@livekit/agents 1.1.0 → 1.2.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (942) hide show
  1. package/dist/audio.cjs +89 -3
  2. package/dist/audio.cjs.map +1 -1
  3. package/dist/audio.d.cts +36 -1
  4. package/dist/audio.d.ts +36 -1
  5. package/dist/audio.d.ts.map +1 -1
  6. package/dist/audio.js +76 -2
  7. package/dist/audio.js.map +1 -1
  8. package/dist/beta/index.cjs +29 -0
  9. package/dist/beta/index.cjs.map +1 -0
  10. package/dist/beta/index.d.cts +2 -0
  11. package/dist/beta/index.d.ts +2 -0
  12. package/dist/beta/index.d.ts.map +1 -0
  13. package/dist/beta/index.js +7 -0
  14. package/dist/beta/index.js.map +1 -0
  15. package/dist/beta/workflows/index.cjs +29 -0
  16. package/dist/beta/workflows/index.cjs.map +1 -0
  17. package/dist/beta/workflows/index.d.cts +2 -0
  18. package/dist/beta/workflows/index.d.ts +2 -0
  19. package/dist/beta/workflows/index.d.ts.map +1 -0
  20. package/dist/beta/workflows/index.js +7 -0
  21. package/dist/beta/workflows/index.js.map +1 -0
  22. package/dist/beta/workflows/task_group.cjs +162 -0
  23. package/dist/beta/workflows/task_group.cjs.map +1 -0
  24. package/dist/beta/workflows/task_group.d.cts +32 -0
  25. package/dist/beta/workflows/task_group.d.ts +32 -0
  26. package/dist/beta/workflows/task_group.d.ts.map +1 -0
  27. package/dist/beta/workflows/task_group.js +138 -0
  28. package/dist/beta/workflows/task_group.js.map +1 -0
  29. package/dist/cli.cjs +44 -46
  30. package/dist/cli.cjs.map +1 -1
  31. package/dist/cli.d.cts +3 -3
  32. package/dist/cli.d.ts +3 -3
  33. package/dist/cli.d.ts.map +1 -1
  34. package/dist/cli.js +45 -47
  35. package/dist/cli.js.map +1 -1
  36. package/dist/connection_pool.cjs +242 -0
  37. package/dist/connection_pool.cjs.map +1 -0
  38. package/dist/connection_pool.d.cts +123 -0
  39. package/dist/connection_pool.d.ts +123 -0
  40. package/dist/connection_pool.d.ts.map +1 -0
  41. package/dist/connection_pool.js +218 -0
  42. package/dist/connection_pool.js.map +1 -0
  43. package/dist/connection_pool.test.cjs +256 -0
  44. package/dist/connection_pool.test.cjs.map +1 -0
  45. package/dist/connection_pool.test.js +255 -0
  46. package/dist/connection_pool.test.js.map +1 -0
  47. package/dist/constants.cjs +30 -0
  48. package/dist/constants.cjs.map +1 -1
  49. package/dist/constants.d.cts +10 -0
  50. package/dist/constants.d.ts +10 -0
  51. package/dist/constants.d.ts.map +1 -1
  52. package/dist/constants.js +20 -0
  53. package/dist/constants.js.map +1 -1
  54. package/dist/cpu.cjs +189 -0
  55. package/dist/cpu.cjs.map +1 -0
  56. package/dist/cpu.d.cts +24 -0
  57. package/dist/cpu.d.ts +24 -0
  58. package/dist/cpu.d.ts.map +1 -0
  59. package/dist/cpu.js +152 -0
  60. package/dist/cpu.js.map +1 -0
  61. package/dist/cpu.test.cjs +227 -0
  62. package/dist/cpu.test.cjs.map +1 -0
  63. package/dist/cpu.test.js +204 -0
  64. package/dist/cpu.test.js.map +1 -0
  65. package/dist/http_server.cjs +9 -6
  66. package/dist/http_server.cjs.map +1 -1
  67. package/dist/http_server.d.cts +5 -1
  68. package/dist/http_server.d.ts +5 -1
  69. package/dist/http_server.d.ts.map +1 -1
  70. package/dist/http_server.js +9 -6
  71. package/dist/http_server.js.map +1 -1
  72. package/dist/index.cjs +24 -9
  73. package/dist/index.cjs.map +1 -1
  74. package/dist/index.d.cts +15 -11
  75. package/dist/index.d.ts +15 -11
  76. package/dist/index.d.ts.map +1 -1
  77. package/dist/index.js +18 -9
  78. package/dist/index.js.map +1 -1
  79. package/dist/inference/api_protos.cjs +70 -2
  80. package/dist/inference/api_protos.cjs.map +1 -1
  81. package/dist/inference/api_protos.d.cts +373 -32
  82. package/dist/inference/api_protos.d.ts +373 -32
  83. package/dist/inference/api_protos.d.ts.map +1 -1
  84. package/dist/inference/api_protos.js +62 -2
  85. package/dist/inference/api_protos.js.map +1 -1
  86. package/dist/inference/index.cjs +8 -0
  87. package/dist/inference/index.cjs.map +1 -1
  88. package/dist/inference/index.d.cts +3 -4
  89. package/dist/inference/index.d.ts +3 -4
  90. package/dist/inference/index.d.ts.map +1 -1
  91. package/dist/inference/index.js +18 -3
  92. package/dist/inference/index.js.map +1 -1
  93. package/dist/inference/interruption/defaults.cjs +81 -0
  94. package/dist/inference/interruption/defaults.cjs.map +1 -0
  95. package/dist/inference/interruption/defaults.d.cts +19 -0
  96. package/dist/inference/interruption/defaults.d.ts +19 -0
  97. package/dist/inference/interruption/defaults.d.ts.map +1 -0
  98. package/dist/inference/interruption/defaults.js +46 -0
  99. package/dist/inference/interruption/defaults.js.map +1 -0
  100. package/dist/inference/interruption/errors.cjs +44 -0
  101. package/dist/inference/interruption/errors.cjs.map +1 -0
  102. package/dist/inference/interruption/errors.d.cts +12 -0
  103. package/dist/inference/interruption/errors.d.ts +12 -0
  104. package/dist/inference/interruption/errors.d.ts.map +1 -0
  105. package/dist/inference/interruption/errors.js +20 -0
  106. package/dist/inference/interruption/errors.js.map +1 -0
  107. package/dist/inference/interruption/http_transport.cjs +163 -0
  108. package/dist/inference/interruption/http_transport.cjs.map +1 -0
  109. package/dist/inference/interruption/http_transport.d.cts +63 -0
  110. package/dist/inference/interruption/http_transport.d.ts +63 -0
  111. package/dist/inference/interruption/http_transport.d.ts.map +1 -0
  112. package/dist/inference/interruption/http_transport.js +137 -0
  113. package/dist/inference/interruption/http_transport.js.map +1 -0
  114. package/dist/inference/interruption/interruption_cache_entry.cjs +58 -0
  115. package/dist/inference/interruption/interruption_cache_entry.cjs.map +1 -0
  116. package/dist/inference/interruption/interruption_cache_entry.d.cts +30 -0
  117. package/dist/inference/interruption/interruption_cache_entry.d.ts +30 -0
  118. package/dist/inference/interruption/interruption_cache_entry.d.ts.map +1 -0
  119. package/dist/inference/interruption/interruption_cache_entry.js +34 -0
  120. package/dist/inference/interruption/interruption_cache_entry.js.map +1 -0
  121. package/dist/inference/interruption/interruption_detector.cjs +198 -0
  122. package/dist/inference/interruption/interruption_detector.cjs.map +1 -0
  123. package/dist/inference/interruption/interruption_detector.d.cts +59 -0
  124. package/dist/inference/interruption/interruption_detector.d.ts +59 -0
  125. package/dist/inference/interruption/interruption_detector.d.ts.map +1 -0
  126. package/dist/inference/interruption/interruption_detector.js +164 -0
  127. package/dist/inference/interruption/interruption_detector.js.map +1 -0
  128. package/dist/inference/interruption/interruption_stream.cjs +368 -0
  129. package/dist/inference/interruption/interruption_stream.cjs.map +1 -0
  130. package/dist/inference/interruption/interruption_stream.d.cts +46 -0
  131. package/dist/inference/interruption/interruption_stream.d.ts +46 -0
  132. package/dist/inference/interruption/interruption_stream.d.ts.map +1 -0
  133. package/dist/inference/interruption/interruption_stream.js +344 -0
  134. package/dist/inference/interruption/interruption_stream.js.map +1 -0
  135. package/dist/inference/interruption/types.cjs +17 -0
  136. package/dist/inference/interruption/types.cjs.map +1 -0
  137. package/dist/inference/interruption/types.d.cts +66 -0
  138. package/dist/inference/interruption/types.d.ts +66 -0
  139. package/dist/inference/interruption/types.d.ts.map +1 -0
  140. package/dist/inference/interruption/types.js +1 -0
  141. package/dist/inference/interruption/types.js.map +1 -0
  142. package/dist/inference/interruption/utils.cjs +130 -0
  143. package/dist/inference/interruption/utils.cjs.map +1 -0
  144. package/dist/inference/interruption/utils.d.cts +41 -0
  145. package/dist/inference/interruption/utils.d.ts +41 -0
  146. package/dist/inference/interruption/utils.d.ts.map +1 -0
  147. package/dist/inference/interruption/utils.js +105 -0
  148. package/dist/inference/interruption/utils.js.map +1 -0
  149. package/dist/inference/interruption/utils.test.cjs +105 -0
  150. package/dist/inference/interruption/utils.test.cjs.map +1 -0
  151. package/dist/inference/interruption/utils.test.js +104 -0
  152. package/dist/inference/interruption/utils.test.js.map +1 -0
  153. package/dist/inference/interruption/ws_transport.cjs +342 -0
  154. package/dist/inference/interruption/ws_transport.cjs.map +1 -0
  155. package/dist/inference/interruption/ws_transport.d.cts +33 -0
  156. package/dist/inference/interruption/ws_transport.d.ts +33 -0
  157. package/dist/inference/interruption/ws_transport.d.ts.map +1 -0
  158. package/dist/inference/interruption/ws_transport.js +308 -0
  159. package/dist/inference/interruption/ws_transport.js.map +1 -0
  160. package/dist/inference/llm.cjs +106 -66
  161. package/dist/inference/llm.cjs.map +1 -1
  162. package/dist/inference/llm.d.cts +65 -43
  163. package/dist/inference/llm.d.ts +65 -43
  164. package/dist/inference/llm.d.ts.map +1 -1
  165. package/dist/inference/llm.js +100 -66
  166. package/dist/inference/llm.js.map +1 -1
  167. package/dist/inference/stt.cjs +319 -170
  168. package/dist/inference/stt.cjs.map +1 -1
  169. package/dist/inference/stt.d.cts +64 -15
  170. package/dist/inference/stt.d.ts +64 -15
  171. package/dist/inference/stt.d.ts.map +1 -1
  172. package/dist/inference/stt.js +319 -170
  173. package/dist/inference/stt.js.map +1 -1
  174. package/dist/inference/stt.test.cjs +218 -0
  175. package/dist/inference/stt.test.cjs.map +1 -0
  176. package/dist/inference/stt.test.js +217 -0
  177. package/dist/inference/stt.test.js.map +1 -0
  178. package/dist/inference/tts.cjs +249 -71
  179. package/dist/inference/tts.cjs.map +1 -1
  180. package/dist/inference/tts.d.cts +55 -16
  181. package/dist/inference/tts.d.ts +55 -16
  182. package/dist/inference/tts.d.ts.map +1 -1
  183. package/dist/inference/tts.js +249 -77
  184. package/dist/inference/tts.js.map +1 -1
  185. package/dist/inference/tts.test.cjs +233 -0
  186. package/dist/inference/tts.test.cjs.map +1 -0
  187. package/dist/inference/tts.test.js +232 -0
  188. package/dist/inference/tts.test.js.map +1 -0
  189. package/dist/inference/utils.cjs +26 -7
  190. package/dist/inference/utils.cjs.map +1 -1
  191. package/dist/inference/utils.d.cts +14 -1
  192. package/dist/inference/utils.d.ts +14 -1
  193. package/dist/inference/utils.d.ts.map +1 -1
  194. package/dist/inference/utils.js +18 -2
  195. package/dist/inference/utils.js.map +1 -1
  196. package/dist/ipc/inference_proc_executor.cjs +6 -3
  197. package/dist/ipc/inference_proc_executor.cjs.map +1 -1
  198. package/dist/ipc/inference_proc_executor.d.ts.map +1 -1
  199. package/dist/ipc/inference_proc_executor.js +6 -3
  200. package/dist/ipc/inference_proc_executor.js.map +1 -1
  201. package/dist/ipc/inference_proc_lazy_main.cjs +13 -1
  202. package/dist/ipc/inference_proc_lazy_main.cjs.map +1 -1
  203. package/dist/ipc/inference_proc_lazy_main.js +13 -1
  204. package/dist/ipc/inference_proc_lazy_main.js.map +1 -1
  205. package/dist/ipc/job_proc_executor.cjs +6 -1
  206. package/dist/ipc/job_proc_executor.cjs.map +1 -1
  207. package/dist/ipc/job_proc_executor.d.ts.map +1 -1
  208. package/dist/ipc/job_proc_executor.js +6 -1
  209. package/dist/ipc/job_proc_executor.js.map +1 -1
  210. package/dist/ipc/job_proc_lazy_main.cjs +89 -17
  211. package/dist/ipc/job_proc_lazy_main.cjs.map +1 -1
  212. package/dist/ipc/job_proc_lazy_main.js +68 -18
  213. package/dist/ipc/job_proc_lazy_main.js.map +1 -1
  214. package/dist/ipc/supervised_proc.cjs +34 -8
  215. package/dist/ipc/supervised_proc.cjs.map +1 -1
  216. package/dist/ipc/supervised_proc.d.cts +8 -0
  217. package/dist/ipc/supervised_proc.d.ts +8 -0
  218. package/dist/ipc/supervised_proc.d.ts.map +1 -1
  219. package/dist/ipc/supervised_proc.js +34 -8
  220. package/dist/ipc/supervised_proc.js.map +1 -1
  221. package/dist/ipc/supervised_proc.test.cjs +145 -0
  222. package/dist/ipc/supervised_proc.test.cjs.map +1 -0
  223. package/dist/ipc/supervised_proc.test.js +122 -0
  224. package/dist/ipc/supervised_proc.test.js.map +1 -0
  225. package/dist/job.cjs +109 -1
  226. package/dist/job.cjs.map +1 -1
  227. package/dist/job.d.cts +14 -0
  228. package/dist/job.d.ts +14 -0
  229. package/dist/job.d.ts.map +1 -1
  230. package/dist/job.js +99 -1
  231. package/dist/job.js.map +1 -1
  232. package/dist/language.cjs +394 -0
  233. package/dist/language.cjs.map +1 -0
  234. package/dist/language.d.cts +15 -0
  235. package/dist/language.d.ts +15 -0
  236. package/dist/language.d.ts.map +1 -0
  237. package/dist/language.js +363 -0
  238. package/dist/language.js.map +1 -0
  239. package/dist/language.test.cjs +43 -0
  240. package/dist/language.test.cjs.map +1 -0
  241. package/dist/language.test.js +49 -0
  242. package/dist/language.test.js.map +1 -0
  243. package/dist/llm/chat_context.cjs +274 -3
  244. package/dist/llm/chat_context.cjs.map +1 -1
  245. package/dist/llm/chat_context.d.cts +86 -2
  246. package/dist/llm/chat_context.d.ts +86 -2
  247. package/dist/llm/chat_context.d.ts.map +1 -1
  248. package/dist/llm/chat_context.js +273 -3
  249. package/dist/llm/chat_context.js.map +1 -1
  250. package/dist/llm/chat_context.test.cjs +574 -0
  251. package/dist/llm/chat_context.test.cjs.map +1 -1
  252. package/dist/llm/chat_context.test.js +574 -0
  253. package/dist/llm/chat_context.test.js.map +1 -1
  254. package/dist/llm/fallback_adapter.cjs +278 -0
  255. package/dist/llm/fallback_adapter.cjs.map +1 -0
  256. package/dist/llm/fallback_adapter.d.cts +73 -0
  257. package/dist/llm/fallback_adapter.d.ts +73 -0
  258. package/dist/llm/fallback_adapter.d.ts.map +1 -0
  259. package/dist/llm/fallback_adapter.js +254 -0
  260. package/dist/llm/fallback_adapter.js.map +1 -0
  261. package/dist/llm/fallback_adapter.test.cjs +176 -0
  262. package/dist/llm/fallback_adapter.test.cjs.map +1 -0
  263. package/dist/llm/fallback_adapter.test.js +175 -0
  264. package/dist/llm/fallback_adapter.test.js.map +1 -0
  265. package/dist/llm/index.cjs +9 -0
  266. package/dist/llm/index.cjs.map +1 -1
  267. package/dist/llm/index.d.cts +4 -3
  268. package/dist/llm/index.d.ts +4 -3
  269. package/dist/llm/index.d.ts.map +1 -1
  270. package/dist/llm/index.js +11 -1
  271. package/dist/llm/index.js.map +1 -1
  272. package/dist/llm/llm.cjs +65 -11
  273. package/dist/llm/llm.cjs.map +1 -1
  274. package/dist/llm/llm.d.cts +13 -2
  275. package/dist/llm/llm.d.ts +13 -2
  276. package/dist/llm/llm.d.ts.map +1 -1
  277. package/dist/llm/llm.js +65 -11
  278. package/dist/llm/llm.js.map +1 -1
  279. package/dist/llm/provider_format/google.cjs +6 -2
  280. package/dist/llm/provider_format/google.cjs.map +1 -1
  281. package/dist/llm/provider_format/google.d.cts +1 -1
  282. package/dist/llm/provider_format/google.d.ts +1 -1
  283. package/dist/llm/provider_format/google.d.ts.map +1 -1
  284. package/dist/llm/provider_format/google.js +6 -2
  285. package/dist/llm/provider_format/google.js.map +1 -1
  286. package/dist/llm/provider_format/google.test.cjs +48 -0
  287. package/dist/llm/provider_format/google.test.cjs.map +1 -1
  288. package/dist/llm/provider_format/google.test.js +54 -1
  289. package/dist/llm/provider_format/google.test.js.map +1 -1
  290. package/dist/llm/provider_format/index.cjs +2 -0
  291. package/dist/llm/provider_format/index.cjs.map +1 -1
  292. package/dist/llm/provider_format/index.d.cts +2 -2
  293. package/dist/llm/provider_format/index.d.ts +2 -2
  294. package/dist/llm/provider_format/index.d.ts.map +1 -1
  295. package/dist/llm/provider_format/index.js +6 -1
  296. package/dist/llm/provider_format/index.js.map +1 -1
  297. package/dist/llm/provider_format/openai.cjs +126 -24
  298. package/dist/llm/provider_format/openai.cjs.map +1 -1
  299. package/dist/llm/provider_format/openai.d.cts +1 -0
  300. package/dist/llm/provider_format/openai.d.ts +1 -0
  301. package/dist/llm/provider_format/openai.d.ts.map +1 -1
  302. package/dist/llm/provider_format/openai.js +124 -23
  303. package/dist/llm/provider_format/openai.js.map +1 -1
  304. package/dist/llm/provider_format/openai.test.cjs +393 -0
  305. package/dist/llm/provider_format/openai.test.cjs.map +1 -1
  306. package/dist/llm/provider_format/openai.test.js +400 -2
  307. package/dist/llm/provider_format/openai.test.js.map +1 -1
  308. package/dist/llm/provider_format/utils.cjs +5 -4
  309. package/dist/llm/provider_format/utils.cjs.map +1 -1
  310. package/dist/llm/provider_format/utils.d.ts.map +1 -1
  311. package/dist/llm/provider_format/utils.js +5 -4
  312. package/dist/llm/provider_format/utils.js.map +1 -1
  313. package/dist/llm/realtime.cjs +3 -0
  314. package/dist/llm/realtime.cjs.map +1 -1
  315. package/dist/llm/realtime.d.cts +15 -1
  316. package/dist/llm/realtime.d.ts +15 -1
  317. package/dist/llm/realtime.d.ts.map +1 -1
  318. package/dist/llm/realtime.js +3 -0
  319. package/dist/llm/realtime.js.map +1 -1
  320. package/dist/llm/remote_chat_context.cjs.map +1 -1
  321. package/dist/llm/remote_chat_context.d.cts +2 -0
  322. package/dist/llm/remote_chat_context.d.ts +2 -0
  323. package/dist/llm/remote_chat_context.d.ts.map +1 -1
  324. package/dist/llm/remote_chat_context.js.map +1 -1
  325. package/dist/llm/tool_context.cjs +50 -2
  326. package/dist/llm/tool_context.cjs.map +1 -1
  327. package/dist/llm/tool_context.d.cts +47 -11
  328. package/dist/llm/tool_context.d.ts +47 -11
  329. package/dist/llm/tool_context.d.ts.map +1 -1
  330. package/dist/llm/tool_context.js +48 -3
  331. package/dist/llm/tool_context.js.map +1 -1
  332. package/dist/llm/tool_context.test.cjs +197 -0
  333. package/dist/llm/tool_context.test.cjs.map +1 -1
  334. package/dist/llm/tool_context.test.js +175 -0
  335. package/dist/llm/tool_context.test.js.map +1 -1
  336. package/dist/llm/utils.cjs +18 -12
  337. package/dist/llm/utils.cjs.map +1 -1
  338. package/dist/llm/utils.d.cts +2 -3
  339. package/dist/llm/utils.d.ts +2 -3
  340. package/dist/llm/utils.d.ts.map +1 -1
  341. package/dist/llm/utils.js +18 -12
  342. package/dist/llm/utils.js.map +1 -1
  343. package/dist/llm/zod-utils.cjs +102 -0
  344. package/dist/llm/zod-utils.cjs.map +1 -0
  345. package/dist/llm/zod-utils.d.cts +65 -0
  346. package/dist/llm/zod-utils.d.ts +65 -0
  347. package/dist/llm/zod-utils.d.ts.map +1 -0
  348. package/dist/llm/zod-utils.js +64 -0
  349. package/dist/llm/zod-utils.js.map +1 -0
  350. package/dist/llm/zod-utils.test.cjs +472 -0
  351. package/dist/llm/zod-utils.test.cjs.map +1 -0
  352. package/dist/llm/zod-utils.test.js +455 -0
  353. package/dist/llm/zod-utils.test.js.map +1 -0
  354. package/dist/log.cjs +45 -14
  355. package/dist/log.cjs.map +1 -1
  356. package/dist/log.d.cts +8 -1
  357. package/dist/log.d.ts +8 -1
  358. package/dist/log.d.ts.map +1 -1
  359. package/dist/log.js +45 -15
  360. package/dist/log.js.map +1 -1
  361. package/dist/metrics/base.cjs.map +1 -1
  362. package/dist/metrics/base.d.cts +75 -19
  363. package/dist/metrics/base.d.ts +75 -19
  364. package/dist/metrics/base.d.ts.map +1 -1
  365. package/dist/metrics/index.cjs +5 -0
  366. package/dist/metrics/index.cjs.map +1 -1
  367. package/dist/metrics/index.d.cts +2 -1
  368. package/dist/metrics/index.d.ts +2 -1
  369. package/dist/metrics/index.d.ts.map +1 -1
  370. package/dist/metrics/index.js +6 -0
  371. package/dist/metrics/index.js.map +1 -1
  372. package/dist/metrics/model_usage.cjs +189 -0
  373. package/dist/metrics/model_usage.cjs.map +1 -0
  374. package/dist/metrics/model_usage.d.cts +92 -0
  375. package/dist/metrics/model_usage.d.ts +92 -0
  376. package/dist/metrics/model_usage.d.ts.map +1 -0
  377. package/dist/metrics/model_usage.js +164 -0
  378. package/dist/metrics/model_usage.js.map +1 -0
  379. package/dist/metrics/model_usage.test.cjs +474 -0
  380. package/dist/metrics/model_usage.test.cjs.map +1 -0
  381. package/dist/metrics/model_usage.test.js +476 -0
  382. package/dist/metrics/model_usage.test.js.map +1 -0
  383. package/dist/metrics/usage_collector.cjs +5 -2
  384. package/dist/metrics/usage_collector.cjs.map +1 -1
  385. package/dist/metrics/usage_collector.d.cts +10 -1
  386. package/dist/metrics/usage_collector.d.ts +10 -1
  387. package/dist/metrics/usage_collector.d.ts.map +1 -1
  388. package/dist/metrics/usage_collector.js +5 -2
  389. package/dist/metrics/usage_collector.js.map +1 -1
  390. package/dist/metrics/utils.cjs +23 -7
  391. package/dist/metrics/utils.cjs.map +1 -1
  392. package/dist/metrics/utils.d.ts.map +1 -1
  393. package/dist/metrics/utils.js +23 -7
  394. package/dist/metrics/utils.js.map +1 -1
  395. package/dist/stream/deferred_stream.cjs +31 -10
  396. package/dist/stream/deferred_stream.cjs.map +1 -1
  397. package/dist/stream/deferred_stream.d.cts +6 -1
  398. package/dist/stream/deferred_stream.d.ts +6 -1
  399. package/dist/stream/deferred_stream.d.ts.map +1 -1
  400. package/dist/stream/deferred_stream.js +31 -10
  401. package/dist/stream/deferred_stream.js.map +1 -1
  402. package/dist/stream/deferred_stream.test.cjs +2 -2
  403. package/dist/stream/deferred_stream.test.cjs.map +1 -1
  404. package/dist/stream/deferred_stream.test.js +2 -2
  405. package/dist/stream/deferred_stream.test.js.map +1 -1
  406. package/dist/stream/index.cjs +3 -0
  407. package/dist/stream/index.cjs.map +1 -1
  408. package/dist/stream/index.d.cts +1 -0
  409. package/dist/stream/index.d.ts +1 -0
  410. package/dist/stream/index.d.ts.map +1 -1
  411. package/dist/stream/index.js +2 -0
  412. package/dist/stream/index.js.map +1 -1
  413. package/dist/stream/multi_input_stream.cjs +139 -0
  414. package/dist/stream/multi_input_stream.cjs.map +1 -0
  415. package/dist/stream/multi_input_stream.d.cts +55 -0
  416. package/dist/stream/multi_input_stream.d.ts +55 -0
  417. package/dist/stream/multi_input_stream.d.ts.map +1 -0
  418. package/dist/stream/multi_input_stream.js +115 -0
  419. package/dist/stream/multi_input_stream.js.map +1 -0
  420. package/dist/stream/multi_input_stream.test.cjs +344 -0
  421. package/dist/stream/multi_input_stream.test.cjs.map +1 -0
  422. package/dist/stream/multi_input_stream.test.js +343 -0
  423. package/dist/stream/multi_input_stream.test.js.map +1 -0
  424. package/dist/stream/stream_channel.cjs +39 -1
  425. package/dist/stream/stream_channel.cjs.map +1 -1
  426. package/dist/stream/stream_channel.d.cts +5 -2
  427. package/dist/stream/stream_channel.d.ts +5 -2
  428. package/dist/stream/stream_channel.d.ts.map +1 -1
  429. package/dist/stream/stream_channel.js +39 -1
  430. package/dist/stream/stream_channel.js.map +1 -1
  431. package/dist/stream/stream_channel.test.cjs +27 -0
  432. package/dist/stream/stream_channel.test.cjs.map +1 -1
  433. package/dist/stream/stream_channel.test.js +27 -0
  434. package/dist/stream/stream_channel.test.js.map +1 -1
  435. package/dist/stt/stream_adapter.cjs +24 -9
  436. package/dist/stt/stream_adapter.cjs.map +1 -1
  437. package/dist/stt/stream_adapter.d.cts +7 -3
  438. package/dist/stt/stream_adapter.d.ts +7 -3
  439. package/dist/stt/stream_adapter.d.ts.map +1 -1
  440. package/dist/stt/stream_adapter.js +24 -9
  441. package/dist/stt/stream_adapter.js.map +1 -1
  442. package/dist/stt/stt.cjs +86 -19
  443. package/dist/stt/stt.cjs.map +1 -1
  444. package/dist/stt/stt.d.cts +60 -5
  445. package/dist/stt/stt.d.ts +60 -5
  446. package/dist/stt/stt.d.ts.map +1 -1
  447. package/dist/stt/stt.js +88 -21
  448. package/dist/stt/stt.js.map +1 -1
  449. package/dist/telemetry/index.cjs +72 -0
  450. package/dist/telemetry/index.cjs.map +1 -0
  451. package/dist/telemetry/index.d.cts +7 -0
  452. package/dist/telemetry/index.d.ts +7 -0
  453. package/dist/telemetry/index.d.ts.map +1 -0
  454. package/dist/telemetry/index.js +37 -0
  455. package/dist/telemetry/index.js.map +1 -0
  456. package/dist/telemetry/logging.cjs +65 -0
  457. package/dist/telemetry/logging.cjs.map +1 -0
  458. package/dist/telemetry/logging.d.cts +21 -0
  459. package/dist/telemetry/logging.d.ts +21 -0
  460. package/dist/telemetry/logging.d.ts.map +1 -0
  461. package/dist/telemetry/logging.js +40 -0
  462. package/dist/telemetry/logging.js.map +1 -0
  463. package/dist/telemetry/otel_http_exporter.cjs +166 -0
  464. package/dist/telemetry/otel_http_exporter.cjs.map +1 -0
  465. package/dist/telemetry/otel_http_exporter.d.cts +63 -0
  466. package/dist/telemetry/otel_http_exporter.d.ts +63 -0
  467. package/dist/telemetry/otel_http_exporter.d.ts.map +1 -0
  468. package/dist/telemetry/otel_http_exporter.js +142 -0
  469. package/dist/telemetry/otel_http_exporter.js.map +1 -0
  470. package/dist/telemetry/pino_otel_transport.cjs +217 -0
  471. package/dist/telemetry/pino_otel_transport.cjs.map +1 -0
  472. package/dist/telemetry/pino_otel_transport.d.cts +58 -0
  473. package/dist/telemetry/pino_otel_transport.d.ts +58 -0
  474. package/dist/telemetry/pino_otel_transport.d.ts.map +1 -0
  475. package/dist/telemetry/pino_otel_transport.js +189 -0
  476. package/dist/telemetry/pino_otel_transport.js.map +1 -0
  477. package/dist/telemetry/trace_types.cjs +233 -0
  478. package/dist/telemetry/trace_types.cjs.map +1 -0
  479. package/dist/telemetry/trace_types.d.cts +74 -0
  480. package/dist/telemetry/trace_types.d.ts +74 -0
  481. package/dist/telemetry/trace_types.d.ts.map +1 -0
  482. package/dist/telemetry/trace_types.js +141 -0
  483. package/dist/telemetry/trace_types.js.map +1 -0
  484. package/dist/telemetry/traces.cjs +484 -0
  485. package/dist/telemetry/traces.cjs.map +1 -0
  486. package/dist/telemetry/traces.d.cts +116 -0
  487. package/dist/telemetry/traces.d.ts +116 -0
  488. package/dist/telemetry/traces.d.ts.map +1 -0
  489. package/dist/telemetry/traces.js +449 -0
  490. package/dist/telemetry/traces.js.map +1 -0
  491. package/dist/telemetry/utils.cjs +86 -0
  492. package/dist/telemetry/utils.cjs.map +1 -0
  493. package/dist/telemetry/utils.d.cts +5 -0
  494. package/dist/telemetry/utils.d.ts +5 -0
  495. package/dist/telemetry/utils.d.ts.map +1 -0
  496. package/dist/telemetry/utils.js +51 -0
  497. package/dist/telemetry/utils.js.map +1 -0
  498. package/dist/tokenize/basic/sentence.cjs +3 -3
  499. package/dist/tokenize/basic/sentence.cjs.map +1 -1
  500. package/dist/tokenize/basic/sentence.js +3 -3
  501. package/dist/tokenize/basic/sentence.js.map +1 -1
  502. package/dist/tokenize/tokenizer.test.cjs +3 -1
  503. package/dist/tokenize/tokenizer.test.cjs.map +1 -1
  504. package/dist/tokenize/tokenizer.test.js +3 -1
  505. package/dist/tokenize/tokenizer.test.js.map +1 -1
  506. package/dist/transcription.cjs.map +1 -1
  507. package/dist/transcription.d.cts +6 -0
  508. package/dist/transcription.d.ts +6 -0
  509. package/dist/transcription.d.ts.map +1 -1
  510. package/dist/transcription.js.map +1 -1
  511. package/dist/tts/fallback_adapter.cjs +466 -0
  512. package/dist/tts/fallback_adapter.cjs.map +1 -0
  513. package/dist/tts/fallback_adapter.d.cts +110 -0
  514. package/dist/tts/fallback_adapter.d.ts +110 -0
  515. package/dist/tts/fallback_adapter.d.ts.map +1 -0
  516. package/dist/tts/fallback_adapter.js +442 -0
  517. package/dist/tts/fallback_adapter.js.map +1 -0
  518. package/dist/tts/index.cjs +3 -0
  519. package/dist/tts/index.cjs.map +1 -1
  520. package/dist/tts/index.d.cts +1 -0
  521. package/dist/tts/index.d.ts +1 -0
  522. package/dist/tts/index.d.ts.map +1 -1
  523. package/dist/tts/index.js +2 -0
  524. package/dist/tts/index.js.map +1 -1
  525. package/dist/tts/stream_adapter.cjs +25 -8
  526. package/dist/tts/stream_adapter.cjs.map +1 -1
  527. package/dist/tts/stream_adapter.d.cts +6 -3
  528. package/dist/tts/stream_adapter.d.ts +6 -3
  529. package/dist/tts/stream_adapter.d.ts.map +1 -1
  530. package/dist/tts/stream_adapter.js +25 -8
  531. package/dist/tts/stream_adapter.js.map +1 -1
  532. package/dist/tts/tts.cjs +189 -57
  533. package/dist/tts/tts.cjs.map +1 -1
  534. package/dist/tts/tts.d.cts +58 -6
  535. package/dist/tts/tts.d.ts +58 -6
  536. package/dist/tts/tts.d.ts.map +1 -1
  537. package/dist/tts/tts.js +191 -59
  538. package/dist/tts/tts.js.map +1 -1
  539. package/dist/types.cjs +24 -32
  540. package/dist/types.cjs.map +1 -1
  541. package/dist/types.d.cts +45 -10
  542. package/dist/types.d.ts +45 -10
  543. package/dist/types.d.ts.map +1 -1
  544. package/dist/types.js +20 -30
  545. package/dist/types.js.map +1 -1
  546. package/dist/utils.cjs +122 -26
  547. package/dist/utils.cjs.map +1 -1
  548. package/dist/utils.d.cts +41 -1
  549. package/dist/utils.d.ts +41 -1
  550. package/dist/utils.d.ts.map +1 -1
  551. package/dist/utils.js +117 -25
  552. package/dist/utils.js.map +1 -1
  553. package/dist/utils.test.cjs +73 -1
  554. package/dist/utils.test.cjs.map +1 -1
  555. package/dist/utils.test.js +74 -10
  556. package/dist/utils.test.js.map +1 -1
  557. package/dist/vad.cjs +35 -15
  558. package/dist/vad.cjs.map +1 -1
  559. package/dist/vad.d.cts +15 -5
  560. package/dist/vad.d.ts +15 -5
  561. package/dist/vad.d.ts.map +1 -1
  562. package/dist/vad.js +35 -15
  563. package/dist/vad.js.map +1 -1
  564. package/dist/version.cjs +1 -1
  565. package/dist/version.cjs.map +1 -1
  566. package/dist/version.d.cts +1 -1
  567. package/dist/version.d.ts +1 -1
  568. package/dist/version.d.ts.map +1 -1
  569. package/dist/version.js +1 -1
  570. package/dist/version.js.map +1 -1
  571. package/dist/voice/agent.cjs +258 -35
  572. package/dist/voice/agent.cjs.map +1 -1
  573. package/dist/voice/agent.d.cts +54 -13
  574. package/dist/voice/agent.d.ts +54 -13
  575. package/dist/voice/agent.d.ts.map +1 -1
  576. package/dist/voice/agent.js +254 -34
  577. package/dist/voice/agent.js.map +1 -1
  578. package/dist/voice/agent.test.cjs +314 -0
  579. package/dist/voice/agent.test.cjs.map +1 -1
  580. package/dist/voice/agent.test.js +316 -2
  581. package/dist/voice/agent.test.js.map +1 -1
  582. package/dist/voice/agent_activity.cjs +1116 -385
  583. package/dist/voice/agent_activity.cjs.map +1 -1
  584. package/dist/voice/agent_activity.d.cts +72 -11
  585. package/dist/voice/agent_activity.d.ts +72 -11
  586. package/dist/voice/agent_activity.d.ts.map +1 -1
  587. package/dist/voice/agent_activity.js +1119 -383
  588. package/dist/voice/agent_activity.js.map +1 -1
  589. package/dist/voice/agent_activity.test.cjs +135 -0
  590. package/dist/voice/agent_activity.test.cjs.map +1 -0
  591. package/dist/voice/agent_activity.test.js +134 -0
  592. package/dist/voice/agent_activity.test.js.map +1 -0
  593. package/dist/voice/agent_session.cjs +550 -90
  594. package/dist/voice/agent_session.cjs.map +1 -1
  595. package/dist/voice/agent_session.d.cts +185 -25
  596. package/dist/voice/agent_session.d.ts +185 -25
  597. package/dist/voice/agent_session.d.ts.map +1 -1
  598. package/dist/voice/agent_session.js +556 -91
  599. package/dist/voice/agent_session.js.map +1 -1
  600. package/dist/voice/audio_recognition.cjs +605 -46
  601. package/dist/voice/audio_recognition.cjs.map +1 -1
  602. package/dist/voice/audio_recognition.d.cts +96 -4
  603. package/dist/voice/audio_recognition.d.ts +96 -4
  604. package/dist/voice/audio_recognition.d.ts.map +1 -1
  605. package/dist/voice/audio_recognition.js +611 -47
  606. package/dist/voice/audio_recognition.js.map +1 -1
  607. package/dist/voice/audio_recognition_span.test.cjs +295 -0
  608. package/dist/voice/audio_recognition_span.test.cjs.map +1 -0
  609. package/dist/voice/audio_recognition_span.test.js +299 -0
  610. package/dist/voice/audio_recognition_span.test.js.map +1 -0
  611. package/dist/voice/avatar/datastream_io.cjs +7 -1
  612. package/dist/voice/avatar/datastream_io.cjs.map +1 -1
  613. package/dist/voice/avatar/datastream_io.d.cts +1 -0
  614. package/dist/voice/avatar/datastream_io.d.ts +1 -0
  615. package/dist/voice/avatar/datastream_io.d.ts.map +1 -1
  616. package/dist/voice/avatar/datastream_io.js +7 -1
  617. package/dist/voice/avatar/datastream_io.js.map +1 -1
  618. package/dist/voice/background_audio.cjs +367 -0
  619. package/dist/voice/background_audio.cjs.map +1 -0
  620. package/dist/voice/background_audio.d.cts +123 -0
  621. package/dist/voice/background_audio.d.ts +123 -0
  622. package/dist/voice/background_audio.d.ts.map +1 -0
  623. package/dist/voice/background_audio.js +343 -0
  624. package/dist/voice/background_audio.js.map +1 -0
  625. package/dist/voice/events.cjs +3 -0
  626. package/dist/voice/events.cjs.map +1 -1
  627. package/dist/voice/events.d.cts +16 -9
  628. package/dist/voice/events.d.ts +16 -9
  629. package/dist/voice/events.d.ts.map +1 -1
  630. package/dist/voice/events.js +3 -0
  631. package/dist/voice/events.js.map +1 -1
  632. package/dist/voice/generation.cjs +205 -41
  633. package/dist/voice/generation.cjs.map +1 -1
  634. package/dist/voice/generation.d.cts +21 -5
  635. package/dist/voice/generation.d.ts +21 -5
  636. package/dist/voice/generation.d.ts.map +1 -1
  637. package/dist/voice/generation.js +215 -43
  638. package/dist/voice/generation.js.map +1 -1
  639. package/dist/voice/generation_tools.test.cjs +236 -0
  640. package/dist/voice/generation_tools.test.cjs.map +1 -0
  641. package/dist/voice/generation_tools.test.js +235 -0
  642. package/dist/voice/generation_tools.test.js.map +1 -0
  643. package/dist/voice/index.cjs +33 -2
  644. package/dist/voice/index.cjs.map +1 -1
  645. package/dist/voice/index.d.cts +8 -2
  646. package/dist/voice/index.d.ts +8 -2
  647. package/dist/voice/index.d.ts.map +1 -1
  648. package/dist/voice/index.js +19 -2
  649. package/dist/voice/index.js.map +1 -1
  650. package/dist/voice/interruption_detection.test.cjs +114 -0
  651. package/dist/voice/interruption_detection.test.cjs.map +1 -0
  652. package/dist/voice/interruption_detection.test.js +113 -0
  653. package/dist/voice/interruption_detection.test.js.map +1 -0
  654. package/dist/voice/io.cjs +66 -6
  655. package/dist/voice/io.cjs.map +1 -1
  656. package/dist/voice/io.d.cts +67 -7
  657. package/dist/voice/io.d.ts +67 -7
  658. package/dist/voice/io.d.ts.map +1 -1
  659. package/dist/voice/io.js +62 -5
  660. package/dist/voice/io.js.map +1 -1
  661. package/dist/voice/recorder_io/index.cjs +23 -0
  662. package/dist/voice/recorder_io/index.cjs.map +1 -0
  663. package/dist/voice/recorder_io/index.d.cts +2 -0
  664. package/dist/voice/recorder_io/index.d.ts +2 -0
  665. package/dist/voice/recorder_io/index.d.ts.map +1 -0
  666. package/dist/voice/recorder_io/index.js +2 -0
  667. package/dist/voice/recorder_io/index.js.map +1 -0
  668. package/dist/voice/recorder_io/recorder_io.cjs +607 -0
  669. package/dist/voice/recorder_io/recorder_io.cjs.map +1 -0
  670. package/dist/voice/recorder_io/recorder_io.d.cts +106 -0
  671. package/dist/voice/recorder_io/recorder_io.d.ts +106 -0
  672. package/dist/voice/recorder_io/recorder_io.d.ts.map +1 -0
  673. package/dist/voice/recorder_io/recorder_io.js +573 -0
  674. package/dist/voice/recorder_io/recorder_io.js.map +1 -0
  675. package/dist/voice/remote_session.cjs +922 -0
  676. package/dist/voice/remote_session.cjs.map +1 -0
  677. package/dist/voice/remote_session.d.cts +108 -0
  678. package/dist/voice/remote_session.d.ts +108 -0
  679. package/dist/voice/remote_session.d.ts.map +1 -0
  680. package/dist/voice/remote_session.js +887 -0
  681. package/dist/voice/remote_session.js.map +1 -0
  682. package/dist/voice/report.cjs +88 -0
  683. package/dist/voice/report.cjs.map +1 -0
  684. package/dist/voice/report.d.cts +49 -0
  685. package/dist/voice/report.d.ts +49 -0
  686. package/dist/voice/report.d.ts.map +1 -0
  687. package/dist/voice/report.js +63 -0
  688. package/dist/voice/report.js.map +1 -0
  689. package/dist/voice/report.test.cjs +121 -0
  690. package/dist/voice/report.test.cjs.map +1 -0
  691. package/dist/voice/report.test.js +120 -0
  692. package/dist/voice/report.test.js.map +1 -0
  693. package/dist/voice/room_io/_input.cjs +40 -7
  694. package/dist/voice/room_io/_input.cjs.map +1 -1
  695. package/dist/voice/room_io/_input.d.cts +5 -2
  696. package/dist/voice/room_io/_input.d.ts +5 -2
  697. package/dist/voice/room_io/_input.d.ts.map +1 -1
  698. package/dist/voice/room_io/_input.js +41 -8
  699. package/dist/voice/room_io/_input.js.map +1 -1
  700. package/dist/voice/room_io/_output.cjs +19 -11
  701. package/dist/voice/room_io/_output.cjs.map +1 -1
  702. package/dist/voice/room_io/_output.d.cts +7 -4
  703. package/dist/voice/room_io/_output.d.ts +7 -4
  704. package/dist/voice/room_io/_output.d.ts.map +1 -1
  705. package/dist/voice/room_io/_output.js +20 -12
  706. package/dist/voice/room_io/_output.js.map +1 -1
  707. package/dist/voice/room_io/room_io.cjs +33 -6
  708. package/dist/voice/room_io/room_io.cjs.map +1 -1
  709. package/dist/voice/room_io/room_io.d.cts +29 -9
  710. package/dist/voice/room_io/room_io.d.ts +29 -9
  711. package/dist/voice/room_io/room_io.d.ts.map +1 -1
  712. package/dist/voice/room_io/room_io.js +33 -7
  713. package/dist/voice/room_io/room_io.js.map +1 -1
  714. package/dist/voice/speech_handle.cjs +22 -4
  715. package/dist/voice/speech_handle.cjs.map +1 -1
  716. package/dist/voice/speech_handle.d.cts +17 -2
  717. package/dist/voice/speech_handle.d.ts +17 -2
  718. package/dist/voice/speech_handle.d.ts.map +1 -1
  719. package/dist/voice/speech_handle.js +21 -4
  720. package/dist/voice/speech_handle.js.map +1 -1
  721. package/dist/voice/testing/fake_llm.cjs +127 -0
  722. package/dist/voice/testing/fake_llm.cjs.map +1 -0
  723. package/dist/voice/testing/fake_llm.d.cts +30 -0
  724. package/dist/voice/testing/fake_llm.d.ts +30 -0
  725. package/dist/voice/testing/fake_llm.d.ts.map +1 -0
  726. package/dist/voice/testing/fake_llm.js +103 -0
  727. package/dist/voice/testing/fake_llm.js.map +1 -0
  728. package/dist/voice/testing/index.cjs +57 -0
  729. package/dist/voice/testing/index.cjs.map +1 -0
  730. package/dist/voice/testing/index.d.cts +21 -0
  731. package/dist/voice/testing/index.d.ts +21 -0
  732. package/dist/voice/testing/index.d.ts.map +1 -0
  733. package/dist/voice/testing/index.js +35 -0
  734. package/dist/voice/testing/index.js.map +1 -0
  735. package/dist/voice/testing/run_result.cjs +817 -0
  736. package/dist/voice/testing/run_result.cjs.map +1 -0
  737. package/dist/voice/testing/run_result.d.cts +385 -0
  738. package/dist/voice/testing/run_result.d.ts +385 -0
  739. package/dist/voice/testing/run_result.d.ts.map +1 -0
  740. package/dist/voice/testing/run_result.js +790 -0
  741. package/dist/voice/testing/run_result.js.map +1 -0
  742. package/dist/voice/testing/types.cjs +46 -0
  743. package/dist/voice/testing/types.cjs.map +1 -0
  744. package/dist/voice/testing/types.d.cts +83 -0
  745. package/dist/voice/testing/types.d.ts +83 -0
  746. package/dist/voice/testing/types.d.ts.map +1 -0
  747. package/dist/voice/testing/types.js +19 -0
  748. package/dist/voice/testing/types.js.map +1 -0
  749. package/dist/voice/transcription/synchronizer.cjs +139 -15
  750. package/dist/voice/transcription/synchronizer.cjs.map +1 -1
  751. package/dist/voice/transcription/synchronizer.d.cts +35 -4
  752. package/dist/voice/transcription/synchronizer.d.ts +35 -4
  753. package/dist/voice/transcription/synchronizer.d.ts.map +1 -1
  754. package/dist/voice/transcription/synchronizer.js +143 -16
  755. package/dist/voice/transcription/synchronizer.js.map +1 -1
  756. package/dist/voice/transcription/synchronizer.test.cjs +151 -0
  757. package/dist/voice/transcription/synchronizer.test.cjs.map +1 -0
  758. package/dist/voice/transcription/synchronizer.test.js +150 -0
  759. package/dist/voice/transcription/synchronizer.test.js.map +1 -0
  760. package/dist/voice/turn_config/endpointing.cjs +33 -0
  761. package/dist/voice/turn_config/endpointing.cjs.map +1 -0
  762. package/dist/voice/turn_config/endpointing.d.cts +30 -0
  763. package/dist/voice/turn_config/endpointing.d.ts +30 -0
  764. package/dist/voice/turn_config/endpointing.d.ts.map +1 -0
  765. package/dist/voice/turn_config/endpointing.js +9 -0
  766. package/dist/voice/turn_config/endpointing.js.map +1 -0
  767. package/dist/voice/turn_config/interruption.cjs +37 -0
  768. package/dist/voice/turn_config/interruption.cjs.map +1 -0
  769. package/dist/voice/turn_config/interruption.d.cts +53 -0
  770. package/dist/voice/turn_config/interruption.d.ts +53 -0
  771. package/dist/voice/turn_config/interruption.d.ts.map +1 -0
  772. package/dist/voice/turn_config/interruption.js +13 -0
  773. package/dist/voice/turn_config/interruption.js.map +1 -0
  774. package/dist/voice/turn_config/turn_handling.cjs +35 -0
  775. package/dist/voice/turn_config/turn_handling.cjs.map +1 -0
  776. package/dist/voice/turn_config/turn_handling.d.cts +36 -0
  777. package/dist/voice/turn_config/turn_handling.d.ts +36 -0
  778. package/dist/voice/turn_config/turn_handling.d.ts.map +1 -0
  779. package/dist/voice/turn_config/turn_handling.js +11 -0
  780. package/dist/voice/turn_config/turn_handling.js.map +1 -0
  781. package/dist/voice/turn_config/utils.cjs +157 -0
  782. package/dist/voice/turn_config/utils.cjs.map +1 -0
  783. package/dist/voice/turn_config/utils.d.cts +37 -0
  784. package/dist/voice/turn_config/utils.d.ts +37 -0
  785. package/dist/voice/turn_config/utils.d.ts.map +1 -0
  786. package/dist/voice/turn_config/utils.js +131 -0
  787. package/dist/voice/turn_config/utils.js.map +1 -0
  788. package/dist/voice/turn_config/utils.test.cjs +128 -0
  789. package/dist/voice/turn_config/utils.test.cjs.map +1 -0
  790. package/dist/voice/turn_config/utils.test.js +127 -0
  791. package/dist/voice/turn_config/utils.test.js.map +1 -0
  792. package/dist/voice/utils.cjs +47 -0
  793. package/dist/voice/utils.cjs.map +1 -0
  794. package/dist/voice/utils.d.cts +4 -0
  795. package/dist/voice/utils.d.ts +4 -0
  796. package/dist/voice/utils.d.ts.map +1 -0
  797. package/dist/voice/utils.js +23 -0
  798. package/dist/voice/utils.js.map +1 -0
  799. package/dist/worker.cjs +44 -52
  800. package/dist/worker.cjs.map +1 -1
  801. package/dist/worker.d.cts +18 -8
  802. package/dist/worker.d.ts +18 -8
  803. package/dist/worker.d.ts.map +1 -1
  804. package/dist/worker.js +43 -43
  805. package/dist/worker.js.map +1 -1
  806. package/package.json +32 -12
  807. package/resources/NOTICE +2 -0
  808. package/resources/keyboard-typing.ogg +0 -0
  809. package/resources/keyboard-typing2.ogg +0 -0
  810. package/resources/office-ambience.ogg +0 -0
  811. package/src/audio.ts +132 -1
  812. package/src/beta/index.ts +9 -0
  813. package/src/beta/workflows/index.ts +9 -0
  814. package/src/beta/workflows/task_group.ts +194 -0
  815. package/src/cli.ts +57 -66
  816. package/src/connection_pool.test.ts +346 -0
  817. package/src/connection_pool.ts +307 -0
  818. package/src/constants.ts +14 -0
  819. package/src/cpu.test.ts +239 -0
  820. package/src/cpu.ts +173 -0
  821. package/src/http_server.ts +18 -6
  822. package/src/index.ts +15 -13
  823. package/src/inference/api_protos.ts +85 -2
  824. package/src/inference/index.ts +32 -4
  825. package/src/inference/interruption/defaults.ts +51 -0
  826. package/src/inference/interruption/errors.ts +25 -0
  827. package/src/inference/interruption/http_transport.ts +206 -0
  828. package/src/inference/interruption/interruption_cache_entry.ts +50 -0
  829. package/src/inference/interruption/interruption_detector.ts +204 -0
  830. package/src/inference/interruption/interruption_stream.ts +467 -0
  831. package/src/inference/interruption/types.ts +84 -0
  832. package/src/inference/interruption/utils.test.ts +132 -0
  833. package/src/inference/interruption/utils.ts +137 -0
  834. package/src/inference/interruption/ws_transport.ts +406 -0
  835. package/src/inference/llm.ts +214 -163
  836. package/src/inference/stt.test.ts +253 -0
  837. package/src/inference/stt.ts +449 -208
  838. package/src/inference/tts.test.ts +267 -0
  839. package/src/inference/tts.ts +377 -115
  840. package/src/inference/utils.ts +30 -2
  841. package/src/ipc/inference_proc_executor.ts +11 -3
  842. package/src/ipc/inference_proc_lazy_main.ts +13 -1
  843. package/src/ipc/job_proc_executor.ts +11 -1
  844. package/src/ipc/job_proc_lazy_main.ts +86 -20
  845. package/src/ipc/supervised_proc.test.ts +153 -0
  846. package/src/ipc/supervised_proc.ts +39 -10
  847. package/src/job.ts +120 -1
  848. package/src/language.test.ts +62 -0
  849. package/src/language.ts +380 -0
  850. package/src/llm/__snapshots__/zod-utils.test.ts.snap +559 -0
  851. package/src/llm/chat_context.test.ts +655 -0
  852. package/src/llm/chat_context.ts +412 -2
  853. package/src/llm/fallback_adapter.test.ts +238 -0
  854. package/src/llm/fallback_adapter.ts +391 -0
  855. package/src/llm/index.ts +11 -0
  856. package/src/llm/llm.ts +77 -12
  857. package/src/llm/provider_format/google.test.ts +72 -1
  858. package/src/llm/provider_format/google.ts +10 -6
  859. package/src/llm/provider_format/index.ts +7 -2
  860. package/src/llm/provider_format/openai.test.ts +480 -2
  861. package/src/llm/provider_format/openai.ts +152 -21
  862. package/src/llm/provider_format/utils.ts +11 -5
  863. package/src/llm/realtime.ts +23 -2
  864. package/src/llm/remote_chat_context.ts +2 -2
  865. package/src/llm/tool_context.test.ts +210 -1
  866. package/src/llm/tool_context.ts +115 -17
  867. package/src/llm/utils.ts +24 -16
  868. package/src/llm/zod-utils.test.ts +577 -0
  869. package/src/llm/zod-utils.ts +153 -0
  870. package/src/log.ts +71 -19
  871. package/src/metrics/base.ts +78 -19
  872. package/src/metrics/index.ts +12 -0
  873. package/src/metrics/model_usage.test.ts +545 -0
  874. package/src/metrics/model_usage.ts +262 -0
  875. package/src/metrics/usage_collector.ts +14 -3
  876. package/src/metrics/utils.ts +27 -7
  877. package/src/stream/deferred_stream.test.ts +3 -3
  878. package/src/stream/deferred_stream.ts +43 -11
  879. package/src/stream/index.ts +1 -0
  880. package/src/stream/multi_input_stream.test.ts +545 -0
  881. package/src/stream/multi_input_stream.ts +172 -0
  882. package/src/stream/stream_channel.test.ts +37 -0
  883. package/src/stream/stream_channel.ts +43 -3
  884. package/src/stt/stream_adapter.ts +30 -9
  885. package/src/stt/stt.ts +131 -22
  886. package/src/telemetry/index.ts +28 -0
  887. package/src/telemetry/logging.ts +55 -0
  888. package/src/telemetry/otel_http_exporter.ts +218 -0
  889. package/src/telemetry/pino_otel_transport.ts +265 -0
  890. package/src/telemetry/trace_types.ts +109 -0
  891. package/src/telemetry/traces.ts +673 -0
  892. package/src/telemetry/utils.ts +61 -0
  893. package/src/tokenize/basic/sentence.ts +3 -3
  894. package/src/tokenize/tokenizer.test.ts +4 -0
  895. package/src/transcription.ts +6 -0
  896. package/src/tts/fallback_adapter.ts +579 -0
  897. package/src/tts/index.ts +1 -0
  898. package/src/tts/stream_adapter.ts +38 -8
  899. package/src/tts/tts.ts +245 -62
  900. package/src/types.ts +62 -33
  901. package/src/utils.test.ts +90 -10
  902. package/src/utils.ts +176 -31
  903. package/src/vad.ts +42 -18
  904. package/src/version.ts +1 -1
  905. package/src/voice/agent.test.ts +347 -2
  906. package/src/voice/agent.ts +346 -44
  907. package/src/voice/agent_activity.test.ts +194 -0
  908. package/src/voice/agent_activity.ts +1457 -388
  909. package/src/voice/agent_session.ts +817 -112
  910. package/src/voice/audio_recognition.ts +845 -70
  911. package/src/voice/audio_recognition_span.test.ts +341 -0
  912. package/src/voice/avatar/datastream_io.ts +9 -1
  913. package/src/voice/background_audio.ts +494 -0
  914. package/src/voice/events.ts +27 -7
  915. package/src/voice/generation.ts +310 -56
  916. package/src/voice/generation_tools.test.ts +268 -0
  917. package/src/voice/index.ts +17 -3
  918. package/src/voice/interruption_detection.test.ts +151 -0
  919. package/src/voice/io.ts +115 -12
  920. package/src/voice/recorder_io/index.ts +4 -0
  921. package/src/voice/recorder_io/recorder_io.ts +783 -0
  922. package/src/voice/remote_session.ts +1083 -0
  923. package/src/voice/report.test.ts +136 -0
  924. package/src/voice/report.ts +140 -0
  925. package/src/voice/room_io/_input.ts +45 -10
  926. package/src/voice/room_io/_output.ts +26 -14
  927. package/src/voice/room_io/room_io.ts +67 -22
  928. package/src/voice/speech_handle.ts +38 -6
  929. package/src/voice/testing/fake_llm.ts +138 -0
  930. package/src/voice/testing/index.ts +52 -0
  931. package/src/voice/testing/run_result.ts +995 -0
  932. package/src/voice/testing/types.ts +118 -0
  933. package/src/voice/transcription/synchronizer.test.ts +206 -0
  934. package/src/voice/transcription/synchronizer.ts +204 -19
  935. package/src/voice/turn_config/endpointing.ts +33 -0
  936. package/src/voice/turn_config/interruption.ts +56 -0
  937. package/src/voice/turn_config/turn_handling.ts +45 -0
  938. package/src/voice/turn_config/utils.test.ts +148 -0
  939. package/src/voice/turn_config/utils.ts +167 -0
  940. package/src/voice/utils.ts +29 -0
  941. package/src/worker.ts +92 -78
  942. package/src/llm/__snapshots__/utils.test.ts.snap +0 -65
@@ -0,0 +1,299 @@
1
+ import { ParticipantKind } from "@livekit/rtc-node";
2
+ import { ROOT_CONTEXT, context as otelContext, trace } from "@opentelemetry/api";
3
+ import {
4
+ InMemorySpanExporter,
5
+ SimpleSpanProcessor
6
+ } from "@opentelemetry/sdk-trace-base";
7
+ import { NodeTracerProvider } from "@opentelemetry/sdk-trace-node";
8
+ import { ReadableStream } from "node:stream/web";
9
+ import { describe, expect, it, vi } from "vitest";
10
+ import { ChatContext } from "../llm/chat_context.js";
11
+ import { initializeLogger } from "../log.js";
12
+ import { SpeechEventType } from "../stt/stt.js";
13
+ import { setTracerProvider, tracer } from "../telemetry/index.js";
14
+ import { VAD, VADEventType } from "../vad.js";
15
+ import { AgentSession } from "./agent_session.js";
16
+ import {
17
+ AudioRecognition
18
+ } from "./audio_recognition.js";
19
+ function setupInMemoryTracing() {
20
+ const exporter = new InMemorySpanExporter();
21
+ const provider = new NodeTracerProvider();
22
+ provider.addSpanProcessor(new SimpleSpanProcessor(exporter));
23
+ provider.register();
24
+ setTracerProvider(provider);
25
+ return { exporter };
26
+ }
27
+ function spanByName(spans, name) {
28
+ return spans.find((s) => s.name === name);
29
+ }
30
+ function createFakeSession(rootSpanContext = ROOT_CONTEXT) {
31
+ return {
32
+ _agentState: "listening",
33
+ _roomIO: {
34
+ linkedParticipant: { sid: "p3", identity: "charlie", kind: ParticipantKind.AGENT }
35
+ },
36
+ _setUserAwayTimer: vi.fn(),
37
+ _cancelUserAwayTimer: vi.fn(),
38
+ _userSpeakingSpan: void 0,
39
+ _userState: "listening",
40
+ emit: vi.fn(),
41
+ rootSpanContext
42
+ };
43
+ }
44
+ class FakeVADStream extends Object {
45
+ // We intentionally avoid extending the real VADStream (it is not exported as a value in JS output
46
+ // in some bundling contexts). Instead we emulate the async iterator shape used by AudioRecognition.
47
+ events;
48
+ idx = 0;
49
+ constructor(events) {
50
+ super();
51
+ this.events = events;
52
+ }
53
+ updateInputStream() {
54
+ }
55
+ detachInputStream() {
56
+ }
57
+ close() {
58
+ }
59
+ [Symbol.asyncIterator]() {
60
+ return this;
61
+ }
62
+ async next() {
63
+ if (this.idx >= this.events.length) {
64
+ return { done: true, value: void 0 };
65
+ }
66
+ const value = this.events[this.idx++];
67
+ return { done: false, value };
68
+ }
69
+ }
70
+ class FakeVAD extends VAD {
71
+ label = "fake-vad";
72
+ events;
73
+ constructor(events) {
74
+ super({ updateInterval: 1 });
75
+ this.events = events;
76
+ }
77
+ stream() {
78
+ return new FakeVADStream(this.events);
79
+ }
80
+ }
81
+ const alwaysTrueTurnDetector = {
82
+ model: "test-turn-detector",
83
+ provider: "test-provider",
84
+ supportsLanguage: async () => true,
85
+ unlikelyThreshold: async () => void 0,
86
+ predictEndOfTurn: async () => 1
87
+ };
88
+ describe("AudioRecognition user_turn span parity", () => {
89
+ initializeLogger({ pretty: false, level: "silent" });
90
+ it("creates user_turn and parents eou_detection under it (stt mode)", async () => {
91
+ const { exporter } = setupInMemoryTracing();
92
+ const hooks = {
93
+ onInterruption: vi.fn(),
94
+ onStartOfSpeech: vi.fn(),
95
+ onVADInferenceDone: vi.fn(),
96
+ onEndOfSpeech: vi.fn(),
97
+ onInterimTranscript: vi.fn(),
98
+ onFinalTranscript: vi.fn(),
99
+ onPreemptiveGeneration: vi.fn(),
100
+ retrieveChatCtx: () => ChatContext.empty(),
101
+ onEndOfTurn: vi.fn(async () => true)
102
+ };
103
+ const sttEvents = [
104
+ { type: SpeechEventType.START_OF_SPEECH },
105
+ {
106
+ type: SpeechEventType.FINAL_TRANSCRIPT,
107
+ alternatives: [
108
+ {
109
+ language: "en",
110
+ text: "hello",
111
+ startTime: 0,
112
+ endTime: 0,
113
+ confidence: 0.9
114
+ }
115
+ ]
116
+ },
117
+ { type: SpeechEventType.END_OF_SPEECH }
118
+ ];
119
+ const sttNode = async () => new ReadableStream({
120
+ start(controller) {
121
+ for (const ev of sttEvents) controller.enqueue(ev);
122
+ controller.close();
123
+ }
124
+ });
125
+ const ar = new AudioRecognition({
126
+ recognitionHooks: hooks,
127
+ stt: sttNode,
128
+ vad: void 0,
129
+ turnDetector: alwaysTrueTurnDetector,
130
+ turnDetectionMode: "stt",
131
+ minEndpointingDelay: 0,
132
+ maxEndpointingDelay: 0,
133
+ sttModel: "deepgram-nova2",
134
+ sttProvider: "deepgram",
135
+ getLinkedParticipant: () => ({ sid: "p1", identity: "bob", kind: ParticipantKind.AGENT })
136
+ });
137
+ await ar.start();
138
+ await new Promise((r) => setTimeout(r, 20));
139
+ await ar.close();
140
+ const spans = exporter.getFinishedSpans();
141
+ const userTurn = spanByName(spans, "user_turn");
142
+ const eou = spanByName(spans, "eou_detection");
143
+ expect(userTurn, "user_turn span missing").toBeTruthy();
144
+ expect(eou, "eou_detection span missing").toBeTruthy();
145
+ if (!userTurn || !eou) {
146
+ throw new Error("expected user_turn and eou_detection spans");
147
+ }
148
+ expect(eou.parentSpanId).toBe(userTurn.spanContext().spanId);
149
+ expect(userTurn.attributes["lk.participant_id"]).toBe("p1");
150
+ expect(userTurn.attributes["lk.participant_identity"]).toBe("bob");
151
+ expect(userTurn.attributes["lk.participant_kind"]).toBe("AGENT");
152
+ expect(userTurn.attributes["gen_ai.request.model"]).toBe("deepgram-nova2");
153
+ expect(userTurn.attributes["gen_ai.provider.name"]).toBe("deepgram");
154
+ expect(userTurn.attributes["lk.user_transcript"]).toContain("hello");
155
+ expect(userTurn.attributes["lk.transcript_confidence"]).toBeGreaterThan(0);
156
+ });
157
+ it("creates user_turn from VAD startTime (vad mode) and keeps same parenting", async () => {
158
+ const { exporter } = setupInMemoryTracing();
159
+ const hooks = {
160
+ onInterruption: vi.fn(),
161
+ onStartOfSpeech: vi.fn(),
162
+ onVADInferenceDone: vi.fn(),
163
+ onEndOfSpeech: vi.fn(),
164
+ onInterimTranscript: vi.fn(),
165
+ onFinalTranscript: vi.fn(),
166
+ onPreemptiveGeneration: vi.fn(),
167
+ retrieveChatCtx: () => ChatContext.empty(),
168
+ onEndOfTurn: vi.fn(async () => true)
169
+ };
170
+ const now = Date.now();
171
+ const vadEvents = [
172
+ {
173
+ type: VADEventType.START_OF_SPEECH,
174
+ samplesIndex: 0,
175
+ timestamp: now,
176
+ speechDuration: 100,
177
+ silenceDuration: 0,
178
+ frames: [],
179
+ probability: 0,
180
+ inferenceDuration: 0,
181
+ speaking: true,
182
+ rawAccumulatedSilence: 0,
183
+ rawAccumulatedSpeech: 0
184
+ },
185
+ {
186
+ type: VADEventType.END_OF_SPEECH,
187
+ samplesIndex: 0,
188
+ timestamp: now + 200,
189
+ speechDuration: 100,
190
+ silenceDuration: 100,
191
+ frames: [],
192
+ probability: 0,
193
+ inferenceDuration: 0,
194
+ speaking: false,
195
+ rawAccumulatedSilence: 0,
196
+ rawAccumulatedSpeech: 0
197
+ }
198
+ ];
199
+ const sttEvents = [
200
+ {
201
+ type: SpeechEventType.FINAL_TRANSCRIPT,
202
+ alternatives: [
203
+ {
204
+ language: "en",
205
+ text: "test",
206
+ startTime: 0,
207
+ endTime: 0,
208
+ confidence: 0.8
209
+ }
210
+ ]
211
+ }
212
+ ];
213
+ const sttNode = async () => new ReadableStream({
214
+ start(controller) {
215
+ for (const ev of sttEvents) controller.enqueue(ev);
216
+ controller.close();
217
+ }
218
+ });
219
+ const ar = new AudioRecognition({
220
+ recognitionHooks: hooks,
221
+ stt: sttNode,
222
+ vad: new FakeVAD(vadEvents),
223
+ turnDetector: alwaysTrueTurnDetector,
224
+ turnDetectionMode: "vad",
225
+ minEndpointingDelay: 0,
226
+ maxEndpointingDelay: 0,
227
+ sttModel: "stt-model",
228
+ sttProvider: "stt-provider",
229
+ getLinkedParticipant: () => ({ sid: "p2", identity: "alice", kind: ParticipantKind.AGENT })
230
+ });
231
+ await ar.start();
232
+ await new Promise((r) => setTimeout(r, 20));
233
+ await ar.close();
234
+ const spans = exporter.getFinishedSpans();
235
+ const userTurn = spanByName(spans, "user_turn");
236
+ const eou = spanByName(spans, "eou_detection");
237
+ expect(userTurn).toBeTruthy();
238
+ expect(eou).toBeTruthy();
239
+ if (!userTurn || !eou) {
240
+ throw new Error("expected user_turn and eou_detection spans");
241
+ }
242
+ expect(eou.parentSpanId).toBe(userTurn.spanContext().spanId);
243
+ expect(hooks.onStartOfSpeech).toHaveBeenCalled();
244
+ expect(hooks.onEndOfSpeech).toHaveBeenCalled();
245
+ });
246
+ it("parents user_speaking under user_turn when an explicit speech context is provided", () => {
247
+ const { exporter } = setupInMemoryTracing();
248
+ const sessionSpan = tracer.startSpan({ name: "agent_session", context: ROOT_CONTEXT });
249
+ const sessionContext = trace.setSpan(ROOT_CONTEXT, sessionSpan);
250
+ const fakeSession = createFakeSession(sessionContext);
251
+ const userTurn = tracer.startSpan({ name: "user_turn", context: sessionContext });
252
+ const userTurnContext = trace.setSpan(sessionContext, userTurn);
253
+ const speakingStartedAt = Date.now() - 100;
254
+ const speakingEndedAt = Date.now();
255
+ otelContext.with(userTurnContext, () => {
256
+ AgentSession.prototype._updateUserState.call(fakeSession, "speaking", {
257
+ lastSpeakingTime: speakingStartedAt,
258
+ otelContext: otelContext.active()
259
+ });
260
+ AgentSession.prototype._updateUserState.call(fakeSession, "listening", {
261
+ lastSpeakingTime: speakingEndedAt,
262
+ otelContext: otelContext.active()
263
+ });
264
+ });
265
+ userTurn.end();
266
+ sessionSpan.end();
267
+ const spans = exporter.getFinishedSpans();
268
+ const userSpeaking = spanByName(spans, "user_speaking");
269
+ const exportedUserTurn = spanByName(spans, "user_turn");
270
+ expect(userSpeaking).toBeTruthy();
271
+ expect(exportedUserTurn).toBeTruthy();
272
+ if (!userSpeaking || !exportedUserTurn) {
273
+ throw new Error("expected user_speaking and user_turn spans");
274
+ }
275
+ expect(userSpeaking.parentSpanId).toBe(exportedUserTurn.spanContext().spanId);
276
+ expect(userSpeaking.attributes["lk.participant_id"]).toBe("p3");
277
+ });
278
+ it("keeps user_speaking attached to the session root without an explicit speech context", () => {
279
+ const { exporter } = setupInMemoryTracing();
280
+ const sessionSpan = tracer.startSpan({ name: "agent_session", context: ROOT_CONTEXT });
281
+ const sessionContext = trace.setSpan(ROOT_CONTEXT, sessionSpan);
282
+ const fakeSession = createFakeSession(sessionContext);
283
+ AgentSession.prototype._updateUserState.call(fakeSession, "speaking", {
284
+ lastSpeakingTime: Date.now() - 100
285
+ });
286
+ AgentSession.prototype._updateUserState.call(fakeSession, "listening", {
287
+ lastSpeakingTime: Date.now()
288
+ });
289
+ sessionSpan.end();
290
+ const spans = exporter.getFinishedSpans();
291
+ const userSpeaking = spanByName(spans, "user_speaking");
292
+ expect(userSpeaking).toBeTruthy();
293
+ if (!userSpeaking) {
294
+ throw new Error("expected user_speaking span");
295
+ }
296
+ expect(userSpeaking.parentSpanId).toBe(sessionSpan.spanContext().spanId);
297
+ });
298
+ });
299
+ //# sourceMappingURL=audio_recognition_span.test.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"sources":["../../src/voice/audio_recognition_span.test.ts"],"sourcesContent":["// SPDX-FileCopyrightText: 2026 LiveKit, Inc.\n//\n// SPDX-License-Identifier: Apache-2.0\nimport { ParticipantKind } from '@livekit/rtc-node';\nimport { ROOT_CONTEXT, context as otelContext, trace } from '@opentelemetry/api';\nimport {\n InMemorySpanExporter,\n type ReadableSpan,\n SimpleSpanProcessor,\n} from '@opentelemetry/sdk-trace-base';\nimport { NodeTracerProvider } from '@opentelemetry/sdk-trace-node';\nimport { ReadableStream } from 'node:stream/web';\nimport { describe, expect, it, vi } from 'vitest';\nimport { ChatContext } from '../llm/chat_context.js';\nimport { initializeLogger } from '../log.js';\nimport { type SpeechEvent, SpeechEventType } from '../stt/stt.js';\nimport { setTracerProvider, tracer } from '../telemetry/index.js';\nimport { VAD, type VADEvent, VADEventType, type VADStream } from '../vad.js';\nimport { AgentSession } from './agent_session.js';\nimport {\n AudioRecognition,\n type RecognitionHooks,\n type _TurnDetector,\n} from './audio_recognition.js';\nimport type { STTNode } from './io.js';\n\nfunction setupInMemoryTracing() {\n const exporter = new InMemorySpanExporter();\n const provider = new NodeTracerProvider();\n provider.addSpanProcessor(new SimpleSpanProcessor(exporter));\n provider.register();\n setTracerProvider(provider);\n return { exporter };\n}\n\nfunction spanByName(spans: ReadableSpan[], name: string) {\n return spans.find((s) => s.name === name);\n}\n\nfunction createFakeSession(rootSpanContext = ROOT_CONTEXT): AgentSession {\n return {\n _agentState: 'listening',\n _roomIO: {\n linkedParticipant: { sid: 'p3', identity: 'charlie', kind: ParticipantKind.AGENT },\n },\n _setUserAwayTimer: vi.fn(),\n _cancelUserAwayTimer: vi.fn(),\n _userSpeakingSpan: undefined,\n _userState: 'listening',\n emit: vi.fn(),\n rootSpanContext,\n } as unknown as AgentSession;\n}\n\nclass FakeVADStream extends (Object as unknown as { new (): VADStream }) {\n // We intentionally avoid extending the real VADStream (it is not exported as a value in JS output\n // in some bundling contexts). Instead we emulate the async iterator shape used by AudioRecognition.\n private events: VADEvent[];\n private idx = 0;\n constructor(events: VADEvent[]) {\n super();\n this.events = events;\n }\n updateInputStream() {}\n detachInputStream() {}\n close() {}\n [Symbol.asyncIterator]() {\n return this;\n }\n async next(): Promise<IteratorResult<VADEvent>> {\n if (this.idx >= this.events.length) {\n return { done: true, value: undefined };\n }\n const value = this.events[this.idx++]!;\n return { done: false, value };\n }\n}\n\nclass FakeVAD extends VAD {\n label = 'fake-vad';\n private events: VADEvent[];\n constructor(events: VADEvent[]) {\n super({ updateInterval: 1 });\n this.events = events;\n }\n stream(): any {\n return new FakeVADStream(this.events);\n }\n}\n\nconst alwaysTrueTurnDetector: _TurnDetector = {\n model: 'test-turn-detector',\n provider: 'test-provider',\n supportsLanguage: async () => true,\n unlikelyThreshold: async () => undefined,\n predictEndOfTurn: async () => 1.0,\n};\n\ndescribe('AudioRecognition user_turn span parity', () => {\n initializeLogger({ pretty: false, level: 'silent' });\n\n it('creates user_turn and parents eou_detection under it (stt mode)', async () => {\n const { exporter } = setupInMemoryTracing();\n\n const hooks: RecognitionHooks = {\n onInterruption: vi.fn(),\n onStartOfSpeech: vi.fn(),\n onVADInferenceDone: vi.fn(),\n onEndOfSpeech: vi.fn(),\n onInterimTranscript: vi.fn(),\n onFinalTranscript: vi.fn(),\n onPreemptiveGeneration: vi.fn(),\n retrieveChatCtx: () => ChatContext.empty(),\n onEndOfTurn: vi.fn(async () => true),\n };\n\n const sttEvents: SpeechEvent[] = [\n { type: SpeechEventType.START_OF_SPEECH },\n {\n type: SpeechEventType.FINAL_TRANSCRIPT,\n alternatives: [\n {\n language: 'en',\n text: 'hello',\n startTime: 0,\n endTime: 0,\n confidence: 0.9,\n },\n ],\n },\n { type: SpeechEventType.END_OF_SPEECH },\n ];\n\n const sttNode: STTNode = async () =>\n new ReadableStream<SpeechEvent | string>({\n start(controller) {\n for (const ev of sttEvents) controller.enqueue(ev);\n controller.close();\n },\n });\n\n const ar = new AudioRecognition({\n recognitionHooks: hooks,\n stt: sttNode,\n vad: undefined,\n turnDetector: alwaysTrueTurnDetector,\n turnDetectionMode: 'stt',\n minEndpointingDelay: 0,\n maxEndpointingDelay: 0,\n sttModel: 'deepgram-nova2',\n sttProvider: 'deepgram',\n getLinkedParticipant: () => ({ sid: 'p1', identity: 'bob', kind: ParticipantKind.AGENT }),\n });\n\n await ar.start();\n // allow background task to drain\n await new Promise((r) => setTimeout(r, 20));\n await ar.close();\n\n const spans = exporter.getFinishedSpans();\n const userTurn = spanByName(spans, 'user_turn');\n const eou = spanByName(spans, 'eou_detection');\n expect(userTurn, 'user_turn span missing').toBeTruthy();\n expect(eou, 'eou_detection span missing').toBeTruthy();\n if (!userTurn || !eou) {\n throw new Error('expected user_turn and eou_detection spans');\n }\n\n expect(eou.parentSpanId).toBe(userTurn.spanContext().spanId);\n\n // creation-time attributes\n expect(userTurn.attributes['lk.participant_id']).toBe('p1');\n expect(userTurn.attributes['lk.participant_identity']).toBe('bob');\n expect(userTurn.attributes['lk.participant_kind']).toBe('AGENT');\n expect(userTurn.attributes['gen_ai.request.model']).toBe('deepgram-nova2');\n expect(userTurn.attributes['gen_ai.provider.name']).toBe('deepgram');\n\n // end-of-turn attributes\n expect(userTurn.attributes['lk.user_transcript']).toContain('hello');\n expect(userTurn.attributes['lk.transcript_confidence']).toBeGreaterThan(0);\n });\n\n it('creates user_turn from VAD startTime (vad mode) and keeps same parenting', async () => {\n const { exporter } = setupInMemoryTracing();\n\n const hooks: RecognitionHooks = {\n onInterruption: vi.fn(),\n onStartOfSpeech: vi.fn(),\n onVADInferenceDone: vi.fn(),\n onEndOfSpeech: vi.fn(),\n onInterimTranscript: vi.fn(),\n onFinalTranscript: vi.fn(),\n onPreemptiveGeneration: vi.fn(),\n retrieveChatCtx: () => ChatContext.empty(),\n onEndOfTurn: vi.fn(async () => true),\n };\n\n const now = Date.now();\n const vadEvents: VADEvent[] = [\n {\n type: VADEventType.START_OF_SPEECH,\n samplesIndex: 0,\n timestamp: now,\n speechDuration: 100,\n silenceDuration: 0,\n frames: [],\n probability: 0,\n inferenceDuration: 0,\n speaking: true,\n rawAccumulatedSilence: 0,\n rawAccumulatedSpeech: 0,\n },\n {\n type: VADEventType.END_OF_SPEECH,\n samplesIndex: 0,\n timestamp: now + 200,\n speechDuration: 100,\n silenceDuration: 100,\n frames: [],\n probability: 0,\n inferenceDuration: 0,\n speaking: false,\n rawAccumulatedSilence: 0,\n rawAccumulatedSpeech: 0,\n },\n ];\n\n const sttEvents: SpeechEvent[] = [\n {\n type: SpeechEventType.FINAL_TRANSCRIPT,\n alternatives: [\n {\n language: 'en',\n text: 'test',\n startTime: 0,\n endTime: 0,\n confidence: 0.8,\n },\n ],\n },\n ];\n\n const sttNode: STTNode = async () =>\n new ReadableStream<SpeechEvent | string>({\n start(controller) {\n for (const ev of sttEvents) controller.enqueue(ev);\n controller.close();\n },\n });\n\n const ar = new AudioRecognition({\n recognitionHooks: hooks,\n stt: sttNode,\n vad: new FakeVAD(vadEvents),\n turnDetector: alwaysTrueTurnDetector,\n turnDetectionMode: 'vad',\n minEndpointingDelay: 0,\n maxEndpointingDelay: 0,\n sttModel: 'stt-model',\n sttProvider: 'stt-provider',\n getLinkedParticipant: () => ({ sid: 'p2', identity: 'alice', kind: ParticipantKind.AGENT }),\n });\n\n await ar.start();\n await new Promise((r) => setTimeout(r, 20));\n await ar.close();\n\n const spans = exporter.getFinishedSpans();\n const userTurn = spanByName(spans, 'user_turn');\n const eou = spanByName(spans, 'eou_detection');\n expect(userTurn).toBeTruthy();\n expect(eou).toBeTruthy();\n if (!userTurn || !eou) {\n throw new Error('expected user_turn and eou_detection spans');\n }\n expect(eou.parentSpanId).toBe(userTurn.spanContext().spanId);\n\n expect(hooks.onStartOfSpeech).toHaveBeenCalled();\n expect(hooks.onEndOfSpeech).toHaveBeenCalled();\n });\n\n it('parents user_speaking under user_turn when an explicit speech context is provided', () => {\n const { exporter } = setupInMemoryTracing();\n const sessionSpan = tracer.startSpan({ name: 'agent_session', context: ROOT_CONTEXT });\n const sessionContext = trace.setSpan(ROOT_CONTEXT, sessionSpan);\n const fakeSession = createFakeSession(sessionContext);\n const userTurn = tracer.startSpan({ name: 'user_turn', context: sessionContext });\n const userTurnContext = trace.setSpan(sessionContext, userTurn);\n const speakingStartedAt = Date.now() - 100;\n const speakingEndedAt = Date.now();\n\n otelContext.with(userTurnContext, () => {\n AgentSession.prototype._updateUserState.call(fakeSession, 'speaking', {\n lastSpeakingTime: speakingStartedAt,\n otelContext: otelContext.active(),\n });\n AgentSession.prototype._updateUserState.call(fakeSession, 'listening', {\n lastSpeakingTime: speakingEndedAt,\n otelContext: otelContext.active(),\n });\n });\n\n userTurn.end();\n sessionSpan.end();\n\n const spans = exporter.getFinishedSpans();\n const userSpeaking = spanByName(spans, 'user_speaking');\n const exportedUserTurn = spanByName(spans, 'user_turn');\n expect(userSpeaking).toBeTruthy();\n expect(exportedUserTurn).toBeTruthy();\n if (!userSpeaking || !exportedUserTurn) {\n throw new Error('expected user_speaking and user_turn spans');\n }\n expect(userSpeaking.parentSpanId).toBe(exportedUserTurn.spanContext().spanId);\n expect(userSpeaking.attributes['lk.participant_id']).toBe('p3');\n });\n\n it('keeps user_speaking attached to the session root without an explicit speech context', () => {\n const { exporter } = setupInMemoryTracing();\n const sessionSpan = tracer.startSpan({ name: 'agent_session', context: ROOT_CONTEXT });\n const sessionContext = trace.setSpan(ROOT_CONTEXT, sessionSpan);\n const fakeSession = createFakeSession(sessionContext);\n\n AgentSession.prototype._updateUserState.call(fakeSession, 'speaking', {\n lastSpeakingTime: Date.now() - 100,\n });\n AgentSession.prototype._updateUserState.call(fakeSession, 'listening', {\n lastSpeakingTime: Date.now(),\n });\n\n sessionSpan.end();\n\n const spans = exporter.getFinishedSpans();\n const userSpeaking = spanByName(spans, 'user_speaking');\n expect(userSpeaking).toBeTruthy();\n if (!userSpeaking) {\n throw new Error('expected user_speaking span');\n }\n expect(userSpeaking.parentSpanId).toBe(sessionSpan.spanContext().spanId);\n });\n});\n"],"mappings":"AAGA,SAAS,uBAAuB;AAChC,SAAS,cAAc,WAAW,aAAa,aAAa;AAC5D;AAAA,EACE;AAAA,EAEA;AAAA,OACK;AACP,SAAS,0BAA0B;AACnC,SAAS,sBAAsB;AAC/B,SAAS,UAAU,QAAQ,IAAI,UAAU;AACzC,SAAS,mBAAmB;AAC5B,SAAS,wBAAwB;AACjC,SAA2B,uBAAuB;AAClD,SAAS,mBAAmB,cAAc;AAC1C,SAAS,KAAoB,oBAAoC;AACjE,SAAS,oBAAoB;AAC7B;AAAA,EACE;AAAA,OAGK;AAGP,SAAS,uBAAuB;AAC9B,QAAM,WAAW,IAAI,qBAAqB;AAC1C,QAAM,WAAW,IAAI,mBAAmB;AACxC,WAAS,iBAAiB,IAAI,oBAAoB,QAAQ,CAAC;AAC3D,WAAS,SAAS;AAClB,oBAAkB,QAAQ;AAC1B,SAAO,EAAE,SAAS;AACpB;AAEA,SAAS,WAAW,OAAuB,MAAc;AACvD,SAAO,MAAM,KAAK,CAAC,MAAM,EAAE,SAAS,IAAI;AAC1C;AAEA,SAAS,kBAAkB,kBAAkB,cAA4B;AACvE,SAAO;AAAA,IACL,aAAa;AAAA,IACb,SAAS;AAAA,MACP,mBAAmB,EAAE,KAAK,MAAM,UAAU,WAAW,MAAM,gBAAgB,MAAM;AAAA,IACnF;AAAA,IACA,mBAAmB,GAAG,GAAG;AAAA,IACzB,sBAAsB,GAAG,GAAG;AAAA,IAC5B,mBAAmB;AAAA,IACnB,YAAY;AAAA,IACZ,MAAM,GAAG,GAAG;AAAA,IACZ;AAAA,EACF;AACF;AAEA,MAAM,sBAAuB,OAA4C;AAAA;AAAA;AAAA,EAG/D;AAAA,EACA,MAAM;AAAA,EACd,YAAY,QAAoB;AAC9B,UAAM;AACN,SAAK,SAAS;AAAA,EAChB;AAAA,EACA,oBAAoB;AAAA,EAAC;AAAA,EACrB,oBAAoB;AAAA,EAAC;AAAA,EACrB,QAAQ;AAAA,EAAC;AAAA,EACT,CAAC,OAAO,aAAa,IAAI;AACvB,WAAO;AAAA,EACT;AAAA,EACA,MAAM,OAA0C;AAC9C,QAAI,KAAK,OAAO,KAAK,OAAO,QAAQ;AAClC,aAAO,EAAE,MAAM,MAAM,OAAO,OAAU;AAAA,IACxC;AACA,UAAM,QAAQ,KAAK,OAAO,KAAK,KAAK;AACpC,WAAO,EAAE,MAAM,OAAO,MAAM;AAAA,EAC9B;AACF;AAEA,MAAM,gBAAgB,IAAI;AAAA,EACxB,QAAQ;AAAA,EACA;AAAA,EACR,YAAY,QAAoB;AAC9B,UAAM,EAAE,gBAAgB,EAAE,CAAC;AAC3B,SAAK,SAAS;AAAA,EAChB;AAAA,EACA,SAAc;AACZ,WAAO,IAAI,cAAc,KAAK,MAAM;AAAA,EACtC;AACF;AAEA,MAAM,yBAAwC;AAAA,EAC5C,OAAO;AAAA,EACP,UAAU;AAAA,EACV,kBAAkB,YAAY;AAAA,EAC9B,mBAAmB,YAAY;AAAA,EAC/B,kBAAkB,YAAY;AAChC;AAEA,SAAS,0CAA0C,MAAM;AACvD,mBAAiB,EAAE,QAAQ,OAAO,OAAO,SAAS,CAAC;AAEnD,KAAG,mEAAmE,YAAY;AAChF,UAAM,EAAE,SAAS,IAAI,qBAAqB;AAE1C,UAAM,QAA0B;AAAA,MAC9B,gBAAgB,GAAG,GAAG;AAAA,MACtB,iBAAiB,GAAG,GAAG;AAAA,MACvB,oBAAoB,GAAG,GAAG;AAAA,MAC1B,eAAe,GAAG,GAAG;AAAA,MACrB,qBAAqB,GAAG,GAAG;AAAA,MAC3B,mBAAmB,GAAG,GAAG;AAAA,MACzB,wBAAwB,GAAG,GAAG;AAAA,MAC9B,iBAAiB,MAAM,YAAY,MAAM;AAAA,MACzC,aAAa,GAAG,GAAG,YAAY,IAAI;AAAA,IACrC;AAEA,UAAM,YAA2B;AAAA,MAC/B,EAAE,MAAM,gBAAgB,gBAAgB;AAAA,MACxC;AAAA,QACE,MAAM,gBAAgB;AAAA,QACtB,cAAc;AAAA,UACZ;AAAA,YACE,UAAU;AAAA,YACV,MAAM;AAAA,YACN,WAAW;AAAA,YACX,SAAS;AAAA,YACT,YAAY;AAAA,UACd;AAAA,QACF;AAAA,MACF;AAAA,MACA,EAAE,MAAM,gBAAgB,cAAc;AAAA,IACxC;AAEA,UAAM,UAAmB,YACvB,IAAI,eAAqC;AAAA,MACvC,MAAM,YAAY;AAChB,mBAAW,MAAM,UAAW,YAAW,QAAQ,EAAE;AACjD,mBAAW,MAAM;AAAA,MACnB;AAAA,IACF,CAAC;AAEH,UAAM,KAAK,IAAI,iBAAiB;AAAA,MAC9B,kBAAkB;AAAA,MAClB,KAAK;AAAA,MACL,KAAK;AAAA,MACL,cAAc;AAAA,MACd,mBAAmB;AAAA,MACnB,qBAAqB;AAAA,MACrB,qBAAqB;AAAA,MACrB,UAAU;AAAA,MACV,aAAa;AAAA,MACb,sBAAsB,OAAO,EAAE,KAAK,MAAM,UAAU,OAAO,MAAM,gBAAgB,MAAM;AAAA,IACzF,CAAC;AAED,UAAM,GAAG,MAAM;AAEf,UAAM,IAAI,QAAQ,CAAC,MAAM,WAAW,GAAG,EAAE,CAAC;AAC1C,UAAM,GAAG,MAAM;AAEf,UAAM,QAAQ,SAAS,iBAAiB;AACxC,UAAM,WAAW,WAAW,OAAO,WAAW;AAC9C,UAAM,MAAM,WAAW,OAAO,eAAe;AAC7C,WAAO,UAAU,wBAAwB,EAAE,WAAW;AACtD,WAAO,KAAK,4BAA4B,EAAE,WAAW;AACrD,QAAI,CAAC,YAAY,CAAC,KAAK;AACrB,YAAM,IAAI,MAAM,4CAA4C;AAAA,IAC9D;AAEA,WAAO,IAAI,YAAY,EAAE,KAAK,SAAS,YAAY,EAAE,MAAM;AAG3D,WAAO,SAAS,WAAW,mBAAmB,CAAC,EAAE,KAAK,IAAI;AAC1D,WAAO,SAAS,WAAW,yBAAyB,CAAC,EAAE,KAAK,KAAK;AACjE,WAAO,SAAS,WAAW,qBAAqB,CAAC,EAAE,KAAK,OAAO;AAC/D,WAAO,SAAS,WAAW,sBAAsB,CAAC,EAAE,KAAK,gBAAgB;AACzE,WAAO,SAAS,WAAW,sBAAsB,CAAC,EAAE,KAAK,UAAU;AAGnE,WAAO,SAAS,WAAW,oBAAoB,CAAC,EAAE,UAAU,OAAO;AACnE,WAAO,SAAS,WAAW,0BAA0B,CAAC,EAAE,gBAAgB,CAAC;AAAA,EAC3E,CAAC;AAED,KAAG,4EAA4E,YAAY;AACzF,UAAM,EAAE,SAAS,IAAI,qBAAqB;AAE1C,UAAM,QAA0B;AAAA,MAC9B,gBAAgB,GAAG,GAAG;AAAA,MACtB,iBAAiB,GAAG,GAAG;AAAA,MACvB,oBAAoB,GAAG,GAAG;AAAA,MAC1B,eAAe,GAAG,GAAG;AAAA,MACrB,qBAAqB,GAAG,GAAG;AAAA,MAC3B,mBAAmB,GAAG,GAAG;AAAA,MACzB,wBAAwB,GAAG,GAAG;AAAA,MAC9B,iBAAiB,MAAM,YAAY,MAAM;AAAA,MACzC,aAAa,GAAG,GAAG,YAAY,IAAI;AAAA,IACrC;AAEA,UAAM,MAAM,KAAK,IAAI;AACrB,UAAM,YAAwB;AAAA,MAC5B;AAAA,QACE,MAAM,aAAa;AAAA,QACnB,cAAc;AAAA,QACd,WAAW;AAAA,QACX,gBAAgB;AAAA,QAChB,iBAAiB;AAAA,QACjB,QAAQ,CAAC;AAAA,QACT,aAAa;AAAA,QACb,mBAAmB;AAAA,QACnB,UAAU;AAAA,QACV,uBAAuB;AAAA,QACvB,sBAAsB;AAAA,MACxB;AAAA,MACA;AAAA,QACE,MAAM,aAAa;AAAA,QACnB,cAAc;AAAA,QACd,WAAW,MAAM;AAAA,QACjB,gBAAgB;AAAA,QAChB,iBAAiB;AAAA,QACjB,QAAQ,CAAC;AAAA,QACT,aAAa;AAAA,QACb,mBAAmB;AAAA,QACnB,UAAU;AAAA,QACV,uBAAuB;AAAA,QACvB,sBAAsB;AAAA,MACxB;AAAA,IACF;AAEA,UAAM,YAA2B;AAAA,MAC/B;AAAA,QACE,MAAM,gBAAgB;AAAA,QACtB,cAAc;AAAA,UACZ;AAAA,YACE,UAAU;AAAA,YACV,MAAM;AAAA,YACN,WAAW;AAAA,YACX,SAAS;AAAA,YACT,YAAY;AAAA,UACd;AAAA,QACF;AAAA,MACF;AAAA,IACF;AAEA,UAAM,UAAmB,YACvB,IAAI,eAAqC;AAAA,MACvC,MAAM,YAAY;AAChB,mBAAW,MAAM,UAAW,YAAW,QAAQ,EAAE;AACjD,mBAAW,MAAM;AAAA,MACnB;AAAA,IACF,CAAC;AAEH,UAAM,KAAK,IAAI,iBAAiB;AAAA,MAC9B,kBAAkB;AAAA,MAClB,KAAK;AAAA,MACL,KAAK,IAAI,QAAQ,SAAS;AAAA,MAC1B,cAAc;AAAA,MACd,mBAAmB;AAAA,MACnB,qBAAqB;AAAA,MACrB,qBAAqB;AAAA,MACrB,UAAU;AAAA,MACV,aAAa;AAAA,MACb,sBAAsB,OAAO,EAAE,KAAK,MAAM,UAAU,SAAS,MAAM,gBAAgB,MAAM;AAAA,IAC3F,CAAC;AAED,UAAM,GAAG,MAAM;AACf,UAAM,IAAI,QAAQ,CAAC,MAAM,WAAW,GAAG,EAAE,CAAC;AAC1C,UAAM,GAAG,MAAM;AAEf,UAAM,QAAQ,SAAS,iBAAiB;AACxC,UAAM,WAAW,WAAW,OAAO,WAAW;AAC9C,UAAM,MAAM,WAAW,OAAO,eAAe;AAC7C,WAAO,QAAQ,EAAE,WAAW;AAC5B,WAAO,GAAG,EAAE,WAAW;AACvB,QAAI,CAAC,YAAY,CAAC,KAAK;AACrB,YAAM,IAAI,MAAM,4CAA4C;AAAA,IAC9D;AACA,WAAO,IAAI,YAAY,EAAE,KAAK,SAAS,YAAY,EAAE,MAAM;AAE3D,WAAO,MAAM,eAAe,EAAE,iBAAiB;AAC/C,WAAO,MAAM,aAAa,EAAE,iBAAiB;AAAA,EAC/C,CAAC;AAED,KAAG,qFAAqF,MAAM;AAC5F,UAAM,EAAE,SAAS,IAAI,qBAAqB;AAC1C,UAAM,cAAc,OAAO,UAAU,EAAE,MAAM,iBAAiB,SAAS,aAAa,CAAC;AACrF,UAAM,iBAAiB,MAAM,QAAQ,cAAc,WAAW;AAC9D,UAAM,cAAc,kBAAkB,cAAc;AACpD,UAAM,WAAW,OAAO,UAAU,EAAE,MAAM,aAAa,SAAS,eAAe,CAAC;AAChF,UAAM,kBAAkB,MAAM,QAAQ,gBAAgB,QAAQ;AAC9D,UAAM,oBAAoB,KAAK,IAAI,IAAI;AACvC,UAAM,kBAAkB,KAAK,IAAI;AAEjC,gBAAY,KAAK,iBAAiB,MAAM;AACtC,mBAAa,UAAU,iBAAiB,KAAK,aAAa,YAAY;AAAA,QACpE,kBAAkB;AAAA,QAClB,aAAa,YAAY,OAAO;AAAA,MAClC,CAAC;AACD,mBAAa,UAAU,iBAAiB,KAAK,aAAa,aAAa;AAAA,QACrE,kBAAkB;AAAA,QAClB,aAAa,YAAY,OAAO;AAAA,MAClC,CAAC;AAAA,IACH,CAAC;AAED,aAAS,IAAI;AACb,gBAAY,IAAI;AAEhB,UAAM,QAAQ,SAAS,iBAAiB;AACxC,UAAM,eAAe,WAAW,OAAO,eAAe;AACtD,UAAM,mBAAmB,WAAW,OAAO,WAAW;AACtD,WAAO,YAAY,EAAE,WAAW;AAChC,WAAO,gBAAgB,EAAE,WAAW;AACpC,QAAI,CAAC,gBAAgB,CAAC,kBAAkB;AACtC,YAAM,IAAI,MAAM,4CAA4C;AAAA,IAC9D;AACA,WAAO,aAAa,YAAY,EAAE,KAAK,iBAAiB,YAAY,EAAE,MAAM;AAC5E,WAAO,aAAa,WAAW,mBAAmB,CAAC,EAAE,KAAK,IAAI;AAAA,EAChE,CAAC;AAED,KAAG,uFAAuF,MAAM;AAC9F,UAAM,EAAE,SAAS,IAAI,qBAAqB;AAC1C,UAAM,cAAc,OAAO,UAAU,EAAE,MAAM,iBAAiB,SAAS,aAAa,CAAC;AACrF,UAAM,iBAAiB,MAAM,QAAQ,cAAc,WAAW;AAC9D,UAAM,cAAc,kBAAkB,cAAc;AAEpD,iBAAa,UAAU,iBAAiB,KAAK,aAAa,YAAY;AAAA,MACpE,kBAAkB,KAAK,IAAI,IAAI;AAAA,IACjC,CAAC;AACD,iBAAa,UAAU,iBAAiB,KAAK,aAAa,aAAa;AAAA,MACrE,kBAAkB,KAAK,IAAI;AAAA,IAC7B,CAAC;AAED,gBAAY,IAAI;AAEhB,UAAM,QAAQ,SAAS,iBAAiB;AACxC,UAAM,eAAe,WAAW,OAAO,eAAe;AACtD,WAAO,YAAY,EAAE,WAAW;AAChC,QAAI,CAAC,cAAc;AACjB,YAAM,IAAI,MAAM,6BAA6B;AAAA,IAC/C;AACA,WAAO,aAAa,YAAY,EAAE,KAAK,YAAY,YAAY,EAAE,MAAM;AAAA,EACzE,CAAC;AACH,CAAC;","names":[]}
@@ -41,9 +41,10 @@ class DataStreamAudioOutput extends import_io.AudioOutput {
41
41
  started = false;
42
42
  lock = new import_mutex.Mutex();
43
43
  startTask;
44
+ firstFrameEmitted = false;
44
45
  #logger = (0, import_log.log)();
45
46
  constructor(opts) {
46
- super(opts.sampleRate, void 0);
47
+ super(opts.sampleRate, void 0, { pause: false });
47
48
  const { room, destinationIdentity, sampleRate, waitRemoteTrack } = opts;
48
49
  this.room = room;
49
50
  this.destinationIdentity = destinationIdentity;
@@ -116,6 +117,10 @@ class DataStreamAudioOutput extends import_io.AudioOutput {
116
117
  }
117
118
  await this.startTask.result;
118
119
  await super.captureFrame(frame);
120
+ if (!this.firstFrameEmitted) {
121
+ this.firstFrameEmitted = true;
122
+ this.onPlaybackStarted(Date.now());
123
+ }
119
124
  if (!this.streamWriter) {
120
125
  this.streamWriter = await this.room.localParticipant.streamBytes({
121
126
  name: (0, import_utils.shortuuid)("AUDIO_"),
@@ -139,6 +144,7 @@ class DataStreamAudioOutput extends import_io.AudioOutput {
139
144
  this.streamWriter.close().finally(() => {
140
145
  this.streamWriter = void 0;
141
146
  });
147
+ this.firstFrameEmitted = false;
142
148
  }
143
149
  clearBuffer() {
144
150
  if (!this.started) return;
@@ -1 +1 @@
1
- {"version":3,"sources":["../../../src/voice/avatar/datastream_io.ts"],"sourcesContent":["// SPDX-FileCopyrightText: 2025 LiveKit, Inc.\n//\n// SPDX-License-Identifier: Apache-2.0\nimport { Mutex } from '@livekit/mutex';\nimport {\n type AudioFrame,\n type ByteStreamWriter,\n type Room,\n RoomEvent,\n type RpcInvocationData,\n type TrackKind,\n} from '@livekit/rtc-node';\nimport { log } from '../../log.js';\nimport {\n Future,\n Task,\n shortuuid,\n waitForParticipant,\n waitForTrackPublication,\n} from '../../utils.js';\nimport { AudioOutput, type PlaybackFinishedEvent } from '../io.js';\n\nconst RPC_CLEAR_BUFFER = 'lk.clear_buffer';\nconst RPC_PLAYBACK_FINISHED = 'lk.playback_finished';\nconst AUDIO_STREAM_TOPIC = 'lk.audio_stream';\n\nexport interface DataStreamAudioOutputOptions {\n room: Room;\n destinationIdentity: string;\n sampleRate?: number;\n waitRemoteTrack?: TrackKind;\n}\n\n/**\n * AudioOutput implementation that streams audio to a remote avatar worker using LiveKit DataStream.\n */\nexport class DataStreamAudioOutput extends AudioOutput {\n static _playbackFinishedRpcRegistered: boolean = false;\n static _playbackFinishedHandlers: Record<string, (data: RpcInvocationData) => string> = {};\n\n private room: Room;\n private destinationIdentity: string;\n private roomConnectedFuture: Future<void>;\n private waitRemoteTrack?: TrackKind;\n private streamWriter?: ByteStreamWriter;\n private pushedDuration: number = 0;\n private started: boolean = false;\n private lock = new Mutex();\n private startTask?: Task<void>;\n\n #logger = log();\n\n constructor(opts: DataStreamAudioOutputOptions) {\n super(opts.sampleRate, undefined);\n\n const { room, destinationIdentity, sampleRate, waitRemoteTrack } = opts;\n this.room = room;\n this.destinationIdentity = destinationIdentity;\n this.sampleRate = sampleRate;\n this.waitRemoteTrack = waitRemoteTrack;\n\n const onRoomConnected = async () => {\n if (this.startTask) return;\n\n await this.roomConnectedFuture.await;\n\n // register the rpc method right after the room is connected\n DataStreamAudioOutput.registerPlaybackFinishedRpc({\n room,\n callerIdentity: this.destinationIdentity,\n handler: (data) => this.handlePlaybackFinished(data),\n });\n\n this.startTask = Task.from(({ signal }) => this._start(signal));\n };\n\n this.roomConnectedFuture = new Future<void>();\n\n this.room.on(RoomEvent.ConnectionStateChanged, (_) => {\n if (room.isConnected && !this.roomConnectedFuture.done) {\n this.roomConnectedFuture.resolve(undefined);\n }\n });\n\n if (this.room.isConnected) {\n this.roomConnectedFuture.resolve(undefined);\n }\n\n onRoomConnected();\n }\n\n private async _start(_abortSignal: AbortSignal) {\n const unlock = await this.lock.lock();\n\n try {\n if (this.started) return;\n\n await this.roomConnectedFuture.await;\n\n this.#logger.debug(\n {\n identity: this.destinationIdentity,\n },\n 'waiting for the remote participant',\n );\n\n await waitForParticipant({\n room: this.room,\n identity: this.destinationIdentity,\n });\n\n if (this.waitRemoteTrack) {\n this.#logger.debug(\n {\n identity: this.destinationIdentity,\n kind: this.waitRemoteTrack,\n },\n 'waiting for the remote track',\n );\n\n await waitForTrackPublication({\n room: this.room,\n identity: this.destinationIdentity,\n kind: this.waitRemoteTrack,\n });\n }\n\n this.#logger.debug(\n {\n identity: this.destinationIdentity,\n },\n 'remote participant ready',\n );\n\n this.started = true;\n } finally {\n unlock();\n }\n }\n\n async captureFrame(frame: AudioFrame): Promise<void> {\n if (!this.startTask) {\n this.startTask = Task.from(({ signal }) => this._start(signal));\n }\n\n await this.startTask.result;\n await super.captureFrame(frame);\n\n if (!this.streamWriter) {\n this.streamWriter = await this.room.localParticipant!.streamBytes({\n name: shortuuid('AUDIO_'),\n topic: AUDIO_STREAM_TOPIC,\n destinationIdentities: [this.destinationIdentity],\n attributes: {\n sample_rate: frame.sampleRate.toString(),\n num_channels: frame.channels.toString(),\n },\n });\n this.pushedDuration = 0;\n }\n\n // frame.data is a Int16Array, write accepts a Uint8Array\n await this.streamWriter.write(new Uint8Array(frame.data.buffer));\n this.pushedDuration += frame.samplesPerChannel / frame.sampleRate;\n }\n\n flush(): void {\n super.flush();\n\n if (this.streamWriter === undefined || !this.started) {\n return;\n }\n\n this.streamWriter.close().finally(() => {\n this.streamWriter = undefined;\n });\n }\n\n clearBuffer(): void {\n if (!this.started) return;\n\n this.room.localParticipant!.performRpc({\n destinationIdentity: this.destinationIdentity,\n method: RPC_CLEAR_BUFFER,\n payload: '',\n });\n }\n\n private handlePlaybackFinished(data: RpcInvocationData): string {\n if (data.callerIdentity !== this.destinationIdentity) {\n this.#logger.warn(\n {\n callerIdentity: data.callerIdentity,\n destinationIdentity: this.destinationIdentity,\n },\n 'playback finished event received from unexpected participant',\n );\n return 'reject';\n }\n\n this.#logger.info(\n {\n callerIdentity: data.callerIdentity,\n },\n 'playback finished event received',\n );\n\n const playbackFinishedEvent = JSON.parse(data.payload) as PlaybackFinishedEvent;\n this.onPlaybackFinished(playbackFinishedEvent);\n return 'ok';\n }\n\n static registerPlaybackFinishedRpc({\n room,\n callerIdentity,\n handler,\n }: {\n room: Room;\n callerIdentity: string;\n handler: (data: RpcInvocationData) => string;\n }) {\n DataStreamAudioOutput._playbackFinishedHandlers[callerIdentity] = handler;\n\n if (DataStreamAudioOutput._playbackFinishedRpcRegistered) {\n return;\n }\n\n const rpcHandler = async (data: RpcInvocationData): Promise<string> => {\n const handler = DataStreamAudioOutput._playbackFinishedHandlers[data.callerIdentity];\n if (!handler) {\n log().warn(\n {\n callerIdentity: data.callerIdentity,\n expectedIdentities: Object.keys(DataStreamAudioOutput._playbackFinishedHandlers),\n },\n 'playback finished event received from unexpected participant',\n );\n\n return 'reject';\n }\n return handler(data);\n };\n\n room.localParticipant?.registerRpcMethod(RPC_PLAYBACK_FINISHED, rpcHandler);\n DataStreamAudioOutput._playbackFinishedRpcRegistered = true;\n }\n}\n"],"mappings":";;;;;;;;;;;;;;;;;;AAAA;AAAA;AAAA;AAAA;AAAA;AAGA,mBAAsB;AACtB,sBAOO;AACP,iBAAoB;AACpB,mBAMO;AACP,gBAAwD;AAExD,MAAM,mBAAmB;AACzB,MAAM,wBAAwB;AAC9B,MAAM,qBAAqB;AAYpB,MAAM,8BAA8B,sBAAY;AAAA,EACrD,OAAO,iCAA0C;AAAA,EACjD,OAAO,4BAAiF,CAAC;AAAA,EAEjF;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA,iBAAyB;AAAA,EACzB,UAAmB;AAAA,EACnB,OAAO,IAAI,mBAAM;AAAA,EACjB;AAAA,EAER,cAAU,gBAAI;AAAA,EAEd,YAAY,MAAoC;AAC9C,UAAM,KAAK,YAAY,MAAS;AAEhC,UAAM,EAAE,MAAM,qBAAqB,YAAY,gBAAgB,IAAI;AACnE,SAAK,OAAO;AACZ,SAAK,sBAAsB;AAC3B,SAAK,aAAa;AAClB,SAAK,kBAAkB;AAEvB,UAAM,kBAAkB,YAAY;AAClC,UAAI,KAAK,UAAW;AAEpB,YAAM,KAAK,oBAAoB;AAG/B,4BAAsB,4BAA4B;AAAA,QAChD;AAAA,QACA,gBAAgB,KAAK;AAAA,QACrB,SAAS,CAAC,SAAS,KAAK,uBAAuB,IAAI;AAAA,MACrD,CAAC;AAED,WAAK,YAAY,kBAAK,KAAK,CAAC,EAAE,OAAO,MAAM,KAAK,OAAO,MAAM,CAAC;AAAA,IAChE;AAEA,SAAK,sBAAsB,IAAI,oBAAa;AAE5C,SAAK,KAAK,GAAG,0BAAU,wBAAwB,CAAC,MAAM;AACpD,UAAI,KAAK,eAAe,CAAC,KAAK,oBAAoB,MAAM;AACtD,aAAK,oBAAoB,QAAQ,MAAS;AAAA,MAC5C;AAAA,IACF,CAAC;AAED,QAAI,KAAK,KAAK,aAAa;AACzB,WAAK,oBAAoB,QAAQ,MAAS;AAAA,IAC5C;AAEA,oBAAgB;AAAA,EAClB;AAAA,EAEA,MAAc,OAAO,cAA2B;AAC9C,UAAM,SAAS,MAAM,KAAK,KAAK,KAAK;AAEpC,QAAI;AACF,UAAI,KAAK,QAAS;AAElB,YAAM,KAAK,oBAAoB;AAE/B,WAAK,QAAQ;AAAA,QACX;AAAA,UACE,UAAU,KAAK;AAAA,QACjB;AAAA,QACA;AAAA,MACF;AAEA,gBAAM,iCAAmB;AAAA,QACvB,MAAM,KAAK;AAAA,QACX,UAAU,KAAK;AAAA,MACjB,CAAC;AAED,UAAI,KAAK,iBAAiB;AACxB,aAAK,QAAQ;AAAA,UACX;AAAA,YACE,UAAU,KAAK;AAAA,YACf,MAAM,KAAK;AAAA,UACb;AAAA,UACA;AAAA,QACF;AAEA,kBAAM,sCAAwB;AAAA,UAC5B,MAAM,KAAK;AAAA,UACX,UAAU,KAAK;AAAA,UACf,MAAM,KAAK;AAAA,QACb,CAAC;AAAA,MACH;AAEA,WAAK,QAAQ;AAAA,QACX;AAAA,UACE,UAAU,KAAK;AAAA,QACjB;AAAA,QACA;AAAA,MACF;AAEA,WAAK,UAAU;AAAA,IACjB,UAAE;AACA,aAAO;AAAA,IACT;AAAA,EACF;AAAA,EAEA,MAAM,aAAa,OAAkC;AACnD,QAAI,CAAC,KAAK,WAAW;AACnB,WAAK,YAAY,kBAAK,KAAK,CAAC,EAAE,OAAO,MAAM,KAAK,OAAO,MAAM,CAAC;AAAA,IAChE;AAEA,UAAM,KAAK,UAAU;AACrB,UAAM,MAAM,aAAa,KAAK;AAE9B,QAAI,CAAC,KAAK,cAAc;AACtB,WAAK,eAAe,MAAM,KAAK,KAAK,iBAAkB,YAAY;AAAA,QAChE,UAAM,wBAAU,QAAQ;AAAA,QACxB,OAAO;AAAA,QACP,uBAAuB,CAAC,KAAK,mBAAmB;AAAA,QAChD,YAAY;AAAA,UACV,aAAa,MAAM,WAAW,SAAS;AAAA,UACvC,cAAc,MAAM,SAAS,SAAS;AAAA,QACxC;AAAA,MACF,CAAC;AACD,WAAK,iBAAiB;AAAA,IACxB;AAGA,UAAM,KAAK,aAAa,MAAM,IAAI,WAAW,MAAM,KAAK,MAAM,CAAC;AAC/D,SAAK,kBAAkB,MAAM,oBAAoB,MAAM;AAAA,EACzD;AAAA,EAEA,QAAc;AACZ,UAAM,MAAM;AAEZ,QAAI,KAAK,iBAAiB,UAAa,CAAC,KAAK,SAAS;AACpD;AAAA,IACF;AAEA,SAAK,aAAa,MAAM,EAAE,QAAQ,MAAM;AACtC,WAAK,eAAe;AAAA,IACtB,CAAC;AAAA,EACH;AAAA,EAEA,cAAoB;AAClB,QAAI,CAAC,KAAK,QAAS;AAEnB,SAAK,KAAK,iBAAkB,WAAW;AAAA,MACrC,qBAAqB,KAAK;AAAA,MAC1B,QAAQ;AAAA,MACR,SAAS;AAAA,IACX,CAAC;AAAA,EACH;AAAA,EAEQ,uBAAuB,MAAiC;AAC9D,QAAI,KAAK,mBAAmB,KAAK,qBAAqB;AACpD,WAAK,QAAQ;AAAA,QACX;AAAA,UACE,gBAAgB,KAAK;AAAA,UACrB,qBAAqB,KAAK;AAAA,QAC5B;AAAA,QACA;AAAA,MACF;AACA,aAAO;AAAA,IACT;AAEA,SAAK,QAAQ;AAAA,MACX;AAAA,QACE,gBAAgB,KAAK;AAAA,MACvB;AAAA,MACA;AAAA,IACF;AAEA,UAAM,wBAAwB,KAAK,MAAM,KAAK,OAAO;AACrD,SAAK,mBAAmB,qBAAqB;AAC7C,WAAO;AAAA,EACT;AAAA,EAEA,OAAO,4BAA4B;AAAA,IACjC;AAAA,IACA;AAAA,IACA;AAAA,EACF,GAIG;AA5NL;AA6NI,0BAAsB,0BAA0B,cAAc,IAAI;AAElE,QAAI,sBAAsB,gCAAgC;AACxD;AAAA,IACF;AAEA,UAAM,aAAa,OAAO,SAA6C;AACrE,YAAMA,WAAU,sBAAsB,0BAA0B,KAAK,cAAc;AACnF,UAAI,CAACA,UAAS;AACZ,4BAAI,EAAE;AAAA,UACJ;AAAA,YACE,gBAAgB,KAAK;AAAA,YACrB,oBAAoB,OAAO,KAAK,sBAAsB,yBAAyB;AAAA,UACjF;AAAA,UACA;AAAA,QACF;AAEA,eAAO;AAAA,MACT;AACA,aAAOA,SAAQ,IAAI;AAAA,IACrB;AAEA,eAAK,qBAAL,mBAAuB,kBAAkB,uBAAuB;AAChE,0BAAsB,iCAAiC;AAAA,EACzD;AACF;","names":["handler"]}
1
+ {"version":3,"sources":["../../../src/voice/avatar/datastream_io.ts"],"sourcesContent":["// SPDX-FileCopyrightText: 2025 LiveKit, Inc.\n//\n// SPDX-License-Identifier: Apache-2.0\nimport { Mutex } from '@livekit/mutex';\nimport {\n type AudioFrame,\n type ByteStreamWriter,\n type Room,\n RoomEvent,\n type RpcInvocationData,\n type TrackKind,\n} from '@livekit/rtc-node';\nimport { log } from '../../log.js';\nimport {\n Future,\n Task,\n shortuuid,\n waitForParticipant,\n waitForTrackPublication,\n} from '../../utils.js';\nimport { AudioOutput, type PlaybackFinishedEvent } from '../io.js';\n\nconst RPC_CLEAR_BUFFER = 'lk.clear_buffer';\nconst RPC_PLAYBACK_FINISHED = 'lk.playback_finished';\nconst AUDIO_STREAM_TOPIC = 'lk.audio_stream';\n\nexport interface DataStreamAudioOutputOptions {\n room: Room;\n destinationIdentity: string;\n sampleRate?: number;\n waitRemoteTrack?: TrackKind;\n}\n\n/**\n * AudioOutput implementation that streams audio to a remote avatar worker using LiveKit DataStream.\n */\nexport class DataStreamAudioOutput extends AudioOutput {\n static _playbackFinishedRpcRegistered: boolean = false;\n static _playbackFinishedHandlers: Record<string, (data: RpcInvocationData) => string> = {};\n\n private room: Room;\n private destinationIdentity: string;\n private roomConnectedFuture: Future<void>;\n private waitRemoteTrack?: TrackKind;\n private streamWriter?: ByteStreamWriter;\n private pushedDuration: number = 0;\n private started: boolean = false;\n private lock = new Mutex();\n private startTask?: Task<void>;\n private firstFrameEmitted: boolean = false;\n\n #logger = log();\n\n constructor(opts: DataStreamAudioOutputOptions) {\n super(opts.sampleRate, undefined, { pause: false });\n\n const { room, destinationIdentity, sampleRate, waitRemoteTrack } = opts;\n this.room = room;\n this.destinationIdentity = destinationIdentity;\n this.sampleRate = sampleRate;\n this.waitRemoteTrack = waitRemoteTrack;\n\n const onRoomConnected = async () => {\n if (this.startTask) return;\n\n await this.roomConnectedFuture.await;\n\n // register the rpc method right after the room is connected\n DataStreamAudioOutput.registerPlaybackFinishedRpc({\n room,\n callerIdentity: this.destinationIdentity,\n handler: (data) => this.handlePlaybackFinished(data),\n });\n\n this.startTask = Task.from(({ signal }) => this._start(signal));\n };\n\n this.roomConnectedFuture = new Future<void>();\n\n this.room.on(RoomEvent.ConnectionStateChanged, (_) => {\n if (room.isConnected && !this.roomConnectedFuture.done) {\n this.roomConnectedFuture.resolve(undefined);\n }\n });\n\n if (this.room.isConnected) {\n this.roomConnectedFuture.resolve(undefined);\n }\n\n onRoomConnected();\n }\n\n private async _start(_abortSignal: AbortSignal) {\n const unlock = await this.lock.lock();\n\n try {\n if (this.started) return;\n\n await this.roomConnectedFuture.await;\n\n this.#logger.debug(\n {\n identity: this.destinationIdentity,\n },\n 'waiting for the remote participant',\n );\n\n await waitForParticipant({\n room: this.room,\n identity: this.destinationIdentity,\n });\n\n if (this.waitRemoteTrack) {\n this.#logger.debug(\n {\n identity: this.destinationIdentity,\n kind: this.waitRemoteTrack,\n },\n 'waiting for the remote track',\n );\n\n await waitForTrackPublication({\n room: this.room,\n identity: this.destinationIdentity,\n kind: this.waitRemoteTrack,\n });\n }\n\n this.#logger.debug(\n {\n identity: this.destinationIdentity,\n },\n 'remote participant ready',\n );\n\n this.started = true;\n } finally {\n unlock();\n }\n }\n\n async captureFrame(frame: AudioFrame): Promise<void> {\n if (!this.startTask) {\n this.startTask = Task.from(({ signal }) => this._start(signal));\n }\n\n await this.startTask.result;\n await super.captureFrame(frame);\n\n if (!this.firstFrameEmitted) {\n this.firstFrameEmitted = true;\n this.onPlaybackStarted(Date.now());\n }\n\n if (!this.streamWriter) {\n this.streamWriter = await this.room.localParticipant!.streamBytes({\n name: shortuuid('AUDIO_'),\n topic: AUDIO_STREAM_TOPIC,\n destinationIdentities: [this.destinationIdentity],\n attributes: {\n sample_rate: frame.sampleRate.toString(),\n num_channels: frame.channels.toString(),\n },\n });\n this.pushedDuration = 0;\n }\n\n // frame.data is a Int16Array, write accepts a Uint8Array\n await this.streamWriter.write(new Uint8Array(frame.data.buffer));\n this.pushedDuration += frame.samplesPerChannel / frame.sampleRate;\n }\n\n flush(): void {\n super.flush();\n\n if (this.streamWriter === undefined || !this.started) {\n return;\n }\n\n this.streamWriter.close().finally(() => {\n this.streamWriter = undefined;\n });\n\n this.firstFrameEmitted = false;\n }\n\n clearBuffer(): void {\n if (!this.started) return;\n\n this.room.localParticipant!.performRpc({\n destinationIdentity: this.destinationIdentity,\n method: RPC_CLEAR_BUFFER,\n payload: '',\n });\n }\n\n private handlePlaybackFinished(data: RpcInvocationData): string {\n if (data.callerIdentity !== this.destinationIdentity) {\n this.#logger.warn(\n {\n callerIdentity: data.callerIdentity,\n destinationIdentity: this.destinationIdentity,\n },\n 'playback finished event received from unexpected participant',\n );\n return 'reject';\n }\n\n this.#logger.info(\n {\n callerIdentity: data.callerIdentity,\n },\n 'playback finished event received',\n );\n\n const playbackFinishedEvent = JSON.parse(data.payload) as PlaybackFinishedEvent;\n this.onPlaybackFinished(playbackFinishedEvent);\n return 'ok';\n }\n\n static registerPlaybackFinishedRpc({\n room,\n callerIdentity,\n handler,\n }: {\n room: Room;\n callerIdentity: string;\n handler: (data: RpcInvocationData) => string;\n }) {\n DataStreamAudioOutput._playbackFinishedHandlers[callerIdentity] = handler;\n\n if (DataStreamAudioOutput._playbackFinishedRpcRegistered) {\n return;\n }\n\n const rpcHandler = async (data: RpcInvocationData): Promise<string> => {\n const handler = DataStreamAudioOutput._playbackFinishedHandlers[data.callerIdentity];\n if (!handler) {\n log().warn(\n {\n callerIdentity: data.callerIdentity,\n expectedIdentities: Object.keys(DataStreamAudioOutput._playbackFinishedHandlers),\n },\n 'playback finished event received from unexpected participant',\n );\n\n return 'reject';\n }\n return handler(data);\n };\n\n room.localParticipant?.registerRpcMethod(RPC_PLAYBACK_FINISHED, rpcHandler);\n DataStreamAudioOutput._playbackFinishedRpcRegistered = true;\n }\n}\n"],"mappings":";;;;;;;;;;;;;;;;;;AAAA;AAAA;AAAA;AAAA;AAAA;AAGA,mBAAsB;AACtB,sBAOO;AACP,iBAAoB;AACpB,mBAMO;AACP,gBAAwD;AAExD,MAAM,mBAAmB;AACzB,MAAM,wBAAwB;AAC9B,MAAM,qBAAqB;AAYpB,MAAM,8BAA8B,sBAAY;AAAA,EACrD,OAAO,iCAA0C;AAAA,EACjD,OAAO,4BAAiF,CAAC;AAAA,EAEjF;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA,iBAAyB;AAAA,EACzB,UAAmB;AAAA,EACnB,OAAO,IAAI,mBAAM;AAAA,EACjB;AAAA,EACA,oBAA6B;AAAA,EAErC,cAAU,gBAAI;AAAA,EAEd,YAAY,MAAoC;AAC9C,UAAM,KAAK,YAAY,QAAW,EAAE,OAAO,MAAM,CAAC;AAElD,UAAM,EAAE,MAAM,qBAAqB,YAAY,gBAAgB,IAAI;AACnE,SAAK,OAAO;AACZ,SAAK,sBAAsB;AAC3B,SAAK,aAAa;AAClB,SAAK,kBAAkB;AAEvB,UAAM,kBAAkB,YAAY;AAClC,UAAI,KAAK,UAAW;AAEpB,YAAM,KAAK,oBAAoB;AAG/B,4BAAsB,4BAA4B;AAAA,QAChD;AAAA,QACA,gBAAgB,KAAK;AAAA,QACrB,SAAS,CAAC,SAAS,KAAK,uBAAuB,IAAI;AAAA,MACrD,CAAC;AAED,WAAK,YAAY,kBAAK,KAAK,CAAC,EAAE,OAAO,MAAM,KAAK,OAAO,MAAM,CAAC;AAAA,IAChE;AAEA,SAAK,sBAAsB,IAAI,oBAAa;AAE5C,SAAK,KAAK,GAAG,0BAAU,wBAAwB,CAAC,MAAM;AACpD,UAAI,KAAK,eAAe,CAAC,KAAK,oBAAoB,MAAM;AACtD,aAAK,oBAAoB,QAAQ,MAAS;AAAA,MAC5C;AAAA,IACF,CAAC;AAED,QAAI,KAAK,KAAK,aAAa;AACzB,WAAK,oBAAoB,QAAQ,MAAS;AAAA,IAC5C;AAEA,oBAAgB;AAAA,EAClB;AAAA,EAEA,MAAc,OAAO,cAA2B;AAC9C,UAAM,SAAS,MAAM,KAAK,KAAK,KAAK;AAEpC,QAAI;AACF,UAAI,KAAK,QAAS;AAElB,YAAM,KAAK,oBAAoB;AAE/B,WAAK,QAAQ;AAAA,QACX;AAAA,UACE,UAAU,KAAK;AAAA,QACjB;AAAA,QACA;AAAA,MACF;AAEA,gBAAM,iCAAmB;AAAA,QACvB,MAAM,KAAK;AAAA,QACX,UAAU,KAAK;AAAA,MACjB,CAAC;AAED,UAAI,KAAK,iBAAiB;AACxB,aAAK,QAAQ;AAAA,UACX;AAAA,YACE,UAAU,KAAK;AAAA,YACf,MAAM,KAAK;AAAA,UACb;AAAA,UACA;AAAA,QACF;AAEA,kBAAM,sCAAwB;AAAA,UAC5B,MAAM,KAAK;AAAA,UACX,UAAU,KAAK;AAAA,UACf,MAAM,KAAK;AAAA,QACb,CAAC;AAAA,MACH;AAEA,WAAK,QAAQ;AAAA,QACX;AAAA,UACE,UAAU,KAAK;AAAA,QACjB;AAAA,QACA;AAAA,MACF;AAEA,WAAK,UAAU;AAAA,IACjB,UAAE;AACA,aAAO;AAAA,IACT;AAAA,EACF;AAAA,EAEA,MAAM,aAAa,OAAkC;AACnD,QAAI,CAAC,KAAK,WAAW;AACnB,WAAK,YAAY,kBAAK,KAAK,CAAC,EAAE,OAAO,MAAM,KAAK,OAAO,MAAM,CAAC;AAAA,IAChE;AAEA,UAAM,KAAK,UAAU;AACrB,UAAM,MAAM,aAAa,KAAK;AAE9B,QAAI,CAAC,KAAK,mBAAmB;AAC3B,WAAK,oBAAoB;AACzB,WAAK,kBAAkB,KAAK,IAAI,CAAC;AAAA,IACnC;AAEA,QAAI,CAAC,KAAK,cAAc;AACtB,WAAK,eAAe,MAAM,KAAK,KAAK,iBAAkB,YAAY;AAAA,QAChE,UAAM,wBAAU,QAAQ;AAAA,QACxB,OAAO;AAAA,QACP,uBAAuB,CAAC,KAAK,mBAAmB;AAAA,QAChD,YAAY;AAAA,UACV,aAAa,MAAM,WAAW,SAAS;AAAA,UACvC,cAAc,MAAM,SAAS,SAAS;AAAA,QACxC;AAAA,MACF,CAAC;AACD,WAAK,iBAAiB;AAAA,IACxB;AAGA,UAAM,KAAK,aAAa,MAAM,IAAI,WAAW,MAAM,KAAK,MAAM,CAAC;AAC/D,SAAK,kBAAkB,MAAM,oBAAoB,MAAM;AAAA,EACzD;AAAA,EAEA,QAAc;AACZ,UAAM,MAAM;AAEZ,QAAI,KAAK,iBAAiB,UAAa,CAAC,KAAK,SAAS;AACpD;AAAA,IACF;AAEA,SAAK,aAAa,MAAM,EAAE,QAAQ,MAAM;AACtC,WAAK,eAAe;AAAA,IACtB,CAAC;AAED,SAAK,oBAAoB;AAAA,EAC3B;AAAA,EAEA,cAAoB;AAClB,QAAI,CAAC,KAAK,QAAS;AAEnB,SAAK,KAAK,iBAAkB,WAAW;AAAA,MACrC,qBAAqB,KAAK;AAAA,MAC1B,QAAQ;AAAA,MACR,SAAS;AAAA,IACX,CAAC;AAAA,EACH;AAAA,EAEQ,uBAAuB,MAAiC;AAC9D,QAAI,KAAK,mBAAmB,KAAK,qBAAqB;AACpD,WAAK,QAAQ;AAAA,QACX;AAAA,UACE,gBAAgB,KAAK;AAAA,UACrB,qBAAqB,KAAK;AAAA,QAC5B;AAAA,QACA;AAAA,MACF;AACA,aAAO;AAAA,IACT;AAEA,SAAK,QAAQ;AAAA,MACX;AAAA,QACE,gBAAgB,KAAK;AAAA,MACvB;AAAA,MACA;AAAA,IACF;AAEA,UAAM,wBAAwB,KAAK,MAAM,KAAK,OAAO;AACrD,SAAK,mBAAmB,qBAAqB;AAC7C,WAAO;AAAA,EACT;AAAA,EAEA,OAAO,4BAA4B;AAAA,IACjC;AAAA,IACA;AAAA,IACA;AAAA,EACF,GAIG;AApOL;AAqOI,0BAAsB,0BAA0B,cAAc,IAAI;AAElE,QAAI,sBAAsB,gCAAgC;AACxD;AAAA,IACF;AAEA,UAAM,aAAa,OAAO,SAA6C;AACrE,YAAMA,WAAU,sBAAsB,0BAA0B,KAAK,cAAc;AACnF,UAAI,CAACA,UAAS;AACZ,4BAAI,EAAE;AAAA,UACJ;AAAA,YACE,gBAAgB,KAAK;AAAA,YACrB,oBAAoB,OAAO,KAAK,sBAAsB,yBAAyB;AAAA,UACjF;AAAA,UACA;AAAA,QACF;AAEA,eAAO;AAAA,MACT;AACA,aAAOA,SAAQ,IAAI;AAAA,IACrB;AAEA,eAAK,qBAAL,mBAAuB,kBAAkB,uBAAuB;AAChE,0BAAsB,iCAAiC;AAAA,EACzD;AACF;","names":["handler"]}
@@ -22,6 +22,7 @@ export declare class DataStreamAudioOutput extends AudioOutput {
22
22
  private started;
23
23
  private lock;
24
24
  private startTask?;
25
+ private firstFrameEmitted;
25
26
  constructor(opts: DataStreamAudioOutputOptions);
26
27
  private _start;
27
28
  captureFrame(frame: AudioFrame): Promise<void>;
@@ -22,6 +22,7 @@ export declare class DataStreamAudioOutput extends AudioOutput {
22
22
  private started;
23
23
  private lock;
24
24
  private startTask?;
25
+ private firstFrameEmitted;
25
26
  constructor(opts: DataStreamAudioOutputOptions);
26
27
  private _start;
27
28
  captureFrame(frame: AudioFrame): Promise<void>;
@@ -1 +1 @@
1
- {"version":3,"file":"datastream_io.d.ts","sourceRoot":"","sources":["../../../src/voice/avatar/datastream_io.ts"],"names":[],"mappings":"AAIA,OAAO,EACL,KAAK,UAAU,EAEf,KAAK,IAAI,EAET,KAAK,iBAAiB,EACtB,KAAK,SAAS,EACf,MAAM,mBAAmB,CAAC;AAS3B,OAAO,EAAE,WAAW,EAA8B,MAAM,UAAU,CAAC;AAMnE,MAAM,WAAW,4BAA4B;IAC3C,IAAI,EAAE,IAAI,CAAC;IACX,mBAAmB,EAAE,MAAM,CAAC;IAC5B,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,eAAe,CAAC,EAAE,SAAS,CAAC;CAC7B;AAED;;GAEG;AACH,qBAAa,qBAAsB,SAAQ,WAAW;;IACpD,MAAM,CAAC,8BAA8B,EAAE,OAAO,CAAS;IACvD,MAAM,CAAC,yBAAyB,EAAE,MAAM,CAAC,MAAM,EAAE,CAAC,IAAI,EAAE,iBAAiB,KAAK,MAAM,CAAC,CAAM;IAE3F,OAAO,CAAC,IAAI,CAAO;IACnB,OAAO,CAAC,mBAAmB,CAAS;IACpC,OAAO,CAAC,mBAAmB,CAAe;IAC1C,OAAO,CAAC,eAAe,CAAC,CAAY;IACpC,OAAO,CAAC,YAAY,CAAC,CAAmB;IACxC,OAAO,CAAC,cAAc,CAAa;IACnC,OAAO,CAAC,OAAO,CAAkB;IACjC,OAAO,CAAC,IAAI,CAAe;IAC3B,OAAO,CAAC,SAAS,CAAC,CAAa;gBAInB,IAAI,EAAE,4BAA4B;YAuChC,MAAM;IAiDd,YAAY,CAAC,KAAK,EAAE,UAAU,GAAG,OAAO,CAAC,IAAI,CAAC;IA0BpD,KAAK,IAAI,IAAI;IAYb,WAAW,IAAI,IAAI;IAUnB,OAAO,CAAC,sBAAsB;IAwB9B,MAAM,CAAC,2BAA2B,CAAC,EACjC,IAAI,EACJ,cAAc,EACd,OAAO,GACR,EAAE;QACD,IAAI,EAAE,IAAI,CAAC;QACX,cAAc,EAAE,MAAM,CAAC;QACvB,OAAO,EAAE,CAAC,IAAI,EAAE,iBAAiB,KAAK,MAAM,CAAC;KAC9C;CA0BF"}
1
+ {"version":3,"file":"datastream_io.d.ts","sourceRoot":"","sources":["../../../src/voice/avatar/datastream_io.ts"],"names":[],"mappings":"AAIA,OAAO,EACL,KAAK,UAAU,EAEf,KAAK,IAAI,EAET,KAAK,iBAAiB,EACtB,KAAK,SAAS,EACf,MAAM,mBAAmB,CAAC;AAS3B,OAAO,EAAE,WAAW,EAA8B,MAAM,UAAU,CAAC;AAMnE,MAAM,WAAW,4BAA4B;IAC3C,IAAI,EAAE,IAAI,CAAC;IACX,mBAAmB,EAAE,MAAM,CAAC;IAC5B,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,eAAe,CAAC,EAAE,SAAS,CAAC;CAC7B;AAED;;GAEG;AACH,qBAAa,qBAAsB,SAAQ,WAAW;;IACpD,MAAM,CAAC,8BAA8B,EAAE,OAAO,CAAS;IACvD,MAAM,CAAC,yBAAyB,EAAE,MAAM,CAAC,MAAM,EAAE,CAAC,IAAI,EAAE,iBAAiB,KAAK,MAAM,CAAC,CAAM;IAE3F,OAAO,CAAC,IAAI,CAAO;IACnB,OAAO,CAAC,mBAAmB,CAAS;IACpC,OAAO,CAAC,mBAAmB,CAAe;IAC1C,OAAO,CAAC,eAAe,CAAC,CAAY;IACpC,OAAO,CAAC,YAAY,CAAC,CAAmB;IACxC,OAAO,CAAC,cAAc,CAAa;IACnC,OAAO,CAAC,OAAO,CAAkB;IACjC,OAAO,CAAC,IAAI,CAAe;IAC3B,OAAO,CAAC,SAAS,CAAC,CAAa;IAC/B,OAAO,CAAC,iBAAiB,CAAkB;gBAI/B,IAAI,EAAE,4BAA4B;YAuChC,MAAM;IAiDd,YAAY,CAAC,KAAK,EAAE,UAAU,GAAG,OAAO,CAAC,IAAI,CAAC;IA+BpD,KAAK,IAAI,IAAI;IAcb,WAAW,IAAI,IAAI;IAUnB,OAAO,CAAC,sBAAsB;IAwB9B,MAAM,CAAC,2BAA2B,CAAC,EACjC,IAAI,EACJ,cAAc,EACd,OAAO,GACR,EAAE;QACD,IAAI,EAAE,IAAI,CAAC;QACX,cAAc,EAAE,MAAM,CAAC;QACvB,OAAO,EAAE,CAAC,IAAI,EAAE,iBAAiB,KAAK,MAAM,CAAC;KAC9C;CA0BF"}
@@ -26,9 +26,10 @@ class DataStreamAudioOutput extends AudioOutput {
26
26
  started = false;
27
27
  lock = new Mutex();
28
28
  startTask;
29
+ firstFrameEmitted = false;
29
30
  #logger = log();
30
31
  constructor(opts) {
31
- super(opts.sampleRate, void 0);
32
+ super(opts.sampleRate, void 0, { pause: false });
32
33
  const { room, destinationIdentity, sampleRate, waitRemoteTrack } = opts;
33
34
  this.room = room;
34
35
  this.destinationIdentity = destinationIdentity;
@@ -101,6 +102,10 @@ class DataStreamAudioOutput extends AudioOutput {
101
102
  }
102
103
  await this.startTask.result;
103
104
  await super.captureFrame(frame);
105
+ if (!this.firstFrameEmitted) {
106
+ this.firstFrameEmitted = true;
107
+ this.onPlaybackStarted(Date.now());
108
+ }
104
109
  if (!this.streamWriter) {
105
110
  this.streamWriter = await this.room.localParticipant.streamBytes({
106
111
  name: shortuuid("AUDIO_"),
@@ -124,6 +129,7 @@ class DataStreamAudioOutput extends AudioOutput {
124
129
  this.streamWriter.close().finally(() => {
125
130
  this.streamWriter = void 0;
126
131
  });
132
+ this.firstFrameEmitted = false;
127
133
  }
128
134
  clearBuffer() {
129
135
  if (!this.started) return;
@@ -1 +1 @@
1
- {"version":3,"sources":["../../../src/voice/avatar/datastream_io.ts"],"sourcesContent":["// SPDX-FileCopyrightText: 2025 LiveKit, Inc.\n//\n// SPDX-License-Identifier: Apache-2.0\nimport { Mutex } from '@livekit/mutex';\nimport {\n type AudioFrame,\n type ByteStreamWriter,\n type Room,\n RoomEvent,\n type RpcInvocationData,\n type TrackKind,\n} from '@livekit/rtc-node';\nimport { log } from '../../log.js';\nimport {\n Future,\n Task,\n shortuuid,\n waitForParticipant,\n waitForTrackPublication,\n} from '../../utils.js';\nimport { AudioOutput, type PlaybackFinishedEvent } from '../io.js';\n\nconst RPC_CLEAR_BUFFER = 'lk.clear_buffer';\nconst RPC_PLAYBACK_FINISHED = 'lk.playback_finished';\nconst AUDIO_STREAM_TOPIC = 'lk.audio_stream';\n\nexport interface DataStreamAudioOutputOptions {\n room: Room;\n destinationIdentity: string;\n sampleRate?: number;\n waitRemoteTrack?: TrackKind;\n}\n\n/**\n * AudioOutput implementation that streams audio to a remote avatar worker using LiveKit DataStream.\n */\nexport class DataStreamAudioOutput extends AudioOutput {\n static _playbackFinishedRpcRegistered: boolean = false;\n static _playbackFinishedHandlers: Record<string, (data: RpcInvocationData) => string> = {};\n\n private room: Room;\n private destinationIdentity: string;\n private roomConnectedFuture: Future<void>;\n private waitRemoteTrack?: TrackKind;\n private streamWriter?: ByteStreamWriter;\n private pushedDuration: number = 0;\n private started: boolean = false;\n private lock = new Mutex();\n private startTask?: Task<void>;\n\n #logger = log();\n\n constructor(opts: DataStreamAudioOutputOptions) {\n super(opts.sampleRate, undefined);\n\n const { room, destinationIdentity, sampleRate, waitRemoteTrack } = opts;\n this.room = room;\n this.destinationIdentity = destinationIdentity;\n this.sampleRate = sampleRate;\n this.waitRemoteTrack = waitRemoteTrack;\n\n const onRoomConnected = async () => {\n if (this.startTask) return;\n\n await this.roomConnectedFuture.await;\n\n // register the rpc method right after the room is connected\n DataStreamAudioOutput.registerPlaybackFinishedRpc({\n room,\n callerIdentity: this.destinationIdentity,\n handler: (data) => this.handlePlaybackFinished(data),\n });\n\n this.startTask = Task.from(({ signal }) => this._start(signal));\n };\n\n this.roomConnectedFuture = new Future<void>();\n\n this.room.on(RoomEvent.ConnectionStateChanged, (_) => {\n if (room.isConnected && !this.roomConnectedFuture.done) {\n this.roomConnectedFuture.resolve(undefined);\n }\n });\n\n if (this.room.isConnected) {\n this.roomConnectedFuture.resolve(undefined);\n }\n\n onRoomConnected();\n }\n\n private async _start(_abortSignal: AbortSignal) {\n const unlock = await this.lock.lock();\n\n try {\n if (this.started) return;\n\n await this.roomConnectedFuture.await;\n\n this.#logger.debug(\n {\n identity: this.destinationIdentity,\n },\n 'waiting for the remote participant',\n );\n\n await waitForParticipant({\n room: this.room,\n identity: this.destinationIdentity,\n });\n\n if (this.waitRemoteTrack) {\n this.#logger.debug(\n {\n identity: this.destinationIdentity,\n kind: this.waitRemoteTrack,\n },\n 'waiting for the remote track',\n );\n\n await waitForTrackPublication({\n room: this.room,\n identity: this.destinationIdentity,\n kind: this.waitRemoteTrack,\n });\n }\n\n this.#logger.debug(\n {\n identity: this.destinationIdentity,\n },\n 'remote participant ready',\n );\n\n this.started = true;\n } finally {\n unlock();\n }\n }\n\n async captureFrame(frame: AudioFrame): Promise<void> {\n if (!this.startTask) {\n this.startTask = Task.from(({ signal }) => this._start(signal));\n }\n\n await this.startTask.result;\n await super.captureFrame(frame);\n\n if (!this.streamWriter) {\n this.streamWriter = await this.room.localParticipant!.streamBytes({\n name: shortuuid('AUDIO_'),\n topic: AUDIO_STREAM_TOPIC,\n destinationIdentities: [this.destinationIdentity],\n attributes: {\n sample_rate: frame.sampleRate.toString(),\n num_channels: frame.channels.toString(),\n },\n });\n this.pushedDuration = 0;\n }\n\n // frame.data is a Int16Array, write accepts a Uint8Array\n await this.streamWriter.write(new Uint8Array(frame.data.buffer));\n this.pushedDuration += frame.samplesPerChannel / frame.sampleRate;\n }\n\n flush(): void {\n super.flush();\n\n if (this.streamWriter === undefined || !this.started) {\n return;\n }\n\n this.streamWriter.close().finally(() => {\n this.streamWriter = undefined;\n });\n }\n\n clearBuffer(): void {\n if (!this.started) return;\n\n this.room.localParticipant!.performRpc({\n destinationIdentity: this.destinationIdentity,\n method: RPC_CLEAR_BUFFER,\n payload: '',\n });\n }\n\n private handlePlaybackFinished(data: RpcInvocationData): string {\n if (data.callerIdentity !== this.destinationIdentity) {\n this.#logger.warn(\n {\n callerIdentity: data.callerIdentity,\n destinationIdentity: this.destinationIdentity,\n },\n 'playback finished event received from unexpected participant',\n );\n return 'reject';\n }\n\n this.#logger.info(\n {\n callerIdentity: data.callerIdentity,\n },\n 'playback finished event received',\n );\n\n const playbackFinishedEvent = JSON.parse(data.payload) as PlaybackFinishedEvent;\n this.onPlaybackFinished(playbackFinishedEvent);\n return 'ok';\n }\n\n static registerPlaybackFinishedRpc({\n room,\n callerIdentity,\n handler,\n }: {\n room: Room;\n callerIdentity: string;\n handler: (data: RpcInvocationData) => string;\n }) {\n DataStreamAudioOutput._playbackFinishedHandlers[callerIdentity] = handler;\n\n if (DataStreamAudioOutput._playbackFinishedRpcRegistered) {\n return;\n }\n\n const rpcHandler = async (data: RpcInvocationData): Promise<string> => {\n const handler = DataStreamAudioOutput._playbackFinishedHandlers[data.callerIdentity];\n if (!handler) {\n log().warn(\n {\n callerIdentity: data.callerIdentity,\n expectedIdentities: Object.keys(DataStreamAudioOutput._playbackFinishedHandlers),\n },\n 'playback finished event received from unexpected participant',\n );\n\n return 'reject';\n }\n return handler(data);\n };\n\n room.localParticipant?.registerRpcMethod(RPC_PLAYBACK_FINISHED, rpcHandler);\n DataStreamAudioOutput._playbackFinishedRpcRegistered = true;\n }\n}\n"],"mappings":"AAGA,SAAS,aAAa;AACtB;AAAA,EAIE;AAAA,OAGK;AACP,SAAS,WAAW;AACpB;AAAA,EACE;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,OACK;AACP,SAAS,mBAA+C;AAExD,MAAM,mBAAmB;AACzB,MAAM,wBAAwB;AAC9B,MAAM,qBAAqB;AAYpB,MAAM,8BAA8B,YAAY;AAAA,EACrD,OAAO,iCAA0C;AAAA,EACjD,OAAO,4BAAiF,CAAC;AAAA,EAEjF;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA,iBAAyB;AAAA,EACzB,UAAmB;AAAA,EACnB,OAAO,IAAI,MAAM;AAAA,EACjB;AAAA,EAER,UAAU,IAAI;AAAA,EAEd,YAAY,MAAoC;AAC9C,UAAM,KAAK,YAAY,MAAS;AAEhC,UAAM,EAAE,MAAM,qBAAqB,YAAY,gBAAgB,IAAI;AACnE,SAAK,OAAO;AACZ,SAAK,sBAAsB;AAC3B,SAAK,aAAa;AAClB,SAAK,kBAAkB;AAEvB,UAAM,kBAAkB,YAAY;AAClC,UAAI,KAAK,UAAW;AAEpB,YAAM,KAAK,oBAAoB;AAG/B,4BAAsB,4BAA4B;AAAA,QAChD;AAAA,QACA,gBAAgB,KAAK;AAAA,QACrB,SAAS,CAAC,SAAS,KAAK,uBAAuB,IAAI;AAAA,MACrD,CAAC;AAED,WAAK,YAAY,KAAK,KAAK,CAAC,EAAE,OAAO,MAAM,KAAK,OAAO,MAAM,CAAC;AAAA,IAChE;AAEA,SAAK,sBAAsB,IAAI,OAAa;AAE5C,SAAK,KAAK,GAAG,UAAU,wBAAwB,CAAC,MAAM;AACpD,UAAI,KAAK,eAAe,CAAC,KAAK,oBAAoB,MAAM;AACtD,aAAK,oBAAoB,QAAQ,MAAS;AAAA,MAC5C;AAAA,IACF,CAAC;AAED,QAAI,KAAK,KAAK,aAAa;AACzB,WAAK,oBAAoB,QAAQ,MAAS;AAAA,IAC5C;AAEA,oBAAgB;AAAA,EAClB;AAAA,EAEA,MAAc,OAAO,cAA2B;AAC9C,UAAM,SAAS,MAAM,KAAK,KAAK,KAAK;AAEpC,QAAI;AACF,UAAI,KAAK,QAAS;AAElB,YAAM,KAAK,oBAAoB;AAE/B,WAAK,QAAQ;AAAA,QACX;AAAA,UACE,UAAU,KAAK;AAAA,QACjB;AAAA,QACA;AAAA,MACF;AAEA,YAAM,mBAAmB;AAAA,QACvB,MAAM,KAAK;AAAA,QACX,UAAU,KAAK;AAAA,MACjB,CAAC;AAED,UAAI,KAAK,iBAAiB;AACxB,aAAK,QAAQ;AAAA,UACX;AAAA,YACE,UAAU,KAAK;AAAA,YACf,MAAM,KAAK;AAAA,UACb;AAAA,UACA;AAAA,QACF;AAEA,cAAM,wBAAwB;AAAA,UAC5B,MAAM,KAAK;AAAA,UACX,UAAU,KAAK;AAAA,UACf,MAAM,KAAK;AAAA,QACb,CAAC;AAAA,MACH;AAEA,WAAK,QAAQ;AAAA,QACX;AAAA,UACE,UAAU,KAAK;AAAA,QACjB;AAAA,QACA;AAAA,MACF;AAEA,WAAK,UAAU;AAAA,IACjB,UAAE;AACA,aAAO;AAAA,IACT;AAAA,EACF;AAAA,EAEA,MAAM,aAAa,OAAkC;AACnD,QAAI,CAAC,KAAK,WAAW;AACnB,WAAK,YAAY,KAAK,KAAK,CAAC,EAAE,OAAO,MAAM,KAAK,OAAO,MAAM,CAAC;AAAA,IAChE;AAEA,UAAM,KAAK,UAAU;AACrB,UAAM,MAAM,aAAa,KAAK;AAE9B,QAAI,CAAC,KAAK,cAAc;AACtB,WAAK,eAAe,MAAM,KAAK,KAAK,iBAAkB,YAAY;AAAA,QAChE,MAAM,UAAU,QAAQ;AAAA,QACxB,OAAO;AAAA,QACP,uBAAuB,CAAC,KAAK,mBAAmB;AAAA,QAChD,YAAY;AAAA,UACV,aAAa,MAAM,WAAW,SAAS;AAAA,UACvC,cAAc,MAAM,SAAS,SAAS;AAAA,QACxC;AAAA,MACF,CAAC;AACD,WAAK,iBAAiB;AAAA,IACxB;AAGA,UAAM,KAAK,aAAa,MAAM,IAAI,WAAW,MAAM,KAAK,MAAM,CAAC;AAC/D,SAAK,kBAAkB,MAAM,oBAAoB,MAAM;AAAA,EACzD;AAAA,EAEA,QAAc;AACZ,UAAM,MAAM;AAEZ,QAAI,KAAK,iBAAiB,UAAa,CAAC,KAAK,SAAS;AACpD;AAAA,IACF;AAEA,SAAK,aAAa,MAAM,EAAE,QAAQ,MAAM;AACtC,WAAK,eAAe;AAAA,IACtB,CAAC;AAAA,EACH;AAAA,EAEA,cAAoB;AAClB,QAAI,CAAC,KAAK,QAAS;AAEnB,SAAK,KAAK,iBAAkB,WAAW;AAAA,MACrC,qBAAqB,KAAK;AAAA,MAC1B,QAAQ;AAAA,MACR,SAAS;AAAA,IACX,CAAC;AAAA,EACH;AAAA,EAEQ,uBAAuB,MAAiC;AAC9D,QAAI,KAAK,mBAAmB,KAAK,qBAAqB;AACpD,WAAK,QAAQ;AAAA,QACX;AAAA,UACE,gBAAgB,KAAK;AAAA,UACrB,qBAAqB,KAAK;AAAA,QAC5B;AAAA,QACA;AAAA,MACF;AACA,aAAO;AAAA,IACT;AAEA,SAAK,QAAQ;AAAA,MACX;AAAA,QACE,gBAAgB,KAAK;AAAA,MACvB;AAAA,MACA;AAAA,IACF;AAEA,UAAM,wBAAwB,KAAK,MAAM,KAAK,OAAO;AACrD,SAAK,mBAAmB,qBAAqB;AAC7C,WAAO;AAAA,EACT;AAAA,EAEA,OAAO,4BAA4B;AAAA,IACjC;AAAA,IACA;AAAA,IACA;AAAA,EACF,GAIG;AA5NL;AA6NI,0BAAsB,0BAA0B,cAAc,IAAI;AAElE,QAAI,sBAAsB,gCAAgC;AACxD;AAAA,IACF;AAEA,UAAM,aAAa,OAAO,SAA6C;AACrE,YAAMA,WAAU,sBAAsB,0BAA0B,KAAK,cAAc;AACnF,UAAI,CAACA,UAAS;AACZ,YAAI,EAAE;AAAA,UACJ;AAAA,YACE,gBAAgB,KAAK;AAAA,YACrB,oBAAoB,OAAO,KAAK,sBAAsB,yBAAyB;AAAA,UACjF;AAAA,UACA;AAAA,QACF;AAEA,eAAO;AAAA,MACT;AACA,aAAOA,SAAQ,IAAI;AAAA,IACrB;AAEA,eAAK,qBAAL,mBAAuB,kBAAkB,uBAAuB;AAChE,0BAAsB,iCAAiC;AAAA,EACzD;AACF;","names":["handler"]}
1
+ {"version":3,"sources":["../../../src/voice/avatar/datastream_io.ts"],"sourcesContent":["// SPDX-FileCopyrightText: 2025 LiveKit, Inc.\n//\n// SPDX-License-Identifier: Apache-2.0\nimport { Mutex } from '@livekit/mutex';\nimport {\n type AudioFrame,\n type ByteStreamWriter,\n type Room,\n RoomEvent,\n type RpcInvocationData,\n type TrackKind,\n} from '@livekit/rtc-node';\nimport { log } from '../../log.js';\nimport {\n Future,\n Task,\n shortuuid,\n waitForParticipant,\n waitForTrackPublication,\n} from '../../utils.js';\nimport { AudioOutput, type PlaybackFinishedEvent } from '../io.js';\n\nconst RPC_CLEAR_BUFFER = 'lk.clear_buffer';\nconst RPC_PLAYBACK_FINISHED = 'lk.playback_finished';\nconst AUDIO_STREAM_TOPIC = 'lk.audio_stream';\n\nexport interface DataStreamAudioOutputOptions {\n room: Room;\n destinationIdentity: string;\n sampleRate?: number;\n waitRemoteTrack?: TrackKind;\n}\n\n/**\n * AudioOutput implementation that streams audio to a remote avatar worker using LiveKit DataStream.\n */\nexport class DataStreamAudioOutput extends AudioOutput {\n static _playbackFinishedRpcRegistered: boolean = false;\n static _playbackFinishedHandlers: Record<string, (data: RpcInvocationData) => string> = {};\n\n private room: Room;\n private destinationIdentity: string;\n private roomConnectedFuture: Future<void>;\n private waitRemoteTrack?: TrackKind;\n private streamWriter?: ByteStreamWriter;\n private pushedDuration: number = 0;\n private started: boolean = false;\n private lock = new Mutex();\n private startTask?: Task<void>;\n private firstFrameEmitted: boolean = false;\n\n #logger = log();\n\n constructor(opts: DataStreamAudioOutputOptions) {\n super(opts.sampleRate, undefined, { pause: false });\n\n const { room, destinationIdentity, sampleRate, waitRemoteTrack } = opts;\n this.room = room;\n this.destinationIdentity = destinationIdentity;\n this.sampleRate = sampleRate;\n this.waitRemoteTrack = waitRemoteTrack;\n\n const onRoomConnected = async () => {\n if (this.startTask) return;\n\n await this.roomConnectedFuture.await;\n\n // register the rpc method right after the room is connected\n DataStreamAudioOutput.registerPlaybackFinishedRpc({\n room,\n callerIdentity: this.destinationIdentity,\n handler: (data) => this.handlePlaybackFinished(data),\n });\n\n this.startTask = Task.from(({ signal }) => this._start(signal));\n };\n\n this.roomConnectedFuture = new Future<void>();\n\n this.room.on(RoomEvent.ConnectionStateChanged, (_) => {\n if (room.isConnected && !this.roomConnectedFuture.done) {\n this.roomConnectedFuture.resolve(undefined);\n }\n });\n\n if (this.room.isConnected) {\n this.roomConnectedFuture.resolve(undefined);\n }\n\n onRoomConnected();\n }\n\n private async _start(_abortSignal: AbortSignal) {\n const unlock = await this.lock.lock();\n\n try {\n if (this.started) return;\n\n await this.roomConnectedFuture.await;\n\n this.#logger.debug(\n {\n identity: this.destinationIdentity,\n },\n 'waiting for the remote participant',\n );\n\n await waitForParticipant({\n room: this.room,\n identity: this.destinationIdentity,\n });\n\n if (this.waitRemoteTrack) {\n this.#logger.debug(\n {\n identity: this.destinationIdentity,\n kind: this.waitRemoteTrack,\n },\n 'waiting for the remote track',\n );\n\n await waitForTrackPublication({\n room: this.room,\n identity: this.destinationIdentity,\n kind: this.waitRemoteTrack,\n });\n }\n\n this.#logger.debug(\n {\n identity: this.destinationIdentity,\n },\n 'remote participant ready',\n );\n\n this.started = true;\n } finally {\n unlock();\n }\n }\n\n async captureFrame(frame: AudioFrame): Promise<void> {\n if (!this.startTask) {\n this.startTask = Task.from(({ signal }) => this._start(signal));\n }\n\n await this.startTask.result;\n await super.captureFrame(frame);\n\n if (!this.firstFrameEmitted) {\n this.firstFrameEmitted = true;\n this.onPlaybackStarted(Date.now());\n }\n\n if (!this.streamWriter) {\n this.streamWriter = await this.room.localParticipant!.streamBytes({\n name: shortuuid('AUDIO_'),\n topic: AUDIO_STREAM_TOPIC,\n destinationIdentities: [this.destinationIdentity],\n attributes: {\n sample_rate: frame.sampleRate.toString(),\n num_channels: frame.channels.toString(),\n },\n });\n this.pushedDuration = 0;\n }\n\n // frame.data is a Int16Array, write accepts a Uint8Array\n await this.streamWriter.write(new Uint8Array(frame.data.buffer));\n this.pushedDuration += frame.samplesPerChannel / frame.sampleRate;\n }\n\n flush(): void {\n super.flush();\n\n if (this.streamWriter === undefined || !this.started) {\n return;\n }\n\n this.streamWriter.close().finally(() => {\n this.streamWriter = undefined;\n });\n\n this.firstFrameEmitted = false;\n }\n\n clearBuffer(): void {\n if (!this.started) return;\n\n this.room.localParticipant!.performRpc({\n destinationIdentity: this.destinationIdentity,\n method: RPC_CLEAR_BUFFER,\n payload: '',\n });\n }\n\n private handlePlaybackFinished(data: RpcInvocationData): string {\n if (data.callerIdentity !== this.destinationIdentity) {\n this.#logger.warn(\n {\n callerIdentity: data.callerIdentity,\n destinationIdentity: this.destinationIdentity,\n },\n 'playback finished event received from unexpected participant',\n );\n return 'reject';\n }\n\n this.#logger.info(\n {\n callerIdentity: data.callerIdentity,\n },\n 'playback finished event received',\n );\n\n const playbackFinishedEvent = JSON.parse(data.payload) as PlaybackFinishedEvent;\n this.onPlaybackFinished(playbackFinishedEvent);\n return 'ok';\n }\n\n static registerPlaybackFinishedRpc({\n room,\n callerIdentity,\n handler,\n }: {\n room: Room;\n callerIdentity: string;\n handler: (data: RpcInvocationData) => string;\n }) {\n DataStreamAudioOutput._playbackFinishedHandlers[callerIdentity] = handler;\n\n if (DataStreamAudioOutput._playbackFinishedRpcRegistered) {\n return;\n }\n\n const rpcHandler = async (data: RpcInvocationData): Promise<string> => {\n const handler = DataStreamAudioOutput._playbackFinishedHandlers[data.callerIdentity];\n if (!handler) {\n log().warn(\n {\n callerIdentity: data.callerIdentity,\n expectedIdentities: Object.keys(DataStreamAudioOutput._playbackFinishedHandlers),\n },\n 'playback finished event received from unexpected participant',\n );\n\n return 'reject';\n }\n return handler(data);\n };\n\n room.localParticipant?.registerRpcMethod(RPC_PLAYBACK_FINISHED, rpcHandler);\n DataStreamAudioOutput._playbackFinishedRpcRegistered = true;\n }\n}\n"],"mappings":"AAGA,SAAS,aAAa;AACtB;AAAA,EAIE;AAAA,OAGK;AACP,SAAS,WAAW;AACpB;AAAA,EACE;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,OACK;AACP,SAAS,mBAA+C;AAExD,MAAM,mBAAmB;AACzB,MAAM,wBAAwB;AAC9B,MAAM,qBAAqB;AAYpB,MAAM,8BAA8B,YAAY;AAAA,EACrD,OAAO,iCAA0C;AAAA,EACjD,OAAO,4BAAiF,CAAC;AAAA,EAEjF;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA,iBAAyB;AAAA,EACzB,UAAmB;AAAA,EACnB,OAAO,IAAI,MAAM;AAAA,EACjB;AAAA,EACA,oBAA6B;AAAA,EAErC,UAAU,IAAI;AAAA,EAEd,YAAY,MAAoC;AAC9C,UAAM,KAAK,YAAY,QAAW,EAAE,OAAO,MAAM,CAAC;AAElD,UAAM,EAAE,MAAM,qBAAqB,YAAY,gBAAgB,IAAI;AACnE,SAAK,OAAO;AACZ,SAAK,sBAAsB;AAC3B,SAAK,aAAa;AAClB,SAAK,kBAAkB;AAEvB,UAAM,kBAAkB,YAAY;AAClC,UAAI,KAAK,UAAW;AAEpB,YAAM,KAAK,oBAAoB;AAG/B,4BAAsB,4BAA4B;AAAA,QAChD;AAAA,QACA,gBAAgB,KAAK;AAAA,QACrB,SAAS,CAAC,SAAS,KAAK,uBAAuB,IAAI;AAAA,MACrD,CAAC;AAED,WAAK,YAAY,KAAK,KAAK,CAAC,EAAE,OAAO,MAAM,KAAK,OAAO,MAAM,CAAC;AAAA,IAChE;AAEA,SAAK,sBAAsB,IAAI,OAAa;AAE5C,SAAK,KAAK,GAAG,UAAU,wBAAwB,CAAC,MAAM;AACpD,UAAI,KAAK,eAAe,CAAC,KAAK,oBAAoB,MAAM;AACtD,aAAK,oBAAoB,QAAQ,MAAS;AAAA,MAC5C;AAAA,IACF,CAAC;AAED,QAAI,KAAK,KAAK,aAAa;AACzB,WAAK,oBAAoB,QAAQ,MAAS;AAAA,IAC5C;AAEA,oBAAgB;AAAA,EAClB;AAAA,EAEA,MAAc,OAAO,cAA2B;AAC9C,UAAM,SAAS,MAAM,KAAK,KAAK,KAAK;AAEpC,QAAI;AACF,UAAI,KAAK,QAAS;AAElB,YAAM,KAAK,oBAAoB;AAE/B,WAAK,QAAQ;AAAA,QACX;AAAA,UACE,UAAU,KAAK;AAAA,QACjB;AAAA,QACA;AAAA,MACF;AAEA,YAAM,mBAAmB;AAAA,QACvB,MAAM,KAAK;AAAA,QACX,UAAU,KAAK;AAAA,MACjB,CAAC;AAED,UAAI,KAAK,iBAAiB;AACxB,aAAK,QAAQ;AAAA,UACX;AAAA,YACE,UAAU,KAAK;AAAA,YACf,MAAM,KAAK;AAAA,UACb;AAAA,UACA;AAAA,QACF;AAEA,cAAM,wBAAwB;AAAA,UAC5B,MAAM,KAAK;AAAA,UACX,UAAU,KAAK;AAAA,UACf,MAAM,KAAK;AAAA,QACb,CAAC;AAAA,MACH;AAEA,WAAK,QAAQ;AAAA,QACX;AAAA,UACE,UAAU,KAAK;AAAA,QACjB;AAAA,QACA;AAAA,MACF;AAEA,WAAK,UAAU;AAAA,IACjB,UAAE;AACA,aAAO;AAAA,IACT;AAAA,EACF;AAAA,EAEA,MAAM,aAAa,OAAkC;AACnD,QAAI,CAAC,KAAK,WAAW;AACnB,WAAK,YAAY,KAAK,KAAK,CAAC,EAAE,OAAO,MAAM,KAAK,OAAO,MAAM,CAAC;AAAA,IAChE;AAEA,UAAM,KAAK,UAAU;AACrB,UAAM,MAAM,aAAa,KAAK;AAE9B,QAAI,CAAC,KAAK,mBAAmB;AAC3B,WAAK,oBAAoB;AACzB,WAAK,kBAAkB,KAAK,IAAI,CAAC;AAAA,IACnC;AAEA,QAAI,CAAC,KAAK,cAAc;AACtB,WAAK,eAAe,MAAM,KAAK,KAAK,iBAAkB,YAAY;AAAA,QAChE,MAAM,UAAU,QAAQ;AAAA,QACxB,OAAO;AAAA,QACP,uBAAuB,CAAC,KAAK,mBAAmB;AAAA,QAChD,YAAY;AAAA,UACV,aAAa,MAAM,WAAW,SAAS;AAAA,UACvC,cAAc,MAAM,SAAS,SAAS;AAAA,QACxC;AAAA,MACF,CAAC;AACD,WAAK,iBAAiB;AAAA,IACxB;AAGA,UAAM,KAAK,aAAa,MAAM,IAAI,WAAW,MAAM,KAAK,MAAM,CAAC;AAC/D,SAAK,kBAAkB,MAAM,oBAAoB,MAAM;AAAA,EACzD;AAAA,EAEA,QAAc;AACZ,UAAM,MAAM;AAEZ,QAAI,KAAK,iBAAiB,UAAa,CAAC,KAAK,SAAS;AACpD;AAAA,IACF;AAEA,SAAK,aAAa,MAAM,EAAE,QAAQ,MAAM;AACtC,WAAK,eAAe;AAAA,IACtB,CAAC;AAED,SAAK,oBAAoB;AAAA,EAC3B;AAAA,EAEA,cAAoB;AAClB,QAAI,CAAC,KAAK,QAAS;AAEnB,SAAK,KAAK,iBAAkB,WAAW;AAAA,MACrC,qBAAqB,KAAK;AAAA,MAC1B,QAAQ;AAAA,MACR,SAAS;AAAA,IACX,CAAC;AAAA,EACH;AAAA,EAEQ,uBAAuB,MAAiC;AAC9D,QAAI,KAAK,mBAAmB,KAAK,qBAAqB;AACpD,WAAK,QAAQ;AAAA,QACX;AAAA,UACE,gBAAgB,KAAK;AAAA,UACrB,qBAAqB,KAAK;AAAA,QAC5B;AAAA,QACA;AAAA,MACF;AACA,aAAO;AAAA,IACT;AAEA,SAAK,QAAQ;AAAA,MACX;AAAA,QACE,gBAAgB,KAAK;AAAA,MACvB;AAAA,MACA;AAAA,IACF;AAEA,UAAM,wBAAwB,KAAK,MAAM,KAAK,OAAO;AACrD,SAAK,mBAAmB,qBAAqB;AAC7C,WAAO;AAAA,EACT;AAAA,EAEA,OAAO,4BAA4B;AAAA,IACjC;AAAA,IACA;AAAA,IACA;AAAA,EACF,GAIG;AApOL;AAqOI,0BAAsB,0BAA0B,cAAc,IAAI;AAElE,QAAI,sBAAsB,gCAAgC;AACxD;AAAA,IACF;AAEA,UAAM,aAAa,OAAO,SAA6C;AACrE,YAAMA,WAAU,sBAAsB,0BAA0B,KAAK,cAAc;AACnF,UAAI,CAACA,UAAS;AACZ,YAAI,EAAE;AAAA,UACJ;AAAA,YACE,gBAAgB,KAAK;AAAA,YACrB,oBAAoB,OAAO,KAAK,sBAAsB,yBAAyB;AAAA,UACjF;AAAA,UACA;AAAA,QACF;AAEA,eAAO;AAAA,MACT;AACA,aAAOA,SAAQ,IAAI;AAAA,IACrB;AAEA,eAAK,qBAAL,mBAAuB,kBAAkB,uBAAuB;AAChE,0BAAsB,iCAAiC;AAAA,EACzD;AACF;","names":["handler"]}