fix #1466: (#1467)

* fix #1466:

* do not send tts streaming events when we are not doing tts streaming

lib/config.js

@@ -119,7 +119,7 @@ const ENCRYPTION_SECRET = process.env.ENCRYPTION_SECRET;
 const HTTP_POOL = process.env.HTTP_POOL && parseInt(process.env.HTTP_POOL);
 const HTTP_POOLSIZE = parseInt(process.env.HTTP_POOLSIZE, 10) || 10;
 const HTTP_PIPELINING = parseInt(process.env.HTTP_PIPELINING, 10) || 1;
-const HTTP_TIMEOUT = 10000;
+const HTTP_TIMEOUT = parseInt(process.env.JAMBONES_HTTP_TIMEOUT, 10) || 10000;
 const HTTP_PROXY_IP = process.env.JAMBONES_HTTP_PROXY_IP;
 const HTTP_PROXY_PORT = process.env.JAMBONES_HTTP_PROXY_PORT;
 const HTTP_PROXY_PROTOCOL = process.env.JAMBONES_HTTP_PROXY_PROTOCOL || 'http';
@@ -139,6 +139,10 @@ const JAMBONES_USE_FREESWITCH_TIMER_FD = process.env.JAMBONES_USE_FREESWITCH_TIM
 const JAMBONES_DIAL_SBC_FOR_REGISTERED_USER = process.env.JAMBONES_DIAL_SBC_FOR_REGISTERED_USER || false;
 const JAMBONES_MEDIA_TIMEOUT_MS = process.env.JAMBONES_MEDIA_TIMEOUT_MS || 0;
 const JAMBONES_MEDIA_HOLD_TIMEOUT_MS = process.env.JAMBONES_MEDIA_HOLD_TIMEOUT_MS || 0;
+
+/* say / tts */
+const JAMBONES_SAY_CHUNK_SIZE = parseInt(process.env.JAMBONES_SAY_CHUNK_SIZE, 10) || 900;
+
 // jambonz
 const JAMBONES_TRANSCRIBE_EP_DESTROY_DELAY_MS =
   process.env.JAMBONES_TRANSCRIBE_EP_DESTROY_DELAY_MS;
@@ -231,5 +235,6 @@ module.exports = {
   JAMBONES_DIAL_SBC_FOR_REGISTERED_USER,
   JAMBONES_MEDIA_TIMEOUT_MS,
   JAMBONES_MEDIA_HOLD_TIMEOUT_MS,
+  JAMBONES_SAY_CHUNK_SIZE,
   JAMBONES_TRANSCRIBE_EP_DESTROY_DELAY_MS,
 };

lib/http-routes/api/create-call.js

@@ -291,7 +291,7 @@ router.post('/',
             }, {
               ...(account.enable_debug_log && {level: 'debug'})
             });
-            app.requestor.logger = app.notifier.logger = sipLogger;
+            app.requestor.logger = app.notifier.logger = restDial.logger = sipLogger;
             const callInfo = new CallInfo({
               direction: CallDirection.Outbound,
               req: inviteReq,

lib/session/call-session.js

@@ -927,7 +927,7 @@ class CallSession extends Emitter {
         this.logger.debug('CallSession:enableBackgroundTtsStream - ttsStream enabled');
       } else {
         this.logger.debug(
-          'CallSession:enableBackgroundTtsStream - ignoring request as call does not have required conditions');
+          'CallSession:enableBackgroundTtsStream - ignoring request; conditions not met (probably not using ws api)');
       }
     }  catch (err) {
       this.logger.info({err, say}, 'CallSession:enableBackgroundTtsStream - Error creating background tts stream task');
@@ -941,9 +941,11 @@ class CallSession extends Emitter {
     }
   }
   clearTtsStream() {
-    this.requestor?.request('tts:streaming-event', '/streaming-event', {event_type: 'user_interruption'})
-      .catch((err) => this.logger.info({err}, 'CallSession:clearTtsStream - Error sending user_interruption'));
-    this.ttsStreamingBuffer?.clear();
+    if (this.isTtsStreamEnabled) {
+      this.requestor?.request('tts:streaming-event', '/streaming-event', {event_type: 'user_interruption'})
+        .catch((err) => this.logger.info({err}, 'CallSession:clearTtsStream - Error sending user_interruption'));
+      this.ttsStreamingBuffer?.clear();
+    }
   }
 
   startTtsStream() {
@@ -951,7 +953,7 @@ class CallSession extends Emitter {
   }
 
   stopTtsStream() {
-    if (this.appIsUsingWebsockets) {
+    if (this.isTtsStreamEnabled) {
       this.requestor?.request('tts:streaming-event', '/streaming-event', {event_type: 'stream_closed'})
         .catch((err) => this.logger.info({err}, 'CallSession:clearTtsStream - Error sending user_interruption'));
       this.ttsStreamingBuffer?.stop();
@@ -1141,7 +1143,6 @@ class CallSession extends Emitter {
           return {
             api_key: credential.api_key,
             model_id: credential.model_id,
-            stt_model_id: credential.stt_model_id,
             api_uri: credential.api_uri,
             options: credential.options
           };

lib/tasks/dial.js

@@ -158,7 +158,7 @@ class TaskDial extends Task {
 
   get canReleaseMedia() {
     const keepAnchor = this.data.anchorMedia ||
-      this.weAreTranscoding ||
+      this.isTranscoding ||
       this.cs.isBackGroundListen ||
       this.cs.onHoldMusic ||
       ANCHOR_MEDIA_ALWAYS ||
@@ -576,7 +576,7 @@ class TaskDial extends Task {
       proxy: `sip:${sbcAddress}`,
       callingNumber: this.callerId || fromUri.user,
       ...(this.callerName && {callingName: this.callerName}),
-      opusFirst: isOpusFirst(this.cs.ep.remote.sdp),
+      opusFirst: isOpusFirst(this.cs.ep.local.sdp),
       isVideoCall: this.cs.ep.remote.sdp.includes('m=video')
     };
 
@@ -773,6 +773,15 @@ class TaskDial extends Task {
   }
 
   async _connectSingleDial(cs, sd) {
+    // start connect with dialed leg, this is the soonest we can identify transcoding
+    if (this.epOther && sd.ep) {
+      const codecA = getLeadingCodec(this.epOther.local.sdp);
+      const codecB = getLeadingCodec(sd.ep.remote.sdp);
+      this.isTranscoding = (codecA !== codecB);
+      if (this.isTranscoding) {
+        this.logger.info(`Dial:_connectSingleDial - transcoding from ${codecA} (A leg) to ${codecB} (B leg)`);
+      }
+    }
     if (!this.bridged && !this.canReleaseMedia) {
       this.logger.debug('Dial:_connectSingleDial bridging endpoints');
       if (this.epOther) {
@@ -930,13 +939,6 @@ class TaskDial extends Task {
         this.logger.info({err}, 'Dial:_selectSingleDial - Error boosting audio signal');
       }
     }
-    /* basic determination to see if call is being transcoded */
-    const codecA = getLeadingCodec(this.epOther.local.sdp);
-    const codecB = getLeadingCodec(this.ep.remote.sdp);
-    this.weAreTranscoding = (codecA !== codecB);
-    if (this.weAreTranscoding) {
-      this.logger.info(`Dial:_selectSingleDial - transcoding from ${codecA} (A leg) to ${codecB} (B leg)`);
-    }
     /* if we can release the media back to the SBC, do so now */
     if (this.canReleaseMedia || this.shouldExitMediaPathEntirely) {
       setTimeout(this._releaseMedia.bind(this, cs, sd, this.shouldExitMediaPathEntirely), 200);

lib/tasks/gather.js

@@ -5,7 +5,6 @@ const {
   AwsTranscriptionEvents,
   AzureTranscriptionEvents,
   DeepgramTranscriptionEvents,
-  ElevenlabsTranscriptionEvents,
   GladiaTranscriptionEvents,
   SonioxTranscriptionEvents,
   CobaltTranscriptionEvents,
@@ -493,17 +492,6 @@ class TaskGather extends SttTask {
         this.addCustomEventListener(ep, DeepgramfluxTranscriptionEvents.Error, this._onVendorError.bind(this, cs, ep));
         break;
 
-      case 'elevenlabs':
-        this.bugname = `${this.bugname_prefix}elevenlabs_transcribe`;
-        this.addCustomEventListener(
-          ep, ElevenlabsTranscriptionEvents.Transcription, this._onTranscription.bind(this, cs, ep));
-        this.addCustomEventListener(
-          ep, ElevenlabsTranscriptionEvents.Connect, this._onVendorConnect.bind(this, cs, ep));
-        this.addCustomEventListener(ep, ElevenlabsTranscriptionEvents.ConnectFailure,
-          this._onVendorConnectFailure.bind(this, cs, ep));
-        this.addCustomEventListener(ep, ElevenlabsTranscriptionEvents.Error, this._onVendorError.bind(this, cs, ep));
-        break;
-
       case 'gladia':
         this.bugname = `${this.bugname_prefix}gladia_transcribe`;
         this.addCustomEventListener(
@@ -893,7 +881,7 @@ class TaskGather extends SttTask {
         this._fillerNoiseOn = false;  // in a race, if we just started audio it may sneak through here
         this.ep.api('uuid_break', this.ep.uuid)
           .catch((err) => this.logger.info(err, 'Error killing audio'));
-        cs.clearTtsStream();
+        if (cs.isTtsStreamEnabled) cs.clearTtsStream();
       }
       return;
     }

lib/tasks/say.js

@@ -1,6 +1,7 @@
 const assert = require('assert');
 const TtsTask = require('./tts-task');
 const {TaskName, TaskPreconditions} = require('../utils/constants');
+const {JAMBONES_SAY_CHUNK_SIZE} = require('../config');
 const pollySSMLSplit = require('polly-ssml-split');
 const { SpeechCredentialError, NonFatalTaskError } = require('../utils/error');
 const { sleepFor } = require('../utils/helpers');
@@ -31,7 +32,7 @@ const isMatchingEvent = (logger, filename, playbackId, evt) => {
 const breakLengthyTextIfNeeded = (logger, text) => {
   // As The text can be used for tts streaming, we need to break lengthy text into smaller chunks
 // HIGH_WATER_BUFFER_SIZE defined in tts-streaming-buffer.js
-  const chunkSize = 900;
+  const chunkSize = JAMBONES_SAY_CHUNK_SIZE;
   const isSSML = text.startsWith('<speak>');
   const options = {
     softLimit: 100,

lib/tasks/tts-task.js

@@ -89,8 +89,9 @@ class TtsTask extends Task {
     // api_key, model_id, api_uri, custom_tts_streaming_url, and auth_token are encoded in the credentials
     // allow them to be overriden via config, using options
     // give preference to options passed in via config
-    const local_options = {...JSON.parse(options), ...this.options};
-    const local_voice_settings = {...JSON.parse(options).voice_settings, ...this.options.voice_settings};
+    const parsed_options = options ? JSON.parse(options) : {};
+    const local_options = {...parsed_options, ...this.options};
+    const local_voice_settings = {...(parsed_options.voice_settings || {}), ...(this.options.voice_settings || {})};
     const local_api_key =  local_options.api_key ?? api_key;
     const local_model_id = local_options.model_id ?? model_id;
     const local_api_uri = local_options.api_uri ?? api_uri;

lib/utils/constants.json

@@ -103,12 +103,6 @@
     "Connect": "deepgramflux_transcribe::connect",
     "Error": "deepgramflux_transcribe::error"
   },
-  "ElevenlabsTranscriptionEvents": {
-    "Transcription": "elevenlabs_transcribe::transcription",
-    "ConnectFailure": "elevenlabs_transcribe::connect_failed",
-    "Connect": "elevenlabs_transcribe::connect",
-    "Error": "elevenlabs_transcribe::error"
-  },
   "GladiaTranscriptionEvents": {
     "Transcription": "gladia_transcribe::transcription",
     "ConnectFailure": "gladia_transcribe::connect_failed",

lib/utils/db-utils.js

@@ -106,7 +106,6 @@ const speechMapper = (cred) => {
       const o = JSON.parse(decrypt(credential));
       obj.api_key = o.api_key;
       obj.model_id = o.model_id;
-      obj.stt_model_id = o.stt_model_id;
       obj.api_uri = o.api_uri;
       obj.options = o.options;
     }

lib/utils/http-requestor.js

@@ -191,7 +191,7 @@ class HttpRequestor extends BaseRequestor {
         method,
         headers: hdrs,
         ...('POST' === method && {body: JSON.stringify(payload)}),
-        timeout: HTTP_TIMEOUT,
+        headersTimeout: HTTP_TIMEOUT,
         followRedirects: false
       };
 

lib/utils/stt-latency-calculator.js

@@ -127,7 +127,6 @@ class SttLatencyCalculator extends Emitter {
 
   calculateLatency() {
     if (!this.isRunning) {
-      this.logger.debug('Latency calculator is not running, cannot calculate latency, returning default values');
       return null;
     }
 

lib/utils/transcription-utils.js

@@ -545,23 +545,6 @@ const normalizeVerbio = (evt, channel, language) => {
   };
 };
 
-const normalizeElevenlabs = (evt, channel, language) => {
-  const copy = JSON.parse(JSON.stringify(evt));
-  return {
-    language_code: language,
-    channel_tag: channel,
-    is_final: evt.message_type === 'committed_transcript',
-    alternatives: [{
-      confidence: 1.0,
-      transcript: evt.text,
-    }],
-    vendor: {
-      name: 'elevenlabs',
-      evt: copy
-    }
-  };
-};
-
 const normalizeMicrosoft = (evt, channel, language, punctuation = true) => {
   const copy = JSON.parse(JSON.stringify(evt));
   const nbest = evt.NBest;
@@ -787,8 +770,6 @@ module.exports = (logger) => {
         return normalizeGladia(evt, channel, language, shortUtterance);
       case 'deepgramflux':
         return normalizeDeepgramFlux(evt, channel, language, shortUtterance);
-      case 'elevenlabs':
-        return normalizeElevenlabs(evt, channel, language);
       case 'microsoft':
         return normalizeMicrosoft(evt, channel, language, punctuation);
       case 'google':
@@ -1104,24 +1085,6 @@ module.exports = (logger) => {
         ...(keyterms && keyterms.length > 0 && {DEEPGRAMFLUX_SPEECH_KEYTERMS: keyterms.join(',')}),
       };
     }
-    else if ('elevenlabs' === vendor) {
-      const {api_key, stt_model_id} = sttCredentials;
-      const {includeTimestamps, commitStrategy, vadSilenceThresholdSecs, vadThreshold,
-        minSpeechDurationMs, minSilenceDurationMs, enableLogging} = rOpts.elevenlabsOptions || {};
-
-      opts = {
-        ...opts,
-        ELEVENLABS_API_KEY: api_key,
-        ELEVENLABS_MODEL_ID: stt_model_id,
-        ELEVENLABS_INCLUDE_TIMESTAMPS: includeTimestamps === true ? true : false,
-        ...(commitStrategy && {ELEVENLABS_COMMIT_STRATEGY: commitStrategy}),
-        ...(vadSilenceThresholdSecs && {ELEVENLABS_VAD_SILENCE_THRESHOLD_SECS: vadSilenceThresholdSecs}),
-        ...(vadThreshold && {ELEVENLABS_VAD_THRESHOLD: vadThreshold}),
-        ...(minSpeechDurationMs && {ELEVENLABS_MIN_SPEECH_DURATION_MS: minSpeechDurationMs}),
-        ...(minSilenceDurationMs && {ELEVENLABS_MIN_SILENCE_DURATION_MS: minSilenceDurationMs}),
-        ...(enableLogging && {ELEVENLABS_ENABLE_LOGGING: enableLogging ? 1 : 0}),
-      };
-    }
     else if ('gladia' === vendor) {
       const {host, path} = sttCredentials;
       opts = {

lib/utils/tts-streaming-buffer.js

@@ -163,7 +163,6 @@ class TtsStreamingBuffer extends Emitter {
   }
 
   clear() {
-    this.logger.debug('TtsStreamingBuffer:clear');
     if (this._connectionStatus !== TtsStreamingConnectionStatus.Connected) return;
     clearTimeout(this.timer);
     this._api(this.ep, [this.ep.uuid, 'clear']).catch((err) =>
@@ -437,7 +436,15 @@ class TtsStreamingBuffer extends Emitter {
 
 const findSentenceBoundary = (text, limit) => {
   // Look for punctuation or double newline that signals sentence end.
-  const sentenceEndRegex = /[.!?](?=\s|$)|\n\n/g;
+  // Includes:
+  //   - ASCII: . ! ?
+  //   - Arabic: ؟ (question mark), ۔ (full stop)
+  //   - Japanese: 。 (full stop), ！, ？ (full-width exclamation/question)
+  //
+  // For languages that use spaces between sentences, we still require
+  // whitespace or end-of-string after the mark. For Japanese (no spaces),
+  // we treat the punctuation itself as a boundary regardless of following char.
+  const sentenceEndRegex = /[.!?؟۔](?=\s|$)|[。！？]|\n\n/g;
   let lastSentenceBoundary = -1;
   let match;
   while ((match = sentenceEndRegex.exec(text)) && match.index < limit) {

package-lock.json

@@ -8752,7 +8752,6 @@
     },
     "node_modules/unix-dgram": {
       "version": "2.0.6",
-      "hasInstallScript": true,
       "license": "ISC",
       "optional": true,
       "dependencies": {