Merge pull request #50 from pointhi/results

Showing publish Date of articles in search results

searx/engines/google_news.py

@@ -2,6 +2,7 @@
 
 from urllib import urlencode
 from json import loads
+from datetime import datetime, timedelta
 
 categories = ['news']
 
@@ -31,7 +32,15 @@ def response(resp):
         return []
 
     for result in search_res['responseData']['results']:
+# S.149 (159), library.pdf
+# datetime.strptime("Mon, 10 Mar 2014 16:26:15 -0700", "%a, %d %b %Y %H:%M:%S %z")
+#        publishedDate = parse(result['publishedDate'])
+        publishedDate = datetime.strptime(str.join(' ',result['publishedDate'].split(None)[0:5]), "%a, %d %b %Y %H:%M:%S")
+        #utc_offset = timedelta(result['publishedDate'].split(None)[5])  # local = utc + offset
+        #publishedDate = publishedDate + utc_offset
+
         results.append({'url': result['unescapedUrl'],
                         'title': result['titleNoFormatting'],
+						'publishedDate': publishedDate,
                         'content': result['content']})
     return results

searx/engines/yahoo_news.py

@@ -4,6 +4,8 @@ from urllib import urlencode
 from lxml import html
 from searx.engines.xpath import extract_text, extract_url
 from searx.engines.yahoo import parse_url
+from datetime import datetime, timedelta
+import re
 
 categories = ['news']
 search_url = 'http://news.search.yahoo.com/search?{query}&b={offset}'
@@ -11,6 +13,7 @@ results_xpath = '//div[@class="res"]'
 url_xpath = './/h3/a/@href'
 title_xpath = './/h3/a'
 content_xpath = './/div[@class="abstr"]'
+publishedDate_xpath = './/span[@class="timestamp"]'
 suggestion_xpath = '//div[@id="satat"]//a'
 
 paging = True
@@ -37,7 +40,22 @@ def response(resp):
         url = parse_url(extract_url(result.xpath(url_xpath), search_url))
         title = extract_text(result.xpath(title_xpath)[0])
         content = extract_text(result.xpath(content_xpath)[0])
-        results.append({'url': url, 'title': title, 'content': content})
+        publishedDate = extract_text(result.xpath(publishedDate_xpath)[0])
+
+        if re.match("^[0-9]+ minute(s|) ago$", publishedDate):
+            publishedDate = datetime.now() - timedelta(minutes=int(re.match(r'\d+', publishedDate).group()))
+        else:
+            if re.match("^[0-9]+ hour(s|), [0-9]+ minute(s|) ago$", publishedDate):
+                timeNumbers = re.findall(r'\d+', publishedDate)
+                publishedDate = datetime.now() - timedelta(hours=int(timeNumbers[0])) - timedelta(minutes=int(timeNumbers[1]))
+            else:
+                # TODO year in string possible?
+                publishedDate = datetime.strptime(publishedDate,"%b %d %H:%M%p")
+
+        if publishedDate.year == 1900:
+            publishedDate = publishedDate.replace(year=datetime.now().year)
+
+        results.append({'url': url, 'title': title, 'content': content,'publishedDate':publishedDate})
 
     if not suggestion_xpath:
         return results

searx/static/css/style.css

@@ -35,6 +35,7 @@ a{text-decoration:none;color:#1a11be}a:visited{color:#8e44ad}
 .result h3{font-size:1em;word-wrap:break-word;margin:5px 0 1px 0;padding:0}
 .result .content{font-size:.8em;margin:0;padding:0;max-width:54em;word-wrap:break-word;line-height:1.24}
 .result .url{font-size:.8em;margin:3px 0 0 0;padding:0;max-width:54em;word-wrap:break-word;color:#c0392b}
+.result .published_date{font-size:.8em;color:#888;margin:5px 20px}
 .engines{color:#888}
 .small_font{font-size:.8em}
 .small p{margin:2px 0}

searx/static/less/definitions.less

@@ -64,6 +64,9 @@
 // Url to result
 @color-result-url-font: #C0392B;
 
+// Publish Date
+@color-result-publishdate-font: #888;
+
 // Images
 @color-result-image-span-background-hover: rgba(0, 0, 0, 0.6);
 @color-result-image-span-font: #FFF;

searx/static/less/style.less

@@ -248,6 +248,12 @@ a {
 		word-wrap:break-word;
 		color: @color-result-url-font;
 	}
+
+	.published_date {
+		font-size: 0.8em;
+		color: @color-result-publishdate-font;
+    	margin: 5px 20px;
+	}
 }
 
 .engines {

searx/templates/result_templates/default.html

@@ -6,6 +6,7 @@
 
   <div>
     <h3 class="result_title"><a href="{{ result.url }}">{{ result.title|safe }}</a></h3>
+	{% if result.publishedDate %}<p class="published_date">{{ result.publishedDate }}</p>{% endif %}
     <p class="content">{% if result.content %}{{ result.content|safe }}<br />{% endif %}</p>
     <p class="url">{{ result.pretty_url }}</p>
   </div>

searx/translations/en/LC_MESSAGES/messages.po

@@ -37,6 +37,14 @@ msgstr ""
 msgid "Errors"
 msgstr ""
 
+#: searx/webapp.py:167
+msgid "{minutes} minute(s) ago"
+msgstr ""
+
+#: searx/webapp.py:169
+msgid "{hours} hour(s), {minutes} minute(s) ago"
+msgstr ""
+
 #: searx/templates/index.html:7
 msgid "about"
 msgstr ""

searx/webapp.py

@@ -26,12 +26,13 @@ import json
 import cStringIO
 import os
 
+from datetime import datetime, timedelta
 from itertools import chain
 from flask import (
     Flask, request, render_template, url_for, Response, make_response,
     redirect, send_from_directory
 )
-from flask.ext.babel import Babel
+from flask.ext.babel import Babel, gettext, ngettext, format_date
 from searx import settings, searx_dir
 from searx.engines import (
     search as do_search, categories, engines, get_engines_stats,
@@ -156,6 +157,19 @@ def index():
             if engine in favicons:
                 result['favicon'] = engine
 
+        # TODO, check if timezone is calculated right
+        if 'publishedDate' in result:
+            if result['publishedDate'] >= datetime.now() - timedelta(days=1):
+                timedifference = datetime.now() - result['publishedDate']
+                minutes = int((timedifference.seconds/60)%60)
+                hours = int(timedifference.seconds/60/60)
+                if hours == 0:
+                    result['publishedDate'] = gettext(u'{minutes} minute(s) ago').format(minutes=minutes)
+                else:
+                    result['publishedDate'] = gettext(u'{hours} hour(s), {minutes} minute(s) ago').format(hours=hours, minutes=minutes)
+            else:
+                result['publishedDate'] = format_date(result['publishedDate'])
+
     if search.request_data.get('format') == 'json':
         return Response(json.dumps({'query': search.query,
                                     'results': search.results}),