searx

commit: f1404ce72d5c789a12211a5ea9f3ea08ebf83986
parent 0549f8c7c8d25d9764c83bc006fa32cdddc13c46
Author: Thomas Pointhuber <thomas.pointhuber@gmx.at>
Date:   Tue,  4 Mar 2014 13:11:53 +0100

Create yahoo_news.py
Diffstat:
A
searx/engines/yahoo_news.py
51
+++++++++++++++++++++++++++++++++++++++++++++++++++

1 file changed, 51 insertions(+), 0 deletions(-)
diff --git a/searx/engines/yahoo_news.py b/searx/engines/yahoo_news.py
@@ -0,0 +1,51 @@
+#!/usr/bin/env python
+
+from urllib import urlencode
+from urlparse import unquote
+from lxml import html
+from searx.engines.xpath import extract_text, extract_url
+
+categories = ['news']
+search_url = 'http://news.search.yahoo.com/search?{query}&b={offset}'
+results_xpath = '//div[@class="res"]'
+url_xpath = './/h3/a/@href'
+title_xpath = './/h3/a'
+content_xpath = './/div[@class="abstr"]'
+suggestion_xpath = '//div[@id="satat"]//a'
+
+paging = True
+
+
+def request(query, params):
+    offset = (params['pageno'] - 1) * 10 + 1
+    if params['language'] == 'all':
+        language = 'en'
+    else:
+        language = params['language'].split('_')[0]
+    params['url'] = search_url.format(offset=offset,
+                                      query=urlencode({'p': query}))
+    params['cookies']['sB'] = 'fl=1&vl=lang_{lang}&sh=1&rw=new&v=1'\
+        .format(lang=language)
+    return params
+
+
+def response(resp):
+    results = []
+    dom = html.fromstring(resp.text)
+
+    for result in dom.xpath(results_xpath):
+        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])
+        content = extract_text(result.xpath(content_xpath)[0])
+        results.append({'url': url, 'title': title, 'content': content})
+
+    if not suggestion_xpath:
+        return results
+
+    for suggestion in dom.xpath(suggestion_xpath):
+        results.append({'suggestion': extract_text(suggestion)})
+
+    return results