1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
|
import requests
import re
import json
import os
import subprocess
from faster_whisper import WhisperModel
class Bili:
def __init__(self, url):
self.url = url
self.headers = {
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 11_2_3) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/89.0.4389.90 Safari/537.36",
"Host": "www.bilibili.com",
"Accept-Encoding": "gzip, deflate, br",
"Accept": "*/*",
"Accept-Language": "zh-CN,zh;q=0.9",
}
self.output_dir = os.path.abspath("../../../codeInputOutput/output/biliVideo2Text")
os.makedirs(self.output_dir, exist_ok=True)
def start(self):
res = requests.get(self.url, headers=self.headers)
res.encoding = "utf-8"
result = re.findall("window.__playinfo__=(.*?)</script>", res.text)
if not result:
print("未找到 playinfo 数据")
return None
playinfo_json = json.loads(result[0])
audio_streams = playinfo_json.get("data", {}).get("dash", {}).get("audio", [])
if not audio_streams:
print("未找到音频流")
return None
print("找到以下音频流：")
for i, stream in enumerate(audio_streams):
print(f"{i+1}. ID: {stream['id']}, Bandwidth: {stream['bandwidth']}, URL: {stream['baseUrl']}")
return audio_streams
def download_audio(self, audio_stream, output_filename="audio.m4s"):
output_path = os.path.join(self.output_dir, output_filename)
audio_url = audio_stream["baseUrl"]
print(f"正在下载音频流: {audio_url}")
response = requests.get(audio_url, headers=self.headers, stream=True)
if response.status_code == 200:
with open(output_path, "wb") as f:
for chunk in response.iter_content(chunk_size=8192):
f.write(chunk)
print(f"音频下载完成: {output_path}")
return output_path
else:
print(f"下载失败，状态码: {response.status_code}")
return None
def convert_to_wav(self, input_file, output_wav="audio.wav"):
output_path = os.path.join(self.output_dir, output_wav)
try:
command = ["ffmpeg", "-i", input_file, "-acodec", "pcm_s16le", "-ar", "44100", output_path, "-y"]
subprocess.run(command, check=True)
print(f"转换完成: {output_path}")
return output_path
except subprocess.CalledProcessError as e:
print(f"转换失败: {e}")
return None
except FileNotFoundError:
print("未找到 ffmpeg，请确保已安装 ffmpeg 并添加到系统路径")
return None
def transcribe_audio(self, audio_file_path, output_txt="transcription.txt", model_size="medium", language="zh"):
output_txt_path = os.path.join(self.output_dir, output_txt)
if not os.path.exists(audio_file_path):
print(f"错误：文件 {audio_file_path} 不存在！")
return None
print(f"加载模型 '{model_size}' 到GPU...")
model = WhisperModel(model_size, device="cuda", compute_type="float16")
print("开始转录音频...")
segments, info = model.transcribe(audio_file_path, language=language)
with open(output_txt_path, "w", encoding="utf-8") as f:
print(f"检测到的语言：{info.language}，概率：{info.language_probability:.2f}")
for segment in segments:
line = f"[{segment.start:.2f}s -> {segment.end:.2f}s] {segment.text}"
print(line)
f.write(line + "\n")
print(f"转录完成，结果已保存到 {output_txt_path}")
return output_txt_path
if __name__ == '__main__':
bili = Bili("https://www.bilibili.com/video/BV1u2oiYqE1a/?spm_id_from=333.1007.tianma.2-1-4.click")
audio_streams = bili.start()
if audio_streams:
best_audio = max(audio_streams, key=lambda x: x["bandwidth"])
print(f"选择最高质量音频流: ID {best_audio['id']}, Bandwidth: {best_audio['bandwidth']}")
m4s_file = bili.download_audio(best_audio)
if m4s_file:
wav_file = bili.convert_to_wav(m4s_file)
if wav_file:
txt_file = bili.transcribe_audio(wav_file, language="zh")
if txt_file:
print(f"\n最终文本文件路径: {os.path.abspath(txt_file)}")
|