app.py

from flask import Flask, request, jsonify, Response
import requests
from google.cloud import speech, texttospeech
from pydub import AudioSegment
import io
import logging
from datetime import datetime  
from google.cloud import texttospeech

logging.basicConfig(level=logging.DEBUG)

processed_sessions = set()
app = Flask(__name__)

# Initialize Google Cloud Speech and Text-to-Speech clients
speech_client = speech.SpeechClient()
tts_client = texttospeech.TextToSpeechClient()

@app.route('/upload_audio', methods=['POST'])
def upload_audio(audio_file=None):
    if not audio_file:
        if 'audio' not in request.files:
            return jsonify({"message": "No audio file in request"}), 400
        audio_file = request.files['audio']

    audio_data = audio_file.read()
    audio_segment = AudioSegment.from_file(io.BytesIO(audio_data)).set_channels(1)

    if audio_segment.sample_width != 2:
        audio_segment = audio_segment.set_sample_width(2)
    if audio_segment.frame_rate != 16000:
        audio_segment = audio_segment.set_frame_rate(16000)

    audio_bytes = audio_segment.raw_data

    audio = speech.RecognitionAudio(content=audio_bytes)
    config = speech.RecognitionConfig(
        encoding=speech.RecognitionConfig.AudioEncoding.LINEAR16,
        sample_rate_hertz=16000,
        language_code="en-US",
        audio_channel_count=1,
        enable_automatic_punctuation=True
    )

    try:
        response = speech_client.recognize(config=config, audio=audio)
        transcript = response.results[0].alternatives[0].transcript
        return jsonify({"transcript": transcript})
    except Exception as e:
        return jsonify({"message": "Error processing audio", "error": str(e)}), 500

language_code_map = {
    'Spanish': 'es',  
    'French': 'fr',   
}

@app.route('/translate_text', methods=['POST'])
def translate_text():
    data = request.get_json(force=True)

    logging.debug(f"Received data for translation: {data}")

    text_to_translate = data.get('text')
    target_language = data.get('language').capitalize()

    if target_language not in language_code_map:
        return jsonify({"message": f"Unsupported target language: {target_language}"}), 400

    target_language_code = language_code_map[target_language]

    logging.debug(f"Target language code: {target_language_code}")

    deepL_api_key = ''  
    url = "https://api-free.deepl.com/v2/translate"
    params = {
        'auth_key': deepL_api_key,
        'text': text_to_translate,
        'target_lang': target_language_code
    }

    response = requests.post(url, data=params)
    logging.debug(f"DeepL API Response: {response.status_code}, {response.text}")

    if response.status_code == 200:
        translated_text = response.json()['translations'][0]['text']
        return jsonify({"translated_text": translated_text})
    else:
        error_msg = f"DeepL API error: {response.status_code} - {response.text}"
        logging.error(error_msg)
        return jsonify({"message": "Translation failed", "error": error_msg}), response.status_code

@app.route('/synthesize_speech', methods=['POST'])
def synthesize_speech():
    data = request.get_json()
    text = data.get('text')
    selected_language = data.get('language', 'es')

    tts_language_voice_map = {
        'es': ('es-ES', 'es-ES-Neural2-F'),  
        'fr': ('fr-FR', 'fr-FR-Neural2-B'),  
    }

    if not text:
        logging.error(f"Received empty 'text' for synthesis, data received: {data}")
        return jsonify({"message": "Error synthesizing speech: 'text' parameter is empty"}), 400
    else:
            logging.debug(f"Synthesizing speech for text: {text}")

    language_code, voice_name = tts_language_voice_map.get(selected_language, ('es-ES', 'es-ES-Neural2-F'))

    synthesis_input = texttospeech.SynthesisInput(text=text)
    voice = texttospeech.VoiceSelectionParams(
        language_code=language_code,
        name=voice_name,
        ssml_gender=texttospeech.SsmlVoiceGender.MALE
    )
    audio_config = texttospeech.AudioConfig(audio_encoding=texttospeech.AudioEncoding.MP3)

    try:
        response = tts_client.synthesize_speech(input=synthesis_input, voice=voice, audio_config=audio_config)
        if len(response.audio_content) == 0:
            logging.error("No audio content generated by the Text-to-Speech API.")
            return jsonify({"message": "Error synthesizing speech: No audio content"}), 500

        return Response(response.audio_content, mimetype='audio/mp3')
    except Exception as e:
        logging.error(f"Error in text-to-speech synthesis: {e}")
        return jsonify({"message": "Error in text-to-speech synthesis", "error": str(e)}), 500


@app.route('/process_audio', methods=['POST'])
def process_audio():
    
    logging.debug(f"Audio in request: {'audio' in request.files}")
    logging.debug(f"Language in request: {'language' in request.form}")

    if 'audio' not in request.files:
        logging.error("No audio file part in the request")
        return jsonify({"message": "Audio file or language not provided"}), 400
    audio_file = request.files['audio']

    selected_language = request.form['language']

    if 'language' not in request.form:
        logging.error("No language part in the request")
        return jsonify({"message": "Audio file or language not provided"}), 400
    selected_language = request.form['language']
    
    logging.debug(f"Processing audio for language: {selected_language}")
    logging.debug(f"Received audio file: {audio_file.filename}")
    logging.debug(f"Selected language: {selected_language}")
    

    try:
        audio_file.seek(0)  
        audio_data = audio_file.read()
        logging.debug(f"Size of audio data: {len(audio_data)} bytes")

        audio_segment = AudioSegment.from_file(io.BytesIO(audio_data)).set_channels(1)
        logging.debug(f"Audio Segment: Channels - {audio_segment.channels}, Frame Rate - {audio_segment.frame_rate}, Sample Width - {audio_segment.sample_width}")

        if audio_segment.sample_width != 2:
            audio_segment = audio_segment.set_sample_width(2)
        if audio_segment.frame_rate != 16000:
            audio_segment = audio_segment.set_frame_rate(16000)

        logging.debug(f"Audio Segment: Channels - {audio_segment.channels}, Frame Rate - {audio_segment.frame_rate}, Sample Width - {audio_segment.sample_width}")

        audio_bytes = audio_segment.raw_data
        audio = speech.RecognitionAudio(content=audio_bytes)
        config = speech.RecognitionConfig(
            encoding=speech.RecognitionConfig.AudioEncoding.LINEAR16,
            sample_rate_hertz=16000,
            language_code="en-US",
            audio_channel_count=1,
            enable_automatic_punctuation=True
        )

        response = speech_client.recognize(config=config, audio=audio)

        if not response.results:
            logging.error("No transcription results.")
            return jsonify({"message": "Error in transcription: No results"}), 500

        transcript = response.results[0].alternatives[0].transcript
        logging.debug(f"Transcript (English): {transcript}")
        
        transcript = response.results[0].alternatives[0].transcript
        logging.debug(f"Transcript (English): {transcript}")

        translation_response = requests.post('http://localhost:5000/translate_text', json={'text': transcript, 'language': selected_language})
        if translation_response.status_code != 200:
            return jsonify({"message": "Error translating text"}), translation_response.status_code
        translated_text = translation_response.json()['translated_text']

        logging.debug(f"Translated text (Target Language - {selected_language}): {translated_text}")

        speech_response = requests.post('http://localhost:5000/synthesize_speech', json={'text': translated_text, 'language': selected_language})
        if speech_response.status_code != 200:
            return jsonify({"message": "Error synthesizing speech"}), speech_response.status_code
        audio_content = speech_response.content

        logging.debug(f"Type of audio content: {type(audio_content)}")
        logging.debug(f"Length of audio content: {len(audio_content)}")
        logging.debug(f"Transcription and translation completed successfully.")

        return jsonify({"transcript": transcript, "translated_text": translated_text})

    except Exception as e:
        logging.error(f"Error in processing audio: {e}")
        return jsonify({"message": "Error in processing audio", "error": str(e)}), 500


@app.route('/')
def index():
    return app.send_static_file('index.html')


if __name__ == '__main__':
    app.run(debug=True)