Merge pull request #811 from ftnext/tweak/google-application-credentials

feat(google-cloud-speech): Delegate to google-auth

README.rst

@@ -155,9 +155,13 @@ Google Cloud Speech Library for Python (for Google Cloud Speech-to-Text API user
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 The library `google-cloud-speech <https://pypi.org/project/google-cloud-speech/>`__ is **required if and only if you want to use Google Cloud Speech-to-Text API** (``recognizer_instance.recognize_google_cloud``).
-
 You can install it with :command:`python3 -m pip install SpeechRecognition[google-cloud]`.
-(ref: `official installation instructions <https://cloud.google.com/speech-to-text/docs/transcribe-client-libraries#install_the_client_library>`__)
+(ref: `official installation instructions <https://cloud.google.com/speech-to-text/docs/transcribe-client-libraries#client-libraries-install-python>`__)
+
+**Prerequisite**: Create local authentication credentials for your Google account
+
+* `Before you begin (Transcribe speech to text by using client libraries) <https://cloud.google.com/speech-to-text/docs/transcribe-client-libraries#before-you-begin>`__
+* Detail: `User credentials (Set up ADC for a local development environment) <https://cloud.google.com/docs/authentication/set-up-adc-local-dev-environment#local-user-cred>`__
 
 Currently only `V1 <https://cloud.google.com/speech-to-text/docs/quickstart>`__ is supported. (`V2 <https://cloud.google.com/speech-to-text/v2/docs/quickstart>`__ is not supported)
 

speech_recognition/recognizers/google_cloud.py

@@ -1,6 +1,5 @@
 from __future__ import annotations
 
-import os
 from urllib.error import URLError
 
 from speech_recognition.audio import AudioData
@@ -42,8 +41,6 @@ def recognize(
     assert isinstance(
         audio_data, AudioData
     ), "``audio_data`` must be audio data"
-    if credentials_json_path is None:
-        assert os.environ.get("GOOGLE_APPLICATION_CREDENTIALS") is not None
     assert isinstance(language, str), "``language`` must be a string"
     assert preferred_phrases is None or all(
         isinstance(preferred_phrases, (type(""), type("")))

tests/recognizers/test_google_cloud.py

@@ -15,11 +15,7 @@
 
 
 @patch("google.cloud.speech.SpeechClient")
-def test_transcribe_with_google_cloud_speech(SpeechClient, monkeypatch):
-    monkeypatch.setenv(
-        "GOOGLE_APPLICATION_CREDENTIALS", "path/to/credentials.json"
-    )
-
+def test_transcribe_with_google_cloud_speech(SpeechClient):
     client = SpeechClient.return_value
     # ref: https://cloud.google.com/speech-to-text/docs/transcribe-gcloud?hl=ja#make_an_audio_transcription_request
     client.recognize.return_value = RecognizeResponse(
@@ -84,11 +80,7 @@ def test_transcribe_with_specified_credentials(SpeechClient):
 
 
 @patch("google.cloud.speech.SpeechClient")
-def test_transcribe_show_all(SpeechClient, monkeypatch):
-    monkeypatch.setenv(
-        "GOOGLE_APPLICATION_CREDENTIALS", "path/to/credentials.json"
-    )
-
+def test_transcribe_show_all(SpeechClient):
     client = SpeechClient.return_value
     client.recognize.return_value = RecognizeResponse(
         results=[
@@ -151,11 +143,7 @@ def test_transcribe_show_all(SpeechClient, monkeypatch):
 
 
 @patch("google.cloud.speech.SpeechClient")
-def test_transcribe_with_specified_api_parameters(SpeechClient, monkeypatch):
-    monkeypatch.setenv(
-        "GOOGLE_APPLICATION_CREDENTIALS", "path/to/credentials.json"
-    )
-
+def test_transcribe_with_specified_api_parameters(SpeechClient):
     client = SpeechClient.return_value
     client.recognize.return_value = RecognizeResponse(
         results=[