},
{
"Id": "PartSketch",
"Name": "[参考图]"
},
{
"Id": "PartEyeBall",
"Name": "眼珠"
},
{
"Id": "ArtMesh55_Skinning",
"Name": "侧发右"
},
{
"Id": "Part4",
"Name": "侧发左(旋转)"
},
{
"Id": "ArtMesh54_Skinning",
"Name": "侧发右(蒙皮)"
},
{
"Id": "Part3",
"Name": "侧发右(旋转)"
},
{
"Id": "ArtMesh61_Skinning",
"Name": "辫子右(蒙皮)"
},
{
"Id": "Part",
"Name": "辫子右(旋转)"
},
{
"Id": "ArtMesh62_Skinning",
"Name": "辫子左(蒙皮)"
},
{
"Id": "Part2",
"Name": "辫子左(旋转)"
}
],
"CombinedParameters": [
[
"ParamAngleX",
"ParamAngleY"
],
[
"ParamMouthForm",
"ParamMouthOpenY"
]
]
}
@@ -0,0 +1,82 @@
{
"Version": 3,
"FileReferences": {
"Moc": "hiyori_pro_t11.moc3",
"Textures": [
"hiyori_pro_t11.2048/texture_00.png",
"hiyori_pro_t11.2048/texture_01.png"
],
"Physics": "hiyori_pro_t11.physics3.json",
"Pose": "hiyori_pro_t11.pose3.json",
"DisplayInfo": "hiyori_pro_t11.cdi3.json",
"Motions": {
"Idle": [
{
"File": "motion/hiyori_m01.motion3.json"
},
{
"File": "motion/hiyori_m02.motion3.json"
},
{
"File": "motion/hiyori_m05.motion3.json"
}
],
"Flick": [
{
"File": "motion/hiyori_m03.motion3.json"
}
],
"FlickDown": [
{
"File": "motion/hiyori_m04.motion3.json"
}
],
"FlickUp": [
{
"File": "motion/hiyori_m06.motion3.json"
}
],
"Tap": [
{
"File": "motion/hiyori_m07.motion3.json"
},
{
"File": "motion/hiyori_m08.motion3.json"
}
],
"Tap@Body": [
{
"File": "motion/hiyori_m09.motion3.json"
}
],
"Flick@Body": [
{
"File": "motion/hiyori_m10.motion3.json"
}
]
}
},
"Groups": [
{
"Target": "Parameter",
"Name": "LipSync",
"Ids": [
"ParamMouthOpenY"
]
},
{
"Target": "Parameter",
"Name": "EyeBlink",
"Ids": [
"ParamEyeLOpen",
"ParamEyeROpen"
]
}
],
"HitAreas": [
{
"Id": "HitArea",
"Name": "Body"
}
]
}
File diff suppressed because it is too large Load Diff
@@ -0,0 +1,15 @@
{
"Type": "Live2D Pose",
"Groups": [
[
{
"Id": "PartArmA",
"Link": []
},
{
"Id": "PartArmB",
"Link": []
}
]
]
}

Some files were not shown because too many files have changed in this diff Show More