v/pol
1
0
Fork 0
mirror of https://github.com/taroved/pol synced 2025-05-18 07:10:09 -07:00
Code Issues Releases Wiki Activity

md5 hash suffix to rss links

Browse Source
This commit is contained in:
Alexandr Nesterenko 2017-05-03 11:48:20 -07:00
parent 1c254e4ca8
commit 51dcd46df6
1 changed files with 12 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
feed.py
View File

@ -14,6 +14,8 @@ import w3lib.url
import w3lib.html import w3lib.html
from lxml import etree from lxml import etree
import re
from hashlib import md5
from feedgenerator import Rss201rev2Feed, Enclosure from feedgenerator import Rss201rev2Feed, Enclosure
import datetime import datetime
@ -21,6 +23,7 @@ import datetime
import MySQLdb import MySQLdb
from settings import DATABASES, DOWNLOADER_USER_AGENT from settings import DATABASES, DOWNLOADER_USER_AGENT
url_hash_regexp = re.compile('(#.*)?$')
def _getPageFactory(url, contextFactory=None, *args, **kwargs): def _getPageFactory(url, contextFactory=None, *args, **kwargs):
""" """
@ -84,12 +87,17 @@ def _buildFeed(response, feed_config):
"Url: " + feed_config['uri'], "Url: " + feed_config['uri'],
language="en", language="en",
) )
for item in items: for item in items:
title = item['title'] if 'title' in item else ''
desc = item['description'] if 'description' in item else ''
if item['title_link']:
link = item['title_link']
else:
link = url_hash_regexp.sub('#' + md5((title+desc).encode('utf-8')).hexdigest(), feed_config['uri'])
feed.add_item( feed.add_item(
title=item['title'] if 'title' in item else '', title = title,
link = item['title_link'] if 'title_link' in item else feed_config['uri'], link = link,
description=item['description'] if 'description' in item else '', description = desc,
#enclosure=Enclosure(fields[4], "32000", "image/jpeg") if 4 in fields else None, #"Image" #enclosure=Enclosure(fields[4], "32000", "image/jpeg") if 4 in fields else None, #"Image"
pubdate=datetime.datetime.now() pubdate=datetime.datetime.now()
) )