318 lines
12 KiB
Python
318 lines
12 KiB
Python
|
"""
|
|||
|
|
使用 Crawl4AI 爬取人民网新闻
|
|||
|
|
"""
|
|||
|
|
from typing import List, Optional
|
|||
|
|
from loguru import logger
|
|||
|
|
import re
|
|||
|
|
from datetime import datetime
|
|||
|
|
from crawler.BaseCrawler import NewsItem
|
|||
|
|
import asyncio
|
|||
|
|
|
|||
|
|
|
|||
|
|
class PeopleNetCrewer:
|
|||
|
|
"""使用 Crawl4AI 爬取人民网新闻"""
|
|||
|
|
|
|||
|
|
def __init__(self, chrome_path: Optional[str] = None):
|
|||
|
|
"""
|
|||
|
|
初始化人民网爬虫
|
|||
|
|
|
|||
|
|
Args:
|
|||
|
|
chrome_path: Chrome 浏览器可执行文件路径，如果为 None 则使用系统默认路径
|
|||
|
|
"""
|
|||
|
|
self.base_url = "http://www.people.com.cn"
|
|||
|
|
self.chrome_path = chrome_path
|
|||
|
|
self.crawler = None
|
|||
|
|
logger.info(f"初始化人民网爬虫 (Crawl4AI)")
|
|||
|
|
|
|||
|
|
async def _get_crawler(self):
|
|||
|
|
"""获取或创建爬虫实例"""
|
|||
|
|
if self.crawler is None:
|
|||
|
|
try:
|
|||
|
|
from crawl4ai import AsyncWebCrawler
|
|||
|
|
except ImportError:
|
|||
|
|
logger.error("请先安装 crawl4ai: pip install crawl4ai")
|
|||
|
|
raise ImportError("crawl4ai 未安装")
|
|||
|
|
|
|||
|
|
# 配置浏览器选项，使用本地 Chrome
|
|||
|
|
browser_config = {
|
|||
|
|
"headless": True,
|
|||
|
|
"verbose": False,
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
# 如果指定了 Chrome 路径，使用指定路径
|
|||
|
|
if self.chrome_path:
|
|||
|
|
browser_config["executable_path"] = self.chrome_path
|
|||
|
|
logger.info(f"使用指定的 Chrome 路径: {self.chrome_path}")
|
|||
|
|
else:
|
|||
|
|
# 使用本地安装的 Chrome（通过 channel 参数）
|
|||
|
|
browser_config["channel"] = "chrome"
|
|||
|
|
logger.info("使用本地安装的 Chrome 浏览器")
|
|||
|
|
|
|||
|
|
try:
|
|||
|
|
self.crawler = AsyncWebCrawler(
|
|||
|
|
browser_type="chromium",
|
|||
|
|
browser_config=browser_config
|
|||
|
|
)
|
|||
|
|
except Exception as e:
|
|||
|
|
logger.error(f"创建爬虫实例失败: {str(e)}")
|
|||
|
|
# 如果使用 channel 失败，尝试使用 executable_path
|
|||
|
|
if not self.chrome_path and "channel" in browser_config:
|
|||
|
|
logger.warning("使用 channel 参数失败，尝试其他方式...")
|
|||
|
|
# 尝试常见的 Chrome 路径
|
|||
|
|
import os
|
|||
|
|
common_paths = [
|
|||
|
|
r"C:\Program Files\Google\Chrome\Application\chrome.exe",
|
|||
|
|
r"C:\Program Files (x86)\Google\Chrome\Application\chrome.exe",
|
|||
|
|
os.path.expanduser(r"~\AppData\Local\Google\Chrome\Application\chrome.exe"),
|
|||
|
|
]
|
|||
|
|
for path in common_paths:
|
|||
|
|
if os.path.exists(path):
|
|||
|
|
browser_config.pop("channel", None)
|
|||
|
|
browser_config["executable_path"] = path
|
|||
|
|
logger.info(f"找到 Chrome 路径: {path}")
|
|||
|
|
try:
|
|||
|
|
self.crawler = AsyncWebCrawler(
|
|||
|
|
browser_type="chromium",
|
|||
|
|
browser_config=browser_config
|
|||
|
|
)
|
|||
|
|
break
|
|||
|
|
except Exception:
|
|||
|
|
continue
|
|||
|
|
else:
|
|||
|
|
raise
|
|||
|
|
else:
|
|||
|
|
raise
|
|||
|
|
return self.crawler
|
|||
|
|
|
|||
|
|
async def crawl(self, category: str = "politics", limit: int = 20) -> List[NewsItem]:
|
|||
|
|
"""
|
|||
|
|
爬取人民网新闻列表
|
|||
|
|
|
|||
|
|
Args:
|
|||
|
|
category: 新闻分类（politics-政治, society-社会, world-国际等）
|
|||
|
|
limit: 爬取数量限制
|
|||
|
|
|
|||
|
|
Returns:
|
|||
|
|
新闻列表
|
|||
|
|
"""
|
|||
|
|
news_list = []
|
|||
|
|
|
|||
|
|
try:
|
|||
|
|
crawler = await self._get_crawler()
|
|||
|
|
|
|||
|
|
# 构建列表页URL
|
|||
|
|
list_url = f"{self.base_url}/{category}/index.html"
|
|||
|
|
logger.info(f"开始爬取人民网新闻: {list_url}")
|
|||
|
|
|
|||
|
|
# 使用 Crawl4AI 爬取页面
|
|||
|
|
result = await crawler.arun(url=list_url)
|
|||
|
|
|
|||
|
|
if not result.success:
|
|||
|
|
logger.error(f"爬取失败: {list_url}")
|
|||
|
|
return news_list
|
|||
|
|
|
|||
|
|
html_content = result.html
|
|||
|
|
if not html_content:
|
|||
|
|
logger.warning("未获取到页面内容")
|
|||
|
|
return news_list
|
|||
|
|
|
|||
|
|
# 解析新闻列表
|
|||
|
|
from bs4 import BeautifulSoup
|
|||
|
|
soup = BeautifulSoup(html_content, 'lxml')
|
|||
|
|
|
|||
|
|
# 根据人民网实际结构调整选择器
|
|||
|
|
# 这里使用通用的新闻列表选择器，实际使用时需要根据网站结构调整
|
|||
|
|
news_items = soup.select('.news-list .news-item, .list-item, .news-item, article')
|
|||
|
|
|
|||
|
|
if not news_items:
|
|||
|
|
# 尝试其他可能的选择器
|
|||
|
|
news_items = soup.select('a[href*="/n1/"], a[href*="/n2/"], a[href*="/n3/"]')
|
|||
|
|
|
|||
|
|
logger.info(f"找到 {len(news_items)} 个新闻链接")
|
|||
|
|
|
|||
|
|
# 提取新闻链接并爬取详情
|
|||
|
|
processed_urls = set()
|
|||
|
|
for item in news_items[:limit * 2]: # 多取一些，因为有些链接可能无效
|
|||
|
|
try:
|
|||
|
|
# 提取链接
|
|||
|
|
if item.name == 'a':
|
|||
|
|
link_tag = item
|
|||
|
|
else:
|
|||
|
|
link_tag = item.select_one('a')
|
|||
|
|
|
|||
|
|
if not link_tag:
|
|||
|
|
continue
|
|||
|
|
|
|||
|
|
news_url = link_tag.get('href', '')
|
|||
|
|
if not news_url:
|
|||
|
|
continue
|
|||
|
|
|
|||
|
|
# 处理相对路径
|
|||
|
|
if not news_url.startswith('http'):
|
|||
|
|
if news_url.startswith('/'):
|
|||
|
|
news_url = self.base_url + news_url
|
|||
|
|
else:
|
|||
|
|
news_url = f"{self.base_url}/{news_url}"
|
|||
|
|
|
|||
|
|
# 跳过重复链接
|
|||
|
|
if news_url in processed_urls:
|
|||
|
|
continue
|
|||
|
|
processed_urls.add(news_url)
|
|||
|
|
|
|||
|
|
# 解析新闻详情
|
|||
|
|
news = await self.parse_news_detail(news_url)
|
|||
|
|
if news:
|
|||
|
|
news_list.append(news)
|
|||
|
|
logger.info(f"成功爬取新闻: {news.title}")
|
|||
|
|
|
|||
|
|
if len(news_list) >= limit:
|
|||
|
|
break
|
|||
|
|
|
|||
|
|
except Exception as e:
|
|||
|
|
logger.error(f"处理新闻项失败: {str(e)}")
|
|||
|
|
continue
|
|||
|
|
|
|||
|
|
logger.info(f"爬取完成，共获取 {len(news_list)} 条新闻")
|
|||
|
|
|
|||
|
|
except Exception as e:
|
|||
|
|
logger.error(f"爬取新闻列表失败: {str(e)}")
|
|||
|
|
|
|||
|
|
return news_list
|
|||
|
|
|
|||
|
|
async def parse_news_detail(self, url: str) -> Optional[NewsItem]:
|
|||
|
|
"""
|
|||
|
|
解析人民网新闻详情
|
|||
|
|
|
|||
|
|
Args:
|
|||
|
|
url: 新闻详情页URL
|
|||
|
|
|
|||
|
|
Returns:
|
|||
|
|
新闻对象
|
|||
|
|
"""
|
|||
|
|
try:
|
|||
|
|
crawler = await self._get_crawler()
|
|||
|
|
|
|||
|
|
# 使用 Crawl4AI 爬取详情页
|
|||
|
|
result = await crawler.arun(url=url)
|
|||
|
|
|
|||
|
|
if not result.success:
|
|||
|
|
logger.warning(f"爬取详情页失败: {url}")
|
|||
|
|
return None
|
|||
|
|
|
|||
|
|
html_content = result.html
|
|||
|
|
if not html_content:
|
|||
|
|
return None
|
|||
|
|
|
|||
|
|
from bs4 import BeautifulSoup
|
|||
|
|
soup = BeautifulSoup(html_content, 'lxml')
|
|||
|
|
|
|||
|
|
# 提取标题
|
|||
|
|
title_tag = soup.select_one('h1, .title, .article-title, .p_title')
|
|||
|
|
title = title_tag.get_text(strip=True) if title_tag else "未知标题"
|
|||
|
|
|
|||
|
|
if title == "未知标题" or not title:
|
|||
|
|
logger.warning(f"无法提取标题: {url}")
|
|||
|
|
return None
|
|||
|
|
|
|||
|
|
# 提取内容
|
|||
|
|
content_tag = soup.select_one(
|
|||
|
|
'.article-content, .content, .text-content, .p_content, '
|
|||
|
|
'.article-body, .article-text, #articleContent'
|
|||
|
|
)
|
|||
|
|
content = ""
|
|||
|
|
if content_tag:
|
|||
|
|
# 移除脚本和样式标签
|
|||
|
|
for script in content_tag(['script', 'style', 'noscript']):
|
|||
|
|
script.decompose()
|
|||
|
|
content = content_tag.get_text(separator='\n', strip=True)
|
|||
|
|
|
|||
|
|
# 提取发布时间
|
|||
|
|
time_tag = soup.select_one(
|
|||
|
|
'.time, .publish-time, .date, .p_time, .article-time, '
|
|||
|
|
'time[datetime], .pubtime'
|
|||
|
|
)
|
|||
|
|
publish_time = None
|
|||
|
|
if time_tag:
|
|||
|
|
time_text = time_tag.get_text(strip=True)
|
|||
|
|
if not time_text:
|
|||
|
|
time_text = time_tag.get('datetime', '')
|
|||
|
|
publish_time = self._parse_time(time_text)
|
|||
|
|
|
|||
|
|
# 提取作者
|
|||
|
|
author_tag = soup.select_one('.author, .writer, .p_author, .article-author')
|
|||
|
|
author = None
|
|||
|
|
if author_tag:
|
|||
|
|
author = author_tag.get_text(strip=True)
|
|||
|
|
|
|||
|
|
# 提取图片
|
|||
|
|
images = []
|
|||
|
|
img_tags = soup.select('.article-content img, .content img, .p_content img')
|
|||
|
|
for img in img_tags:
|
|||
|
|
img_url = img.get('src', '') or img.get('data-src', '')
|
|||
|
|
if img_url and not img_url.startswith('data:'):
|
|||
|
|
if not img_url.startswith('http'):
|
|||
|
|
if img_url.startswith('/'):
|
|||
|
|
img_url = self.base_url + img_url
|
|||
|
|
else:
|
|||
|
|
img_url = f"{self.base_url}/{img_url}"
|
|||
|
|
images.append(img_url)
|
|||
|
|
|
|||
|
|
# 创建新闻对象
|
|||
|
|
news = NewsItem(
|
|||
|
|
title=title,
|
|||
|
|
content=content,
|
|||
|
|
url=url,
|
|||
|
|
publish_time=publish_time,
|
|||
|
|
author=author,
|
|||
|
|
source="人民网",
|
|||
|
|
category="时政",
|
|||
|
|
images=images
|
|||
|
|
)
|
|||
|
|
|
|||
|
|
return news
|
|||
|
|
|
|||
|
|
except Exception as e:
|
|||
|
|
logger.error(f"解析新闻详情失败 [{url}]: {str(e)}")
|
|||
|
|
return None
|
|||
|
|
|
|||
|
|
def _parse_time(self, time_text: str) -> Optional[str]:
|
|||
|
|
"""
|
|||
|
|
解析时间字符串
|
|||
|
|
|
|||
|
|
Args:
|
|||
|
|
time_text: 时间文本
|
|||
|
|
|
|||
|
|
Returns:
|
|||
|
|
标准化的时间字符串
|
|||
|
|
"""
|
|||
|
|
if not time_text:
|
|||
|
|
return None
|
|||
|
|
|
|||
|
|
try:
|
|||
|
|
# 尝试匹配常见的时间格式
|
|||
|
|
patterns = [
|
|||
|
|
r'(\d{4})-(\d{2})-(\d{2})\s+(\d{2}):(\d{2}):(\d{2})',
|
|||
|
|
r'(\d{4})年(\d{2})月(\d{2})日\s+(\d{2}):(\d{2})',
|
|||
|
|
r'(\d{4})/(\d{2})/(\d{2})\s+(\d{2}):(\d{2})',
|
|||
|
|
r'(\d{4})-(\d{2})-(\d{2})',
|
|||
|
|
]
|
|||
|
|
|
|||
|
|
for pattern in patterns:
|
|||
|
|
match = re.search(pattern, time_text)
|
|||
|
|
if match:
|
|||
|
|
return time_text.strip()
|
|||
|
|
|
|||
|
|
return time_text.strip()
|
|||
|
|
|
|||
|
|
except Exception as e:
|
|||
|
|
logger.warning(f"时间解析失败: {str(e)}")
|
|||
|
|
return None
|
|||
|
|
|
|||
|
|
async def close(self):
|
|||
|
|
"""关闭爬虫，释放资源"""
|
|||
|
|
if self.crawler:
|
|||
|
|
await self.crawler.close()
|
|||
|
|
self.crawler = None
|
|||
|
|
logger.info("爬虫已关闭")
|
|||
|
|
|