K2CrO4/NeoBot
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

chore: 整理配置与功能，优化B站解析流程

Browse Source 
1. 从.gitignore移除config.example.toml并新增该示例配置文件
2. 新增项目规则文档说明开发环境要求
3. 修复config加载时的编码缺失问题
4. 重写bili_login.py，优化扫码登录流程与凭证输出
5. 重构B站解析器：简化下载逻辑，改用bilibili_api内置下载，优化音视频合并流程
This commit is contained in:
K2CRO4
2026-05-12 12:38:34 +08:00
parent dcfb5d4892
commit 2cb55992f9
6 changed files with 254 additions and 235 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
src/neobot/core/config_loader.py
View File

@@ -216,8 +216,8 @@ class Config:
self.logger.error(f"示例配置文件 {example_path} 不存在，无法生成配置")
raise ConfigNotFoundError(message=f"示例配置文件 {example_path} 不存在")
content = example_path.read_text()
self.path.write_text(content)
content = example_path.read_text(encoding='utf-8')
self.path.write_text(content, encoding='utf-8')
# 通过属性访问配置
@property

316
src/neobot/plugins/web_parser/parsers/bili.py
View File

@@ -2,6 +2,7 @@
import asyncio
import re
import os
import shutil
import subprocess
import tempfile
from typing import Optional, Dict, Any, List, Union
@@ -11,17 +12,17 @@ from neobot.models import MessageEvent, MessageSegment
from ..base import BaseParser
from ..utils import format_duration
from bilibili_api import video, select_client, Credential
from bilibili_api import video, select_client, Credential, get_client, HEADERS
from bilibili_api.exceptions import ResponseCodeException
from neobot.core.config_loader import global_config
from neobot.core.services.local_file_server import download_to_local
from neobot.core.services.local_file_server import download_to_local, get_local_file_server
try:
import aiohttp
AIOHTTP_AVAILABLE = True
except ImportError:
AIOHTTP_AVAILABLE = False
logger.warning("[B站解析器] aiohttp 未安装，音视频合并功能将不可用")
logger.warning("[B站解析器] aiohttp 未安装，备用解析功能将不可用")
# bilibili_api-python 可用性标志
BILI_API_AVAILABLE = True
@@ -284,263 +285,130 @@ class BiliParser(BaseParser):
try:
credential = self._get_credential()
v = video.Video(bvid=bvid, credential=credential)
# 先获取视频信息以获取 cid
info = await v.get_info()
cid = info.get('cid', 0)
if not cid:
return None
# 获取下载链接数据，使用 html5=True 获取网页格式（通常包含合并的音视频）
download_url_data = await v.get_download_url(cid=cid, html5=True)
# 使用 VideoDownloadURLDataDetecter 解析数据
download_url_data = await v.get_download_url(cid=cid)
detecter = video.VideoDownloadURLDataDetecter(data=download_url_data)
# 尝试获取 MP4 格式的合并流（包含音视频）
streams = detecter.detect_best_streams()
# 如果没有获取到流，尝试其他格式
if not streams:
logger.warning(f"[{self.name}] 无法获取 html5 格式，尝试获取其他格式...")
download_url_data = await v.get_download_url(cid=cid, html5=False)
detecter = video.VideoDownloadURLDataDetecter(data=download_url_data)
if detecter.check_flv_mp4_stream():
streams = detecter.detect_best_streams()
if streams:
# 获取视频直链
video_direct_url = streams[0].url
# 检查是否是分离的 m4s 流（可能没有声音）
is_m4s_stream = '.m4s' in video_direct_url
if is_m4s_stream:
logger.warning(f"[{self.name}] 检测到分离的 m4s 流，B站 API 返回的 m4s 流通常是分离的视频和音频，需要客户端合并才能有声音")
logger.info(f"[{self.name}] 建议: 使用支持合并 m4s 流的下载工具（如 ffmpeg）合并视频和音频")
logger.info(f"[{self.name}] 获取到视频直链，开始下载到本地...")
# B站下载需要 Referer 和 User-Agent
headers = {
"Referer": "https://www.bilibili.com",
"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36"
}
# 调试：打印 download_url_data 结构
logger.debug(f"[{self.name}] download_url_data 类型: {type(download_url_data)}")
if isinstance(download_url_data, dict):
logger.debug(f"[{self.name}] download_url_data keys: {list(download_url_data.keys())}")
# 如果是 m4s 流且 ffmpeg 可用，先保存 download_url_data 供合并使用
if is_m4s_stream and FFMPEG_AVAILABLE and AIOHTTP_AVAILABLE:
local_url = await self._download_and_merge_m4s(video_direct_url, headers, bvid, download_url_data)
else:
# 使用本地文件服务器下载
local_url = await download_to_local(video_direct_url, timeout=120, headers=headers)
if local_url:
logger.success(f"[{self.name}] 视频已下载到本地: {local_url}")
return local_url
else:
logger.error(f"[{self.name}] 下载到本地失败")
if not streams:
return None
logger.info(f"[{self.name}] 检测到合并音视频流，直接下载...")
return await download_to_local(streams[0].url, timeout=120, headers=HEADERS)
if not FFMPEG_AVAILABLE:
logger.warning(f"[{self.name}] ffmpeg 不可用，无法合并音视频，仅下载视频流（无声音）")
streams = detecter.detect_best_streams()
if streams and streams[0]:
return await download_to_local(streams[0].url, timeout=120, headers=HEADERS)
return None
return await self._download_and_merge_m4s(detecter, bvid)
except (aiohttp.ClientError, asyncio.TimeoutError, ValueError, ResponseCodeException) as e:
logger.error(f"[{self.name}] 获取视频直链失败: {e}")
return None
async def _download_and_merge_m4s(self, video_url: str, headers: Dict[str, str], bvid: str, download_url_data: Dict) -> Optional[str]:
async def _download_and_merge_m4s(self, detecter: video.VideoDownloadURLDataDetecter, bvid: str) -> Optional[str]:
"""
下载并合并 m4s 视频和音频流
Args:
video_url (str): 视频流 URL
headers (Dict[str, str]): 请求头
detecter (VideoDownloadURLDataDetecter): 视频流检测器
bvid (str): BV号
download_url_data (Dict): 下载 URL 数据
Returns:
Optional[str]: 合并后的本地视频 URL，如果失败则返回None
"""
if not FFMPEG_AVAILABLE:
logger.warning("[B站解析器] ffmpeg 不可用，无法合并音视频")
return None
if not AIOHTTP_AVAILABLE:
logger.warning("[B站解析器] aiohttp 不可用，无法合并音视频")
streams = detecter.detect_best_streams()
if not streams or not streams[0]:
logger.error(f"[{self.name}] 未检测到可用的视频流")
return None
video_stream = streams[0]
audio_stream = streams[1] if len(streams) > 1 else None
video_file = None
audio_file = None
merged_file = None
try:
logger.info(f"[{self.name}] 开始下载并合并 m4s 音视频...")
# 创建共享的 ClientSession 用于下载
async with aiohttp.ClientSession() as session:
# 下载视频流
video_file = tempfile.NamedTemporaryFile(suffix='.m4s', delete=False)
video_file.close()
async with session.get(video_url, headers=headers, timeout=60) as response:
if response.status != 200:
logger.error(f"[{self.name}] 下载视频流失败: HTTP {response.status}")
return None
with open(video_file.name, 'wb') as f:
while True:
chunk = await response.content.read(8192)
if not chunk:
break
f.write(chunk)
logger.info(f"[{self.name}] 视频流下载完成: {video_file.name}")
# 从 download_url_data 中提取音频 URL
# B站的 dash 格式包含视频和音频流
audio_url = None
if isinstance(download_url_data, dict):
# 尝试 dash 格式（推荐）
if 'dash' in download_url_data and isinstance(download_url_data['dash'], dict):
dash = download_url_data['dash']
if 'audio' in dash and isinstance(dash['audio'], list) and len(dash['audio']) > 0:
# 获取第一个音频流
audio_item = dash['audio'][0]
audio_url = audio_item.get('baseUrl') or audio_item.get('url') or audio_item.get('backupUrl')
logger.debug(f"[{self.name}] 从 dash.audio 提取音频 URL: {audio_url is not None}")
elif 'audio' in dash and isinstance(dash['audio'], dict):
audio_url = dash['audio'].get('baseUrl') or dash['audio'].get('url')
logger.debug(f"[{self.name}] 从 dash.audio (dict) 提取音频 URL: {audio_url is not None}")
# 尝试 durl 格式（非分段流）
elif 'durl' in download_url_data:
if isinstance(download_url_data['durl'], list) and len(download_url_data['durl']) > 0:
main_url = download_url_data['durl'][0].get('url') or download_url_data['durl'][0].get('baseUrl')
if main_url:
video_url = main_url
logger.debug(f"[{self.name}] 使用 durl 主 URL: {video_url}")
if not audio_url and not video_url.startswith('http'):
logger.warning(f"[{self.name}] 无法从 download_url_data 中提取音频 URL")
logger.debug(f"[{self.name}] download_url_data 结构: {download_url_data}")
os.unlink(video_file.name)
return None
# 下载音频流
audio_file = tempfile.NamedTemporaryFile(suffix='.m4s', delete=False)
audio_file.close()
async with session.get(audio_url, headers=headers, timeout=60) as response:
if response.status != 200:
logger.error(f"[{self.name}] 下载音频流失败: HTTP {response.status}")
os.unlink(video_file.name)
return None
with open(audio_file.name, 'wb') as f:
while True:
chunk = await response.content.read(8192)
if not chunk:
break
f.write(chunk)
logger.info(f"[{self.name}] 音频流下载完成: {audio_file.name}")
# 使用 ffmpeg 合并视频和音频
video_file = tempfile.NamedTemporaryFile(suffix='.m4s', delete=False)
video_file.close()
dwn_id = await get_client().download_create(video_stream.url, HEADERS)
tot = get_client().download_content_length(dwn_id)
with open(video_file.name, 'wb') as f:
while True:
chunk = await get_client().download_chunk(dwn_id)
f.write(chunk)
if f.tell() >= tot:
break
await get_client().download_close(cnt=dwn_id)
if not audio_stream:
logger.warning(f"[{self.name}] 未检测到音频流，仅返回视频")
return await download_to_local(video_stream.url, timeout=120, headers=HEADERS)
audio_file = tempfile.NamedTemporaryFile(suffix='.m4s', delete=False)
audio_file.close()
dwn_id = await get_client().download_create(audio_stream.url, HEADERS)
tot = get_client().download_content_length(dwn_id)
with open(audio_file.name, 'wb') as f:
while True:
chunk = await get_client().download_chunk(dwn_id)
f.write(chunk)
if f.tell() >= tot:
break
await get_client().download_close(cnt=dwn_id)
merged_file = tempfile.NamedTemporaryFile(suffix='.mp4', delete=False)
merged_file.close()
# ffmpeg命令：使用ffmpeg -i多次输入，然后合并
# 先转换视频流（移除音频），然后添加音频流
ffmpeg_cmd = [
'ffmpeg', '-y', '-i', video_file.name, '-i', audio_file.name,
'-c:v', 'libx264', '-c:a', 'aac',
'-shortest', merged_file.name
'ffmpeg', '-y',
'-i', video_file.name,
'-i', audio_file.name,
'-c:v', 'copy',
'-c:a', 'copy',
merged_file.name
]
logger.debug(f"[{self.name}] ffmpeg命令: {' '.join(ffmpeg_cmd)}")
result = subprocess.run(ffmpeg_cmd, capture_output=True, text=True)
# 详细记录ffmpeg输出
if result.stdout:
logger.debug(f"[{self.name}] ffmpeg stdout: {result.stdout}")
if result.stderr:
logger.debug(f"[{self.name}] ffmpeg stderr: {result.stderr}")
if result.returncode != 0:
logger.error(f"[{self.name}] ffmpeg 合并失败: {result.stderr}")
os.unlink(video_file.name)
os.unlink(audio_file.name)
return None
# 验证输出文件
merged_size = os.path.getsize(merged_file.name)
logger.debug(f"[{self.name}] 合并文件大小: {merged_size} bytes")
if merged_size == 0:
logger.error(f"[{self.name}] ffmpeg生成了空文件，命令可能有问题")
logger.error(f"[{self.name}] ffmpeg命令: {' '.join(ffmpeg_cmd)}")
if result.stderr:
logger.error(f"[{self.name}] ffmpeg错误输出: {result.stderr}")
os.unlink(video_file.name)
os.unlink(audio_file.name)
return None
logger.info(f"[{self.name}] 音视频合并成功: {merged_file.name} ({merged_size} bytes)")
# 上传合并后的文件到本地文件服务器
from neobot.core.services.local_file_server import get_local_file_server
subprocess.run(ffmpeg_cmd, capture_output=True, check=True)
server = get_local_file_server()
if server:
try:
file_id = server._generate_file_id(f'file://{merged_file.name}')
dest_path = server.download_dir / file_id
# 获取合并文件大小
merged_size = os.path.getsize(merged_file.name)
logger.debug(f"[{self.name}] 合并文件大小: {merged_size} bytes")
if merged_size == 0:
logger.error(f"[{self.name}] 合并文件为空，ffmpeg可能失败了")
merged_url = None
else:
# 复制本地文件到服务器目录
import shutil
shutil.copy2(merged_file.name, dest_path)
server.file_map[file_id] = dest_path
# 验证复制后的文件
if dest_path.exists():
dest_size = dest_path.stat().st_size
logger.debug(f"[{self.name}] 复制后文件大小: {dest_size} bytes")
if dest_size == merged_size:
merged_url = f"http://127.0.0.1:{server.port}/download?id={file_id}"
logger.success(f"[{self.name}] 合并后的视频已上传到本地服务器: {merged_url}")
else:
logger.error(f"[{self.name}] 文件大小不匹配: 原始 {merged_size} vs 复制 {dest_size}")
merged_url = None
else:
logger.error(f"[{self.name}] 文件复制失败: {dest_path} 不存在")
merged_url = None
except Exception as e:
logger.error(f"[{self.name}] 上传合并文件失败: {e}")
merged_url = None
else:
merged_url = None
# 清理临时文件
try:
os.unlink(video_file.name)
os.unlink(audio_file.name)
os.unlink(merged_file.name)
except (OSError, PermissionError) as e:
logger.warning(f"[{self.name}] 清理临时文件失败: {e}")
if merged_url:
logger.success(f"[{self.name}] 合并后的视频已上传到本地服务器: {merged_url}")
return merged_url
except (aiohttp.ClientError, asyncio.TimeoutError, ValueError, OSError, subprocess.CalledProcessError) as e:
file_id = f"bili_{bvid}"
dest_path = server.download_dir / file_id
shutil.copy2(merged_file.name, str(dest_path))
server.file_map[file_id] = dest_path
logger.success(f"[{self.name}] 合并后的视频已注册到本地文件服务器")
return f"http://{server.host}:{server.port}/download?id={file_id}"
logger.warning(f"[{self.name}] 本地文件服务器不可用")
return None
except Exception as e:
logger.error(f"[{self.name}] 合并音视频失败: {e}")
return None
return None
finally:
for f in [video_file, audio_file, merged_file]:
if f and os.path.exists(f.name):
try:
os.unlink(f.name)
except OSError:
pass
async def format_response(self, event: MessageEvent, data: Dict[str, Any]) -> List[Any]:
"""