优化文字识别

docker-compose.yaml

@@ -7,6 +7,10 @@ services:
         restart: always
         container_name: douyin
         image: registry.cn-hangzhou.aliyuncs.com/yinzhou_docker_hub/douyin:yz1
+        #环境变量需要替换成自己的
+        environment:
+            DASHSCOPE_API_KEY: sk-063b48fffb914d558ddcddc5166fd34d
+
         ports:
             - "1314:1314"
         depends_on:

main.py

@@ -2,6 +2,7 @@ from fastapi import FastAPI, Query
 import uvicorn
 import requests
 from utils.response import success_response, error_400_response, error_503_response
+from utils.music_analysis import music_analysis
 
 app = FastAPI(title="douyin文档", swagger_ui_parameters={
     "defaultModelsExpandDepth": -1
@@ -11,6 +12,36 @@ DEFAULT_PC_VIDEO_URL = 'https://v.douyin.com/-NvlqBdIJo4/'
 DEFAULT_mobile_VIDEO_URL = 'https://v.douyin.com/BCfMrTFPYGQ/'
 
 
+@app.get('/douyin_video', tags=["抖音"], summary="返回视屏信息,最全的接口了")
+def douyin_video(video_url: str = Query(DEFAULT_mobile_VIDEO_URL, min_length=10)):
+    print(video_url)
+    if not video_url:
+        return {"code": 400, "message": "An error occurred.", "data": "请指定video_url"}
+
+    # 获取get请求参数
+    url = "http://douyin_tiktok_download_api:9579/api/hybrid/video_data"
+
+
+    querystring = {"url": video_url, "minimal": "true"}
+    response = requests.request("GET", url, params=querystring)
+    print(response.text)
+    print(response.json()['data']['music']['play_url'])
+
+
+    try:
+        if response.json()['code'] == 200:
+            video_url=response.json()['data']['video_data']['wm_video_url']
+            video_music=response.json()['data']['music']['play_url']['url_list'][0]
+            content=music_analysis(video_music)
+            return success_response({"video_url": video_url,"video_music": video_music,"content": content
+                                     })
+
+    except Exception as e:
+        print(e)
+        return error_400_response({"data": "解析字段失败."})
+
+    return error_503_response({"data": "抖音风控稍后请求."})
+
 @app.get('/douyin_content', tags=["抖音"], summary="手机和pc获取文案")
 def douyin_content(video_url: str = Query(DEFAULT_mobile_VIDEO_URL, min_length=10)):
     print(video_url)
@@ -69,12 +100,13 @@ def douyin_phone(video_url: str = Query(DEFAULT_mobile_VIDEO_URL, min_length=10)
     # 获取get请求参数
     url = "http://douyin_tiktok_download_api:9579/api/hybrid/video_data"
 
+
     querystring = {"url": video_url, "minimal": "true"}
     response = requests.request("GET", url, params=querystring)
-    print(response.json()['data']['author']['avatar_thumb'])
+    print(response.text)
     try:
         if response.json()['code'] == 200:
-            return success_response({"phone_url": response.json()['data']['author']['avatar_thumb']['url_list']
+            return success_response({"phone_url": response.json()['data']['video_data']['wm_video_url']
                                      })
 
     except Exception as e:

utils/music_analysis.py

@@ -0,0 +1,40 @@
+from http import HTTPStatus
+from dashscope.audio.asr import Transcription
+import dashscope
+import requests
+
+
+
+def music_analysis(music_url):
+    transcribe_response = Transcription.async_call(
+        model='paraformer-v2',
+        file_urls=[music_url],
+        language_hints=['zh', 'en']  # “language_hints”只支持paraformer-v2模型
+    )
+
+    while True:
+        if transcribe_response.output.task_status == 'SUCCEEDED' or transcribe_response.output.task_status == 'FAILED':
+            break
+        transcribe_response = Transcription.fetch(task=transcribe_response.output.task_id)
+
+    if transcribe_response.status_code == HTTPStatus.OK:
+        url=transcribe_response.output['results'][0]['transcription_url']
+        print(url)
+
+        # 发送GET请求
+        response = requests.get(url)
+        text = ''
+
+        # 验证响应状态
+        if response.status_code == 200:
+            # 解析JSON数据
+            data = response.json()
+            for transcripts in data['transcripts']:
+                text += transcripts['text']
+
+        else:
+            print(f"请求失败，状态码：{response.status_code}")
+    return text
+
+if __name__ == '__main__':
+    music_analysis('https://lf26-music-east.douyinstatic.com/obj/ies-music-hj/7494207652008839996.mp3')