FazBrowse GitHub Viewer | Trending |
URL:
| Home
Tools: [Download Repo ZIP]   [Original HTTPS Page]

Merge pull request #2426 from daspecster/add-speech-async · googleapis/google-cloud-python@4fd0eca · GitHub

/ google-cloud-python Public

Commit 4fd0eca

Browse files
daspecster
authored
Merge pull request #2426 from daspecster/add-speech-async
Add speech asynchronous recognize support.
2 parents 2de8865 + 5fe1235 commit 4fd0eca

12 files changed

Lines changed: 747 additions & 67 deletions

File tree

‎packages/google-cloud-speech/google/cloud/speech/client.py‎

Lines changed: 181 additions & 57 deletions
Original file line numberDiff line numberDiff line change
@@ -19,30 +19,8 @@
1919
from google.cloud._helpers import _to_bytes
2020
from google.cloud import client as client_module
2121
from google.cloud.speech.connection import Connection
22-
23-
24-
class Encoding(object):
25-
"""Audio encoding types.
26-
27-
See:
28-
https://cloud.google.com/speech/reference/rest/v1beta1/\
29-
RecognitionConfig#AudioEncoding
30-
"""
31-
32-
LINEAR16 = 'LINEAR16'
33-
"""LINEAR16 encoding type."""
34-
35-
FLAC = 'FLAC'
36-
"""FLAC encoding type."""
37-
38-
MULAW = 'MULAW'
39-
"""MULAW encoding type."""
40-
41-
AMR = 'AMR'
42-
"""AMR encoding type."""
43-
44-
AMR_WB = 'AMR_WB'
45-
"""AMR_WB encoding type."""
22+
from google.cloud.speech.encoding import Encoding
23+
from google.cloud.speech.operation import Operation
4624

4725

4826
class Client(client_module.Client):
@@ -68,6 +46,81 @@ class Client(client_module.Client):
6846

6947
_connection_class = Connection
7048

49+
def async_recognize(self, content, source_uri, encoding, sample_rate,
50+
language_code=None, max_alternatives=None,
51+
profanity_filter=None, speech_context=None):
52+
"""Asychronous Recognize request to Google Speech API.
53+
54+
.. _async_recognize: https://cloud.google.com/speech/reference/\
55+
rest/v1beta1/speech/asyncrecognize
56+
57+
See `async_recognize`_.
58+
59+
:type content: bytes
60+
:param content: Byte stream of audio.
61+
62+
:type source_uri: str
63+
:param source_uri: URI that points to a file that contains audio
64+
data bytes as specified in RecognitionConfig.
65+
Currently, only Google Cloud Storage URIs are
66+
supported, which must be specified in the following
67+
format: ``gs://bucket_name/object_name``.
68+
69+
:type encoding: str
70+
:param encoding: encoding of audio data sent in all RecognitionAudio
71+
messages, can be one of: :attr:`~.Encoding.LINEAR16`,
72+
:attr:`~.Encoding.FLAC`, :attr:`~.Encoding.MULAW`,
73+
:attr:`~.Encoding.AMR`, :attr:`~.Encoding.AMR_WB`
74+
75+
:type sample_rate: int
76+
:param sample_rate: Sample rate in Hertz of the audio data sent in all
77+
requests. Valid values are: 8000-48000. For best
78+
results, set the sampling rate of the audio source
79+
to 16000 Hz. If that's not possible, use the
80+
native sample rate of the audio source (instead of
81+
re-sampling).
82+
83+
:type language_code: str
84+
:param language_code: (Optional) The language of the supplied audio as
85+
BCP-47 language tag. Example: ``'en-GB'``.
86+
If omitted, defaults to ``'en-US'``.
87+
88+
:type max_alternatives: int
89+
:param max_alternatives: (Optional) Maximum number of recognition
90+
hypotheses to be returned. The server may
91+
return fewer than maxAlternatives.
92+
Valid values are 0-30. A value of 0 or 1
93+
will return a maximum of 1. Defaults to 1
94+
95+
:type profanity_filter: bool
96+
:param profanity_filter: If True, the server will attempt to filter
97+
out profanities, replacing all but the
98+
initial character in each filtered word with
99+
asterisks, e.g. ``'f***'``. If False or
100+
omitted, profanities won't be filtered out.
101+
102+
:type speech_context: list
103+
:param speech_context: A list of strings (max 50) containing words and
104+
phrases "hints" so that the speech recognition
105+
is more likely to recognize them. This can be
106+
used to improve the accuracy for specific words
107+
and phrases. This can also be used to add new
108+
words to the vocabulary of the recognizer.
109+
110+
:rtype: `~google.cloud.speech.operation.Operation`
111+
:returns: ``Operation`` for asynchronous request to Google Speech API.
112+
"""
113+
114+
data = _build_request_data(content, source_uri, encoding,
115+
sample_rate, language_code,
116+
max_alternatives, profanity_filter,
117+
speech_context)
118+
119+
api_response = self.connection.api_request(
120+
method='POST', path='speech:asyncrecognize', data=data)
121+
122+
return Operation.from_api_repr(self, api_response)
123+
71124
def sync_recognize(self, content, source_uri, encoding, sample_rate,
72125
language_code=None, max_alternatives=None,
73126
profanity_filter=None, speech_context=None):
@@ -139,44 +192,115 @@ def sync_recognize(self, content, source_uri, encoding, sample_rate,
139192
between 0 and 1.
140193
"""
141194

142-
if content is None and source_uri is None:
143-
raise ValueError('content and source_uri cannot be both '
144-
'equal to None')
145-
146-
if content is not None and source_uri is not None:
147-
raise ValueError('content and source_uri cannot be both '
148-
'different from None')
195+
data = _build_request_data(content, source_uri, encoding,
196+
sample_rate, language_code,
197+
max_alternatives, profanity_filter,
198+
speech_context)
149199

150-
if encoding is None:
151-
raise ValueError('encoding cannot be None')
152-
if sample_rate is None:
153-
raise ValueError('sample_rate cannot be None')
200+
api_response = self.connection.api_request(
201+
method='POST', path='speech:syncrecognize', data=data)
154202

155-
if content is not None:
156-
audio = {'content': b64encode(_to_bytes(content))}
203+
if len(api_response['results']) == 1:
204+
return api_response['results'][0]['alternatives']
157205
else:
158-
audio = {'uri': source_uri}
206+
raise ValueError('result in api should have length 1')
159207

160-
config = {'encoding': encoding, 'sampleRate': sample_rate}
161208

162-
if language_code is not None:
163-
config['languageCode'] = language_code
164-
if max_alternatives is not None:
165-
config['maxAlternatives'] = max_alternatives
166-
if profanity_filter is not None:
167-
config['profanityFilter'] = profanity_filter
168-
if speech_context is not None:
169-
config['speechContext'] = {'phrases': speech_context}
209+
def _build_request_data(content, source_uri, encoding, sample_rate,
210+
language_code=None, max_alternatives=None,
211+
profanity_filter=None, speech_context=None):
212+
"""Builds the request data before making API request.
213+
214+
:type content: bytes
215+
:param content: Byte stream of audio.
216+
217+
:type source_uri: str
218+
:param source_uri: URI that points to a file that contains audio
219+
data bytes as specified in RecognitionConfig.
220+
Currently, only Google Cloud Storage URIs are
221+
supported, which must be specified in the following
222+
format: ``gs://bucket_name/object_name``.
223+
224+
:type encoding: str
225+
:param encoding: encoding of audio data sent in all RecognitionAudio
226+
messages, can be one of: :attr:`~.Encoding.LINEAR16`,
227+
:attr:`~.Encoding.FLAC`, :attr:`~.Encoding.MULAW`,
228+
:attr:`~.Encoding.AMR`, :attr:`~.Encoding.AMR_WB`
229+
230+
:type sample_rate: int
231+
:param sample_rate: Sample rate in Hertz of the audio data sent in all
232+
requests. Valid values are: 8000-48000. For best
233+
results, set the sampling rate of the audio source
234+
to 16000 Hz. If that's not possible, use the
235+
native sample rate of the audio source (instead of
236+
re-sampling).
237+
238+
:type language_code: str
239+
:param language_code: (Optional) The language of the supplied audio as
240+
BCP-47 language tag. Example: ``'en-GB'``.
241+
If omitted, defaults to ``'en-US'``.
242+
243+
:type max_alternatives: int
244+
:param max_alternatives: (Optional) Maximum number of recognition
245+
hypotheses to be returned. The server may
246+
return fewer than maxAlternatives.
247+
Valid values are 0-30. A value of 0 or 1
248+
will return a maximum of 1. Defaults to 1
249+
250+
:type profanity_filter: bool
251+
:param profanity_filter: If True, the server will attempt to filter
252+
out profanities, replacing all but the
253+
initial character in each filtered word with
254+
asterisks, e.g. ``'f***'``. If False or
255+
omitted, profanities won't be filtered out.
256+
257+
:type speech_context: list
258+
:param speech_context: A list of strings (max 50) containing words and
259+
phrases "hints" so that the speech recognition
260+
is more likely to recognize them. This can be
261+
used to improve the accuracy for specific words
262+
and phrases. This can also be used to add new
263+
words to the vocabulary of the recognizer.
264+
265+
:rtype: dict
266+
:returns: Dictionary with required data for Google Speech API.
267+
"""
268+
if content is None and source_uri is None:
269+
raise ValueError('content and source_uri cannot be both '
270+
'equal to None')
170271

171-
data = {
172-
'audio': audio,
173-
'config': config,
174-
}
272+
if content is not None and source_uri is not None:
273+
raise ValueError('content and source_uri cannot be both '
274+
'different from None')
175275

176-
api_response = self.connection.api_request(
177-
method='POST', path='syncrecognize', data=data)
276+
if encoding is None:
277+
raise ValueError('encoding cannot be None')
178278

179-
if len(api_response['results']) == 1:
180-
return api_response['results'][0]['alternatives']
181-
else:
182-
raise ValueError('result in api should have length 1')
279+
encoding_value = getattr(Encoding, encoding)
280+
281+
if sample_rate is None:
282+
raise ValueError('sample_rate cannot be None')
283+
284+
if content is not None:
285+
audio = {'content': b64encode(_to_bytes(content))}
286+
else:
287+
audio = {'uri': source_uri}
288+
289+
config = {'encoding': encoding_value,
290+
'sampleRate': sample_rate}
291+
292+
if language_code is not None:
293+
config['languageCode'] = language_code
294+
if max_alternatives is not None:
295+
config['maxAlternatives'] = max_alternatives
296+
if profanity_filter is not None:
297+
config['profanityFilter'] = profanity_filter
298+
if speech_context is not None:
299+
config['speechContext'] = {'phrases': speech_context}
300+
301+
data = {
302+
'audio': audio,
303+
'config': config,
304+
}
305+
306+
return data

‎packages/google-cloud-speech/google/cloud/speech/connection.py‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -26,7 +26,7 @@ class Connection(base_connection.JSONConnection):
2626
API_VERSION = 'v1beta1'
2727
"""The version of the API, used in building the API call's URL."""
2828

29-
API_URL_TEMPLATE = '{api_base_url}/{api_version}/speech:{path}'
29+
API_URL_TEMPLATE = '{api_base_url}/{api_version}/{path}'
3030
"""A template for the URL of a particular API call."""
3131

3232
SCOPE = ('https://www.googleapis.com/auth/cloud-platform',)

‎packages/google-cloud-speech/google/cloud/speech/encoding.py‎

Lines changed: 39 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,39 @@
1+
# Copyright 2016 Google Inc. All rights reserved.
2+
#
3+
# Licensed under the Apache License, Version 2.0 (the "License");
4+
# you may not use this file except in compliance with the License.
5+
# You may obtain a copy of the License at
6+
#
7+
# http://www.apache.org/licenses/LICENSE-2.0
8+
#
9+
# Unless required by applicable law or agreed to in writing, software
10+
# distributed under the License is distributed on an "AS IS" BASIS,
11+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
12+
# See the License for the specific language governing permissions and
13+
# limitations under the License.
14+
15+
"""Encodings used by the Google Cloud Speech API."""
16+
17+
18+
class Encoding(object):
19+
"""Audio encoding types.
20+
21+
See:
22+
https://cloud.google.com/speech/reference/rest/v1beta1/\
23+
RecognitionConfig#AudioEncoding
24+
"""
25+
26+
LINEAR16 = 'LINEAR16'
27+
"""LINEAR16 encoding type."""
28+
29+
FLAC = 'FLAC'
30+
"""FLAC encoding type."""
31+
32+
MULAW = 'MULAW'
33+
"""MULAW encoding type."""
34+
35+
AMR = 'AMR'
36+
"""AMR encoding type."""
37+
38+
AMR_WB = 'AMR_WB'
39+
"""AMR_WB encoding type."""

‎packages/google-cloud-speech/google/cloud/speech/metadata.py‎

Lines changed: 78 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,78 @@
1+
# Copyright 2016 Google Inc.
2+
#
3+
# Licensed under the Apache License, Version 2.0 (the "License");
4+
# you may not use this file except in compliance with the License.
5+
# You may obtain a copy of the License at
6+
#
7+
# http://www.apache.org/licenses/LICENSE-2.0
8+
#
9+
# Unless required by applicable law or agreed to in writing, software
10+
# distributed under the License is distributed on an "AS IS" BASIS,
11+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
12+
# See the License for the specific language governing permissions and
13+
# limitations under the License.
14+
15+
"""Metadata representation from Google Speech API"""
16+
17+
from google.cloud._helpers import _rfc3339_to_datetime
18+
19+
20+
class Metadata(object):
21+
"""Representation of metadata from a Google Speech API Operation.
22+
23+
:type last_update: datetime
24+
:param last_update: When the Speech operation was last updated.
25+
26+
:type start_time: datetime
27+
:param start_time: When the Speech operation was started.
28+
29+
:type progress_percent: int
30+
:param progress_percent: Percentage of operation that has been completed.
31+
"""
32+
def __init__(self, last_update, start_time, progress_percent):
33+
self._last_update = last_update
34+
self._start_time = start_time
35+
self._progress_percent = progress_percent
36+
37+
@classmethod
38+
def from_api_repr(cls, response):
39+
"""Factory: construct representation of operation metadata.
40+
41+
:type response: dict
42+
:param response: Dictionary containing operation metadata.
43+
44+
:rtype: :class:`~google.cloud.speech.metadata.Metadata`
45+
:returns: Instance of operation Metadata.
46+
"""
47+
last_update = _rfc3339_to_datetime(response['lastUpdateTime'])
48+
start_time = _rfc3339_to_datetime(response['startTime'])
49+
progress_percent = response['progressPercent']
50+
51+
return cls(last_update, start_time, progress_percent)
52+
53+
@property
54+
def last_update(self):
55+
"""Last time operation was updated.
56+
57+
:rtype: datetime
58+
:returns: Datetime when operation was last updated.
59+
"""
60+
return self._last_update
61+
62+
@property
63+
def start_time(self):
64+
"""Start time of operation.
65+
66+
:rtype: datetime
67+
:returns: Datetime when operation was started.
68+
"""
69+
return self._start_time
70+
71+
@property
72+
def progress_percent(self):
73+
"""Progress percentage completed of operation.
74+
75+
:rtype: int
76+
:returns: Percentage of operation completed.
77+
"""
78+
return self._progress_percent

0 commit comments

Comments
 (0)
Back | FazBrowse Home | New Git URL