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

[mod][fix] startpage engine updates

This commit is contained in:
asciimoo 2014-02-18 16:14:31 +01:00
parent 6f86b96a35
commit 7db4558de7
1 changed files with 4 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
searx/engines/startpage.py
View File

@ -29,9 +29,12 @@ def response(resp):
dom = html.fromstring(resp.content) dom = html.fromstring(resp.content)
# ads xpath //div[@id="results"]/div[@id="sponsored"]//div[@class="result"] # ads xpath //div[@id="results"]/div[@id="sponsored"]//div[@class="result"]
# not ads: div[@class="result"] are the direct childs of div[@id="results"] # not ads: div[@class="result"] are the direct childs of div[@id="results"]
for result in dom.xpath('//div[@id="results"]/div[@class="result"]'): for result in dom.xpath('//div[@class="result"]'):
link = result.xpath('.//h3/a')[0] link = result.xpath('.//h3/a')[0]
url = link.attrib.get('href') url = link.attrib.get('href')
if url.startswith('http://www.google.')\
or url.startswith('https://www.google.'):
continue
title = link.text_content() title = link.text_content()
content = '' content = ''