Merge pull request #232 from chughts/sttupdate

STT diarization support

Author: chughts

README.md

@@ -7,6 +7,10 @@ Node-RED Watson Nodes for IBM Bluemix
 
 <a href="https://cla-assistant.io/watson-developer-cloud/node-red-node-watson"><img src="https://cla-assistant.io/readme/badge/watson-developer-cloud/node-red-node-watson" alt="CLA assistant" /></a>
 
+### New in version 0.4.32
+- Added diarization support to STT Node via the parameter speaker_labels
+- STT and TTS node use url based services utility to detect bound service.
+
 ### New in version 0.4.31
 - New V1 Query Builder Node for the Discovery Node
 

package.json

@@ -1,6 +1,6 @@
 {
   "name": "node-red-node-watson",
-  "version": "0.4.31",
+  "version": "0.4.32",
   "description": "A collection of Node-RED nodes for IBM Watson services",
   "dependencies": {
     "alchemy-api": "^1.3.0",

services/language_translator_identify/v2.html

@@ -35,7 +35,7 @@
 </script>
 
 <script type="text/x-red" data-help-name="watson-language-translator-identify">
-    <p>Packaged in with release 0.4.31 of node-red-node-watson</p>
+    <p>Packaged in with release 0.4.32 of node-red-node-watson</p>
     <p>The Watson Language Translator service can be used to identify languages used in a text input. <p>
     <p>Node input : </p>
     <ul>

services/speech_to_text/v1.html

@@ -62,6 +62,12 @@
         <input type="checkbox" id="node-input-continuous"></input>
         </select>
     </div>
+    <div class="form-row">
+        <label for="node-input-speakerlabels"><i class="fa fa-audio-o"></i> Speaker Labels</label>
+        <input type="checkbox" id="node-input-speakerlabels"></input>
+        </select>
+    </div>
+
 
 </script>
 
@@ -82,6 +88,8 @@
     source sampling rates, <b>Narrowband (8KHz) and Broadband (16Khz)</b>. Higher sample rates will be downsampled
     automatically but the is not true in reverse.</p>
     <p>Use the continuous property to choose whether the decoding should stop at the first pause or wait until the end of the file.</p>
+    <p>Use the speaker label property to  identify which individuals spoke which words in a multi-participant exchange.
+    </p>
     <p>The returned audio transcription will be returned on <code>msg.transcription</code>.</p>
     <p>The full response, including alternative transcriptions can be found on
     <code>msg.fullresult</code>.</p>
@@ -126,6 +134,7 @@
       $('#credentials-not-found').show();
       $('label#node-label-message').parent().hide();
       $('input#node-input-continuous').parent().hide();
+      $('input#node-input-speakerlabels').parent().hide();
       $('select#node-input-lang').parent().hide();
       $('select#node-input-band').parent().hide();
     }
@@ -137,11 +146,13 @@
     if (stt.models) {
       $('label#node-label-message').parent().hide();
       $('input#node-input-continuous').parent().show();
+      $('input#node-input-speakerlabels').parent().show();
       $('select#node-input-lang').parent().show();
       $('select#node-input-band').parent().show();
     } else {
       $('label#node-label-message').parent().hide();
       $('input#node-input-continuous').parent().hide();
+      $('input#node-input-speakerlabels').parent().hide();
       $('select#node-input-lang').parent().hide();
       $('select#node-input-band').parent().hide();
     }
@@ -323,6 +334,7 @@
             defaults: {
                 name: {value: ''},
                 continuous: {value: true},
+                speakerlabels: {value: true},
                 lang: {value: ''},
                 langhidden: {value: ''},
                 band: {value: ''},

services/speech_to_text/v1.js

@@ -15,16 +15,16 @@
  **/
 
 module.exports = function (RED) {
-  var request = require('request');
-  var cfenv = require('cfenv');
-  var temp = require('temp');
-  var url = require('url');
-  var fs = require('fs');
-  var fileType = require('file-type');
-  //var watson = require('watson-developer-cloud');
-  var sttV1 = require('watson-developer-cloud/speech-to-text/v1');
-
-  var service = cfenv.getAppEnv().getServiceCreds(/speech to text/i);
+  const SERVICE_IDENTIFIER = 'speech-to-text';
+  var request = require('request'),
+    cfenv = require('cfenv'),
+    temp = require('temp'),
+    url = require('url'),
+    fs = require('fs'),
+    fileType = require('file-type'),
+    serviceutils = require('../../utilities/service-utils'),
+    sttV1 = require('watson-developer-cloud/speech-to-text/v1'),
+    service = serviceutils.getServiceCreds(SERVICE_IDENTIFIER);
 
   // Require the Cloud Foundry Module to pull credentials from bound service
   // If they are found then the username and password will be stored in
@@ -174,7 +174,8 @@ module.exports = function (RED) {
           audio: audio,
           content_type: 'audio/' + format,
           model: model,
-          continuous: config.continuous
+          continuous: config.continuous ? config.continuous : false,
+          speaker_labels: config.speakerlabels ? config.speakerlabels : false
         };
 
         node.status({fill:'blue', shape:'dot', text:'requesting'});

services/text_to_speech/v1.html

@@ -93,6 +93,7 @@
                     'de-DE': 'German',
                     'zh-CN': 'Mandarin',
                     'es-ES': 'Spanish',
+                    'es-LA': 'Latin American Spanish',
                     'es-US': 'US Spanish',
                     'ar-AR': 'Arablic',
                     'ja-JP': 'Japanese'

services/text_to_speech/v1.js

@@ -15,8 +15,10 @@
  **/
 
 module.exports = function(RED) {
+  const SERVICE_IDENTIFIER = 'text-to-speech';
   var cfenv = require('cfenv');
-  var watson = require('watson-developer-cloud');
+  var TextToSpeechV1 = require('watson-developer-cloud/text-to-speech/v1');
+  var serviceutils = require('../../utilities/service-utils');
 
   // Require the Cloud Foundry Module to pull credentials from bound service
   // If they are found then the username and password will be stored in
@@ -30,7 +32,8 @@ module.exports = function(RED) {
 
   var username, password, sUsername, sPassword;
 
-  var service = cfenv.getAppEnv().getServiceCreds(/text to speech/i)
+  //var service = cfenv.getAppEnv().getServiceCreds(/text to speech/i)
+  var service = serviceutils.getServiceCreds(SERVICE_IDENTIFIER);
 
   if (service) {
     sUsername = service.username;
@@ -44,11 +47,9 @@ module.exports = function(RED) {
 
   // API used by widget to fetch available models
   RED.httpAdmin.get('/watson-text-to-speech/voices', function (req, res) {
-    var tts = watson.text_to_speech({
+    var tts = new TextToSpeechV1({
       username: sUsername ? sUsername : req.query.un,
-      password: sPassword ? sPassword : req.query.pwd,
-      version: 'v1',
-      url: 'https://stream.watsonplatform.net/text-to-speech/api'
+      password: sPassword ? sPassword : req.query.pwd
     });
 
     tts.voices({}, function(err, voices){
@@ -75,19 +76,17 @@ module.exports = function(RED) {
       }
 
       username = sUsername || this.credentials.username;
-      password = sPassword || this.credentials.password || config.password; 
+      password = sPassword || this.credentials.password || config.password;
 
       if (!username || !password) {
         var message = 'Missing Speech To Text service credentials';
         node.error(message, msg);
         return;
       }
 
-      var text_to_speech = watson.text_to_speech({
+      var text_to_speech = new TextToSpeechV1({
         username: username,
-        password: password,
-        version: 'v1',
-        url: 'https://stream.watsonplatform.net/text-to-speech/api'
+        password: password 
       });
 
       var params = {