def test_transcribe_gcs_word_time_offsets(capsys):
transcribe_async.transcribe_gcs(
'gs://python-docs-samples-tests/speech/audio.flac')
out, err = capsys.readouterr()
match = re.search(r'Bridge, start_time: ([0-9.]+)', out, re.DOTALL | re.I)
time = float(match.group(1))
assert time > 0
def edit(filename):
audio_urls = []
files = crud.blobs_list(bucket)
audio_files = [x for x, y in files if (y == '.wav' and filename in x)]
for af in audio_files:
audio_filename = "{}.wav".format(af)
audio_urls.append("https://storage.cloud.google.com/{}/{}".format(
bucket, audio_filename))
text_filename = "{}.json".format(filename)
text_url = "https://storage.cloud.google.com/{}/{}".format(
bucket, text_filename)
print(audio_urls)
response = ""
if crud.blob_exists(bucket, text_filename):
item = crud.blob_info(bucket, text_filename)
response = json.loads(item.download_as_string())
if request.method == 'POST':
if request.form['button'] == 'save':
response = request.form['text']
with open('copy.json', 'w', encoding='utf-8') as f:
json.dump(response, f, ensure_ascii=False)
metadata = {'status': 0} # set status to in progress
crud.upload_blob(bucket, "copy.json", text_filename, metadata)
if request.form['button'] == 'transcribe':
transcript = transcribe.transcribe_gcs("gs://{}/{}".format(
bucket, audio_filename))
response = " ".join(transcript)
if request.form['button'] == 'done':
response = request.form['text']
blob = crud.blob_info(response)
blob.metadata = {'status': 1} # set status to in progress
blob.patch()
return render_template('editor.html',
audio_urls=audio_urls,
response=response)
description=__doc__,
formatter_class=argparse.RawDescriptionHelpFormatter)
parser.add_argument(
'lecture_name', help='Name of the lecture')
parser.add_argument(
'path', help='File or GCS path for audio file to be recognized')
args = parser.parse_args()
transcript_filename = args.lecture_name + ".txt"
content_filename = args.lecture_name + "_content.txt"
subtitle_filename = args.lecture_name + ".vtt"
summary_filename = args.lecture_name + "_summary.txt"
# generate transcript using google api
sys.stdout = open(transcript_filename, 'w+')
content = transcribe_async.transcribe_gcs(args.path)
open(content_filename, 'w+').write(content)
# generate vtt file from transcript
sys.stdout = original_stdout
if not os.path.exists(transcript_filename):
print("transcript file does not exist.\n")
exit(-1)
vtt_gen.generate_subtitles(transcript_filename, subtitle_filename)
# generate summary
# textsum.generate_summary(content_filename, summary_filename)
# entity analysis
entity_analysis.entities_text(content)
def test_transcribe_gcs(capsys):
transcribe_async.transcribe_gcs(
'gs://python-docs-samples-tests/speech/audio.flac')
out, err = capsys.readouterr()
assert re.search(r'how old is the Brooklyn Bridge', out, re.DOTALL | re.I)
def test_transcribe_gcs(resource, capsys):
transcribe_async.transcribe_gcs(
'gs://python-docs-samples-tests/speech/audio.flac')
out, err = capsys.readouterr()
assert re.search(r'how old is the Brooklyn Bridge', out, re.DOTALL | re.I)