Python利用Requests库实现XML文档解析并提取摘要

pip install requests lxml

import requests
from requests.adapters import HTTPAdapter
from urllib3.util import Retry

def create_retry_session(retries=3, backoff_factor=0.3, status_forcelist=(500, 502, 504)):
    """
    创建一个带有重试机制的requests Session。
    :param retries: 最大重试次数
    :param backoff_factor: 退避因子
    :param status_forcelist: 触发重试的HTTP状态码
    :return: requests.Session对象
    """
    session = requests.Session()
    retry = Retry(
        total=retries,
        read=retries,
        connect=retries,
        backoff_factor=backoff_factor,
        status_forcelist=status_forcelist,
        allowed_methods=["GET"],  # 只对GET请求重试
        raise_on_status=False
    )
    adapter = HTTPAdapter(max_retries=retry)
    session.mount('http://', adapter)
    session.mount('https://', adapter)
    return session

import re

def simple_summarize(text, max_sentences=2):
    """
    极简摘要：按中文或英文句号、感叹号、问号分割句子，取前max_sentences句。
    注意：实际应用中可替换为更复杂的算法。
    """
    if not text:
        return ""
    # 分割句子，保留标点
    sentences = re.split(r'(?<=[。！？!?])\s*', text.strip())
    summary = ''.join(sentences[:max_sentences])
    return summary

def fetch_and_summarize_xml(url, xpath_for_text, max_sentences=2, timeout=10):
    """
    获取XML文档，提取指定XPath的文本，生成摘要。
    :param url: XML文档URL
    :param xpath_for_text: 用于提取待摘要文本的XPath表达式（应返回字符串列表）
    :param max_sentences: 摘要的最大句子数
    :param timeout: 请求超时时间（秒）
    :return: 摘要字符串，失败时返回None
    """
    session = create_retry_session()
    try:
        response = session.get(url, timeout=timeout)
        response.raise_for_status()  # 触发HTTPError异常（4xx或5xx）
    except requests.exceptions.RequestException as e:
        logging.error(f"请求失败: {e}")
        return None

    # 解析XML
    try:
        root = etree.fromstring(response.content)
    except etree.XMLSyntaxError as e:
        logging.error(f"XML解析失败: {e}")
        return None

    # 提取文本
    text_parts = root.xpath(xpath_for_text)
    if not text_parts:
        logging.warning("未找到匹配文本，无法生成摘要")
        return ""

    full_text = ' '.join(text_parts).strip()
    if not full_text:
        return ""

    # 生成摘要
    summary = simple_summarize(full_text, max_sentences)
    return summary

import logging
from lxml import etree

logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')

def process_rss_item(item):
    """处理单个RSS条目，返回标题和摘要"""
    title = item.findtext('title', default='')
    description = item.findtext('description', default='')
    full_text = f"{title} {description}".strip()
    if not full_text:
        return None
    summary = simple_summarize(full_text, max_sentences=2)
    return {'title': title, 'summary': summary}

def main():
    rss_url = "https://example.com/feed.xml"  # 替换为实际RSS地址
    session = create_retry_session()

    try:
        resp = session.get(rss_url, timeout=15)
        resp.raise_for_status()
    except Exception as e:
        logging.error(f"请求RSS失败: {e}")
        return

    # 解析XML
    try:
        root = etree.fromstring(resp.content)
    except etree.XMLSyntaxError as e:
        logging.error(f"XML解析错误: {e}")
        return

    # 查找所有item元素（忽略命名空间）
    items = root.xpath("//*[local-name()='item']")
    if not items:
        # 尝试另一种路径：channel下的item
        items = root.xpath("//*[local-name()='channel']/*[local-name()='item']")

    if not items:
        logging.warning("未找到任何item元素")
        return

    for item in items:
        result = process_rss_item(item)
        if result:
            print(f"标题: {result['title']}")
            print(f"摘要: {result['summary']}")
            print("-" * 50)

if __name__ == "__main__":
    main()

<rss version="2.0" xmlns="http://purl.org/rss/1.0/">

namespaces = {'rss': 'http://purl.org/rss/1.0/'}
items = root.xpath('//rss:item', namespaces=namespaces)

from transformers import pipeline

summarizer = pipeline("summarization", model="facebook/bart-large-cnn")

def summarize_with_transformers(text):
    result = summarizer(text, max_length=50, min_length=10, do_sample=False)
    return result[0]['summary_text']