From a0617660f5d7bfbe1f5a47eb017d5e07b60866bb Mon Sep 17 00:00:00 2001
From: Adam Reed <adam@adamjreed.com>
Date: Mon, 28 Mar 2016 18:44:27 +0000
Subject: [PATCH] Add IBM Watson as a STT option

---
 client/stt.py | 145 ++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 145 insertions(+)

diff --git a/client/stt.py b/client/stt.py
index a48696099..eae42a372 100644
--- a/client/stt.py
+++ b/client/stt.py
@@ -427,6 +427,151 @@ def is_available(cls):
         return diagnose.check_network_connection()
 
 
+class WatsonSTT(AbstractSTTEngine):
+    """
+    Speech-To-Text implementation which relies on the IBM Watson Speech-To-Text
+    API. This requires an IBM Bluemix account, but the first 1000 minutes of
+    transcribing per month are free.
+
+    To obtain a login:
+    1. Register for IBM Bluemix here:
+       https://console.ng.bluemix.net/registration/
+    2. Once you've logged in, click the "Use Services & APIs" link on the
+       dashboard
+    3. Click the "Speech To Text" icon
+    4. In the form on the right, leave all options as defaults and click Create
+    5. You'll now have a new service listed on your dashboard. If you click
+       that service there will be a navigation option for "Service Credentials"
+       in the left hand nav. Find your username and password there.
+
+    Excerpt from sample profile.yml:
+
+        ...
+        timezone: US/Pacific
+        stt_engine: watson
+        watson:
+            username: $YOUR_USERNAME_HERE
+            password: $YOUR_PASSWORD_HERE
+
+    """
+
+    SLUG = 'watson'
+
+    def __init__(self, username=None, password=None, language='en-us'):
+        # FIXME: get init args from config
+        """
+        Arguments:
+        username - the watson api username credential
+        password - the watson api password credential
+        """
+        self._logger = logging.getLogger(__name__)
+        self._username = None
+        self._password = None
+        self._http = requests.Session()
+        self.username = username
+        self.password = password
+
+    @property
+    def request_url(self):
+        return self._request_url
+
+    @property
+    def username(self):
+        return self._username
+
+    @username.setter
+    def username(self, value):
+        self._username = value
+
+    @property
+    def password(self):
+        return self._password
+
+    @password.setter
+    def password(self, value):
+        self._password = value
+
+    @classmethod
+    def get_config(cls):
+        # FIXME: Replace this as soon as we have a config module
+        config = {}
+        # HMM dir
+        # Try to get hmm_dir from config
+        profile_path = jasperpath.config('profile.yml')
+        if os.path.exists(profile_path):
+            with open(profile_path, 'r') as f:
+                profile = yaml.safe_load(f)
+                if 'watson' in profile:
+                    if 'username' in profile['watson']:
+                        config['username'] = profile['watson']['username']
+                    if 'password' in profile['watson']:
+                        config['password'] = profile['watson']['password']
+        return config
+
+    def transcribe(self, fp):
+        """
+        Performs STT via the Watson Speech-to-Text API, transcribing an audio
+        file and returning an English string.
+
+        Arguments:
+        fp -- the path to the .wav file to be transcribed
+        """
+
+        if not self.username:
+            self._logger.critical('Username missing, transcription request ' +
+                                  'aborted.')
+            return []
+        elif not self.password:
+            self._logger.critical('Password missing, transcription ' +
+                                  'request aborted.')
+            return []
+
+        wav = wave.open(fp, 'rb')
+        frame_rate = wav.getframerate()
+        wav.close()
+        data = fp.read()
+
+        headers = {'content-type':
+                   'audio/l16; rate=%s; channels=1' % frame_rate}
+        r = self._http.post(
+            'https://stream.watsonplatform.net/' +
+            'speech-to-text/api/v1/recognize?continuous=true',
+            data=data, headers=headers, auth=(self.username, self.password)
+        )
+        try:
+            r.raise_for_status()
+        except requests.exceptions.HTTPError as e:
+            self._logger.critical('Request failed with http status %d',
+                                  r.status_code)
+            if r.status_code == requests.codes['forbidden']:
+                self._logger.warning('Status 403 is probably caused by ' +
+                                     'invalid credentials.')
+            return []
+        r.encoding = 'utf-8'
+        try:
+            response = r.json()
+            if len(response['results']) == 0:
+                # Response result is empty
+                raise ValueError('Nothing has been transcribed.')
+            results = [alt['transcript'] for alt
+                       in response['results'][0]['alternatives']]
+        except ValueError as e:
+            self._logger.warning('Empty response: %s', e.args[0])
+            results = []
+        except (KeyError, IndexError):
+            self._logger.warning('Cannot parse response.', exc_info=True)
+            results = []
+        else:
+            # Convert all results to uppercase
+            results = tuple(result.strip().upper() for result in results)
+            self._logger.info('Transcribed: %r', results)
+        return results
+
+    @classmethod
+    def is_available(cls):
+        return diagnose.check_network_connection()
+
+
 class AttSTT(AbstractSTTEngine):
     """
     Speech-To-Text implementation which relies on the AT&T Speech API.