Beispiel #1
0
    def get(self):
        """Fetches NLP API results from GCS and sends to frontend.

        Args:
            text: String representing section of transcription to get results.
            file_name: String of audio file to retrieve chunk of text from.

        Returns:
            List of entity objects holding type, name, and sentiment score.
        """
        try:
            text = request.args['text']
            file_name = request.args['file_name']
            gcs_client = gcs_transcript_utils.authenticate_gcs()
            bucket_list = list(gcs_client.list_buckets())
            nlp_bucket = gcs_transcript_utils.find_bucket_with_prefix(
                bucket_list, 'output-files')
            nlp_path = f'nlp-files/{file_name}'
            nlp_json = gcs_transcript_utils.get_gcs_object(
                gcs_client, nlp_bucket, nlp_path)
            if nlp_json:
                text_section = [
                    section for section in nlp_json if section['text'] == text
                ]
                sentiments = text_section[0]['nlp_response']
                sentiments.sort(key=lambda entity: entity['score'])
                return jsonify(sentiment_result=sentiments)
            else:
                raise NotFound

        except NotFound as e:
            logging.error('Fetching entity failed.')
            logging.error(e)
            return jsonify(e.to_dict())
Beispiel #2
0
    def get(self):
        """Fetches transcript and toxicity from GCS to send to frontend.

        Args:
            'file_name': String holding audio file name.

        Returns:
            File_name: String of audio file.
            transcript: String holding entire audio file transcription.
            per_segment_toxicity: Array of objects holding toxicity probability
              per segment, start_time, end_time.
        """
        try:
            file_name = request.args['file_name']
            gcs_client = gcs_transcript_utils.authenticate_gcs()
            bucket_list = list(gcs_client.list_buckets())
            transcript_bucket = gcs_transcript_utils.find_bucket_with_prefix(
                bucket_list, 'transcript')
            transcript_per_segment = gcs_transcript_utils.get_gcs_object(
                gcs_client, transcript_bucket, file_name)
            if transcript_per_segment:
                transcript_json = transcript_per_segment['json_payload']
                transcript = gcs_transcript_utils.extract_full_transcript(
                    transcript_json)
                output_bucket = gcs_transcript_utils.find_bucket_with_prefix(
                    bucket_list, 'output-files')
                toxicity_path = f'toxicity-files/{file_name}'
                toxicity = gcs_transcript_utils.get_gcs_object(
                    gcs_client, output_bucket, toxicity_path)

                if toxicity:
                    toxicity.sort(key=lambda text: text['toxicity'],
                                  reverse=True)
                    return jsonify(file_name=file_name,
                                   transcript=transcript,
                                   per_segment_toxicity=toxicity)
                else:
                    raise NotFound

            else:
                raise NotFound

        except NotFound as e:
            logging.error(e)
            logging.error('Failed to retrieve analysis.')
            return jsonify(e.to_dict())
Beispiel #3
0
    def get(self):
        """Fetches list of files from GCS to send to front-end.

        Returns:
            Array of objects holding object name and type for audio files.
        """
        try:
            gcs_client = gcs_transcript_utils.authenticate_gcs()
            bucket_list = list(gcs_client.list_buckets())
            processed_audio_bucket = gcs_transcript_utils.find_bucket_with_prefix(
                bucket_list, 'processed-audio-files')
            files = gcs_transcript_utils.get_files(gcs_client,
                                                   processed_audio_bucket)
            return jsonify(files=files)

        except NotFound as e:
            logging.error(e)
            return jsonify(e.to_dict())