mirror of
https://github.com/jambonz/jambonz-feature-server.git
synced 2026-02-13 09:49:30 +00:00
Compare commits
12 Commits
v0.8.5-6
...
feat/aws-s
| Author | SHA1 | Date | |
|---|---|---|---|
|
|
c71a58dcb5 | ||
|
|
8022f9d16c | ||
|
|
c4dcb051be | ||
|
|
29dc2f7052 | ||
|
|
7718f50877 | ||
|
|
11283edf6f | ||
|
|
6043921067 | ||
|
|
8ad947c0fd | ||
|
|
63c925c731 | ||
|
|
e8647b2b55 | ||
|
|
dab83423cf | ||
|
|
864a673ea0 |
@@ -116,9 +116,12 @@ const customSanitizeFunction = (value) => {
|
||||
/* trims characters at the beginning and at the end of a string */
|
||||
value = value.trim();
|
||||
|
||||
/* Verify strings including 'http' via new URL */
|
||||
/* We don't escape URLs but verify them via new URL */
|
||||
if (value.includes('http')) {
|
||||
value = new URL(value).toString();
|
||||
} else {
|
||||
/* replaces <, >, &, ', " and / with their corresponding HTML entities */
|
||||
value = escape(value);
|
||||
}
|
||||
}
|
||||
} catch (error) {
|
||||
|
||||
@@ -160,7 +160,7 @@ module.exports = function(srf, logger) {
|
||||
};
|
||||
logger.info({callId, metadata, sdp}, 'successfully parsed SIPREC payload');
|
||||
} catch (err) {
|
||||
logger.info({err, callId}, 'Error parsing multipart payload');
|
||||
logger.info({callId}, 'Error parsing multipart payload');
|
||||
return res.send(503);
|
||||
}
|
||||
}
|
||||
|
||||
@@ -170,18 +170,6 @@ class CallSession extends Emitter {
|
||||
return this.application.notifier;
|
||||
}
|
||||
|
||||
/**
|
||||
* syntheizer
|
||||
*/
|
||||
|
||||
get synthesizer() {
|
||||
return this._synthesizer;
|
||||
}
|
||||
|
||||
set synthesizer(synth) {
|
||||
this._synthesizer = synth;
|
||||
}
|
||||
|
||||
/**
|
||||
* default vendor to use for speech synthesis if not provided in the app
|
||||
*/
|
||||
@@ -264,16 +252,6 @@ class CallSession extends Emitter {
|
||||
set fallbackSpeechRecognizerVendor(vendor) {
|
||||
this.application.fallback_speech_recognizer_vendor = vendor;
|
||||
}
|
||||
/**
|
||||
* recognizer
|
||||
*/
|
||||
get recognizer() {
|
||||
return this._recognizer;
|
||||
}
|
||||
|
||||
set recognizer(rec) {
|
||||
this._recognizer = rec;
|
||||
}
|
||||
/**
|
||||
* default vendor to use for speech recognition if not provided in the app
|
||||
*/
|
||||
|
||||
@@ -109,7 +109,6 @@ class TaskConfig extends Task {
|
||||
});
|
||||
|
||||
if (this.hasSynthesizer) {
|
||||
cs.synthesizer = this.synthesizer;
|
||||
cs.speechSynthesisVendor = this.synthesizer.vendor !== 'default'
|
||||
? this.synthesizer.vendor
|
||||
: cs.speechSynthesisVendor;
|
||||
@@ -139,7 +138,6 @@ class TaskConfig extends Task {
|
||||
this.logger.info({synthesizer: this.synthesizer}, 'Config: updated synthesizer');
|
||||
}
|
||||
if (this.hasRecognizer) {
|
||||
cs.recognizer = this.recognizer;
|
||||
cs.speechRecognizerVendor = this.recognizer.vendor !== 'default'
|
||||
? this.recognizer.vendor
|
||||
: cs.speechRecognizerVendor;
|
||||
|
||||
@@ -103,7 +103,7 @@ class TaskGather extends SttTask {
|
||||
|
||||
async exec(cs, {ep}) {
|
||||
this.logger.debug({options: this.data}, 'Gather:exec');
|
||||
await super.exec(cs, {ep});
|
||||
await super.exec(cs);
|
||||
const {updateSpeechCredentialLastUsed} = require('../utils/db-utils')(this.logger, cs.srf);
|
||||
|
||||
if (cs.hasGlobalSttHints && !this.maskGlobalSttHints) {
|
||||
@@ -141,6 +141,59 @@ class TaskGather extends SttTask {
|
||||
this.interim = true;
|
||||
this.logger.debug('Gather:exec - early hints match enabled');
|
||||
}
|
||||
|
||||
this.ep = ep;
|
||||
if ('default' === this.vendor || !this.vendor) {
|
||||
this.vendor = cs.speechRecognizerVendor;
|
||||
if (this.data.recognizer) this.data.recognizer.vendor = this.vendor;
|
||||
}
|
||||
if ('default' === this.language || !this.language) {
|
||||
this.language = cs.speechRecognizerLanguage;
|
||||
if (this.data.recognizer) this.data.recognizer.language = this.language;
|
||||
}
|
||||
if ('default' === this.label || !this.label) {
|
||||
this.label = cs.speechRecognizerLabel;
|
||||
if (this.data.recognizer) this.data.recognizer.label = this.label;
|
||||
}
|
||||
// Fallback options
|
||||
if ('default' === this.fallbackVendor || !this.fallbackVendor) {
|
||||
this.fallbackVendor = cs.fallbackSpeechRecognizerVendor;
|
||||
if (this.data.recognizer) this.data.recognizer.fallbackVendor = this.fallbackVendor;
|
||||
}
|
||||
if ('default' === this.fallbackLanguage || !this.fallbackLanguage) {
|
||||
this.fallbackLanguage = cs.fallbackSpeechRecognizerLanguage;
|
||||
if (this.data.recognizer) this.data.recognizer.fallbackLanguage = this.fallbackLanguage;
|
||||
}
|
||||
if ('default' === this.fallbackLabel || !this.fallbackLabel) {
|
||||
this.fallbackLabel = cs.fallbackSpeechRecognizerLabel;
|
||||
if (this.data.recognizer) this.data.recognizer.fallbackLabel = this.fallbackLabel;
|
||||
}
|
||||
if (!this.data.recognizer.vendor) {
|
||||
this.data.recognizer.vendor = this.vendor;
|
||||
}
|
||||
if (this.vendor === 'cobalt' && !this.data.recognizer.model) {
|
||||
// By default, application saves cobalt model in language
|
||||
this.data.recognizer.model = cs.speechRecognizerLanguage;
|
||||
}
|
||||
|
||||
if (this.needsStt && !this.sttCredentials) {
|
||||
try {
|
||||
this.sttCredentials = await this._initSpeechCredentials(cs, this.vendor, this.label);
|
||||
} catch (error) {
|
||||
if (this.fallbackVendor && this.isHandledByPrimaryProvider) {
|
||||
await this._fallback();
|
||||
} else {
|
||||
throw error;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/* when using cobalt model is required */
|
||||
if (this.vendor === 'cobalt' && !this.data.recognizer.model) {
|
||||
this.notifyError({ msg: 'ASR error', details:'Cobalt requires a model to be specified'});
|
||||
throw new Error('Cobalt requires a model to be specified');
|
||||
}
|
||||
|
||||
const startListening = async(cs, ep) => {
|
||||
this._startTimer();
|
||||
if (this.isContinuousAsr && 0 === this.timeout) this._startAsrTimer();
|
||||
@@ -266,13 +319,6 @@ class TaskGather extends SttTask {
|
||||
this._resolve('dtmf-terminator-key');
|
||||
}
|
||||
else if (this.input.includes('digits')) {
|
||||
if (this.digitBuffer.length === 0 && this.needsStt) {
|
||||
// DTMF is higher priority than STT.
|
||||
this.removeSpeechListeners(ep);
|
||||
ep.stopTranscription({vendor: this.vendor})
|
||||
.catch((err) => this.logger.error({err},
|
||||
` Received DTMF, Error stopping transcription for vendor ${this.vendor}`));
|
||||
}
|
||||
this.digitBuffer += evt.dtmf;
|
||||
const len = this.digitBuffer.length;
|
||||
if (len === this.numDigits || len === this.maxDigits) {
|
||||
@@ -482,9 +528,7 @@ class TaskGather extends SttTask {
|
||||
this._clearTimer();
|
||||
this._timeoutTimer = setTimeout(() => {
|
||||
if (this.isContinuousAsr) this._startAsrTimer();
|
||||
else if (this.interDigitTimeout <= 0 ||
|
||||
this.digitBuffer.length < this.minDigits ||
|
||||
this.needsStt && this.digitBuffer.length === 0) {
|
||||
else if (this.interDigitTimeout <= 0 || this.digitBuffer.length < this.minDigits || this.needsStt) {
|
||||
this._resolve(this.digitBuffer.length >= this.minDigits ? 'dtmf-num-digits' : 'timeout');
|
||||
}
|
||||
}, this.timeout);
|
||||
@@ -780,13 +824,14 @@ class TaskGather extends SttTask {
|
||||
if (code === 413 && error === 'Too much speech') return this._resolve('timeout');
|
||||
}
|
||||
this.logger.info({evt}, 'TaskGather:_onJambonzError');
|
||||
const errMessage = evt.error || evt.Message;
|
||||
writeAlerts({
|
||||
account_sid: cs.accountSid,
|
||||
alert_type: AlertType.STT_FAILURE,
|
||||
message: `Custom speech vendor ${this.vendor} error: ${evt.error}`,
|
||||
message: `Speech vendor ${this.vendor} error: ${errMessage}`,
|
||||
vendor: this.vendor,
|
||||
}).catch((err) => this.logger.info({err}, 'Error generating alert for jambonz custom connection failure'));
|
||||
this.notifyError({msg: 'ASR error', details:`Custom speech vendor ${this.vendor} error: ${evt.error}`});
|
||||
this.notifyError({msg: 'ASR error', details:`Speech vendor ${this.vendor} error: ${evt.error}`});
|
||||
}
|
||||
|
||||
_onVendorConnectFailure(cs, _ep, evt) {
|
||||
|
||||
@@ -52,81 +52,8 @@ class SttTask extends Task {
|
||||
|
||||
}
|
||||
|
||||
async exec(cs, {ep, ep2}) {
|
||||
super.exec(cs);
|
||||
this.ep = ep;
|
||||
this.ep2 = ep2;
|
||||
if ('default' === this.vendor || !this.vendor) {
|
||||
this.vendor = cs.speechRecognizerVendor;
|
||||
if (this.data.recognizer) this.data.recognizer.vendor = this.vendor;
|
||||
}
|
||||
if ('default' === this.language || !this.language) {
|
||||
this.language = cs.speechRecognizerLanguage;
|
||||
if (this.data.recognizer) this.data.recognizer.language = this.language;
|
||||
}
|
||||
if ('default' === this.label || !this.label) {
|
||||
this.label = cs.speechRecognizerLabel;
|
||||
if (this.data.recognizer) this.data.recognizer.label = this.label;
|
||||
}
|
||||
// Fallback options
|
||||
if ('default' === this.fallbackVendor || !this.fallbackVendor) {
|
||||
this.fallbackVendor = cs.fallbackSpeechRecognizerVendor;
|
||||
if (this.data.recognizer) this.data.recognizer.fallbackVendor = this.fallbackVendor;
|
||||
}
|
||||
if ('default' === this.fallbackLanguage || !this.fallbackLanguage) {
|
||||
this.fallbackLanguage = cs.fallbackSpeechRecognizerLanguage;
|
||||
if (this.data.recognizer) this.data.recognizer.fallbackLanguage = this.fallbackLanguage;
|
||||
}
|
||||
if ('default' === this.fallbackLabel || !this.fallbackLabel) {
|
||||
this.fallbackLabel = cs.fallbackSpeechRecognizerLabel;
|
||||
if (this.data.recognizer) this.data.recognizer.fallbackLabel = this.fallbackLabel;
|
||||
}
|
||||
if (!this.data.recognizer.vendor) {
|
||||
this.data.recognizer.vendor = this.vendor;
|
||||
}
|
||||
if (this.vendor === 'cobalt' && !this.data.recognizer.model) {
|
||||
// By default, application saves cobalt model in language
|
||||
this.data.recognizer.model = cs.speechRecognizerLanguage;
|
||||
}
|
||||
|
||||
if (cs.recognizer) {
|
||||
for (const k in cs.recognizer) {
|
||||
if (Array.isArray(this.data.recognizer[k]) ||
|
||||
Array.isArray(cs.recognizer[k]) ||
|
||||
typeof this.data.recognizer[k] === 'object' ||
|
||||
typeof cs.recognizer[k] === 'object'
|
||||
) {
|
||||
this.data.recognizer[k] = {
|
||||
...this.data.recognizer[k],
|
||||
...cs.recognizer[k]
|
||||
};
|
||||
} else {
|
||||
this.data.recognizer[k] = cs.recognizer[k] || this.data.recognizer[k];
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
if (!this.sttCredentials) {
|
||||
try {
|
||||
this.sttCredentials = await this._initSpeechCredentials(this.cs, this.vendor, this.label);
|
||||
} catch (error) {
|
||||
if (this.fallbackVendor && this.isHandledByPrimaryProvider) {
|
||||
await this._fallback();
|
||||
} else {
|
||||
throw error;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/* when using cobalt model is required */
|
||||
if (this.vendor === 'cobalt' && !this.data.recognizer.model) {
|
||||
this.notifyError({ msg: 'ASR error', details:'Cobalt requires a model to be specified'});
|
||||
throw new Error('Cobalt requires a model to be specified');
|
||||
}
|
||||
}
|
||||
|
||||
async _initSpeechCredentials(cs, vendor, label) {
|
||||
const {getNuanceAccessToken, getIbmAccessToken} = cs.srf.locals.dbHelpers;
|
||||
const {getNuanceAccessToken, getIbmAccessToken, getAwsAuthToken} = this.cs.srf.locals.dbHelpers;
|
||||
let credentials = cs.getSpeechCredentials(vendor, 'stt', label);
|
||||
|
||||
if (!credentials) {
|
||||
@@ -160,6 +87,15 @@ class SttTask extends Task {
|
||||
this.logger.debug({stt_api_key}, `got ibm access token ${servedFromCache ? 'from cache' : ''}`);
|
||||
credentials = {...credentials, access_token, stt_region};
|
||||
}
|
||||
else if (vendor == 'aws') {
|
||||
/* get AWS access token */
|
||||
const {accessKeyId, secretAccessKey, securityToken, region } = credentials;
|
||||
if (!securityToken) {
|
||||
const { servedFromCache, ...newCredentials} = await getAwsAuthToken(accessKeyId, secretAccessKey, region);
|
||||
this.logger.debug({newCredentials}, `got aws security token ${servedFromCache ? 'from cache' : ''}`);
|
||||
credentials = {...newCredentials, region};
|
||||
}
|
||||
}
|
||||
return credentials;
|
||||
}
|
||||
|
||||
|
||||
@@ -45,7 +45,7 @@ class TaskTranscribe extends SttTask {
|
||||
get name() { return TaskName.Transcribe; }
|
||||
|
||||
async exec(cs, {ep, ep2}) {
|
||||
await super.exec(cs, {ep, ep2});
|
||||
super.exec(cs);
|
||||
const {updateSpeechCredentialLastUsed} = require('../utils/db-utils')(this.logger, cs.srf);
|
||||
|
||||
if (cs.hasGlobalSttHints) {
|
||||
@@ -63,6 +63,60 @@ class TaskTranscribe extends SttTask {
|
||||
if (cs.hasGlobalSttPunctuation && !this.data.recognizer.punctuation) {
|
||||
this.data.recognizer.punctuation = cs.globalSttPunctuation;
|
||||
}
|
||||
|
||||
this.ep = ep;
|
||||
this.ep2 = ep2;
|
||||
if ('default' === this.vendor || !this.vendor) {
|
||||
this.vendor = cs.speechRecognizerVendor;
|
||||
if (this.data.recognizer) this.data.recognizer.vendor = this.vendor;
|
||||
}
|
||||
if ('default' === this.language || !this.language) {
|
||||
this.language = cs.speechRecognizerLanguage;
|
||||
if (this.data.recognizer) this.data.recognizer.language = this.language;
|
||||
}
|
||||
if ('default' === this.label || !this.label) {
|
||||
this.label = cs.speechRecognizerLabel;
|
||||
if (this.data.recognizer) this.data.recognizer.label = this.label;
|
||||
}
|
||||
// Fallback options
|
||||
if ('default' === this.fallbackVendor || !this.fallbackVendor) {
|
||||
this.fallbackVendor = cs.fallbackSpeechRecognizerVendor;
|
||||
if (this.data.recognizer) this.data.recognizer.fallbackVendor = this.fallbackVendor;
|
||||
}
|
||||
if ('default' === this.fallbackLanguage || !this.fallbackLanguage) {
|
||||
this.fallbackLanguage = cs.fallbackSpeechRecognizerLanguage;
|
||||
if (this.data.recognizer) this.data.recognizer.fallbackLanguage = this.fallbackLanguage;
|
||||
}
|
||||
if ('default' === this.fallbackLabel || !this.fallbackLabel) {
|
||||
this.fallbackLabel = cs.fallbackSpeechRecognizerLabel;
|
||||
if (this.data.recognizer) this.data.recognizer.fallbackLabel = this.fallbackLabel;
|
||||
}
|
||||
if (!this.data.recognizer.vendor) {
|
||||
this.data.recognizer.vendor = this.vendor;
|
||||
}
|
||||
if (this.vendor === 'cobalt' && !this.data.recognizer.model) {
|
||||
// By default, application saves cobalt model in language
|
||||
this.data.recognizer.model = cs.speechRecognizerLanguage;
|
||||
}
|
||||
|
||||
if (!this.sttCredentials) {
|
||||
try {
|
||||
this.sttCredentials = await this._initSpeechCredentials(cs, this.vendor, this.label);
|
||||
} catch (error) {
|
||||
if (this.fallbackVendor && this.isHandledByPrimaryProvider) {
|
||||
await this._fallback();
|
||||
} else {
|
||||
throw error;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/* when using cobalt model is required */
|
||||
if (this.vendor === 'cobalt' && !this.data.recognizer.model) {
|
||||
this.notifyError({ msg: 'ASR error', details:'Cobalt requires a model to be specified'});
|
||||
throw new Error('Cobalt requires a model to be specified');
|
||||
}
|
||||
|
||||
try {
|
||||
await this._startTranscribing(cs, ep, 1);
|
||||
if (this.separateRecognitionPerChannel && ep2) {
|
||||
|
||||
@@ -178,7 +178,8 @@ function installSrfLocals(srf, logger) {
|
||||
synthAudio,
|
||||
getNuanceAccessToken,
|
||||
getIbmAccessToken,
|
||||
} = require('@jambonz/speech-utils')({}, logger);
|
||||
getAwsAuthToken
|
||||
} = require('@jambonz/speech-utils')({redis_client: client}, logger);
|
||||
const {
|
||||
writeAlerts,
|
||||
AlertType
|
||||
@@ -231,6 +232,7 @@ function installSrfLocals(srf, logger) {
|
||||
getListPosition,
|
||||
getNuanceAccessToken,
|
||||
getIbmAccessToken,
|
||||
getAwsAuthToken,
|
||||
addToSortedSet,
|
||||
retrieveFromSortedSet,
|
||||
retrieveByPatternSortedSet,
|
||||
|
||||
@@ -60,7 +60,13 @@ const stickyVars = {
|
||||
aws: [
|
||||
'AWS_VOCABULARY_NAME',
|
||||
'AWS_VOCABULARY_FILTER_METHOD',
|
||||
'AWS_VOCABULARY_FILTER_NAME'
|
||||
'AWS_VOCABULARY_FILTER_NAME',
|
||||
'AWS_LANGUAGE_MODEL_NAME',
|
||||
'AWS_ACCESS_KEY_ID',
|
||||
'AWS_SECRET_ACCESS_KEY',
|
||||
'AWS_REGION',
|
||||
'AWS_SECURITY_TOKEN',
|
||||
'AWS_PII_ENTITY_TYPES'
|
||||
],
|
||||
nuance: [
|
||||
'NUANCE_ACCESS_TOKEN',
|
||||
@@ -368,11 +374,19 @@ const normalizeMicrosoft = (evt, channel, language) => {
|
||||
|
||||
const normalizeAws = (evt, channel, language) => {
|
||||
const copy = JSON.parse(JSON.stringify(evt));
|
||||
const alternatives = evt.Transcript?.Results[0]?.Alternatives.map((alt) => {
|
||||
const items = alt.Items.filter((item) => item.Type === 'pronunciation' && 'Confidence' in item);
|
||||
const confidence = items.reduce((acc, item) => acc + item.Confidence, 0) / items.length;
|
||||
return {
|
||||
transcript: alt.Transcript,
|
||||
confidence
|
||||
};
|
||||
});
|
||||
return {
|
||||
language_code: language,
|
||||
channel_tag: channel,
|
||||
is_final: evt[0].is_final,
|
||||
alternatives: evt[0].alternatives,
|
||||
is_final: evt.Transcript?.Results[0].IsPartial === false,
|
||||
alternatives,
|
||||
vendor: {
|
||||
name: 'aws',
|
||||
evt: copy
|
||||
@@ -483,16 +497,29 @@ module.exports = (logger) => {
|
||||
};
|
||||
}
|
||||
else if (['aws', 'polly'].includes(vendor)) {
|
||||
const {awsOptions = {}} = rOpts;
|
||||
const vocabularyName = awsOptions.vocabularyName || rOpts.vocabularyName;
|
||||
const vocabularyFilterName = awsOptions.vocabularyFilterName || rOpts.vocabularyFilterName;
|
||||
const filterMethod = awsOptions.vocabularyFilterMethod || rOpts.filterMethod;
|
||||
opts = {
|
||||
...opts,
|
||||
...(rOpts.vocabularyName && {AWS_VOCABULARY_NAME: rOpts.vocabularyName}),
|
||||
...(rOpts.vocabularyFilterName && {AWS_VOCABULARY_FILTER_NAME: rOpts.vocabularyFilterName}),
|
||||
...(rOpts.filterMethod && {AWS_VOCABULARY_FILTER_METHOD: rOpts.filterMethod}),
|
||||
...(vocabularyName && {AWS_VOCABULARY_NAME: vocabularyName}),
|
||||
...(vocabularyFilterName && {AWS_VOCABULARY_FILTER_NAME: vocabularyFilterName}),
|
||||
...(filterMethod && {AWS_VOCABULARY_FILTER_METHOD: filterMethod}),
|
||||
...(sttCredentials && {
|
||||
AWS_ACCESS_KEY_ID: sttCredentials.accessKeyId,
|
||||
AWS_SECRET_ACCESS_KEY: sttCredentials.secretAccessKey,
|
||||
AWS_REGION: sttCredentials.region
|
||||
AWS_REGION: sttCredentials.region,
|
||||
AWS_SECURITY_TOKEN: sttCredentials.securityToken
|
||||
}),
|
||||
...(awsOptions.accessKey && {AWS_ACCESS_KEY_ID: awsOptions.accessKey}),
|
||||
...(awsOptions.secretKey && {AWS_SECRET_ACCESS_KEY: awsOptions.secretKey}),
|
||||
...(awsOptions.region && {AWS_REGION: awsOptions.region}),
|
||||
...(awsOptions.securityToken && {AWS_SECURITY_TOKEN: awsOptions.securityToken}),
|
||||
...(awsOptions.languageModelName && {AWS_LANGUAGE_MODEL_NAME: awsOptions.languageModelName}),
|
||||
...(awsOptions.piiEntityTypes?.length && {AWS_PII_ENTITY_TYPES: awsOptions.piiEntityTypes.join(',')}),
|
||||
...(awsOptions.piiIdentifyEntities && {AWS_PII_IDENTIFY_ENTITIES: true}),
|
||||
...(awsOptions.languageModelName && {AWS_LANGUAGE_MODEL_NAME: awsOptions.languageModelName}),
|
||||
};
|
||||
}
|
||||
else if ('microsoft' === vendor) {
|
||||
@@ -801,6 +828,17 @@ module.exports = (logger) => {
|
||||
if (clientId && secret) return {client_id: clientId, secret};
|
||||
if (kryptonEndpoint) return {nuance_stt_uri: kryptonEndpoint};
|
||||
}
|
||||
else if (recognizer.vendor === 'aws') {
|
||||
const {accessKey, secretKey, region, securityToken} = recognizer.awsOptions || {};
|
||||
if (accessKey || secretKey || region || securityToken) {
|
||||
return {
|
||||
accessKeyId: accessKey,
|
||||
secretAccessKey: secretKey,
|
||||
region,
|
||||
securityToken
|
||||
};
|
||||
}
|
||||
}
|
||||
else if (recognizer.vendor === 'nvidia') {
|
||||
const {rivaUri} = recognizer.nvidiaOptions || {};
|
||||
if (rivaUri) return {riva_uri: rivaUri};
|
||||
|
||||
812
package-lock.json
generated
812
package-lock.json
generated
File diff suppressed because it is too large
Load Diff
22
package.json
22
package.json
@@ -31,19 +31,19 @@
|
||||
"@jambonz/http-health-check": "^0.0.1",
|
||||
"@jambonz/mw-registrar": "^0.2.4",
|
||||
"@jambonz/realtimedb-helpers": "^0.8.7",
|
||||
"@jambonz/speech-utils": "^0.0.31",
|
||||
"@jambonz/speech-utils": "^0.0.30",
|
||||
"@jambonz/stats-collector": "^0.1.9",
|
||||
"@jambonz/time-series": "^0.2.8",
|
||||
"@jambonz/verb-specifications": "^0.0.46",
|
||||
"@opentelemetry/api": "^1.4.0",
|
||||
"@opentelemetry/exporter-jaeger": "^1.9.0",
|
||||
"@opentelemetry/exporter-trace-otlp-http": "^0.35.0",
|
||||
"@opentelemetry/exporter-zipkin": "^1.9.0",
|
||||
"@opentelemetry/instrumentation": "^0.35.0",
|
||||
"@opentelemetry/resources": "^1.9.0",
|
||||
"@opentelemetry/sdk-trace-base": "^1.9.0",
|
||||
"@opentelemetry/sdk-trace-node": "^1.9.0",
|
||||
"@opentelemetry/semantic-conventions": "^1.9.0",
|
||||
"@jambonz/verb-specifications": "^0.0.49",
|
||||
"@opentelemetry/api": "^1.7.0",
|
||||
"@opentelemetry/exporter-jaeger": "^1.18.1",
|
||||
"@opentelemetry/exporter-trace-otlp-http": "^0.45.1",
|
||||
"@opentelemetry/exporter-zipkin": "^1.18.1",
|
||||
"@opentelemetry/instrumentation": "^0.45.1",
|
||||
"@opentelemetry/resources": "^1.18.1",
|
||||
"@opentelemetry/sdk-trace-base": "^1.18.1",
|
||||
"@opentelemetry/sdk-trace-node": "^1.18.1",
|
||||
"@opentelemetry/semantic-conventions": "^1.18.1",
|
||||
"bent": "^7.3.12",
|
||||
"debug": "^4.3.4",
|
||||
"deepcopy": "^2.1.0",
|
||||
|
||||
@@ -137,64 +137,6 @@ test('\'transcribe\' test - aws', async(t) => {
|
||||
}
|
||||
});
|
||||
|
||||
|
||||
test('\'transcribe\' test - deepgram config options', async(t) => {
|
||||
if (!DEEPGRAM_API_KEY ) {
|
||||
t.pass('skipping deepgram tests');
|
||||
return t.end();
|
||||
}
|
||||
clearModule.all();
|
||||
const {srf, disconnect} = require('../app');
|
||||
|
||||
try {
|
||||
await connect(srf);
|
||||
// GIVEN
|
||||
let verbs = [
|
||||
{
|
||||
"verb": "config",
|
||||
"recognizer": {
|
||||
"vendor": "deepgram",
|
||||
"language": "en-US",
|
||||
"deepgramOptions": {
|
||||
"model": "2-ea",
|
||||
"tier": "nova",
|
||||
"numerals": true,
|
||||
"ner": true,
|
||||
"vadTurnoff": 10,
|
||||
"keywords": [
|
||||
"CPT"
|
||||
]
|
||||
}
|
||||
}
|
||||
},
|
||||
{
|
||||
"verb": "transcribe",
|
||||
"transcriptionHook": "/transcriptionHook",
|
||||
"recognizer": {
|
||||
"vendor": "deepgram",
|
||||
"hints": ["customer support", "sales", "human resources", "HR"],
|
||||
"deepgramOptions": {
|
||||
"apiKey": DEEPGRAM_API_KEY,
|
||||
}
|
||||
}
|
||||
}
|
||||
];
|
||||
let from = "gather_success";
|
||||
await provisionCallHook(from, verbs);
|
||||
// THEN
|
||||
await sippUac('uac-gather-account-creds-success.xml', '172.38.0.10', from);
|
||||
let obj = await getJSON(`http://127.0.0.1:3100/lastRequest/${from}_actionHook`);
|
||||
t.ok(obj.body.speech.alternatives[0].transcript.toLowerCase().includes('like to speak to customer support'),
|
||||
'transcribe: succeeds when using deepgram credentials');
|
||||
|
||||
disconnect();
|
||||
} catch (err) {
|
||||
console.log(`error received: ${err}`);
|
||||
disconnect();
|
||||
t.error(err);
|
||||
}
|
||||
});
|
||||
|
||||
test('\'transcribe\' test - deepgram', async(t) => {
|
||||
if (!DEEPGRAM_API_KEY ) {
|
||||
t.pass('skipping deepgram tests');
|
||||
|
||||
Reference in New Issue
Block a user