From 454d53d68aa6d3bf665aaa8f85a12a6563b881cc Mon Sep 17 00:00:00 2001
From: j <j@mailb.org>
Date: Wed, 26 Jul 2017 13:18:10 +0200
Subject: [PATCH] fix ddg results

---
 ox/web/duckduckgo.py | 16 ++++++++++++----
 1 file changed, 12 insertions(+), 4 deletions(-)

diff --git a/ox/web/duckduckgo.py b/ox/web/duckduckgo.py
index a8f7869..b4b3494 100644
--- a/ox/web/duckduckgo.py
+++ b/ox/web/duckduckgo.py
@@ -6,17 +6,25 @@ from six.moves import urllib
 import ox
 from ox import strip_tags, decode_html
 from ox.cache import read_url
+import lxml.html
 
 
 def find(query, timeout=ox.cache.cache_timeout):
+    """
+    Returns tuples with title, url, description
+    """
     if not isinstance(query, bytes):
         query = query.encode('utf-8')
     params = urllib.parse.urlencode({'q': query})
     url = 'http://duckduckgo.com/html/?' + params
     data = read_url(url, timeout=timeout).decode('utf-8')
+    doc = lxml.html.document_fromstring(data)
     results = []
-    regex = '<a .*?class="large" href="(.+?)">(.*?)</a>.*?<div class="snippet">(.*?)</div>'
-    for r in re.compile(regex, re.DOTALL).findall(data):
-        results.append((strip_tags(decode_html(r[1])), r[0], strip_tags(decode_html(r[2]))))
+    for e in doc.xpath("//a[contains(@class, 'result__a')]"):
+        url = e.attrib['href']
+        if 'uddg=' in url:
+            url = urllib.parse.unquote(url.split('&uddg=')[-1])
+        title = e.text_content()
+        description = ''
+        results.append((title, url, description))
     return results
-