@wildix/wilma-voice-client 0.0.43 → 0.0.45

package/dist-cjs/models/enums.js

@@ -1,10 +1,14 @@
 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
-exports.MediaVadFrameType = exports.MediaStartFrameType = exports.MediaRunPurpose = exports.MediaAudioEncoding = exports.MediaPlaybackFrameType = exports.MediaPlaybackEvent = exports.MediaPlaybackStopCause = exports.MediaFinishFrameType = exports.MediaFinishReason = exports.MediaDtmfFrameType = exports.MediaTransferFrameType = exports.MediaTranscriptionFrameType = exports.MediaParty = exports.MediaStopFrameType = exports.MediaStartedFrameType = exports.MediaRestoreFrameType = exports.MediaLanguageFrameType = exports.MediaHangupFrameType = exports.MediaFinishedFrameType = exports.MediaErrorFrameType = exports.MediaDuckFrameType = exports.MediaAudioEndFrameType = exports.MediaAudioBeginFrameType = exports.VoiceCommandType = exports.VoiceInputMode = exports.VoiceAudioEncoding = exports.VoiceTranscriptionDropReason = exports.VoiceCallToolPhase = exports.VoiceCallEndReason = exports.VoiceControlAction = exports.VoiceEventType = exports.VoicePronunciationEncoding = void 0;
+exports.MediaVadFrameType = exports.MediaStartFrameType = exports.MediaRunPurpose = exports.MediaAudioEncoding = exports.MediaPlaybackFrameType = exports.MediaPlaybackEvent = exports.MediaPlaybackStopCause = exports.MediaFinishFrameType = exports.MediaFinishReason = exports.MediaDtmfFrameType = exports.MediaTransferFrameType = exports.MediaTranscriptionFrameType = exports.MediaParty = exports.MediaStopFrameType = exports.MediaStartedFrameType = exports.MediaRestoreFrameType = exports.MediaLanguageFrameType = exports.MediaHangupFrameType = exports.MediaFinishedFrameType = exports.MediaErrorFrameType = exports.MediaDuckFrameType = exports.MediaAudioEndFrameType = exports.MediaAudioBeginFrameType = exports.VoiceCommandType = exports.VoiceInputMode = exports.VoiceAudioEncoding = exports.VoiceTranscriptionDropReason = exports.VoiceCallToolPhase = exports.VoiceCallEndReason = exports.VoiceControlAction = exports.VoiceEventType = exports.VoiceLowConfidenceAction = exports.VoicePronunciationEncoding = void 0;
 exports.VoicePronunciationEncoding = {
     ALIAS: "alias",
     IPA: "ipa",
 };
+exports.VoiceLowConfidenceAction = {
+    ANSWER: "answer",
+    ASK_TO_REPEAT: "ask_to_repeat",
+};
 exports.VoiceEventType = {
     AGENT_TRANSCRIPTION: "agentTranscription",
     AUDIO: "audio",

package/dist-cjs/schemas/schemas_0.js

@@ -201,6 +201,7 @@ const _in = "interruptions";
 const _int = "interruptible";
 const _inte = "interruption";
 const _l = "language";
+const _lC = "lowConfidence";
 const _la = "languages";
 const _m = "message";
 const _mC = "minConfidence";
@@ -233,10 +234,9 @@ const _rAMe = "restoreAfterMs";
 const _rI = "replyId";
 const _rM = "rampMs";
 const _rWE = "replyWebsocketEvent";
-const _re = "recogniser";
+const _re = "restore";
 const _rep = "reply";
-const _res = "restore";
-const _resu = "resume";
+const _res = "resume";
 const _ret = "retryable";
 const _s = "smithy.ts.sdk.synthetic.wildix.wilma.voice";
 const _sEWE = "sessionEndedWebsocketEvent";
@@ -382,8 +382,8 @@ exports.GetVoiceDefaultsInput$ = [3, n1, _GVDI,
 ];
 exports.GetVoiceDefaultsOutput$ = [3, n1, _GVDO,
     0,
-    [_re, _vD, _in, _tE],
-    [0, () => exports.VoiceDetectionDefaults$, () => exports.VoiceInterruptionDefaults$, () => exports.VoiceTurnEndDefaults$], 4
+    [_vD, _in, _tE],
+    [() => exports.VoiceDetectionDefaults$, () => exports.VoiceInterruptionDefaults$, () => exports.VoiceTurnEndDefaults$], 3
 ];
 exports.HangupCommandData$ = [3, n1, _HCD,
     0,
@@ -432,7 +432,7 @@ exports.MediaAudioFormat$ = [3, n1, _MAF,
 ];
 exports.MediaCloudFrames$ = [3, n1, _MCF,
     0,
-    [_sta, _tr, _aB, _au, _aE, _sto, _du, _res, _l, _tra, _h, _e, _fi],
+    [_sta, _tr, _aB, _au, _aE, _sto, _du, _re, _l, _tra, _h, _e, _fi],
     [() => exports.MediaStartedFrame$, () => exports.MediaTranscriptionFrame$, () => exports.MediaAudioBeginFrame$, 21, () => exports.MediaAudioEndFrame$, () => exports.MediaStopFrame$, () => exports.MediaDuckFrame$, () => exports.MediaRestoreFrame$, () => exports.MediaLanguageFrame$, () => exports.MediaTransferFrame$, () => exports.MediaHangupFrame$, () => exports.MediaErrorFrame$, () => exports.MediaFinishedFrame$]
 ];
 exports.MediaDtmfFrame$ = [3, n1, _MDF,
@@ -512,7 +512,7 @@ exports.MediaStartedFrame$ = [3, n1, _MSF,
 ];
 exports.MediaStartFrame$ = [3, n1, _MSFe,
     0,
-    [_t, _pu, _cI, _fI, _sIe, _resu, _rAM, _au, _cal, _ag, _la, _pL, _aD, _dN, _var],
+    [_t, _pu, _cI, _fI, _sIe, _res, _rAM, _au, _cal, _ag, _la, _pL, _aD, _dN, _var],
     [0, 0, 0, 1, 0, 2, 1, () => exports.MediaAudioFormat$, () => exports.MediaStartCaller$, () => exports.MediaStartAgent$, 64 | 0, 0, 2, 0, 0], 9
 ];
 exports.MediaStopFrame$ = [3, n1, _MSFed,
@@ -782,8 +782,8 @@ exports.VoiceErrorEvent$ = [3, n1, _VEE,
 ];
 exports.VoiceInterruptionDefaults$ = [3, n1, _VID,
     0,
-    [_mW, _mC, _dAM, _dVD, _rAMe],
-    [1, 1, 1, 1, 1], 5
+    [_mW, _dAM, _dVD, _rAMe],
+    [1, 1, 1, 1], 4
 ];
 exports.VoiceSpeechPayload$ = [3, n1, _VSP,
     0,
@@ -792,8 +792,8 @@ exports.VoiceSpeechPayload$ = [3, n1, _VSP,
 ];
 exports.VoiceTurnEndDefaults$ = [3, n1, _VTED,
     0,
-    [_hM, _mWM],
-    [128 | 1, 1], 2
+    [_hM, _mWM, _mC, _lC],
+    [1, 1, 1, 0], 4
 ];
 var LanguageCodeList = 64 | 0;
 var VoicePronunciationList = [1, n2, _VPL,
@@ -803,7 +803,6 @@ var MediaStartLanguageList = 64 | 0;
 var VoiceList = [1, n1, _VL,
     0, () => exports.Voice$
 ];
-var VoiceHoldByRecogniser = 128 | 1;
 exports.VoiceCallEvent$ = [4, n1, _VCE,
     0,
     [_sta, _tr, _aT, _rep, _sa, _pl, _inte, _l, _too, _end],

package/dist-es/models/enums.js

@@ -2,6 +2,10 @@ export const VoicePronunciationEncoding = {
     ALIAS: "alias",
     IPA: "ipa",
 };
+export const VoiceLowConfidenceAction = {
+    ANSWER: "answer",
+    ASK_TO_REPEAT: "ask_to_repeat",
+};
 export const VoiceEventType = {
     AGENT_TRANSCRIPTION: "agentTranscription",
     AUDIO: "audio",

package/dist-es/schemas/schemas_0.js

@@ -196,6 +196,7 @@ const _in = "interruptions";
 const _int = "interruptible";
 const _inte = "interruption";
 const _l = "language";
+const _lC = "lowConfidence";
 const _la = "languages";
 const _m = "message";
 const _mC = "minConfidence";
@@ -228,10 +229,9 @@ const _rAMe = "restoreAfterMs";
 const _rI = "replyId";
 const _rM = "rampMs";
 const _rWE = "replyWebsocketEvent";
-const _re = "recogniser";
+const _re = "restore";
 const _rep = "reply";
-const _res = "restore";
-const _resu = "resume";
+const _res = "resume";
 const _ret = "retryable";
 const _s = "smithy.ts.sdk.synthetic.wildix.wilma.voice";
 const _sEWE = "sessionEndedWebsocketEvent";
@@ -377,8 +377,8 @@ export var GetVoiceDefaultsInput$ = [3, n1, _GVDI,
 ];
 export var GetVoiceDefaultsOutput$ = [3, n1, _GVDO,
     0,
-    [_re, _vD, _in, _tE],
-    [0, () => VoiceDetectionDefaults$, () => VoiceInterruptionDefaults$, () => VoiceTurnEndDefaults$], 4
+    [_vD, _in, _tE],
+    [() => VoiceDetectionDefaults$, () => VoiceInterruptionDefaults$, () => VoiceTurnEndDefaults$], 3
 ];
 export var HangupCommandData$ = [3, n1, _HCD,
     0,
@@ -427,7 +427,7 @@ export var MediaAudioFormat$ = [3, n1, _MAF,
 ];
 export var MediaCloudFrames$ = [3, n1, _MCF,
     0,
-    [_sta, _tr, _aB, _au, _aE, _sto, _du, _res, _l, _tra, _h, _e, _fi],
+    [_sta, _tr, _aB, _au, _aE, _sto, _du, _re, _l, _tra, _h, _e, _fi],
     [() => MediaStartedFrame$, () => MediaTranscriptionFrame$, () => MediaAudioBeginFrame$, 21, () => MediaAudioEndFrame$, () => MediaStopFrame$, () => MediaDuckFrame$, () => MediaRestoreFrame$, () => MediaLanguageFrame$, () => MediaTransferFrame$, () => MediaHangupFrame$, () => MediaErrorFrame$, () => MediaFinishedFrame$]
 ];
 export var MediaDtmfFrame$ = [3, n1, _MDF,
@@ -507,7 +507,7 @@ export var MediaStartedFrame$ = [3, n1, _MSF,
 ];
 export var MediaStartFrame$ = [3, n1, _MSFe,
     0,
-    [_t, _pu, _cI, _fI, _sIe, _resu, _rAM, _au, _cal, _ag, _la, _pL, _aD, _dN, _var],
+    [_t, _pu, _cI, _fI, _sIe, _res, _rAM, _au, _cal, _ag, _la, _pL, _aD, _dN, _var],
     [0, 0, 0, 1, 0, 2, 1, () => MediaAudioFormat$, () => MediaStartCaller$, () => MediaStartAgent$, 64 | 0, 0, 2, 0, 0], 9
 ];
 export var MediaStopFrame$ = [3, n1, _MSFed,
@@ -777,8 +777,8 @@ export var VoiceErrorEvent$ = [3, n1, _VEE,
 ];
 export var VoiceInterruptionDefaults$ = [3, n1, _VID,
     0,
-    [_mW, _mC, _dAM, _dVD, _rAMe],
-    [1, 1, 1, 1, 1], 5
+    [_mW, _dAM, _dVD, _rAMe],
+    [1, 1, 1, 1], 4
 ];
 export var VoiceSpeechPayload$ = [3, n1, _VSP,
     0,
@@ -787,8 +787,8 @@ export var VoiceSpeechPayload$ = [3, n1, _VSP,
 ];
 export var VoiceTurnEndDefaults$ = [3, n1, _VTED,
     0,
-    [_hM, _mWM],
-    [128 | 1, 1], 2
+    [_hM, _mWM, _mC, _lC],
+    [1, 1, 1, 0], 4
 ];
 var LanguageCodeList = 64 | 0;
 var VoicePronunciationList = [1, n2, _VPL,
@@ -798,7 +798,6 @@ var MediaStartLanguageList = 64 | 0;
 var VoiceList = [1, n1, _VL,
     0, () => Voice$
 ];
-var VoiceHoldByRecogniser = 128 | 1;
 export var VoiceCallEvent$ = [4, n1, _VCE,
     0,
     [_sta, _tr, _aT, _rep, _sa, _pl, _inte, _l, _too, _end],

package/dist-types/commands/GetVoiceDefaultsCommand.d.ts

@@ -40,7 +40,6 @@ declare const GetVoiceDefaultsCommand_base: {
  * const command = new GetVoiceDefaultsCommand(input);
  * const response = await client.send(command);
  * // { // GetVoiceDefaultsOutput
- * //   recogniser: "STRING_VALUE", // required
  * //   voiceDetection: { // VoiceDetectionDefaults
  * //     silenceMs: Number("int"), // required
  * //     silenceLevelDb: Number("int"), // required
@@ -48,16 +47,15 @@ declare const GetVoiceDefaultsCommand_base: {
  * //   },
  * //   interruptions: { // VoiceInterruptionDefaults
  * //     minWords: Number("int"), // required
- * //     minConfidence: Number("double"), // required
  * //     duckAfterMs: Number("int"), // required
  * //     duckVolumeDb: Number("int"), // required
  * //     restoreAfterMs: Number("int"), // required
  * //   },
  * //   turnEnd: { // VoiceTurnEndDefaults
- * //     holdMs: { // VoiceHoldByRecogniser // required
- * //       "<keys>": Number("int"),
- * //     },
+ * //     holdMs: Number("int"), // required
  * //     maxWaitMs: Number("int"), // required
+ * //     minConfidence: Number("double"), // required
+ * //     lowConfidence: "answer" || "ask_to_repeat", // required
  * //   },
  * // };
  *

package/dist-types/models/enums.d.ts

@@ -16,6 +16,24 @@ export declare const VoicePronunciationEncoding: {
  * @public
  */
 export type VoicePronunciationEncoding = (typeof VoicePronunciationEncoding)[keyof typeof VoicePronunciationEncoding];
+/**
+ * @public
+ * @enum
+ */
+export declare const VoiceLowConfidenceAction: {
+    /**
+     * The agent answers the words as they were recognised.
+     */
+    readonly ANSWER: "answer";
+    /**
+     * The agent asks the caller to repeat.
+     */
+    readonly ASK_TO_REPEAT: "ask_to_repeat";
+};
+/**
+ * @public
+ */
+export type VoiceLowConfidenceAction = (typeof VoiceLowConfidenceAction)[keyof typeof VoiceLowConfidenceAction];
 /**
  * @public
  * @enum
@@ -435,7 +453,7 @@ export declare const MediaRunPurpose: {
      */
     readonly AGENT: "agent";
     /**
-     * The cloud transcribes the call and sends no audio.
+     * Reserved: the cloud transcribes the call and sends no audio. The cloud does not serve it at this time and closes the socket with 4400.
      */
     readonly TRANSCRIPTION: "transcription";
 };

package/dist-types/models/models_0.d.ts

@@ -1,5 +1,5 @@
 import type { DocumentType as __DocumentType, StreamingBlobTypes } from "@smithy/types";
-import type { MediaAudioBeginFrameType, MediaAudioEncoding, MediaAudioEndFrameType, MediaDtmfFrameType, MediaDuckFrameType, MediaErrorFrameType, MediaFinishedFrameType, MediaFinishFrameType, MediaFinishReason, MediaHangupFrameType, MediaLanguageFrameType, MediaParty, MediaPlaybackEvent, MediaPlaybackFrameType, MediaPlaybackStopCause, MediaRestoreFrameType, MediaRunPurpose, MediaStartedFrameType, MediaStartFrameType, MediaStopFrameType, MediaTranscriptionFrameType, MediaTransferFrameType, MediaVadFrameType, VoiceAudioEncoding, VoiceCallEndReason, VoiceCallToolPhase, VoiceCommandType, VoiceControlAction, VoiceEventType, VoiceInputMode, VoicePronunciationEncoding, VoiceTranscriptionDropReason } from "./enums";
+import type { MediaAudioBeginFrameType, MediaAudioEncoding, MediaAudioEndFrameType, MediaDtmfFrameType, MediaDuckFrameType, MediaErrorFrameType, MediaFinishedFrameType, MediaFinishFrameType, MediaFinishReason, MediaHangupFrameType, MediaLanguageFrameType, MediaParty, MediaPlaybackEvent, MediaPlaybackFrameType, MediaPlaybackStopCause, MediaRestoreFrameType, MediaRunPurpose, MediaStartedFrameType, MediaStartFrameType, MediaStopFrameType, MediaTranscriptionFrameType, MediaTransferFrameType, MediaVadFrameType, VoiceAudioEncoding, VoiceCallEndReason, VoiceCallToolPhase, VoiceCommandType, VoiceControlAction, VoiceEventType, VoiceInputMode, VoiceLowConfidenceAction, VoicePronunciationEncoding, VoiceTranscriptionDropReason } from "./enums";
 /**
  * A pronunciation rule for one phrase. Belongs to a LANGUAGE, not to the agent: 'Wildix' is
  *     respelled differently for an Italian and a German voice, so one list shared ACROSS languages
@@ -352,7 +352,7 @@ export interface VoiceCallTranscriptionEvent {
      */
     time?: number | undefined;
     /**
-     * Present on a final that did not go to the agent as a caller turn. It came while the agent spoke and the cloud took it for noise or echo, or for another speaker.
+     * Present on a final that did not go to the agent as a caller turn. It came while the agent spoke, and its recognition confidence was below the minimum of the call or the recogniser labelled it with another speaker.
      * @public
      */
     dropped?: VoiceTranscriptionDropReason | undefined;
@@ -900,7 +900,7 @@ export interface MediaVoiceDetection {
      */
     silenceMs: number;
     /**
-     * Level in dBFS below which caller audio is silence.
+     * Mean frame level in dBFS below which caller audio is silence.
      * @public
      */
     silenceLevelDb: number;
@@ -1163,7 +1163,7 @@ export interface MediaStartAgent {
     version?: number | undefined;
 }
 /**
- * Format of the binary PBX audio (`MediaPbxAudio`). Channel 0 is the caller; channel 1, when present, is the callee. Samples interleave.
+ * Format of the binary PBX audio (`MediaPbxAudio`). Channel 0 is the caller; channel 1, when present, is the callee. Samples interleave. The cloud accepts only 8000 Hz and one channel at this time; another format closes the socket with 4400.
  * @public
  */
 export interface MediaAudioFormat {
@@ -1220,7 +1220,7 @@ export interface MediaStartFrame {
      */
     sessionId: string;
     /**
-     * True when an earlier connection of this run got `started`: the cloud continues that session without a new greeting.
+     * True when an earlier connection of this run got agent audio (`audio.begin`): the cloud continues that session without a new greeting.
      * @public
      */
     resume: boolean;
@@ -1230,7 +1230,7 @@ export interface MediaStartFrame {
      */
     resumeAtMs: number;
     /**
-     * Format of the binary PBX audio (`MediaPbxAudio`). Channel 0 is the caller; channel 1, when present, is the callee. Samples interleave.
+     * Format of the binary PBX audio (`MediaPbxAudio`). Channel 0 is the caller; channel 1, when present, is the callee. Samples interleave. The cloud accepts only 8000 Hz and one channel at this time; another format closes the socket with 4400.
      * @public
      */
     audio: MediaAudioFormat;
@@ -1879,11 +1879,6 @@ export interface VoiceInterruptionDefaults {
      * @public
      */
     minWords: number;
-    /**
-     * Minimum recognition confidence, from 0 to 1, of the caller words that count as an interruption.
-     * @public
-     */
-    minConfidence: number;
     /**
      * Caller speech in ms before the agent's reply is lowered.
      * @public
@@ -1906,15 +1901,25 @@ export interface VoiceInterruptionDefaults {
  */
 export interface VoiceTurnEndDefaults {
     /**
-     * Default of `holdMs` for each recogniser, by the URI scheme of the recogniser, for example `soniox` and `aws`.
+     * Default of `holdMs`.
      * @public
      */
-    holdMs: Record<string, number>;
+    holdMs: number;
     /**
      * Default of `maxWaitMs`.
      * @public
      */
     maxWaitMs: number;
+    /**
+     * Default of `minConfidence`. 0: off.
+     * @public
+     */
+    minConfidence: number;
+    /**
+     * Default of `lowConfidence`.
+     * @public
+     */
+    lowConfidence: VoiceLowConfidenceAction;
 }
 /**
  * Defaults of `VoiceChannelSettings.voiceDetection`.
@@ -1927,7 +1932,7 @@ export interface VoiceDetectionDefaults {
      */
     silenceMs: number;
     /**
-     * Level in dBFS below which caller audio is silence.
+     * Mean frame level in dBFS below which caller audio is silence.
      * @public
      */
     silenceLevelDb: number;
@@ -1941,11 +1946,6 @@ export interface VoiceDetectionDefaults {
  * @public
  */
 export interface GetVoiceDefaultsOutput {
-    /**
-     * URI scheme of the recogniser that a call uses when its agent sets none, for example `soniox`.
-     * @public
-     */
-    recogniser: string;
     /**
      * Defaults of `VoiceChannelSettings.voiceDetection`.
      * @public

package/package.json

@@ -1,7 +1,7 @@
 {
     "name": "@wildix/wilma-voice-client",
     "description": "@wildix/wilma-voice-client client",
-    "version": "0.0.43",
+    "version": "0.0.45",
     "scripts": {
         "build": "concurrently 'yarn:build:cjs' 'yarn:build:es' 'yarn:build:types'",
         "build:cjs": "tsc -p tsconfig.cjs.json",