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

[fix] yahoo url mods

This commit is contained in:
Adam Tauber 2014-02-27 10:59:39 +01:00
parent 8c8d9d10cb
commit 4627788834
1 changed files with 5 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
searx/engines/yahoo.py
View File

@ -1,6 +1,7 @@
#!/usr/bin/env python #!/usr/bin/env python
from urllib import urlencode from urllib import urlencode
from urlparse import unquote
from lxml import html from lxml import html
from searx.engines.xpath import extract_text, extract_url from searx.engines.xpath import extract_text, extract_url
@ -33,7 +34,10 @@ def response(resp):
dom = html.fromstring(resp.text) dom = html.fromstring(resp.text)
for result in dom.xpath(results_xpath): for result in dom.xpath(results_xpath):
url = extract_url(result.xpath(url_xpath), search_url) url_string = extract_url(result.xpath(url_xpath), search_url)
start = url_string.find('/RU=')+4
end = url_string.rfind('/RS')
url = unquote(url_string[start:end])
title = extract_text(result.xpath(title_xpath)[0]) title = extract_text(result.xpath(title_xpath)[0])
content = extract_text(result.xpath(content_xpath)[0]) content = extract_text(result.xpath(content_xpath)[0])
results.append({'url': url, 'title': title, 'content': content}) results.append({'url': url, 'title': title, 'content': content})