zaclys
/
searxng
mirror of https://github.com/searxng/searxng.git
1
0
Fork 1
Code Issues Packages Projects Releases Wiki Activity

[fix] urljoin removed

This commit is contained in:
asciimoo 2013-10-24 23:43:39 +02:00
parent 7ddf7d43ba
commit cf6a92a996
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
searx/engines/startpage.py
View File

@ -1,6 +1,6 @@
from urllib import quote from urllib import quote
from lxml import html from lxml import html
from urlparse import urljoin, urlparse from urlparse import urlparse
from cgi import escape from cgi import escape
base_url = 'https://startpage.com/' base_url = 'https://startpage.com/'
@ -21,7 +21,7 @@ def response(resp):
dom = html.fromstring(resp.content) dom = html.fromstring(resp.content)
for result in dom.xpath('//div[@class="result"]'): for result in dom.xpath('//div[@class="result"]'):
link = result.xpath('.//h3/a')[0] link = result.xpath('.//h3/a')[0]
url = urljoin(base_url, link.attrib.get('href')) url = link.attrib.get('href')
parsed_url = urlparse(url) parsed_url = urlparse(url)
# TODO better google link detection # TODO better google link detection
if parsed_url.netloc.find('www.google.com') >= 0: if parsed_url.netloc.find('www.google.com') >= 0: