cloudlanguagetools 11.3.4__tar.gz → 11.4.0__tar.gz

{cloudlanguagetools-11.3.4 → cloudlanguagetools-11.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: cloudlanguagetools
-Version: 11.3.4
+Version: 11.4.0
 Summary: Interface with various cloud APIs for language processing such as translation, text to speech
 Home-page: https://github.com/Language-Tools/cloud-language-tools-core
 Author: Luc

{cloudlanguagetools-11.3.4 → cloudlanguagetools-11.4.0}/cloudlanguagetools/amazon.py

@@ -6,6 +6,7 @@ import boto3
 import botocore.exceptions
 import contextlib
 
+
 import cloudlanguagetools.service
 import cloudlanguagetools.constants
 import cloudlanguagetools.options
@@ -14,6 +15,9 @@ import cloudlanguagetools.ttsvoice
 import cloudlanguagetools.translationlanguage
 import cloudlanguagetools.transliterationlanguage
 import cloudlanguagetools.errors
+import cloudlanguagetools.audio_processing
+
+from cloudlanguagetools.options import AudioFormat
 
 DEFAULT_VOICE_PITCH = 0
 DEFAULT_VOICE_RATE = 100
@@ -73,6 +77,7 @@ class AmazonVoice(cloudlanguagetools.ttsvoice.TtsVoice):
                 'values': [
                     cloudlanguagetools.options.AudioFormat.mp3.name,
                     cloudlanguagetools.options.AudioFormat.ogg_vorbis.name,
+                    cloudlanguagetools.options.AudioFormat.wav.name
                 ],
                 'default': cloudlanguagetools.options.AudioFormat.mp3.name
             }            
@@ -114,13 +119,14 @@ class AmazonService(cloudlanguagetools.service.Service):
         return result.get('TranslatedText')
 
     def get_tts_audio(self, text, voice_key, options):
-        audio_format_str = options.get(cloudlanguagetools.options.AUDIO_FORMAT_PARAMETER, cloudlanguagetools.options.AudioFormat.mp3.name)
-        audio_format = cloudlanguagetools.options.AudioFormat[audio_format_str]
+        response_format_parameter, audio_format = self.get_request_audio_format({
+            AudioFormat.mp3: 'mp3',
+            AudioFormat.ogg_vorbis: 'ogg_vorbis',
+            AudioFormat.wav: 'pcm'
+        }, options, AudioFormat.mp3)
 
-        audio_format_map = {
-            cloudlanguagetools.options.AudioFormat.mp3: 'mp3',
-            cloudlanguagetools.options.AudioFormat.ogg_vorbis: 'ogg_vorbis'
-        }
+        # wav, we need to convert as described here:
+        # https://aws.amazon.com/blogs/machine-learning/integrating-amazon-polly-with-legacy-ivr-systems-by-converting-output-to-wav-format/
 
         output_temp_file = tempfile.NamedTemporaryFile()
         output_temp_filename = output_temp_file.name
@@ -143,7 +149,15 @@ class AmazonService(cloudlanguagetools.service.Service):
 </speak>"""
 
         try:
-            response = self.polly_client.synthesize_speech(Text=ssml_str, TextType="ssml", OutputFormat=audio_format_map[audio_format], VoiceId=voice_key['voice_id'], Engine=voice_key['engine'])
+            if audio_format == cloudlanguagetools.options.AudioFormat.wav:
+                response = self.polly_client.synthesize_speech(Text=ssml_str, 
+                    TextType="ssml", 
+                    OutputFormat=response_format_parameter,
+                    VoiceId=voice_key['voice_id'], 
+                    Engine=voice_key['engine'],
+                    SampleRate="16000")
+            else:
+                response = self.polly_client.synthesize_speech(Text=ssml_str, TextType="ssml", OutputFormat=response_format_parameter, VoiceId=voice_key['voice_id'], Engine=voice_key['engine'])
         except (botocore.exceptions.BotoCoreError, botocore.exceptions.ClientError) as error:
             raise cloudlanguagetools.errors.RequestError(str(error))
 
@@ -155,6 +169,12 @@ class AmazonService(cloudlanguagetools.service.Service):
             with contextlib.closing(response["AudioStream"]) as stream:
                 with open(output_temp_filename, 'wb') as audio:
                     audio.write(stream.read())
+
+                if audio_format == cloudlanguagetools.options.AudioFormat.wav:
+                    return cloudlanguagetools.audio_processing.wrap_pcm_data_wave(output_temp_file, 
+                        num_channels=1, 
+                        sample_width=2, 
+                        framerate=16000)
                 return output_temp_file
 
         else:

cloudlanguagetools-11.4.0/cloudlanguagetools/audio_processing.py

@@ -0,0 +1,29 @@
+import wave
+import tempfile
+
+# audio utilities
+
+
+# take PCM audio and wrap it in a WAV container
+def wrap_pcm_data_wave(audio_temp_file: tempfile.NamedTemporaryFile,
+        num_channels,
+        sample_width,
+        framerate) -> tempfile.NamedTemporaryFile:
+    # read the audio data
+    f = open(audio_temp_file.name, 'rb')
+    data = f.read()
+    f.close()
+
+    wav_frames = []
+    wav_frames.append(data)
+
+    wav_audio_temp_file = tempfile.NamedTemporaryFile(prefix='clt_wav_audio_', suffix='.wav')
+
+    WAVEFORMAT = wave.open(wav_audio_temp_file.name,'wb')
+    WAVEFORMAT.setnchannels(num_channels) # one channel, mono
+    WAVEFORMAT.setsampwidth(sample_width) # Polly's output is a stream of 16-bits (2 bytes) samples
+    WAVEFORMAT.setframerate(framerate)
+    WAVEFORMAT.writeframes(b''.join(wav_frames))
+    WAVEFORMAT.close()    
+    
+    return wav_audio_temp_file

{cloudlanguagetools-11.3.4 → cloudlanguagetools-11.4.0}/cloudlanguagetools/azure.py

@@ -6,6 +6,8 @@ import operator
 import pydub
 import logging
 import pprint
+import time
+import cachetools
 from typing import List
 
 import cloudlanguagetools.service
@@ -17,6 +19,7 @@ import cloudlanguagetools.translationlanguage
 import cloudlanguagetools.transliterationlanguage
 import cloudlanguagetools.dictionarylookup
 import cloudlanguagetools.errors
+from cloudlanguagetools.options import AudioFormat
 
 
 import azure.cognitiveservices.speech
@@ -52,6 +55,7 @@ VOICE_OPTIONS = {
                 'values': [
                     cloudlanguagetools.options.AudioFormat.mp3.name,
                     cloudlanguagetools.options.AudioFormat.ogg_opus.name,
+                    cloudlanguagetools.options.AudioFormat.wav.name,
                 ],
                 'default': cloudlanguagetools.options.AudioFormat.mp3.name
             }
@@ -258,22 +262,18 @@ class AzureService(cloudlanguagetools.service.Service):
         return headers        
 
     def get_tts_audio(self, text, voice_key, options):
-
-        audio_format_str = options.get(cloudlanguagetools.options.AUDIO_FORMAT_PARAMETER, cloudlanguagetools.options.AudioFormat.mp3.name)
-        audio_format = cloudlanguagetools.options.AudioFormat[audio_format_str]
-
         # https://learn.microsoft.com/en-us/azure/ai-services/speech-service/rest-text-to-speech?tabs=streaming#audio-outputs
         # https://learn.microsoft.com/en-us/python/api/azure-cognitiveservices-speech/azure.cognitiveservices.speech.speechsynthesisoutputformat?view=azure-python
-        audio_format_map = {
-            cloudlanguagetools.options.AudioFormat.mp3: 'Audio24Khz96KBitRateMonoMp3',
-            cloudlanguagetools.options.AudioFormat.ogg_opus: 'Ogg48Khz16BitMonoOpus',
-            cloudlanguagetools.options.AudioFormat.wav: 'Riff48Khz16BitMonoPcm',
-        }
+        response_format_parameter, audio_format = self.get_request_audio_format({
+            AudioFormat.mp3: 'Audio24Khz96KBitRateMonoMp3',
+            AudioFormat.ogg_opus: 'Ogg48Khz16BitMonoOpus',
+            AudioFormat.wav: 'Riff48Khz16BitMonoPcm'
+        }, options, AudioFormat.mp3)
 
         output_temp_file = tempfile.NamedTemporaryFile(prefix=f'cloudlanguage_tools_{self.__class__.__name__}_audio', suffix=f'.{audio_format.name}')
         output_temp_filename = output_temp_file.name
         speech_config = azure.cognitiveservices.speech.SpeechConfig(subscription=self.key, region=self.region)
-        speech_config.set_speech_synthesis_output_format(azure.cognitiveservices.speech.SpeechSynthesisOutputFormat[audio_format_map[audio_format]])
+        speech_config.set_speech_synthesis_output_format(azure.cognitiveservices.speech.SpeechSynthesisOutputFormat[response_format_parameter])
         synthesizer = azure.cognitiveservices.speech.SpeechSynthesizer(speech_config=speech_config, audio_config=None)
 
         default_pitch = 0
@@ -330,17 +330,18 @@ class AzureService(cloudlanguagetools.service.Service):
         headers = {
             'Authorization': 'Bearer ' + token,
         }        
-        response = requests.get(constructed_url, headers=headers)
-        if response.status_code == 200:
-            voice_list = json.loads(response.content)
-            result = []
-            for voice_data in voice_list:
-                # print(voice_data['Status'])
-                try:
-                    result.append(AzureVoice(voice_data))
-                except KeyError:
-                    logging.error(f'could not process voice for {voice_data}', exc_info=True)
-            return result
+        response = requests.get(constructed_url, headers=headers, 
+            timeout=cloudlanguagetools.constants.RequestTimeout)
+        response.raise_for_status()
+        voice_list = json.loads(response.content)
+        result = []
+        for voice_data in voice_list:
+            # print(voice_data['Status'])
+            try:
+                result.append(AzureVoice(voice_data))
+            except KeyError:
+                logging.error(f'could not process voice for {voice_data}', exc_info=True)
+        return result
 
     def get_tts_voice_list_v3(self) -> List[cloudlanguagetools.ttsvoice.TtsVoice_v3]:
         # returns list of TtsVoice_v3
@@ -353,17 +354,18 @@ class AzureService(cloudlanguagetools.service.Service):
         headers = {
             'Authorization': 'Bearer ' + token,
         }        
-        response = requests.get(constructed_url, headers=headers)
-        if response.status_code == 200:
-            voice_list = json.loads(response.content)
-            result = []
-            for voice_data in voice_list:
-                # print(voice_data['Status'])
-                try:
-                    result.append(build_tts_voice_v3(voice_data))
-                except:
-                    logger.exception(f'could not process voice for {voice_data}')
-            return result            
+        response = requests.get(constructed_url, headers=headers, 
+            timeout=cloudlanguagetools.constants.RequestTimeout)
+        response.raise_for_status()
+        voice_list = json.loads(response.content)
+        result = []
+        for voice_data in voice_list:
+            # print(voice_data['Status'])
+            try:
+                result.append(build_tts_voice_v3(voice_data))
+            except:
+                logger.exception(f'could not process voice for {voice_data}')
+        return result            
 
     def get_translation(self, text, from_language_key, to_language_key):
         base_url = f'{self.url_translator_base}/translate?api-version=3.0'
@@ -415,7 +417,10 @@ class AzureService(cloudlanguagetools.service.Service):
         return result
 
 
+    @cachetools.cached(cache=cachetools.TTLCache(maxsize=1024, ttl=cloudlanguagetools.constants.TTLCacheTimeout))
     def get_supported_languages(self):
+        max_retries = 3
+        retry_delay = 5  # seconds
         url = 'https://api.cognitive.microsofttranslator.com/languages?api-version=3.0'
 
         # If you encounter any issues with the base_url or path, make sure
@@ -427,10 +432,20 @@ class AzureService(cloudlanguagetools.service.Service):
             'X-ClientTraceId': str(uuid.uuid4())
         }
 
-        request = requests.get(url, headers=headers)
-        response = request.json()
-
-        return response
+        for attempt in range(max_retries):
+            try:
+                request = requests.get(url, headers=headers, 
+                    timeout=cloudlanguagetools.constants.RequestTimeoutLong)
+                request.raise_for_status()
+                response = request.json()
+                return response
+            except requests.exceptions.RequestException as e:
+                if attempt < max_retries - 1:  # If not the last attempt
+                    logger.warning(f"Request failed. Retrying in {retry_delay} seconds... (Attempt {attempt + 1}/{max_retries})")
+                    time.sleep(retry_delay)
+                else:
+                    logger.error(f"Max retries reached. Unable to get supported languages.")
+                    raise  # Re-raise the last exception if all retries failed
 
         # print(json.dumps(response, sort_keys=True, indent=4, ensure_ascii=False, separators=(',', ': ')))        
 
@@ -475,6 +490,8 @@ class AzureService(cloudlanguagetools.service.Service):
             sound = pydub.AudioSegment.from_mp3(mp3_filepath)
         elif audio_format in [cloudlanguagetools.options.AudioFormat.ogg_opus, cloudlanguagetools.options.AudioFormat.ogg_vorbis]:
             sound = pydub.AudioSegment.from_ogg(mp3_filepath)
+        elif audio_format == cloudlanguagetools.options.AudioFormat.wav:
+            sound = pydub.AudioSegment.from_wav(mp3_filepath)
         wav_filepath = tempfile.NamedTemporaryFile(suffix='.wav').name
         sound.export(wav_filepath, format="wav")
 
@@ -647,4 +664,4 @@ class AzureService(cloudlanguagetools.service.Service):
 
         # return response[0]['translations'][0]['text']    
 
-    
+    

{cloudlanguagetools-11.3.4 → cloudlanguagetools-11.4.0}/cloudlanguagetools/constants.py

@@ -55,6 +55,8 @@ class APIVersion(enum.Enum):
 # ======================================
 
 RequestTimeout = 10 # 10 seconds max
+# need to change timeout for long requests, such as retrieving list of services
+RequestTimeoutLong = 20 # 10 seconds max
 ReadTimeout = 3 # 3 seconds read timeout
 
 TTLCacheTimeout = 86400 # 24 hours

{cloudlanguagetools-11.3.4 → cloudlanguagetools-11.4.0}/cloudlanguagetools/elevenlabs.py

@@ -5,6 +5,7 @@ import tempfile
 import os
 import contextlib
 import logging
+import urllib.parse
 from typing import List
 
 import cloudlanguagetools.service
@@ -15,6 +16,7 @@ import cloudlanguagetools.ttsvoice
 import cloudlanguagetools.translationlanguage
 import cloudlanguagetools.transliterationlanguage
 import cloudlanguagetools.errors
+from cloudlanguagetools.options import AudioFormat
 
 logger = logging.getLogger(__name__)
 
@@ -39,7 +41,15 @@ VOICE_OPTIONS = {
                 'min': 0.0,
                 'max': 1.0,
                 'default': DEFAULT_SIMILARITY_BOOST
-            },                        
+            },
+            cloudlanguagetools.options.AUDIO_FORMAT_PARAMETER: {
+                'type': cloudlanguagetools.options.ParameterType.list.name,
+                'values': [
+                    cloudlanguagetools.options.AudioFormat.mp3.name,
+                    cloudlanguagetools.options.AudioFormat.wav.name
+                ],
+                'default': cloudlanguagetools.options.AudioFormat.mp3.name
+            }
 }
 
 class ElevenLabsVoice(cloudlanguagetools.ttsvoice.TtsVoice):
@@ -84,6 +94,11 @@ class ElevenLabsService(cloudlanguagetools.service.Service):
         voice_id = voice_key['voice_id']
         url = f'https://api.elevenlabs.io/v1/text-to-speech/{voice_id}'
 
+        response_format_parameter, audio_format = self.get_request_audio_format({
+            AudioFormat.mp3: 'mp3_44100_128',
+            AudioFormat.wav: 'pcm_44100'
+        }, options, AudioFormat.mp3)
+
         headers = self.get_headers()
         headers['Accept'] = "audio/mpeg"
 
@@ -96,7 +111,17 @@ class ElevenLabsService(cloudlanguagetools.service.Service):
             }
         }
 
-        return self.get_tts_audio_base_post_request(url, json=data, headers=headers)
+        query_params = {
+            'output_format': response_format_parameter
+        }
+        full_url = f'{url}?{urllib.parse.urlencode(query_params)}'
+
+        if audio_format == cloudlanguagetools.options.AudioFormat.wav:
+            return cloudlanguagetools.audio_processing.wrap_pcm_data_wave(self.get_tts_audio_base_post_request(full_url, json=data, headers=headers), 
+                num_channels=1,
+                sample_width=2, 
+                framerate=44100) # pcm_44100 - PCM format (S16LE) with 44.1kHz sample rate. 
+        return self.get_tts_audio_base_post_request(full_url, json=data, headers=headers)
 
 
 

{cloudlanguagetools-11.3.4 → cloudlanguagetools-11.4.0}/cloudlanguagetools/google.py

@@ -74,6 +74,7 @@ class GoogleVoice(cloudlanguagetools.ttsvoice.TtsVoice):
                 'values': [
                     cloudlanguagetools.options.AudioFormat.mp3.name,
                     cloudlanguagetools.options.AudioFormat.ogg_opus.name,
+                    cloudlanguagetools.options.AudioFormat.wav.name
                 ],
                 'default': cloudlanguagetools.options.AudioFormat.mp3.name
             }            
@@ -146,7 +147,8 @@ class GoogleService(cloudlanguagetools.service.Service):
 
         audio_format_map = {
             cloudlanguagetools.options.AudioFormat.mp3: google.cloud.texttospeech.AudioEncoding.MP3,
-            cloudlanguagetools.options.AudioFormat.ogg_opus: google.cloud.texttospeech.AudioEncoding.OGG_OPUS
+            cloudlanguagetools.options.AudioFormat.ogg_opus: google.cloud.texttospeech.AudioEncoding.OGG_OPUS,
+            cloudlanguagetools.options.AudioFormat.wav: google.cloud.texttospeech.AudioEncoding.LINEAR16
         }
 
         try:

cloudlanguagetools-11.4.0/cloudlanguagetools/keys.py

@@ -0,0 +1 @@
+KEYS='gAAAAABnCcYzERlg56peIVaJoWPgT7m06v9-gZ-xuDco6PWxZEu-BRxgHR9UaqN3g8HzKcfVSS-j1rBf-x0SHZlFQwcRlIQIY1S9mE3Wm8I6_xCGkFfqeuLkc8c9ScDIFnMt5MtLevVchJWklkl86HH6qoUZyzmzpQm5ihwdShaCwqryQZ95UDzuFis0rPDLX63gvmdWvXSiYRjXb4hO53ajE380bmjjp9E9mbH01Y6aKAz1hBpGWjEmepnajkkopM3uw-mCz0ABq9Ysdhs0yd9UklwvbOgUgdKL1BC_st0xt2piEShHtu0lQL5LOB1ZXBYHzFhk8wLlpftvDhYcFXapoZYBlg6d3tTKoOonPumN78CSIjQdxvTcG6mcA0p1VlAurcN4-YgtbDvtObsN5clpmDXXPupwMwf7Wddmut4sGGvMJXEba4EP5G6R_3Se8F6GRfawOgLVBAJ-SSpdKszJBu9w-BPIOzXkZ0D_ddQwC1CKURM4Kpjv0phHbqulykHgHe1l8mDr8ycC87647jowIz55An2ZttpRo9mdJztPGeuEjoz5RvQHJQvjmhAT7vajWnXv7sB2y50qUnWTF5ip1rde3GfDgmNzrOd7jSe8dxonAtFziKBCiCqc3pFfh7YgeHzWCR7lJcF7wIZvGJ-goLFiUT9uayXXVAVbPFku6et1DnUpDU48nyrO7lUvmwqRprTM706UWbbdZk8dBn7n91KzzjR4AbQhQ3hl6QuvA4WoJYdpzAaUN_4LUiWHNFWtd3uBLjHqTBLOHUlAETkkPkssm48pBqqALSJBWOrNSBIcLhAQI-JcJoer_fgH4KG4pDaeKgmGF7X26xi4rE48GsYwd7Tf2kRMQrHfyLefYDocMTCqH2I7NHs4i66LDvj_Zn9nmkSyuAbO1o-cnWvGmmrEymA39Jn1oiEG-f4gTv7kTkURSZu-FtBie9dKXTWG-EHy8qr7acFitEyuLnBfUpA75OWCEJG7wB4D5DfNti3zagUTq_DSjTBmTGSLRIlS0TVBEiyQ9Vjfpy4QaCd5AFXSXSI-sUzH_wIbY5kDysM8OJ_8rD1MeTDuXRJWLMJmokcOkjR8AXpVPxWCBvDmFnYhwjyA1AkvqpIUCw0H2W59ZQFG-k_AmKJXc3ZGMwEQQuE-fSlZr9jbwAGNTWJPvOVvZWil_yoT8Rb6P_KuA5px6CtfIERCEf-lpag_j6Rn9dGXAFtPrY1kCFCL6pFG9pgi8pHzSHpIkzCfYN1sGytUz6iiWeEV_ua9OCOyoTmuwHJBmMzD2SNMXaNmH3tcniIalbG817u0RUTgsgc79FlMnJV8fgub5kxw4Kedh9S8J6P4P2Z7JyF7x44IQxc9nGPLfT37wb8LibO_esJ-d8hI4XMyApoupt-hzCRNvPbmHRGyzoty9ubtkx_1mGLOMLjCDAEZqBpI1utKYL3j-U1eJDOJGBBugkvTCok9HkbMwWqsEUwMITfwpgz1Oc2ce-DK65Hg75KO55larQlJUU5QkjU65MRduSreCNT-8kLdWTSpyeReUGA1YhbmPhGNd3yzmEVnJV0VDxhrVjljabRpQIbYJ-13OXyRsrNWv386BDW5Wt8RH3MgyluTPh3azuaQkVn3u9C37cy1CgV67BaW0VPevpKFzAclVTQHcFI2zT4byccJSS8sOztnQ2Gzwl1ACFMGwAoso68JENQhEE8jMBLbI6Q-xENbSEaXj7E24oiTFTKsI--b8e1-dyLlTYl1aa-sdj6KKyx0Yr1MTWTZfzqo0Ep2-9JRJAeOBR_5_QMQBoRRwjJXmUeQGMSDynXaExHdn-lkggGxF6Rx_1iXwsvY_7n6uSM_Qg96KHS1eDFhyev1vBZ34MIgRFiYo8Ehr4uGJeXZqwAwFBQERkYR_xFPPgJ1ESk3J5xDa-WtvTgr7iEIulNfvbmx8rCZxvDwh0t6GhUmuLDf2xV0JS_hi7wRGSOLZ1Q4VtbDUr3q11FbKYIvoD_pHYKdqUovVn0oYUmwysWIKaZ-05GUF7Tu9SYluv5JB4cKf8r-Wi2GUQBC-KUnlfYsrGcLlje_puy8XGsU09NUzDDYeRkfeB3M6ZEk3agyfRf15_3813L05MoswWpQ2Kt5UUgY1SVgwxKAJltwN4tHzO2SUrxxalf7eBNV8PO8bPcIXMJwNh_jq6jPqzYbSBk2LRSuUWz_oQl2yUo2uBoL8kPAcR8WMp2uGymgHwPz9-msey2Vk1rfYcwK3yQY3P9KYN55heTdEua-QnS9d_rREqtsdToUH6WtPzMB9zwTrXe7wThP6u9FbqYkgzlCEqzdG-5OnOTZdRcQGhroxpL6fy17_B5RsTvBcjqvBiuR9X4C0mVgae1lwMQWF95ljH3j81oeUhFESdWUy94bluNAL-dHfTRoBwJQ9oaSWAT84sS6GyW98TUIM1M0pOphxLkFKndaD0lsHVxKQObKML0xb3QAACwt3lbcCAqA6_r33zBxreL4_GUGcPkG3eu-dTTztT-aQeL4DqgaOskOPdPuSi3ygUYMBLZ1slWpq6R9sT-pk9LtRMl192vy77fenhW2EEaPp3lgbwAzTsSmwL8Abtd_vEQ9gIAKHdZq0wBy0KUKQjPmho4Ry6tfsjJs2X-qpX0c-fsP_CCqZ8EMuMFVLd1xqtlA59Utp7lj9K-AObYX2ohAALIO4X146csvOyV0BgUoHu5ycNk6T93JcyeocSvOlgLjgn9YgAvgHw0tFSq8EyYGUb9ewrKyjNHuzFffjzvwgfupVJ1PiA8tO7TZuLMSY1NYnSQE3yLluKQNlH2oq9JN2XQ4wtr71yQq2R47GvcqpVqPn_VPCdcKnWqNRo4qlB8dRtfzyACWYRkLzeqWtQ5gEU5vipViECyBnQurpCBNVuMOckfff_dnBhMSkuhsvGbnecX7ThH5As8gDQG23TWb1agL-cnOKS6OPvZty1MiL0lB0jcvTNSHsgE2X1c9d_B_-yII9_EKnEoOh36y7YEWHN66sZE0qG2muawG31EnaYCtBU8FBUILcOI7xKwLZfZm2LxIdSd4oNd6fEWaRtYD-z4V0MkOhpWh2oUuxtpeB9tVMKv8UHBFifRrR87-3xykOEUSBDvTUb33tahvfSeDsEk24FEwa0ZJfq0NXma4a-MdpnvHDSblDPCrpw7AKY1wWWg-oEy4vMsO14qAIcFKgl6A4vWlGPO_rkunrgm9JlTXISRUqHDkgjxdFxdwlsDu-x1oIXvLpTbWveFM9E066Z7Rk0VKvaw8EPOw6FgqYH81yaHhhMzY8BMYiPgzteqtEwUceGoU2H3LXOK5-xw6_24c14nftqLjk0ujhY23HC8I3C8gA42LZsX6gD21rTVafSaS-XD4x2RMOfuhqsKvwODTu7tBx-HRBEekZRhVe_HNYHTi24XyDhQsCfLRZ4OtkI3j_-uGL34Am_qtapr3aQLTs0jLtBkeqmtpRVhw6Z3t2jcoHAzjhzlPUyTn_Pjc9RXWZ_mBNk_Fjl-waZZNii0kRqqvr99yjNZqopeweKfLtVLjHObpyL5DC9KTw7VUIAQgywFxFj7pgWmlPIdjQsBBIsqGFElOZmS-7UZ-ti79A2PEboqJYmDKOQXLMAdOt07O0GPornqeh-IYmaXWQ-BWdaMuaXVdq0Z1aRdi-mryOpcvy02N3L__vUdLXBQZIXG7iPA-Lu2t3S1ixn5ytVLPvCTmWPCl9BmCSZzy5vJK2BvSTF8spGW4sO2EAS_JktsvdnqH2RbTtenf7QcZP1c6aqyQTz44aW0KDnEMyTDBrwGwGesA4iVAZv6oxwSmD-d075paAXb5F9lu_9h8wkjllZyVDsYxpu2tRXNBynROQzSF4nvhdy3PGGSXKz0bPf5Ia9iZuUBg5ndH2SXgtFzW23lgbjNujgFfFi5bVqikyh8xJuIhMeRio8YwZP8IqSZwnezsIdoDEvihcclssrsyK4D9z9jx4UEHgK3dYTc5IpINmv4AnzKCEmMYgeHPnpTNWE7z7wk42rTQ--DuDe3mdwvAdqYfxSYMFu8MM-N1ZOqG2f_UljIbYnFQpYCiiXt8peffWYNx0NvKgRTuBzPZDJlZwGoi5H5Do_sht5kxX3fY4K9UegHcnnUPYvSm6aXOTNkHBC4fy3KGw-qxCrFpW8tx8twc0m2aYMXTR2XmHReLWOjrM6wFvGqf6JyDAieQEAJLCmAFjAgmeDHPpUymRvs4Cd9Xj821aCXoqDbE1MKNmr5OtIx8l8ktJzzzArrnfebu-S138_yD5ZtxddIx4D6jrUrdeTAb_IpSq5Ov9R4u7xZLhl19WhXDUQFcafVS19eyAZkh-JWkQKuVHuXg66rPCw4zFxeAHGnmu-Y7p7nsN303DGMD5b1ug4ameHGc9JpvGnfNEpw7dbbevmFwuAjkBOkutC1v2kVvmO4tnR5t3kEgJyBbgRMT4IldfFhWpMO0fdVjLAJCcTuJqm1jtE9xwbzL-OVmnNpjrG4Nh-kuOFEww8MjF9OgH8I1cGUqZpd9QSi6QfoVDA7Rb6DYcpRjgeSM2LVxD1nZT01ta1mD7AW-aprakQWk7MqVtU9RUbprmkNnL1td5AT7y9JoRukC0QhB-QtM8z2VwrX5k92bk7SoXXbmBpVysm-FKrzuY8yhYXCuyK996Pp7orkuZLbtAma1mRVOboRxx25GDKwPwZQN3vzBGxgcrTmr5Bw7uzaRIkPIO6kCLrRfgPHE0BPJcAI8jlDws1SSGrLZ1kpeOMwpoVtXij9jUFttwQWx0RGfXLYKK30H9cvIxrgaAwkIaq7J5-BeUmzPvZxZENTgyAgqFQWf_CbKgCDr8BTCtjCCsXsxk0QdgRL4B7ceBZ4orFjeCgWh2fEHg9YVNgegTL5vR5YH8hCnz4vBjA2diOU4ZMUvMdHQsIIbSK4Y29Lz3L27qsK8V55Go16-1u_ExHC5o5cV25cce4gzzJUugoHctrLWa73C8K5dlqh16W9gP_jekpfUPj52sfoi-aF9CteSvB8gyBjRtYNCEBjxNgAx_rTNZHL1jhEnZiGUmvs2B8ZSzdX8eWzr-NL5WN5xrysnMgPqCMBwnIRZrqgaccUGsuYJ9qag5WUcvhVS2QIwUj7_be41soNj1_5tA6Q-3dGdVNS8P7iPi79tSgP3FVcGCnW9-P2t8b6nBSn7FsbHLH-ytSNOVRcKyCt0jEa8x0-PRsnza67tWz_logEmPrRPrSaC4r7eCD04IIoZuSM4EcqaFWMIXEm_BNU9vcHSTRfZioxafIQAR7sMjZy0vxUZPIAsaodSu-QfmqJfH0eSXaEpfCngFiidD0YUz2pB8lQMRAvl_eXPynXccgyk7PsjtrMAjh-l48_C3t7hw1BYAPl8BLQzvOBB5__GOAofxItQa50VIx_FK2aT95n_6u-CcYSYP37y_y84AzuTEmPafqN0Y-0m-pjlq56P63RJXmH84VOHv1cfxV-5v4wf7vbrpU2WZOVrgudO0uvsU4dDjuzMkjiTSvqGNSnvp4HUrpZ0YSq8pwegHYTx8TpTFA8oo5UsM0jJY4nTDk9zPe2hRH_RDomNCJKYoTKNjcWaTn0apqq-0LyNtOBx2qAtYN9HKny-WMlI4EGIO4ZoU2j_Krln9NGXSo_CwuW58NBirfQ7RkMmByJFml7KymuhuCaN5SzKb3rTaOgU7Y-RI16c14YgC__8QDzDJeXKNDnlKs83qQWoFijrBAuSFw49G-OUpYEpx3pt7bmYpirL_NOaKfJ7_simblVpMJp6B3bbBnmg5YcnTNHwLOcW99CwTIximeASfGYuMqELYKnLw1nPTFOuZbw8M5IjacSwMiIWD0Oj0ZgiRaNzUWpubLXpRsk5Ap8fzVmIuQrN2BoyKdmvrvN3Ibwx8LolfWjE0I72R2bliQndIDbYkMVKUsZidigbCszLeQTm2wb8DOi4sryp3m7o6R-kEcw0aoFHnWezRqjA-h7oLUzwyM-Vd7o33HwxrxCiq5LbGvNmtQ4Y0uLndqWy4BaiIPKLXPx8hicexxyG_nb0TZ7MX2iDl0nmYvwe_-owxmw-0GQ7dVZfeLCJhjcR6R1ltjSe6KAeoOWBPymHwf2y7O_K31vw1VXfBG_Z8IP9LjMtNz9xNbDMat0V_MTCxDCUNLpKfRjjCr3jQOZQ8cEaBnIUC7ZXuqI6uZVNrAbc0zNSyC_PESFDV2-f1aQ426GpMZqXX2yFI5ieF2G9cPbpWgDez_J12EHTBFiUPoHxFi_0dSVz3qCWOgUCnbgBlQqyy1VXl38gJnKGSED6knhOrNZz0Y3CT1b_B1110EScTeePvzQRIx7R39ob1sxN'

{cloudlanguagetools-11.3.4 → cloudlanguagetools-11.4.0}/cloudlanguagetools/naver.py

@@ -13,6 +13,7 @@ import cloudlanguagetools.ttsvoice
 import cloudlanguagetools.translationlanguage
 import cloudlanguagetools.transliterationlanguage
 import cloudlanguagetools.errors
+from cloudlanguagetools.options import AudioFormat
 
 NAVER_VOICE_SPEED_DEFAULT = 0
 NAVER_VOICE_PITCH_DEFAULT = 0
@@ -48,6 +49,14 @@ class NaverVoice(cloudlanguagetools.ttsvoice.TtsVoice):
                 'min': -5,
                 'max': 5,
                 'default': NAVER_VOICE_PITCH_DEFAULT
+            },
+            cloudlanguagetools.options.AUDIO_FORMAT_PARAMETER: {
+                'type': cloudlanguagetools.options.ParameterType.list.name,
+                'values': [
+                    cloudlanguagetools.options.AudioFormat.mp3.name,
+                    cloudlanguagetools.options.AudioFormat.wav.name
+                ],
+                'default': cloudlanguagetools.options.AudioFormat.mp3.name
             }
         }
 
@@ -63,7 +72,7 @@ class NaverTranslationLanguage(cloudlanguagetools.translationlanguage.Translatio
 
 class NaverService(cloudlanguagetools.service.Service):
     def __init__(self):
-        pass
+        self.service = cloudlanguagetools.constants.Service.Naver
 
     def configure(self, config):
         self.client_id = config['client_id']
@@ -93,8 +102,10 @@ class NaverService(cloudlanguagetools.service.Service):
 
 
     def get_tts_audio(self, text, voice_key, options):
-        output_temp_file = tempfile.NamedTemporaryFile()
-        output_temp_filename = output_temp_file.name
+        response_format_parameter, audio_format = self.get_request_audio_format({
+            AudioFormat.mp3: 'mp3',
+            AudioFormat.wav: 'wav'
+        }, options, AudioFormat.mp3)
 
         url = 'https://naveropenapi.apigw.ntruss.com/tts-premium/v1/tts'
         headers = {
@@ -107,19 +118,13 @@ class NaverService(cloudlanguagetools.service.Service):
             'text': text,
             'speaker': voice_key['name'],
             'speed': options.get('speed', NAVER_VOICE_SPEED_DEFAULT),
-            'pitch': options.get('pitch', NAVER_VOICE_PITCH_DEFAULT)
+            'pitch': options.get('pitch', NAVER_VOICE_PITCH_DEFAULT),
+            'format': response_format_parameter
         }
+        if audio_format == cloudlanguagetools.options.AudioFormat.wav:
+            data['sampling-rate'] = 48000
 
-        # alternate_data = 'speaker=clara&text=vehicle&volume=0&speed=0&pitch=0&format=mp3'
-        response = requests.post(url, data=data, headers=headers, timeout=cloudlanguagetools.constants.RequestTimeout)
-        if response.status_code == 200:
-            with open(output_temp_filename, 'wb') as audio:
-                audio.write(response.content)
-            return output_temp_file
-
-        response_data = response.json()
-        error_message = f'Status code: {response.status_code}: {response_data}'
-        raise cloudlanguagetools.errors.RequestError(error_message)
+        return self.get_tts_audio_base_post_request(url, data=data, headers=headers)
 
     def get_tts_voice_list(self):
         # returns list of TtSVoice

{cloudlanguagetools-11.3.4 → cloudlanguagetools-11.4.0}/cloudlanguagetools/openai.py

@@ -12,6 +12,7 @@ import cloudlanguagetools.languages
 import cloudlanguagetools.options
 
 from cloudlanguagetools.languages import AudioLanguage
+from cloudlanguagetools.options import AudioFormat
 
 logger = logging.getLogger(__name__)
 
@@ -29,6 +30,7 @@ VOICE_OPTIONS = {
                 'values': [
                     cloudlanguagetools.options.AudioFormat.mp3.name,
                     cloudlanguagetools.options.AudioFormat.ogg_opus.name,
+                    cloudlanguagetools.options.AudioFormat.wav.name
                 ],
                 'default': cloudlanguagetools.options.AudioFormat.mp3.name
             }
@@ -217,16 +219,17 @@ class OpenAIService(cloudlanguagetools.service.Service):
         output_temp_file = tempfile.NamedTemporaryFile()
 
         speed = options.get('speed', DEFAULT_TTS_SPEED)
-        response_format = options.get(cloudlanguagetools.options.AUDIO_FORMAT_PARAMETER, 
-            cloudlanguagetools.options.AudioFormat.mp3.name)
-        if response_format == cloudlanguagetools.options.AudioFormat.ogg_opus.name:
-            response_format = 'opus'
+        response_format_parameter, audio_format = self.get_request_audio_format({
+            AudioFormat.mp3: 'mp3',
+            AudioFormat.ogg_opus: 'opus',
+            AudioFormat.wav: 'wav'
+        }, options, AudioFormat.mp3)
 
         response = self.client.audio.speech.create(
             model='tts-1-hd',
             voice=voice_key['name'],
             input=text,
-            response_format=response_format,
+            response_format=response_format_parameter,
             speed=speed
         )
         response.stream_to_file(output_temp_file.name)

{cloudlanguagetools-11.3.4 → cloudlanguagetools-11.4.0}/cloudlanguagetools/service.py

@@ -1,10 +1,11 @@
 import requests
 import tempfile
 import logging
-from typing import List
+from typing import List, Dict
 
 import cloudlanguagetools.constants
 import cloudlanguagetools.ttsvoice
+import cloudlanguagetools.options
 
 logger = logging.getLogger(__name__)
 
@@ -21,9 +22,10 @@ class Service():
 
     def get_tts_audio_base_post_request(self, url, **kwargs):
         try:
+            kwargs['timeout'] = cloudlanguagetools.constants.RequestTimeout
             response = self.post_request(url, **kwargs)
             response.raise_for_status()
-            output_temp_file = tempfile.NamedTemporaryFile()
+            output_temp_file = tempfile.NamedTemporaryFile(prefix='clt_audio_')
             output_temp_filename = output_temp_file.name            
             with open(output_temp_filename, 'wb') as audio:
                 audio.write(response.content)
@@ -35,6 +37,13 @@ class Service():
             logger.exception(error_message)
             raise cloudlanguagetools.errors.RequestError(error_message)
 
+    def get_request_audio_format(self, format_map: Dict, options: Dict, default_format: cloudlanguagetools.options.AudioFormat):
+        response_format_str = options.get(cloudlanguagetools.options.AUDIO_FORMAT_PARAMETER, 
+            default_format.name)
+        response_format = cloudlanguagetools.options.AudioFormat[response_format_str]
+        response_format_parameter = format_map[response_format]
+        return response_format_parameter, response_format
+
     # used for pre-loading models
     def load_data(self):
         pass

{cloudlanguagetools-11.3.4 → cloudlanguagetools-11.4.0}/cloudlanguagetools/watson.py

@@ -11,6 +11,7 @@ import cloudlanguagetools.ttsvoice
 import cloudlanguagetools.translationlanguage
 import cloudlanguagetools.transliterationlanguage
 import cloudlanguagetools.errors
+from cloudlanguagetools.options import AudioFormat
 
 logger = logging.getLogger(__name__)
 
@@ -71,7 +72,18 @@ class WatsonVoice(cloudlanguagetools.ttsvoice.TtsVoice):
         return self.description.split(':')[0] + is_dnn
 
     def get_options(self):
-        return {}
+        return {
+            cloudlanguagetools.options.AUDIO_FORMAT_PARAMETER: {
+                'type': cloudlanguagetools.options.ParameterType.list.name,
+                'values': [
+                    cloudlanguagetools.options.AudioFormat.mp3.name,
+                    cloudlanguagetools.options.AudioFormat.ogg_opus.name,
+                    cloudlanguagetools.options.AudioFormat.ogg_vorbis.name,
+                    cloudlanguagetools.options.AudioFormat.wav.name
+                ],
+                'default': cloudlanguagetools.options.AudioFormat.mp3.name
+            }
+        }
 
 class WatsonService(cloudlanguagetools.service.Service):
     def __init__(self):
@@ -124,6 +136,13 @@ class WatsonService(cloudlanguagetools.service.Service):
         return result
 
     def get_tts_audio(self, text, voice_key, options):
+        response_format_parameter, audio_format = self.get_request_audio_format({
+            AudioFormat.mp3: 'audio/mp3',
+            AudioFormat.ogg_opus: 'audio/ogg;codecs=opus',
+            AudioFormat.ogg_vorbis: 'audio/ogg;codecs=vorbis',
+            AudioFormat.wav: 'audio/wav;rate=48000'
+        }, options, AudioFormat.mp3)
+
         output_temp_file = tempfile.NamedTemporaryFile()
         output_temp_filename = output_temp_file.name
 
@@ -133,7 +152,7 @@ class WatsonService(cloudlanguagetools.service.Service):
         constructed_url = base_url + url_path + f'?voice={voice_name}'
         headers = {
             'Content-Type': 'application/json',
-            'Accept': 'audio/mp3'
+            'Accept': response_format_parameter
         }
 
         data = {

{cloudlanguagetools-11.3.4 → cloudlanguagetools-11.4.0}/cloudlanguagetools.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: cloudlanguagetools
-Version: 11.3.4
+Version: 11.4.0
 Summary: Interface with various cloud APIs for language processing such as translation, text to speech
 Home-page: https://github.com/Language-Tools/cloud-language-tools-core
 Author: Luc

{cloudlanguagetools-11.3.4 → cloudlanguagetools-11.4.0}/cloudlanguagetools.egg-info/SOURCES.txt

@@ -4,6 +4,7 @@ setup.py
 cloudlanguagetools/__init__.py
 cloudlanguagetools/amazon.py
 cloudlanguagetools/argostranslate.py
+cloudlanguagetools/audio_processing.py
 cloudlanguagetools/azure.py
 cloudlanguagetools/cereproc.py
 cloudlanguagetools/chatapi.py

{cloudlanguagetools-11.3.4 → cloudlanguagetools-11.4.0}/setup.py

@@ -6,7 +6,7 @@ from setuptools.command.install import install
 # twine upload dist/*
 
 setup(name='cloudlanguagetools',
-      version='11.3.4',
+      version='11.4.0',
       description='Interface with various cloud APIs for language processing such as translation, text to speech',
       long_description=open('README.rst', encoding='utf-8').read(),
       url='https://github.com/Language-Tools/cloud-language-tools-core',

{cloudlanguagetools-11.3.4 → cloudlanguagetools-11.4.0}/tests/test_audio.py

@@ -11,6 +11,7 @@ import pytest
 import json
 import time
 import pprint
+import functools
 
 import audio_utils
 
@@ -33,13 +34,26 @@ def get_manager():
 
     return manager
 
+
+
+def skip_unreliable_clt_test():
+    def decorator(func):
+        @functools.wraps(func)
+        def wrapper(*args, **kwargs):
+            if not CLOUDLANGUAGETOOLS_CORE_TEST_UNRELIABLE:
+                pytest.skip(f'you must set CLOUDLANGUAGETOOLS_CORE_TEST_UNRELIABLE=yes')
+            return func(*args, **kwargs)
+        return wrapper
+    return decorator
+
+
 class TestAudio(unittest.TestCase):
 
     ENGLISH_INPUT_TEXT = 'This is the best restaurant in town.'
-    FRENCH_INPUT_TEXT = "On a volé mes affaires."
-    JAPANESE_INPUT_TEXT = 'おはようございます'
-    CHINESE_INPUT_TEXT = '老人家'
-    KOREAN_INPUT_TEXT = '여보세요'
+    FRENCH_INPUT_TEXT = 'Bonjour'
+    JAPANESE_INPUT_TEXT = 'こんにちは'
+    CHINESE_INPUT_TEXT = '你好'
+    KOREAN_INPUT_TEXT = '안녕하세요'
 
     @classmethod
     def setUpClass(cls):
@@ -73,6 +87,18 @@ class TestAudio(unittest.TestCase):
         subset = [x for x in self.voice_list if x['audio_language_code'] == audio_language.name and x['service'] == service.name]
         return subset        
 
+    def get_voice_by_service_and_name(self, service: Service, voice_name) -> cloudlanguagetools.ttsvoice.TtsVoice_v3:
+        subset = [x for x in self.voice_list_v3 if voice_name in x.name and x.service == service]
+        self.assertEqual(len(subset), 1)
+        return subset[0]
+
+    def get_voice_by_lambda(self, service: Service, filter_func, assert_unique=True):
+        service_voices = [x for x in self.voice_list_v3 if x.service == service]
+        subset = [x for x in service_voices if filter_func(x)]
+        if assert_unique:
+            self.assertEqual(len(subset), 1, pprint.pformat(subset))
+        return subset[0]
+
     def verify_voice(self, voice, text, recognition_language):
         return self.verify_voice_internal(voice['voice_key'], voice['service'], text, recognition_language)
 
@@ -204,11 +230,11 @@ class TestAudio(unittest.TestCase):
         self.verify_service_audio_language(source_text, Service.Amazon, AudioLanguage.fr_FR, 'fr-FR')        
 
     def test_mandarin_google(self):
-        source_text = '老人家'
+        source_text = self.CHINESE_INPUT_TEXT
         self.verify_service_audio_language(source_text, Service.Google, AudioLanguage.zh_CN, 'zh-CN')
 
     def test_mandarin_azure(self):
-        source_text = '你好'
+        source_text = self.CHINESE_INPUT_TEXT
         self.verify_service_audio_language(source_text, Service.Azure, AudioLanguage.zh_CN, 'zh-CN') 
 
     def test_azure_standard_voice_deprecated(self):
@@ -285,16 +311,15 @@ class TestAudio(unittest.TestCase):
         source_text = '老人家'
         self.verify_service_audio_language(source_text, Service.CereProc, AudioLanguage.zh_CN, 'zh-CN')
 
+    @skip_unreliable_clt_test()
     def test_mandarin_vocalware(self):
         # pytest test_audio.py -k test_mandarin_vocalware
-        if not CLOUDLANGUAGETOOLS_CORE_TEST_UNRELIABLE:
-            pytest.skip('you must set CLOUDLANGUAGETOOLS_CORE_TEST_UNRELIABLE=yes')
         
         source_text = '你好'
         self.verify_service_audio_language(source_text, Service.VocalWare, AudioLanguage.zh_CN, 'zh-CN')
 
     def test_cantonese_google(self):
-        source_text = '天氣預報'
+        source_text = self.CHINESE_INPUT_TEXT
         self.verify_service_audio_language(source_text, Service.Google, AudioLanguage.zh_HK, 'zh-HK')
 
     def test_cantonese_azure(self):
@@ -354,6 +379,7 @@ class TestAudio(unittest.TestCase):
         source_text = 'hello <break time="50ms"/>world'
         self.verify_service_audio_language(source_text, Service.Google, AudioLanguage.en_US, 'en-US')
 
+    @skip_unreliable_clt_test()
     def test_ssml_english_azure(self):
         # pytest tests/test_audio.py -k test_ssml_english_azure
         source_text = 'hello <break time="200ms"/>world'
@@ -475,6 +501,50 @@ class TestAudio(unittest.TestCase):
         audio_text = audio_utils.speech_to_text(self.manager, audio_temp_file, 'fr-FR', audio_format=cloudlanguagetools.options.AudioFormat.ogg_opus)
         self.assertEqual(audio_utils.sanitize_recognized_text(source_text), audio_utils.sanitize_recognized_text(audio_text))        
 
+    def verify_wav_voice(self, voice: cloudlanguagetools.ttsvoice.TtsVoice_v3, text: str, recognition_language: str):
+        # assert that the wav format is in the list of supported formats
+        self.assertTrue(cloudlanguagetools.options.AudioFormat.wav.name in 
+                        voice.options[cloudlanguagetools.options.AUDIO_FORMAT_PARAMETER]['values'])
+        options = {cloudlanguagetools.options.AUDIO_FORMAT_PARAMETER: cloudlanguagetools.options.AudioFormat.wav.name}
+        audio_temp_file = self.manager.get_tts_audio(text, voice.service, voice.voice_key, options)
+        audio_utils.assert_is_wav_format(self, audio_temp_file.name)
+        audio_text = audio_utils.speech_to_text(self.manager, audio_temp_file, recognition_language, audio_format=cloudlanguagetools.options.AudioFormat.wav)
+        self.assertEqual(audio_utils.sanitize_recognized_text(text), audio_utils.sanitize_recognized_text(audio_text))
+
+    def test_azure_format_wav(self):
+        fr_voice = self.get_voice_by_service_and_name(Service.Azure, 'Denise')
+        self.verify_wav_voice(fr_voice, self.FRENCH_INPUT_TEXT, 'fr-FR')
+
+    def test_amazon_format_wav(self):
+        fr_voice = self.get_voice_by_service_and_name(Service.Amazon, 'Mathieu')
+        self.verify_wav_voice(fr_voice, self.FRENCH_INPUT_TEXT, 'fr-FR')
+
+    def test_elevenlabs_format_wav(self):
+        fr_voice = self.get_voice_by_lambda(Service.ElevenLabs, 
+            lambda x: 'Charlotte' in x.name and x.voice_key['model_id'] == 'eleven_multilingual_v2')
+        self.verify_wav_voice(fr_voice, self.FRENCH_INPUT_TEXT, 'fr-FR')        
+
+    def test_google_format_wav(self):
+        fr_voice = self.get_voice_by_lambda(Service.Google, 
+            lambda x: AudioLanguage.fr_FR in x.audio_languages, assert_unique=False)
+        self.verify_wav_voice(fr_voice, self.FRENCH_INPUT_TEXT, 'fr-FR')
+
+    def test_naver_format_wav(self):
+        ko_voice = self.get_voice_by_lambda(Service.Naver,
+            lambda x: AudioLanguage.ko_KR in x.audio_languages, assert_unique=False)
+        self.verify_wav_voice(ko_voice, self.KOREAN_INPUT_TEXT, 'ko-KR')
+
+    def test_openai_format_wav(self):
+        ko_voice = self.get_voice_by_lambda(Service.OpenAI,
+            lambda x: AudioLanguage.en_US in x.audio_languages, assert_unique=False)
+        self.verify_wav_voice(ko_voice, self.ENGLISH_INPUT_TEXT, 'en-US')
+
+    def test_watson_format_wav(self):
+        en_voice = self.get_voice_by_lambda(Service.Watson,
+            lambda x: AudioLanguage.en_US in x.audio_languages, assert_unique=False)
+        self.verify_wav_voice(en_voice, self.ENGLISH_INPUT_TEXT, 'en-US')
+
+
     def test_google_voice_journey(self):
         service = 'Google'
         source_text = self.ENGLISH_INPUT_TEXT
@@ -592,12 +662,15 @@ class TestAudio(unittest.TestCase):
         source_text = 'Ich mag das Essen nicht.'
         self.verify_service_audio_language(source_text, Service.ElevenLabs, AudioLanguage.de_DE, 'de-DE')        
 
+    @skip_unreliable_clt_test()
     def test_elevenlabs_japanese(self):
         self.verify_service_japanese(Service.ElevenLabs)
 
+    @skip_unreliable_clt_test()
     def test_elevenlabs_chinese(self):
         self.verify_service_chinese(Service.ElevenLabs)
 
+    @skip_unreliable_clt_test()
     def test_elevenlabs_korean(self):
         self.verify_service_korean(Service.ElevenLabs)
 

{cloudlanguagetools-11.3.4 → cloudlanguagetools-11.4.0}/tests/test_translation.py

@@ -5,6 +5,8 @@ import unittest
 import pytest
 import json
 import pprint
+import time
+import requests
 
 sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
 
@@ -21,11 +23,24 @@ def get_manager():
 
 class TestTranslation(unittest.TestCase):
     def setUp(self):
-        self.manager = get_manager()
-        self.language_list = self.manager.get_language_list()
-        self.translation_language_list = self.manager.get_translation_language_list_json()
-        self.transliteration_language_list = self.manager.get_transliteration_language_list_json()
-        self.tokenization_options = self.manager.get_tokenization_options_json()
+        max_retries = 3
+        retry_delay = 5  # seconds
+
+        for attempt in range(max_retries):
+            try:
+                self.manager = get_manager()
+                self.language_list = self.manager.get_language_list()
+                self.translation_language_list = self.manager.get_translation_language_list_json()
+                self.transliteration_language_list = self.manager.get_transliteration_language_list_json()
+                self.tokenization_options = self.manager.get_tokenization_options_json()
+                break  # If successful, break out of the retry loop
+            except requests.exceptions.RequestException as e:
+                if attempt < max_retries - 1:  # If not the last attempt
+                    logging.warning(f"Request failed. Retrying in {retry_delay} seconds... (Attempt {attempt + 1}/{max_retries})")
+                    time.sleep(retry_delay)
+                else:
+                    logging.error(f"Max retries reached. Unable to set up the test environment.")
+                    raise  # Re-raise the last exception if all retries failed
 
     def test_language_list(self):
         self.assertTrue(len(self.language_list) > 0)
@@ -140,7 +155,7 @@ class TestTranslation(unittest.TestCase):
         self.assertRaises(cloudlanguagetools.errors.RequestError, self.translate_text, Service.Naver, 'Veuillez parler lentement.', Language.fr, Language.th, 'Please speak slowly.')
 
     def test_translate_deepl(self):
-        # pytest test_translation.py -rPP -k test_translate_deepl
+        # pytest tests/test_translation.py -rPP -k test_translate_deepl
         self.translate_text(Service.DeepL, 'Please speak slowly', Language.en, Language.fr, 'Veuillez parler lentement')
         self.translate_text(Service.DeepL, 'Je ne suis pas intéressé.', Language.fr, Language.en, ["""I'm not interested.""", 'i am not interested'])
         self.translate_text(Service.DeepL, '送外卖的人', Language.zh_cn, Language.en, ['delivery person', 'takeaway delivery people'])

cloudlanguagetools-11.3.4/cloudlanguagetools/keys.py

@@ -1 +0,0 @@
-KEYS='gAAAAABkk5vaZm3izjIqmeoWzO8UiDRp8Riz0F2BmB5_tFmcc2Fgu0W-MXvC7y4-7vfN5ggR0hgVRKJVC2UbyJpNBt3yDuJAfBoccZ1327aDb2KfXb69akWXWrTnIqHTpfl1XbpVIbgMk3MT777H_0cz88GM4cgVtH9WuMQC20wjr3AQgHEhWUlDowIdHPbyb_UX-bom3bEF0DfZ__EkYS7tIBmMOH_B6oSayUSibUG1SJc7xfGvJqW4zXUeESJtBujRqVvbr5z0zf46eicNAMiQXEm9IpYIRnrldsUID-6KGw4eUCYx32PPYjldGkvapelHnpAv5WqqHQ6M_p4lu95CbgWwfQIF1qLW68cOOiAZBgjHZrPbY0DLqT-eEfB_xrek0U7ZqlpOisIfc3Q8QfoyGaW-mSchB5GYqXfp38xLlPGI3OSYnJl08MuJ6p7M5cxRTRkjvTj8UimShfvmPm5ltM-XW43uIsuy5rAkrHIbsKCkiJ-tXK1eEIySR092cl0qpZ3U2ztySbzX1eXsgzynYVfGdD7V4tGNpTMPhekOTpFLNKw6swA3EK1tC_7aWWvnqec6ccew7nAPHKE00Li7trYpvZeBdLQvOdVzZj77TTLdBE39_syeB4EDZfV6WXFu8_DhAuNiPF2LJHFIMWCA6RiAo1HtPUC8l3IDU3uSBljyZ1QTGNUBg975SljdrgLWocm-Hmyfvq0WfnTk19JnXFAp_hiIsPtsulyLkuGLXXO3aUHl2xcGy6u9cQE3NxtslnSXy8IRnv9MqCae54NwzNGkWiCiN29pvNvEppuHG-riexmh9MhUvHVOV68kwuB5Et2gm9kFgpKJlV0ZkUWDdJUZDYrgQ2oQ7TmBhez43duyJZ2k2uYHaZ8MFFcFxjgT-WkvfkI8jIUZ0ZIwZMKjcM5OCNLPiL-EbxVO2grithjZ_bCbnpxsahKLvWA4QYVy-3eWhhqutmNpsFIig05No1I4tXNQmcxJcFANGIG2p73DbgBIfDrJToxYIPw_bX2Ii93cKX0FNhzv1be0uUP3RZcuni_ESYnvrXgdNFZ_fIf-lci6lZcWiITFxejDFe1pqb9rd_BLEyJNIFfkvFou_eC8wWT0G-h8yPckuVCYlOEWAPfhL5BjFxQhFvnwyTm4vNHBLnhoOhsURc1NTwJ9EL1_ksYLCktLnlFjjg21Z-93aV4G6T6X2bwIdFpqebo5G-SIz-xKVo7Wln105uzUr2CJEliaufNOWZC6by2N2HJ2jqsajiJX2ZNOJCjkYtjH_-jln9taKF3Q5BFSlOeraam-ppZkBsABMewrUZ8d7lqqdyN7ntDHWiASjFm11KL_25hRT6tMhx4BpKlP40ovfY7XoJhRT-_Ss7Aeze6jBj-Qe58dRDEiIzZcko2F1SEvIZFNVPbN3FuEjhlrOP9N1IvhCYn_8NjXENetbjIVhkQx5nKGaLVeT9toRc4qwhg-4XdN620TnIxPusva9ncXUYXVz1iDVFtN5-RuJRAppuBj7_nfJOG8MngntJOV7B0o6fVDK8McVi3TQ9lyCl1GMPuoeB41xCKyY3Lc-lqzgXhRhiWbuSBg1G45jQ3BE8bWciFjQCXuMeGXmDZzEFtzJMPQWwxzDkuOB2Y74ITEzsbleSvFXhlgUpa692ckDK1z_mK2UPrLbAa629GzPsrdjVGzQ6vWiYiKkuJFHQUkxpBscIH7y9rPW6zc0jg7tVmYwuV94UCS61g2g-s1-UodQog9vcD4JGyPgr2AcetoBzSh-wAsBBzra3A_E-cWdD2uzB5Mv60hgOlInLbi8QUBppCig89IwnFLdXAP7QFamWtijflUpqRECMvokxaImhtKSwPy3BlyJAMr7T3lFC3XY74fA-3k2wKAC_0LLYxedED68-GtklIpfgSwQ9e5oxEzv927oF_uEibZizjJusNtnk4Wz76YMd-VHq-wPJcxFcRqkKdZCgyToLY2z4YP81tWVxB_9F5aR_yF9cvCPbAycRAxny9Qsno44VPjZ5v7zLj5G7xqhEza11EzrUdPk657s_a1M6G8ARehCblBChTUD6Gxz1Y2-dZ0ejpB81kGFqICPEw24dF1xq44GAwNJsx_zYd_0y5GldZSMCYpmLbJNDFx1wGOJwLQq9h0JaqRT5jAYl3uacpmApYLj8jMCdn1HAdJwKSIw-s17RyEZHBg4JHC3tAMzZhhco5k7gHLCCRI_jep9fAusZTjo3mdaJD_tJfAZsoQXx4kq9HJ8k71nyxYz18pwzIfT0hJgFsSbS-UyrhAk8mskTcw5UPKGcgzMOwafh0Giy69Bn4KfU_nHUDeHng_b-KCN7G_fF43xX7l3EB3qIyN9vUtPYrHteJt0LYS0yXtUM_Dc0jxeEdZwcpKYxbkMlTbK1DM6JZ2JitNGAPDvNv8ZjRQxPmKjjgVdvs8OzJZ4kdPzAfEVBvhYx3gKmggcnsQBgxFki3FOhXhPrCAyxXO5-_glom5KzEGqG2VQgJWU3ycCtmFypoNmx38wyd3eH7_M2ydfkIE6e350g8y7sFayK6vqO4GdANxTRbHThF7HfI-3nWtPJ4NpILiu71CtpuaU5MAhqD8qAJjQkUUifANeY6i-aqQZXBPDWw7VaYpQWGAoiPBuFJc-a906yU-FzViRToZIyBbunOgJO0GOy4iJRJSnAQY3Cd3Xrk7VenUnBuYBLpUUNiVCZw6MnWTAhY1BujnDf-Pxrj8jJYf2Y0KipSxkCK4Hkjnc13bTkBovA8YM7HECTIVkkIlsjvV440eyqLpO1r4yD_lecvExcSNsF-EBwSbp0hRps3wTynY0yeuWPM3NTQ4dEmO6fW1ewpsBKicDu0sHWAhK_WOpYiydTT3rAGSgDolNSKKYUpFlHS92g5-Nb_nLW4qwOV5F8yDKbfYtaBKylCNrmDu8aXc5GIdIeK7eGC5JF8uaUCsidPOh7JuOpn-hhKym4Aq6Hb5FlnLf9eNkimCYDm3F1a91CcJwhPRN_FZNLVfpBqw8pOfFK9RkBEttStWxvgWvW-C5K5EzTiBGlZBR2j4LHazDJR8_bjJ6uBCw-8fFQsPWGgliic4_k1ioRVwhJDNfw6VN0mR6iqt9D7duTxU8nuNC3u5idCJuwDC4aJC9vgvpbU2TledADfaqeBujUn8DhkOEhGJsE3WEcwaidsqI3NFSKjtoxiDEOfWmXVmFAjDo_tENvjRnriFPrWR5vMvmluAlncg16ok0qqpLBpgnGYDYhfv4IC3HsYfSwmyCDrmoMUBqv5dUtbEBMbQB1BcWH1hOahCN60AO3Emedq2jsM853UI6_EuZA19TIhLreNMpEL1sSWbB3cVlLfGEvoSzga1MdtKsxCZmBIWa2_Eez9lBtdpuqG4phfpFs5sM6mWKza0ZLr8v5N4-oIRqApAAr2JNscXeehiZPhWxjaXmA_69N0mdmzkjwznH6sQ3bkX7b9UWOQPblEwPfk4iXktxiIOBGxDh0SgWMpEvtEO9MxFnOYoIsmeyTVNOG73iFKQ6gybMoAiCFBeGSq24scq1Tf8ydBkpaM7Z3SzNj3b4u5feRTb1dZCVsjQ3kCP657INEpRWx6GfVAZBH2-GGVuDNrr7-YY7XcLees9h5jSI5JdelsuFwfI8U1D3gH2F_gZenLaR5RlFpH-yma-XPOuCpbflzs-MoEO8A55L2zs6EZTxcjEGt8v-J8rKEIPt8sr0u80YklSYs0m5q8AwDjgzwRG9UPoaZJYclGo4U5N_P-fixUtF2cr7rHWxDZLzuK7jF8KNUW8pWRgYn5Zn6BfKN8ro-49xbqnDQ7CORKEiFxjcSUP4D0ocyFIhHf3V4YiyS860KHqbhkyRm4d765skBe5DqrJn2EeGwh_ehBMw30uqYai-ws1920xFGPYCGuHC2uY5cgY_8QF1HWsDsr-Ei4jtsE24_AA8oJTx2wIbchQIu7oJl9gElYbIRXaoHHyuaFIkxpV4CMpTXtiWTwIjQeCBbQmESnj9s5vBdKvqbqRDZQ0LigkqwCoO8nizJ8kNse601UN5Nci9UkH5vcOUmmc1RCj8kQLiaCWMTQ_lT3bzGXgUxRgUdrgVyKYbSSigWZ98WzT_YjsbjLufIY5uz1ix8e_mz7JK2MxFywSt_JJ2O-ry6NNVOUHig10XJBUEIj2hE4X7ziecWMAH93axfiwk9GxLy_vFxA5Nl0KbZGbvwixN4rrAH1yFChh83vF819kjhe0H4oCsGWvxq3OrxdVSbuF95p6lhxhpcy0hPM4rEQ5EcM3Gd_iFUxuZMBgZzvnjDgvvwGbX97Sx6WjIqEfNHFLbvID1nmUMXROPoMmnr_P14QptBvDCgE1SxDhavEOr8EMTxxzyrBEFeKwZ2Sw8QukuOceDMpwmR5ksAk0QiiWgvdzu_LkgUgayKhp8hQMmtt0wsTlmizkG5tATqGCri1dENCcKM5TqCqd3FEAPlm07VqPJrjwq2OHt5xR90HF8uizO9pdBXyY5_M5asChPsUyeBm5dYLfhl7JKLwVdj8wcZLsVojlunIUKppFdOSx9RW_jzqa20jvq1i7-f43kcMsjzuklAY4a3QoTh9gYOrOxzecko6lij6W0jVC0EuplsbUi75rwNzYVjRV1s_3Ze96S9Ym-dNf3murhzPTOvCL1v3Ugb0vtW5cEK0MwMEM7jN-K1GNprILjVMky15f5hiJEKlmsNuvILi6-F911Zu3s6Rl_a2P6-7zWa7crlrgsrCUoCmGWLpVW6fsscfGyBgSp74NV1qOz_062V9Fg7-LBcYWgR3VILw3VbFQ1ZvLggLuEpxbxwOIWgb_jhWjo5wbQ9pdGziHY1bR7ejh3n_Dc1YwfIKEo0yNG6nnLchcThTggRZhLozHpx-ElamXM7hMgotm1XUbjLlPEgDHfAQjazFPI0h_i_4N2MiJ_OSXvF3IN2U8nKivvfcCn3lVrWLzzHtiAxBoSOfGrwD6S46Um1cSV0_cRCJQAvbUsn2iT8CNVbOli6xzCceWkUdVjpZYhklwV8SMs839PTNN2CU9Poxm2t2PIeQcT_A-KFBqC6MYwGy_wLO_uibXHY1oxcL8FKl2JUD7JbXEWItFydDLQVz3KpumP43D_zJ3pK7kin5sA3CX6ultD94PKZegIpe1gc2SUh2bte31HwGZh96IuFNoEuYVumlSMrJhtpOvpKC_ScSW6uw_f1WHCnBotYXWX_kkmmig6H-Dzv4zOgLmMcSE_R8QhYc3XrUXL_0lpl85xsJHv2sWx-evCWhGeIF-LkYmP8mwCHXkfcigbJV5bPVX8XRXx6X4_ZKSTKmKmsyRaBLmxpWtg6bZRi8D2d1Oz9sNTSISWxJlChqcL5VVqKN2EAaGAupGqmRvn7dPt5JQqGARx14PGQGOcHygif0eCrGv3x78sG2Vi_mNZKbBszJNu-yePTDStZ_LPjKya87DEyIghwe-S4fFrx8X-oSz6PYd26bhCUfYfkPfsSShcfJ1aHXCY6b9In8EXd0-G5um6Ng0eIYlc6Y0Mr--cpc0xXEubTslRvUhwLumZ59Lr-4oN75NFra427iErxLrzhhFCSbcVc7mh-qYa-z_6mQqdbUfiPVwfE2iDEtGiz8-keZUfEysxh_nkKQSJO-dmie8031UpF2TEJTMQ4onnyKNmLLSFltmqT5xa6CRpHXPrhuoCx7J-lH9L5_qbOFiCNwNhMtEKuhMp-CBRQsK3KPLtzKx01haclk4uR36_fXLdLkY3GPG2jrkEoc3sWIlibQA5EtcNAT5Y7vVHVm1hW_P9xVOjGe_uokHLdln9rAOZayJkJid1hqikXjQ2uYdpk9rSNtxeWw-9YB6fGL0o3IFL01MngtMvSfvwXSWK2fbWkPEzSgPv8qpqQuDa8QzIh5IQ_r5IH7jHBY1zwn4IPhb6qxU7WvCmQ0zD0mt3b2XxNsL5qNbmGM0L63Pc-6ICwN5HMIxzo7T5rVzdqZn1sWu00T6pRQzzOXfhJIqS_WvmZe8X1dcTUo9mdrSjjYj5F5GR9k4M6MDih3M6ndfKJWUdE1-l7eD6BvES5faIuDicoU0HCOrU8oK82v7NApultvIu9LkZTBNQc2mv2Tc7nNx0jjoO1N3Kw4LPaFJxfSp3xEEn0eL1R7c-bKmVCCOT1rtbbJx2QCB_LWXqJnQtgZbX7x72jTokiqooKQTBlEc_YiqV4GTDCHQL-mW6XDm96YzhYhY6RxCfcABDx6Ot22uqZQ1g9gbLXV1raTUXC2zowuV49ehW8RWKDq6vVR94l9doLLjhO29cGk5LuQ2oYLA8pO9dDs6HMWtILj5msrcKYyjWR7hYQnnEyG4'