searx

commit: 1e99cf2a0e541a3d2df0104d64fadf955bbccc20
parent: c23db1b2bfb2c9233816fc378927c49b67eeffaf
Author: Adam Tauber <asciimoo@gmail.com>
Date:   Sun,  7 Sep 2014 19:12:05 +0200

Merge pull request #93 from dalf/master

yahoo, bing_new and dailymotion fixes
Diffstat:
M
searx/engines/bing_news.py
23
++++++++++++++++++-----
M
searx/engines/dailymotion.py
7
++++++-
M
searx/engines/yahoo.py
8
+++++---

3 files changed, 29 insertions(+), 9 deletions(-)
diff --git a/searx/engines/bing_news.py b/searx/engines/bing_news.py
@@ -56,10 +56,14 @@ def response(resp):
         link = result.xpath('.//div[@class="newstitle"]/a')[0]
         url = link.attrib.get('href')
         title = ' '.join(link.xpath('.//text()'))
-        content = escape(' '.join(result.xpath('.//div[@class="sn_txt"]/div//span[@class="sn_snip"]//text()')))
-        
+        contentXPath = result.xpath('.//div[@class="sn_txt"]/div//span[@class="sn_snip"]//text()')
+        if contentXPath != None:
+            content = escape(' '.join(contentXPath))
+            
         # parse publishedDate
-        publishedDate = escape(' '.join(result.xpath('.//div[@class="sn_txt"]/div//span[@class="sn_ST"]//span[@class="sn_tm"]//text()')))
+        publishedDateXPath = result.xpath('.//div[@class="sn_txt"]/div//span[contains(@class,"sn_ST")]//span[contains(@class,"sn_tm")]//text()')
+        if publishedDateXPath != None:
+            publishedDate = escape(' '.join(publishedDateXPath))
 
         if re.match("^[0-9]+ minute(s|) ago$", publishedDate):
             timeNumbers = re.findall(r'\d+', publishedDate)
@@ -74,9 +78,18 @@ def response(resp):
             publishedDate = datetime.now()\
                 - timedelta(hours=int(timeNumbers[0]))\
                 - timedelta(minutes=int(timeNumbers[1]))
+        elif re.match("^[0-9]+ day(s|) ago$", publishedDate):
+            timeNumbers = re.findall(r'\d+', publishedDate)
+            publishedDate = datetime.now()\
+                - timedelta(days=int(timeNumbers[0]))
         else:
-            publishedDate = parser.parse(publishedDate)  
-
+            try:
+                # FIXME use params['language'] to parse either mm/dd or dd/mm
+                publishedDate = parser.parse(publishedDate, dayfirst=False)
+            except TypeError:
+                # FIXME
+                publishedDate = datetime.now()
+                
         # append result
         results.append({'url': url, 
                         'title': title, 
diff --git a/searx/engines/dailymotion.py b/searx/engines/dailymotion.py
@@ -16,8 +16,8 @@ from lxml import html
 
 # engine dependent config
 categories = ['videos']
-locale = 'en_US'
 paging = True
+language_support = True
 
 # search-url
 # see http://www.dailymotion.com/doc/api/obj-video.html
@@ -26,6 +26,11 @@ search_url = 'https://api.dailymotion.com/videos?fields=title,description,durati
 
 # do search-request
 def request(query, params):
+    if params['language'] == 'all':
+        locale = 'en-US'
+    else:
+        locale = params['language']
+
     params['url'] = search_url.format(
         query=urlencode({'search': query, 'localization': locale}),
         pageno=params['pageno'])
diff --git a/searx/engines/yahoo.py b/searx/engines/yahoo.py
@@ -40,9 +40,11 @@ def parse_url(url_string):
         if endpos > -1:
             endpositions.append(endpos)
 
-    end = min(endpositions)
-
-    return unquote(url_string[start:end])
+    if start==0 or len(endpositions) == 0:
+        return url_string        
+    else:
+        end = min(endpositions)
+        return unquote(url_string[start:end])
 
 
 # do search-request