def changeAudio(file_path, sample):
access_token = '02UMSGJJclAaCjsV2JhZn2cVfRiFTo4NBlQtKycPT6k5IxaGNumOSU7a33Bc-uaGXjnXKbo8y1IaVjMHOchz6ua68Lb7U'
# Create client with your access token
client = apiclient.RevAiAPIClient(access_token)
file_job = client.submit_job_local_file(
filename=file_path,
metadata="This_is_some_job_metadata",
callback_url="",
skip_diarization=False)
job_id = file_job.id
while file_job.status == JobStatus.IN_PROGRESS:
file_job = client.get_job_details(job_id)
mylist = []
file = client.get_transcript_json(job_id)
end_time = 0.0
for i in file['monologues']:
for j in i['elements']:
if j['type'] == "text":
if j["value"] == sample:
new_tuple = (j["ts"], j["end_ts"])
mylist.append(new_tuple)
end_time = j["end_ts"]
print(end_time)
mergetwoaudio(mylist, end_time)
def get_transcript(audio_fpath):
client = apiclient.RevAiAPIClient(
"023nGm1xNAOR_e8BecAb-8UNWoHzpqh-xBOOgr85kiy1f3y95NYZLo_r634vB8GDBGhlPBhqX7Am-4Mu6-WUcRWtqSfU4"
)
job = client.submit_job_local_file(os.path.realpath(audio_fpath))
job_id = client.get_job_details(job.id)
time.sleep(2)
attempts = 20
for i in range(attempts):
try:
transcript_text = client.get_transcript_text(job.id)
return transcript_text
except:
time.sleep(1)
else:
return "Operation timed out"
def upload(request):
if request.method == 'GET': ## For default view
return render(request, 'upload.html')
elif request.method == 'POST':
uploadedFile = request.FILES['file']
fileName = "static/videos/" + uploadedFile.name
# Write File mp4
with open(fileName, "wb") as f:
for chunk in uploadedFile.chunks():
f.write(chunk)
# Convert to mp3
mp3fileName = fileName[:-4] + ".mp3"
clip = mp.VideoFileClip(fileName).subclip(0)
mp3fileNameNoFolder = uploadedFile.name[:-4] + ".mp3"
clip.audio.write_audiofile(mp3fileName)
# rev_ai request
# Create client with your access token
client = apiclient.RevAiAPIClient(settings.REV_ACCESS_TOKEN)
callback_url = "http://35.239.24.77/vidhub/apis/revai/"
# Submit from local file
file_job = client.submit_job_local_file(
filename=mp3fileName,
callback_url=callback_url,
metadata="This_is_some_job_metadata",
skip_diarization=False)
video_url = "http://35.239.24.77/static/videos/" + uploadedFile.name
video = Video.objects.create(title=request.POST['title'],
author=request.POST['author'],
video_url=video_url,
rev_id=file_job.id,
subtitle_url='')
return render(request, 'uploaded.html', {'id': video.id})
else:
return HttpResponse("WTF")
def transcriber(recording_path):
client = apiclient.RevAiAPIClient(accesscode)
# you can send a local file
job = client.submit_job_local_file(recording_path)
while True:
# Obtains details of a job in json format
job_details = client.get_job_details(job.id)
status = job_details.status.name
print("Job Status : {}".format(status))
# Checks if the job has been transcribed
if status == "IN_PROGRESS":
time.sleep(5)
continue
elif status == "FAILED":
print("Job Failed : {}".format(job_details.failure_detail))
break
if status == "TRANSCRIBED":
# Getting a list of current jobs connected with your account
# The optional parameters limits limits the length of the list. Starting_after
# Cuts off the beginning x jobs off the list returned
list_of_jobs = client.get_list_of_jobs(limit=None,
starting_after=None)
# obtain transcript text as a string for the job.
transcript_text = client.get_transcript_text(job.id)
print(transcript_text)
# Write to txt file
f = open("transcript.txt", "w+")
f.write(transcript_text)
f.close()
break
def revai_callback(request):
json_result = (request.body).decode("utf-8")
print(json_result)
job = json.loads(json_result)['job']
print(job)
video = Video.objects.get(rev_id=job['id'])
if job['status'] == 'transcribed':
# Create client with your access token
client = apiclient.RevAiAPIClient(settings.REV_ACCESS_TOKEN)
fileName = "static/videos/" + job['name'][:-4] + ".vtt"
caption = client.get_captions(job['id'], CaptionType.VTT)
with open(fileName, "w") as f:
f.write(caption)
# TRanslate caption
content_list = caption.splitlines()
print(content_list)
client = translate.Client(target_language="th") ## client to Google
for i in range(4, len(content_list), 4):
print("Send:", content_list[i])
returnDict = client.translate(content_list[i])
print("Return:", returnDict)
returnString = returnDict['translatedText']
content_list[i] = returnString
thaifileName = "static/videos/" + job['name'][:-4] + ".vtt.th"
with open(thaifileName, "w") as f:
for line in content_list:
f.write(line + "\n")
video.subtitle_url = 'http://35.239.24.77/' + fileName
video.save()
else:
video.subtitle_url = 'failed'
video.save()
return HttpResponse("Callback received.")
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
"""
import time
from rev_ai import apiclient
# String containing your access token
access_token = "your_access_token"
# Create your api client
client = apiclient.RevAiAPIClient(access_token)
# Submitting a job through a local file. The optional parameters are shown below.
# job = client.submit_job_local_file("your_local_file_path",
# metadata=None,
# callback_url=None,
# skip_diarization=False,
# custom_vocabularies=None)
# Submitting a job with a link to the file you want transcribed
# Change url to your url
url = "https://www.rev.ai/FTC_Sample_1.mp3"
job = client.submit_job_url(url,
metadata=None,
callback_url=None,
skip_diarization=False,
from rev_ai import apiclient
import json
import time
import os
import datetime
client = apiclient.RevAiAPIClient(
"02-SktQcq_WILkfcQ4ZFPC0A7B0jy57Mdef1Y1dv6I7dB-C_BQGduPIR03QxNomQVjX5nOt0uap-Zn35qDpDUcjwPN1_k")
def convertTime(t):
return str(datetime.timedelta(seconds=t))
def S2T(audioFile):
print("Sending ", audioFile, " for speech2text")
job = client.submit_job_local_file(audioFile)
id = job.id
# id = "NIva4qKmLUt1"
# id = "5xprc0oRG1id"
# job_details = client.get_job_details(id)
while True:
time.sleep(2)
print("Trying to fetch speech2text output")
job_details = client.get_job_details(id)
if str(job_details.status) == "JobStatus.TRANSCRIBED":
print("Ready, reading json, txt, Txt files")
speech = client.get_transcript_text(id)
def __init__(self):
# create your client
self.client = apiclient.RevAiAPIClient(ACCESS_TOKEN)
def getname(list, dict):
for i in list:
if i in dict:
print (i + ":" + dict[i][0])
#image = Image.open(dict[i][1])
#image = image.resize((240, 320))
#disp.display(image)
Image.open(dict[i][1]).show()
elif i[0:len(i)-1] in dict:
print (i[0:len(i)-1] + ":" + dict[i[0:len(i)-1]][0])
Image.open(dict[i[0:len(i)-1]][1]).show()
#Access Token: 02a5E3uxq-KgH_z07nuWDriS-CI8mDsK5whI1DFcel6r5jfjOeSyKSKmjhXgsy7oEYezKicH5VUZW8S1PyhqAj9Fqn4TE
# create your client
client = apiclient.RevAiAPIClient("02a5E3uxq-KgH_z07nuWDriS-CI8mDsK5whI1DFcel6r5jfjOeSyKSKmjhXgsy7oEYezKicH5VUZW8S1PyhqAj9Fqn4TE")
def transcribe(location):
job = client.submit_job_local_file(location)
job_details = client.get_job_details(job.id)
status = job_details.status.name
i = 0
while status != "TRANSCRIBED":
job_details = client.get_job_details(job.id)
status = job_details.status.name
print(status, " Time to complete:",i,"s")
i+=1
time.sleep(1)
'--file',
required=True,
help='The name of the .mp3 file to convert to text')
return parser.parse_args().file
if __name__ == "__main__":
# Get the desired audio file from cmd line
filename = get_file_name()
if '.mp3' not in filename:
print("Error: file type is not mp3")
sys.exit()
# Create client with given access token
client = apiclient.RevAiAPIClient(
"028f4jLg6e4JCROQXE8vVZbkDlf_5oSfzH-HWyn1ehu6yOiTrVKtwtvLYnVzZZTAayHlalZFqD0zsKajkhYKjO7p8L37k"
)
# Create a job to translate a given mp3 file
job = client.submit_job_local_file(filename)
# Check if the job is done transcribing
while True:
# Obtains details of a job in json format
job_details = client.get_job_details(job.id)
status = job_details.status.name
print("Job Status : {}".format(status))
# Checks if the job has been transcribed
if status == "IN_PROGRESS":
from rev_ai import apiclient
import sounddevice as sd # python -m pip install sounddevice --user
from scipy.io.wavfile import write # pip install scipy
from pydub import AudioSegment # convert WAV to mp3
rev_ai_token = "02g01w-kZhbjrtz5qXsiT71gchuHv7SyFFqnRTOAjYYTrJlxV7KA9DkYehOZofxVvOZl_qfQD10LlsiMsu6G--WT7ai2I"
client = apiclient.RevAiAPIClient(rev_ai_token)
SentithinkAPI = "[[AZURE WEB API ENDPOINT]]"
Sentithink_auth = "code=[[AUTH TOKEN]]"
x = float(input("Enter microphone X location: "))
y = float(input("Enter microphone Y location: "))
name = input("Enter microphone name: ")
# Capture the sound snippet and save it to file here
frameRate = 44100 # Sample rate
seconds = 10 # Duration of recording
myrecording = sd.rec(int(seconds * frameRate),
samplerate=frameRate,
channels=2)
sd.wait() # Wait until recording is finished
write('voiceRecord.wav', frameRate, myrecording) # Save as WAV file
convertedAudio = AudioSegment.from_wav('voiceRecord.wav')
convertedAudio.export('voiceRecord.mp3', format='mp3')
rev_ai_callback_url = SentithinkAPI + Sentithink_auth + "&X=" + str(
x) + "&Y=" + str(y) + "&name=" + name
job = client.submit_job_local_file(
from rev_ai import apiclient
# create your client
client = apiclient.RevAiAPIClient("02kiXlheeVr0xkYA-oZTevxPeuEWjwVGtyogDlUHyfqmIgFMtR4eX8etW2qyBGTyAoBfjrY6uPBIgZmLuyA4F0nf_g5Hc")
# you can send a local file
job = client.submit_job_local_file("recording.wav")
job_details = client.get_job_details(job.id)
print(repr(job_details), job.id)
# transcript_text = client.get_transcript_text(job.id)
# print(transcript_text)
# client.delete_job(job.id)
def getRevAiAPIClient(): configFileContents = getConfigFileContent() ACCESS_TOKEN = configFileContents["rev-ai-access-token"] return api.RevAiAPIClient(ACCESS_TOKEN)
from rev_ai import apiclient, JobStatus
API_TOKEN = 'SET_ACCESS_TOKEN_HERE'
REV_CLIENT = apiclient.RevAiAPIClient(API_TOKEN)
from glob import glob
from tqdm import tqdm
from time import sleep
import sys, os
from os.path import basename, isfile
def schedule_jobs(audio_files):
fid2jid = {}
for audio_file in tqdm(audio_files, desc='Scheduling jobs'):
file_id = basename(audio_file).replace('.wav', '')
job = REV_CLIENT.submit_job_local_file(audio_file)
if job.status == JobStatus.FAILED:
print('Job Schedule failed for:', audio_file)
continue
fid2jid[file_id] = job.id
return fid2jid
def libri_transcribe(audio_folder, output_folder, samples_per_batch=100):
os.makedirs(output_folder, exist_ok=True)
audio_files = sorted(glob(os.path.join(audio_folder, '*.wav')))
total_batches = (len(audio_files) + samples_per_batch -
1) / samples_per_batch
i = 0
while i < len(audio_files):
from rev_ai import apiclient
import time
from apihelper import *
# create your client
client = apiclient.RevAiAPIClient(key)
job = client.submit_job_local_file("./recording.m4a")
jobs = client.get_list_of_jobs()
time.sleep(35)
transcript_text = client.get_transcript_text(job.id)
print(transcript_text)
