RSS 提要手动解析图像？

Question

使用 python 上的 feedparser 模块，我能够从这个 XML 文件中解析标题、描述、链接和日期：https://www.espn.com/espn/rss/nba/news

在 XML 文件中，每篇文章肯定都有一张图片，但 feedparser 无法识别。它显示的键如下：

dict_keys(['title', 'title_detail', 'summary', 'summary_detail', 'links', 'link', 'published', 'published_parsed', 'id', 'guidislink'])

有没有办法手动解析图像？这是我当前的脚本：

url = 'https://www.espn.com/espn/rss/' + league + '/news'
    
parser = feedparser.parse(url)

newsInfo = []

for entry in parser.entries:
        newEntry = {
            'title': entry.title,
            'description': entry.description,
            'link': entry.link,
            'date': entry.published_parsed,
            'displayDate': entry.published
        }
        newsInfo.append(newEntry)
return newsInfo

我想保留相应文章中的图片。

Answer 1

def newsParse(league):
    url = 'https://www.espn.com/espn/rss/' + league + '/news'

    def seperate_description_and_image(s):
        if '\n' in s:
            return s.rsplit('\n', 1)
        else:
            return [s, ""]

    # Need access to feed before parsing
    # parser = feedparser.parse(url)
    response = requests.get(url)
    # error check response.status_code here...

    # join sub-element <description> and unknown sub-element <image> 
    # in RSS 2.0 <item> and seperate with \n
    raw = response.text
    raw = raw.replace("</description><image>", "\n")
    raw = raw.replace(".jpg]]></image>", ".jpg]]>\n</description>")
    
    # now parse
    parser = feedparser.parse(raw)

    newsInfo = []
    for entry in parser.entries:
        try:
            newEntry = {
            'title': entry.title,
            'description': seperate_description_and_image(entry.description)[0],
            'image': seperate_description_and_image(entry.description)[1],
            'link': entry.link,
            'date': entry.published_parsed,
            'displayDate': entry.published
            }
            newsInfo.append(newEntry)
        except AttributeError:
            continue
    
    return newsInfo

Answer 2

基于@RudyGoburt 的答案，我有一个更通用的解决方案。由于feedparser 不处理<item> 字段内的<image> ... </image>，我解析原始html 并提取图像标签内的url 并将其放置在<enclosure> 标签内。 Feedparser 可以处理附件，以便提取图像 url。

下面的解决方案还进行了一些检查，以便能够处理各种其他 rss 提要。请注意，如果可用，也可以从media_content 提取图像。

import requests
import re
import time
import feedparser
from pprint import pprint


def generalised_parse(url):
    raw = ''
    try:
        response = requests.get(url)

        if response and response.status_code == 200:
            raw = response.text

        else:
            pass

    except (requests.exceptions.ConnectionError, requests.exceptions.MissingSchema):
        pass

    # if raw <item>'s have a '<image> ... </image>' pattern extract the image url and
    # put this image url in an <enclosure /> tag which can be handled by feedparser
    raw = re.sub(r'(<item>.*?)<image>.*?(http.*?jpg|png|gif).*?</image>(.*?</item>)',
                 r'\1<enclosure url="\2" />\3', raw)
    
    # some url give an empty raw string, in that case parse with the url instead of 
    # the raw string
    if raw:
        parser = feedparser.parse(raw)

    else:
        parser = feedparser.parse(url)

    newsInfo = []
    for entry in parser.entries:
        # looks like entry has always an attribute enclosures, although maybe an 
        # empty list
        if entry.enclosures:
            enclosure_href = entry.enclosures[0]['href']
        
        else:
            enclosure_href = ''

        # check if there is media_content
        try:
            media_url = entry.media_content[0]['url']
        
        except AttributeError:
            media_url = ''

        # some entry have no attribute published, in that case check for attribute
        # updated, if that does not exist give default date of 1970-1-1
        try: 
            published = entry.published
            published_parsed = entry.updated_parsed
        
        except AttributeError:
            published = entry.get('updated', '1970-01-01')
            published_parsed = entry.get('updated_parsed', 
                time.struct_time((1970, 1, 1, 0, 0, 0, 0, 0, 0)))

        # use get method so that in case attribute does not exist we can give a default
        # value
        newEntry = {
            'title': entry.get('title', ''),
            'description': entry.get('summary', ''),
            'link': entry.get('link', ''),
            'enclosure' : enclosure_href,
            'media': media_url,
            'date': published_parsed,
            'displayDate': published,
        }
        newsInfo.append(newEntry)
    
    return newsInfo


def main():
    # you can try out with various rss feeds
    url_espn_rpm = 'https://www.espn.com/espn/rss/rpm/news'
    url_espn_pl = 'http://www.espnfc.com/english-premier-league/23/rss'
    url_cnn = 'http://rss.cnn.com/rss/edition.rss'
    url_reddit = 'http://www.reddit.com/.rss'
    url_bbc = 'http://feeds.bbci.co.uk/news/rss.xml'

    news_info = generalised_parse(url_espn_pl)
    pprint(news_info)


if __name__ == '__main__':
    main()