88 lines
2.7 KiB
Python
88 lines
2.7 KiB
Python
|
from flask import Flask, request, jsonify, send_file
|
||
|
|
import subprocess
|
||
|
|
import os
|
||
|
|
import logging
|
||
|
|
import uuid
|
||
|
|
from dotenv import load_dotenv
|
||
|
|
|
||
|
|
# .envファイルを読み込む
|
||
|
|
load_dotenv()
|
||
|
|
|
||
|
|
app = Flask(__name__)
|
||
|
|
|
||
|
|
# 環境変数から基本設定を取得
|
||
|
|
VOICEPEAK_PATH = os.getenv('VOICEPEAK_PATH')
|
||
|
|
PROJECT_PATH = os.getenv('PROJECT_PATH')
|
||
|
|
|
||
|
|
def _build_command_args(args):
|
||
|
|
return [
|
||
|
|
args['voicepeak_path'],
|
||
|
|
"-s", args['script'],
|
||
|
|
"-n", args['narrator'],
|
||
|
|
"-o", args['output_path'],
|
||
|
|
"-e", args['emotions'],
|
||
|
|
"--pitch", str(args['pitch']),
|
||
|
|
]
|
||
|
|
|
||
|
|
def _synthesize_with_voicepeak(script, narrator, emotions, pitch, output_path, retries=3):
|
||
|
|
args = {
|
||
|
|
'voicepeak_path': VOICEPEAK_PATH,
|
||
|
|
'script': script,
|
||
|
|
'narrator': narrator,
|
||
|
|
'output_path': output_path,
|
||
|
|
'emotions': emotions,
|
||
|
|
'pitch': pitch,
|
||
|
|
}
|
||
|
|
command_args = _build_command_args(args)
|
||
|
|
|
||
|
|
for attempt in range(retries):
|
||
|
|
try:
|
||
|
|
subprocess.run(command_args, check=True)
|
||
|
|
if os.path.exists(output_path):
|
||
|
|
return output_path
|
||
|
|
except subprocess.CalledProcessError as e:
|
||
|
|
error_message = f"VoicePeak synthesis failed on attempt {attempt + 1}: {e}"
|
||
|
|
logging.error(error_message)
|
||
|
|
if attempt < retries - 1:
|
||
|
|
logging.info(f"Retrying... ({attempt + 2}/{retries})")
|
||
|
|
else:
|
||
|
|
raise RuntimeError(error_message) from e
|
||
|
|
|
||
|
|
error_message = f"Expected audio file not found after {retries} attempts: {output_path}"
|
||
|
|
logging.error(error_message)
|
||
|
|
raise FileNotFoundError(error_message)
|
||
|
|
|
||
|
|
@app.route('/generate_voice', methods=['POST'])
|
||
|
|
def generate_voice():
|
||
|
|
try:
|
||
|
|
data = request.json
|
||
|
|
text = data.get('text')
|
||
|
|
narrator = data.get('narrator', '1')
|
||
|
|
emotion_happy = data.get('emotion_happy', '0')
|
||
|
|
emotion_sad = data.get('emotion_sad', '0')
|
||
|
|
emotion_angry = data.get('emotion_angry', '0')
|
||
|
|
emotion_fun = data.get('emotion_fun', '0')
|
||
|
|
pitch = data.get('pitch', '1')
|
||
|
|
|
||
|
|
if not text:
|
||
|
|
return jsonify({'error': 'Text is required'}), 400
|
||
|
|
|
||
|
|
emotions = f"happy={emotion_happy},sad={emotion_sad},angry={emotion_angry},fun={emotion_fun}"
|
||
|
|
|
||
|
|
unique_id = str(uuid.uuid4())
|
||
|
|
output_file = os.path.join(PROJECT_PATH, f'{unique_id}.wav')
|
||
|
|
|
||
|
|
_synthesize_with_voicepeak(text, narrator, emotions, pitch, output_file)
|
||
|
|
|
||
|
|
response = send_file(output_file, mimetype='audio/wav')
|
||
|
|
|
||
|
|
os.remove(output_file)
|
||
|
|
|
||
|
|
return response
|
||
|
|
|
||
|
|
except Exception as e:
|
||
|
|
return jsonify({'error': str(e)}), 500
|
||
|
|
|
||
|
|
if __name__ == '__main__':
|
||
|
|
app.run(host='0.0.0.0', port=5000)
|