018a144 · searx · git.iko.soy

git.iko.soy/searx

Merge pull request #57 from pointhi/results

improving publishDate extraction and output of it

commit018a14431bd3612db4e8840ce24f3e60026ece0f authorAdam Tauber <asciimoo@gmail.com> dateTue 18 Mar 2014, 18:20 (13 years ago) parentfaed14b 993271b

9 files changed+26−15

compare with master

unifiedsplit

requirements.txt+1 searx/engines/google_news.py+4−10 searx/engines/vimeo.py+6 searx/engines/yahoo_news.py+2−3 searx/engines/youtube.py+7 searx/templates/opensearch_response_rss.xml+1 searx/templates/result_templates/videos.html+1 searx/webapp.py+3−2 setup.py+1

requirements.txt+1

@@ -3,3 +3,4 @@ flask-babel

33grequests

44lxml

55pyyaml

6+python-dateutil

searx/engines/google_news.py+4−10

@@ -2,6 +2,7 @@

33from urllib import urlencode

44from json import loads

5+from dateutil import parser

56from datetime import datetime

78categories = ['news']

@@ -32,16 +33,9 @@ def response(resp):

3233 return []

3334

3435 for result in search_res['responseData']['results']:

35−# S.149 (159), library.pdf

36−# datetime.strptime("Mon, 10 Mar 2014 16:26:15 -0700",

37−# "%a, %d %b %Y %H:%M:%S %z")

38−# publishedDate = parse(result['publishedDate'])

39− publishedDate = datetime.strptime(

40− str.join(' ', result['publishedDate'].split(None)[0:5]),

41− "%a, %d %b %Y %H:%M:%S")

42− #utc_offset = timedelta(result['publishedDate'].split(None)[5])

43− # local = utc + offset

44− #publishedDate = publishedDate + utc_offset

36+

37+# Mon, 10 Mar 2014 16:26:15 -0700

38+ publishedDate = parser.parse(result['publishedDate'])

4539

4640 results.append({'url': result['unescapedUrl'],

4741 'title': result['titleNoFormatting'],

searx/engines/vimeo.py+6

@@ -2,6 +2,8 @@ from urllib import urlencode

22from HTMLParser import HTMLParser

33from lxml import html

44from xpath import extract_text

5+from datetime import datetime

6+from dateutil import parser

68base_url = 'http://vimeo.com'

79search_url = base_url + '/search?{query}'

@@ -10,6 +12,7 @@ content_xpath = None

1012title_xpath = None

1113results_xpath = ''

1214content_tpl = '<a href="{0}"> <img src="{2}"/> </a>'

15+publishedDate_xpath = './/p[@class="meta"]//attribute::datetime'

1316

1417# the cookie set by vimeo contains all the following values,

1518# but only __utma seems to be requiered

@@ -40,9 +43,12 @@ def response(resp):

4043 url = base_url + result.xpath(url_xpath)[0]

4144 title = p.unescape(extract_text(result.xpath(title_xpath)))

4245 thumbnail = extract_text(result.xpath(content_xpath)[0])

46+ publishedDate = parser.parse(extract_text(result.xpath(publishedDate_xpath)[0]))

47+

4348 results.append({'url': url,

4449 'title': title,

4550 'content': content_tpl.format(url, title, thumbnail),

4651 'template': 'videos.html',

52+ 'publishedDate': publishedDate,

4753 'thumbnail': thumbnail})

4854 return results

searx/engines/yahoo_news.py+2−3

@@ -6,6 +6,7 @@ from searx.engines.xpath import extract_text, extract_url

66from searx.engines.yahoo import parse_url

77from datetime import datetime, timedelta

88import re

9+from dateutil import parser

910

1011categories = ['news']

1112search_url = 'http://news.search.yahoo.com/search?{query}&b={offset}'

@@ -52,9 +53,7 @@ def response(resp):

5253 - timedelta(hours=int(timeNumbers[0]))\

5354 - timedelta(minutes=int(timeNumbers[1]))

5455 else:

55− # TODO year in string possible?

56− publishedDate = datetime.strptime(publishedDate,

57− "%b %d %H:%M%p")

56+ publishedDate = parser.parse(publishedDate)

5857

5958 if publishedDate.year == 1900:

6059 publishedDate = publishedDate.replace(year=datetime.now().year)

searx/engines/youtube.py+7

@@ -1,5 +1,7 @@

11from json import loads

22from urllib import urlencode

3+from dateutil import parser

4+from datetime import datetime

46categories = ['videos']

@@ -35,6 +37,10 @@ def response(resp):

3537 content = ''

3638 thumbnail = ''

3739

40+#"2013-12-31T15:22:51.000Z"

41+ pubdate = result['published']['$t']

42+ publishedDate = parser.parse(pubdate)

43+

3844 if result['media$group']['media$thumbnail']:

3945 thumbnail = result['media$group']['media$thumbnail'][0]['url']

4046 content += '<a href="{0}" title="{0}" ><img src="{1}" /></a>'.format(url, thumbnail) # noqa

@@ -48,6 +54,7 @@ def response(resp):

4854 'title': title,

4955 'content': content,

5056 'template': 'videos.html',

57+ 'publishedDate': publishedDate,

5158 'thumbnail': thumbnail})

5259

5360 return results

searx/templates/opensearch_response_rss.xml+1

@@ -16,6 +16,7 @@

1616 <title>{{ r.title }}</title>

1717 <link>{{ r.url }}</link>

1818 <description>{{ r.content }}</description>

19+ {% if r.pubdate %}<pubDate>{{ r.pubdate }}</pubDate>{% endif %}

1920 </item>

2021 {% endfor %}

2122 </channel>

searx/templates/result_templates/videos.html+1

@@ -5,6 +5,7 @@

66 <p>

77 <h3 class="result_title"><a href="{{ result.url }}">{{ result.title|safe }}</a></h3>

8+ {% if result.publishedDate %}<p class="published_date">{{ result.publishedDate }}</p>{% endif %}

89 <a href="{{ result.url }}"><img width="400px" src="{{ result.thumbnail }}" title={{ result.title }} alt=" {{ result.title }}"/></a>

910 <p class="url">{{ result.url }}</p>

1011 </p>

searx/webapp.py+3−2

@@ -159,8 +159,8 @@ def index():

159159

160160 # TODO, check if timezone is calculated right

161161 if 'publishedDate' in result:

162− if result['publishedDate'] >= datetime.now() - timedelta(days=1):

163− timedifference = datetime.now() - result['publishedDate']

162+ if result['publishedDate'].replace(tzinfo=None) >= datetime.now() - timedelta(days=1):

163+ timedifference = datetime.now() - result['publishedDate'].replace(tzinfo=None)

164164 minutes = int((timedifference.seconds / 60) % 60)

165165 hours = int(timedifference.seconds / 60 / 60)

166166 if hours == 0:

@@ -168,6 +168,7 @@ def index():

168168 else:

169169 result['publishedDate'] = gettext(u'{hours} hour(s), {minutes} minute(s) ago').format(hours=hours, minutes=minutes) # noqa

170170 else:

171+ result['pubdate'] = result['publishedDate'].strftime('%a, %d %b %Y %H:%M:%S %z')

171172 result['publishedDate'] = format_date(result['publishedDate'])

172173

173174 if search.request_data.get('format') == 'json':

setup.py+1

@@ -35,6 +35,7 @@ setup(

3535 'lxml',

3636 'pyyaml',

3737 'setuptools',

38+ 'python-dateutil',

3839 ],

3940 extras_require={

4041 'test': [