Add Speech Streaming API. by daspecster · Pull Request #2523 · googleapis/google-cloud-python

daspecster · 2016-10-10T23:02:48Z

One step closer to #2522.

Sorry this one is kind of rough.

I'm really not sure how to handle the interim_results for this. It ends up making for some weird object handling.

Update:
This was streamlined a lot. Results are just returned to the end user as they come in so for interim_results, you get the extra results.

dhermes · 2016-10-11T20:06:05Z

😱

daspecster · 2016-10-11T20:07:44Z

Yeah I know :( A good number of them are DTOs and tests for the DTOs. I can try and figure out a way to break this up if you want.

Sign in to view

@@ -0,0 +1,33 @@
+Speech StreamingResponseContainer


Sign in to view

+ :undoc-members:
+ :show-inheritance:
+
+


Sign in to view

+
+ >>> import io
+ >>> from google.cloud import speech
+ >>> from google.cloud.speech.encoding import Encoding


Sign in to view

+ <google.cloud.speech.streaming.container.StreamingResponseContainer object at 0x10538ee10>
+ >>> print(stream_container.responses)
+ {0: <google.cloud.speech.streaming.response.StreamingSpeechResponse object at 0x10f9ac9d0>}
+ >>> print(stream_container.responses[0].results[0].alternatives[0].confidence)


Sign in to view

+ 0.698092460632
+ >>> print(stream_container.is_finished)
+ True
+ >>> print stream_container.get_full_text()


Sign in to view

+ text = ''
+ for response in self.responses.values():
+ for result in response.results:
+ text += result.alternatives[0].transcript


Sign in to view

+ yield StreamingRecognizeRequest(audio_content=buff)
+ buff = data
+ else:
+ buff += data


Sign in to view

+ buff += data
+
+ # Clear final contents of buffer.
+ yield StreamingRecognizeRequest(audio_content=buff)


Sign in to view

@@ -0,0 +1,183 @@
+# Copyright 2016 Google Inc.


Sign in to view

+ """
+ def __init__(self, error, endpointer_type, results, result_index):
+ self._error = error
+ self._endpointer_type = endpointer_type # Should be enum.


daspecster · 2016-10-17T17:35:44Z

@dhermes @tseaver PTAL when you have a chance.

dhermes · 2016-10-17T17:38:13Z

Reviewing now

Sign in to view

+ ... sample_rate=16000)
+ ... for response in client.stream_recognize(sample):
+ ... print(response.transcript)
+ hello


Sign in to view

+
+By setting ``interim_results`` to true, interim results (tentative hypotheses)
+may be returned as they become available (these interim results are indicated
+with the is_final=false flag). If false or omitted, only is_final=true


Sign in to view

+ True
+
+
+By setting ``interim_results`` to true, interim results (tentative hypotheses)


Sign in to view

+ ... print(response.transcript)
+ hello
+ ... print(response.is_final)
+ True


Sign in to view

+ interim_results=interim_results)
+
+ for response in self.speech_api.streaming_recognize(requests):
+ if hasattr(response, 'results') or interim_results:


Sign in to view

+ speech_context=None, single_utterance=False,
+ interim_results=False):
+ """Streaming speech recognition.
+


Sign in to view

+ and phrases. This can also be used to add new
+ words to the vocabulary of the recognizer.
+
+ :type single_utterance: boolean


Sign in to view

+ words to the vocabulary of the recognizer.
+
+ :type single_utterance: boolean
+ :param single_utterance: [Optional] If false or omitted, the recognizer


Sign in to view

+ data = sample.stream.read(sample.chunk_size)
+ if not data:
+ break
+ # Optimize the request data size to around 100ms.


Sign in to view

+ :param sample: Instance of ``Sample`` containing audio information.
+
+ :type language_code: str
+ :param language_code: (Optional) The language of the supplied audio as


daspecster · 2016-10-19T13:28:15Z

@dhermes LMKWYT when you get a chance. I think I got everything you mentioned.

daspecster · 2016-10-20T14:14:34Z

Sign in to view

+ ====Response====
+ hell
+ False
+ ====Repsonse====


Sign in to view

+ google.cloud.speech.v1beta1#endpointertype
+ """
+ REVERSE_MAP = {v: k for k, v
+ in vars(StreamingRecognizeResponse.EndpointerType).items()}


Sign in to view

+ """
+ if self.results and self.results[0].alternatives:
+ top_alternative = self.results[0].alternatives[0]
+ return top_alternative.transcript


Sign in to view

+ ... print(response.transcript)
+ ... print(response.is_final)
+ hello
+ True


Sign in to view

+ results = results or []
+ self._error = error
+ self._endpointer_type = EndpointerType.REVERSE_MAP.get(
+ endpointer_type, None)


Sign in to view

+ :param error: Instance of ``Status``
+
+ :type endpointer_type: :class:`~EndpointerType`
+ :param endpointer_type: Enum of endpointer event.


Sign in to view

+ top_alternative = self.results[0].alternatives[0]
+ return top_alternative.confidence
+ else:
+ return 0.0


Sign in to view

 'google.cloud.pubsub.__init__',
 'google.cloud.resource_manager.__init__',
 'google.cloud.speech.__init__',
+ 'google.cloud.speech.streaming.__init__',


* Tables Notebooks [(#2090)](GoogleCloudPlatform/python-docs-samples#2090) * initial commit * update census * update notebooks * remove the reference to a bug [(#2100)](GoogleCloudPlatform/python-docs-samples#2100) as the bug has been fixed in the public client lib * delete this file. [(#2102)](GoogleCloudPlatform/python-docs-samples#2102) * rename file name [(#2103)](GoogleCloudPlatform/python-docs-samples#2103) * trying to fix images [(#2101)](GoogleCloudPlatform/python-docs-samples#2101) * remove typo in installation [(#2110)](GoogleCloudPlatform/python-docs-samples#2110) * Rename census_income_prediction.ipynb to getting_started_notebook.ipynb [(#2115)](GoogleCloudPlatform/python-docs-samples#2115) renaming the notebooks as Getting Started (will be in sync with the doc). It will be great if the folder could be renamed too * added back missing file package import [(#2150)](GoogleCloudPlatform/python-docs-samples#2150) * added back missing file import [(#2145)](GoogleCloudPlatform/python-docs-samples#2145) * remove incorrect reference to Iris dataset [(#2203)](GoogleCloudPlatform/python-docs-samples#2203) * conversion to jupyter/colab [(#2340)](GoogleCloudPlatform/python-docs-samples#2340) plus bug fixes * updated for the Jupyter support [(#2337)](GoogleCloudPlatform/python-docs-samples#2337) * updated readme for support Jupyter [(#2336)](GoogleCloudPlatform/python-docs-samples#2336) to approve with the updated notebook supporting jupyter * conversion to jupyer/colab [(#2339)](GoogleCloudPlatform/python-docs-samples#2339) plus bug fixes * conversion of notebook for jupyter/Colab [(#2338)](GoogleCloudPlatform/python-docs-samples#2338) conversion of the notebook to support both Jupyter and Colab + bug fixes * [BLOCKED] AutoML Tables: Docs samples updated to use new (pending) client [(#2276)](GoogleCloudPlatform/python-docs-samples#2276) * AutoML Tables: Docs samples updated to use new (pending) client * Linter warnings * add product recommendation for automl tables notebook [(#2257)](GoogleCloudPlatform/python-docs-samples#2257) * added colab filtering notebook * update to tables client * update readme * tell user to restart kernel for automl * AutoML Tables: Notebook samples updated to use new tables client [(#2424)](GoogleCloudPlatform/python-docs-samples#2424) * fix users bug and emphasize kernal restart [(#2407)](GoogleCloudPlatform/python-docs-samples#2407) * fix problems with automl docs [(#2501)](GoogleCloudPlatform/python-docs-samples#2501) Today when we try to use the function `batch_predict` follow the docs we receive and error saying: `the paramaters should be a pandas.Dataframe` it’s happens because the first parameter of the function `batch_predict` is a pandas.Dataframe. To solve this problem we need to use de named parameters of python. * Fix typo in GCS URI parameter [(#2459)](GoogleCloudPlatform/python-docs-samples#2459) * fix: fix tables notebook links and bugs [(#2601)](GoogleCloudPlatform/python-docs-samples#2601) * feat(tables): update samples to show explainability [(#2523)](GoogleCloudPlatform/python-docs-samples#2523) * show xai * local feature importance * use updated client * use fixed library * use new model * Auto-update dependencies. [(#2005)](GoogleCloudPlatform/python-docs-samples#2005) * Auto-update dependencies. * Revert update of appengine/flexible/datastore. * revert update of appengine/flexible/scipy * revert update of bigquery/bqml * revert update of bigquery/cloud-client * revert update of bigquery/datalab-migration * revert update of bigtable/quickstart * revert update of compute/api * revert update of container_registry/container_analysis * revert update of dataflow/run_template * revert update of datastore/cloud-ndb * revert update of dialogflow/cloud-client * revert update of dlp * revert update of functions/imagemagick * revert update of functions/ocr/app * revert update of healthcare/api-client/fhir * revert update of iam/api-client * revert update of iot/api-client/gcs_file_to_device * revert update of iot/api-client/mqtt_example * revert update of language/automl * revert update of run/image-processing * revert update of vision/automl * revert update testing/requirements.txt * revert update of vision/cloud-client/detect * revert update of vision/cloud-client/product_search * revert update of jobs/v2/api_client * revert update of jobs/v3/api_client * revert update of opencensus * revert update of translate/cloud-client * revert update to speech/cloud-client Co-authored-by: Kurtis Van Gent <31518063+kurtisvg@users.noreply.github.com> Co-authored-by: Doug Mahugh <dmahugh@gmail.com> * Update dependency google-cloud-automl to v0.10.0 [(#3033)](GoogleCloudPlatform/python-docs-samples#3033) Co-authored-by: Bu Sun Kim <8822365+busunkim96@users.noreply.github.com> Co-authored-by: Leah E. Cole <6719667+leahecole@users.noreply.github.com> * Simplify noxfile setup. [(#2806)](GoogleCloudPlatform/python-docs-samples#2806) * chore(deps): update dependency requests to v2.23.0 * Simplify noxfile and add version control. * Configure appengine/standard to only test Python 2.7. * Update Kokokro configs to match noxfile. * Add requirements-test to each folder. * Remove Py2 versions from everything execept appengine/standard. * Remove conftest.py. * Remove appengine/standard/conftest.py * Remove 'no-sucess-flaky-report' from pytest.ini. * Add GAE SDK back to appengine/standard tests. * Fix typo. * Roll pytest to python 2 version. * Add a bunch of testing requirements. * Remove typo. * Add appengine lib directory back in. * Add some additional requirements. * Fix issue with flake8 args. * Even more requirements. * Readd appengine conftest.py. * Add a few more requirements. * Even more Appengine requirements. * Add webtest for appengine/standard/mailgun. * Add some additional requirements. * Add workaround for issue with mailjet-rest. * Add responses for appengine/standard/mailjet. Co-authored-by: Renovate Bot <bot@renovateapp.com> * chore: some lint fixes [(#3750)](GoogleCloudPlatform/python-docs-samples#3750) * automl: tables code sample clean-up [(#3571)](GoogleCloudPlatform/python-docs-samples#3571) * delete unused tables_dataset samples * delete args code associated with unused automl_tables samples * delete tests associated with unused automl_tables samples * restore get_dataset method/yargs without region tagging * Restore update_dataset methodsa without region tagging Co-authored-by: Takashi Matsuo <tmatsuo@google.com> Co-authored-by: Leah E. Cole <6719667+leahecole@users.noreply.github.com> * add example of creating AutoML Tables client with non-default endpoint ('new' sdk) [(#3929)](GoogleCloudPlatform/python-docs-samples#3929) * add example of creating client with non-default endpoint * more test file cleanup * move connectivity print stmt out of test fn Co-authored-by: Leah E. Cole <6719667+leahecole@users.noreply.github.com> Co-authored-by: Torry Yang <sirtorry@users.noreply.github.com> * Replace GCLOUD_PROJECT with GOOGLE_CLOUD_PROJECT. [(#4022)](GoogleCloudPlatform/python-docs-samples#4022) * chore(deps): update dependency google-cloud-automl to v1 [(#4127)](GoogleCloudPlatform/python-docs-samples#4127) This PR contains the following updates: | Package | Update | Change | |---|---|---| | [google-cloud-automl](https://togithub.com/googleapis/python-automl) | major | `==0.10.0` -> `==1.0.1` | --- ### Release Notes <details> <summary>googleapis/python-automl</summary> ### [`v1.0.1`](https://togithub.com/googleapis/python-automl/blob/master/CHANGELOG.md#101-httpswwwgithubcomgoogleapispython-automlcomparev100v101-2020-06-18) [Compare Source](https://togithub.com/googleapis/python-automl/compare/v0.10.0...v1.0.1) </details> --- ### Renovate configuration :date: **Schedule**: At any time (no schedule defined). :vertical_traffic_light: **Automerge**: Disabled by config. Please merge this manually once you are satisfied. :recycle: **Rebasing**: Never, or you tick the rebase/retry checkbox. :no_bell: **Ignore**: Close this PR and you won't be reminded about this update again. --- - [ ] If you want to rebase/retry this PR, check this box --- This PR has been generated by [WhiteSource Renovate](https://renovate.whitesourcesoftware.com). View repository job log [here](https://app.renovatebot.com/dashboard#GoogleCloudPlatform/python-docs-samples). * [tables/automl] fix: update the csv file and the dataset name [(#4188)](GoogleCloudPlatform/python-docs-samples#4188) fixes#4177fixes#4178 * samples: Automl table batch test [(#4267)](GoogleCloudPlatform/python-docs-samples#4267) * added rtest req.txt * samples: added automl batch predict test * added missing package * Update tables/automl/batch_predict_test.py Co-authored-by: Bu Sun Kim <8822365+busunkim96@users.noreply.github.com> Co-authored-by: Bu Sun Kim <8822365+busunkim96@users.noreply.github.com> * samples: fixed wrong format on GCS input Uri [(#4270)](GoogleCloudPlatform/python-docs-samples#4270) ## Description Current predict sample indicates that it can multiples GCS URI inputs but it should be singular. ## Checklist - [X] Please **merge** this PR for me once it is approved. * chore(deps): update dependency pytest to v5.4.3 [(#4279)](GoogleCloudPlatform/python-docs-samples#4279) * chore(deps): update dependency pytest to v5.4.3 * specify pytest for python 2 in appengine Co-authored-by: Leah Cole <coleleah@google.com> * Update automl_tables_predict.py with batch_predict_bq sample [(#4142)](GoogleCloudPlatform/python-docs-samples#4142) Added a new method `batch_predict_bq` demonstrating running batch_prediction using BigQuery. Added notes in comments about asynchronicity for `batch_predict` method. The region `automl_tables_batch_predict_bq` will be used on cloud.google.com (currently both sections for GCS and BigQuery use the same sample code which is incorrect). Fixes#4141 Note: It's a good idea to open an issue first for discussion. - [x] Please **merge** this PR for me once it is approved. * Update dependency pytest to v6 [(#4390)](GoogleCloudPlatform/python-docs-samples#4390) * chore: exclude notebooks * chore: update templates * chore: add codeowners and fix tests * chore: ignore warnings from sphinx * chore: fix tables client * test: fix unit tests Co-authored-by: Torry Yang <sirtorry@users.noreply.github.com> Co-authored-by: florencep <florenceperot@google.com> Co-authored-by: Mike Burton <mb-github@niskala.org> Co-authored-by: Lars Wander <lwander@users.noreply.github.com> Co-authored-by: Michael Hu <Michael.an.hu@gmail.com> Co-authored-by: Michael Hu <michaelanhu@gmail.com> Co-authored-by: Alefh Sousa <alefh.sousa@gmail.com> Co-authored-by: DPEBot <dpebot@google.com> Co-authored-by: Kurtis Van Gent <31518063+kurtisvg@users.noreply.github.com> Co-authored-by: Doug Mahugh <dmahugh@gmail.com> Co-authored-by: WhiteSource Renovate <bot@renovateapp.com> Co-authored-by: Leah E. Cole <6719667+leahecole@users.noreply.github.com> Co-authored-by: Takashi Matsuo <tmatsuo@google.com> Co-authored-by: Anthony <wens.ajw@gmail.com> Co-authored-by: Amy <amy@infosleuth.net> Co-authored-by: Mike <45373284+munkhuushmgl@users.noreply.github.com> Co-authored-by: Leah Cole <coleleah@google.com> Co-authored-by: Sergei Dorogin <github@dorogin.com>

Fixes internal issue 418025765 🦕

daspecster added do not merge Indicates a pull request not ready for merge, due to either quality or timing. api: speech Issues related to the Speech-to-Text API. labels Oct 10, 2016

googlebot added the cla: yes This human has signed the Contributor License Agreement. label Oct 10, 2016

daspecster assigned tseaver and dhermes Oct 10, 2016

daspecster removed the do not merge Indicates a pull request not ready for merge, due to either quality or timing. label Oct 10, 2016

daspecster mentioned this pull request Oct 11, 2016
Tracking Speech API development. #2522
Closed
14 tasks

dhermes reviewed Oct 11, 2016
View reviewed changes

daspecster force-pushed the speech-streaming branch 4 times, most recently from ea0cb87 to 7acd247Compare October 13, 2016 20:42

daspecster mentioned this pull request Oct 14, 2016
Add gapic speech sync_recognize() #2546
Closed

dhermes reviewed Oct 17, 2016
View reviewed changes

daspecster added 5 commits October 19, 2016 11:23

Add Speech Streaming API.
bc31b96

Make API surface more usable.
0fef321

Alternate strategy,to just pass back responses.
a6cc2c7

Feedback updates.
0a04d5f

More verbose access to alternatives.
f6ff654

daspecster force-pushed the speech-streaming branch from 5844cad to f6ff654Compare October 19, 2016 15:23

dhermes reviewed Oct 20, 2016
View reviewed changes

dhermes reviewed Oct 21, 2016
View reviewed changes

Comment threadscripts/verify_included_modules.py
'google.cloud.pubsub.__init__',
'google.cloud.resource_manager.__init__',
'google.cloud.speech.__init__',
'google.cloud.speech.streaming.__init__',
This comment was marked as spam.
Sign in to view

Remote response calculating confidence and transcript.
4e536ab

daspecster closed this Nov 2, 2016

daspecster deleted the speech-streaming branch January 24, 2017 15:17

theacodes unassigned dhermes Sep 28, 2018

chalmerlowe pushed a commit that referenced this pull request Mar 31, 2026
refactor: use sqlglot to build create_external_table ddl (#2523)
5519f4c
Fixes internal issue 418025765 🦕

		True


		By setting ``interim_results`` to true, interim results (tentative hypotheses)

Uh oh!

Conversation

daspecster commented Oct 10, 2016• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dhermes commented Oct 11, 2016

Uh oh!

daspecster commented Oct 11, 2016

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

daspecster commented Oct 17, 2016

Uh oh!

dhermes commented Oct 17, 2016

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

This comment was marked as spam.

Uh oh!

daspecster commented Oct 19, 2016

Uh oh!

daspecster commented Oct 20, 2016

daspecster commented Oct 10, 2016•
edited
Loading