python-base--/tts/main.py

from flask import Flask, request, send_file, jsonify
from flask_cors import CORS
import dashscope
from dashscope.audio.tts_v2 import SpeechSynthesizer
import hashlib
import os

app = Flask(__name__)
CORS(app)  # 允许所有域名跨域访问

# 配置
CACHE_FOLDER = 'audio_cache'
MODEL = "cosyvoice-v1"
VOICE = "longxiaochun"

# 确保缓存文件夹存在
os.makedirs(CACHE_FOLDER, exist_ok=True)


def get_md5(text):
    """生成文本的MD5哈希值"""
    return hashlib.md5(text.encode('utf-8')).hexdigest()


def generate_audio(text, output_path):
    """使用DashScope生成语音文件"""
    synthesizer = SpeechSynthesizer(model=MODEL, voice=VOICE)
    audio = synthesizer.call(text)
    with open(output_path, 'wb') as f:
        f.write(audio)
    return synthesizer.get_last_request_id()


@app.route('/tts', methods=['POST', 'OPTIONS'])
def text_to_speech():
    if request.method == 'OPTIONS':
        return jsonify({}), 200

    # 获取请求数据
    data = request.get_json()
    if not data or 'text' not in data:
        return jsonify({'error': 'Missing text parameter'}), 400

    text = data['text']

    # 生成文件名
    filename = get_md5(text) + '.mp3'
    filepath = os.path.join(CACHE_FOLDER, filename)

    # 检查缓存
    if not os.path.exists(filepath):
        try:
            request_id = generate_audio(text, filepath)
            print(f'Generated new audio file with requestId: {request_id}')
        except Exception as e:
            return jsonify({'error': f'Failed to generate audio: {str(e)}'}), 500

    # 返回音频文件
    try:
        return send_file(
            filepath,
            mimetype='audio/mpeg',
            as_attachment=True,
            download_name=f'tts_{filename}'
        )
    except Exception as e:
        return jsonify({'error': f'Failed to send audio file: {str(e)}'}), 500


if __name__ == '__main__':
    app.run(host='0.0.0.0', port=15001, debug=True)