searxng/searx/engines/google_images.py

70 lines
1.7 KiB
Python
Raw Normal View History

"""
Google (Images)
@website https://www.google.com
@provide-api yes (https://developers.google.com/custom-search/)
2015-12-09 00:23:05 +00:00
@using-api no
@results HTML chunks with JSON inside
@stable no
@parse url, title, img_src
"""
2013-10-19 20:19:14 +00:00
from urllib import urlencode
from urlparse import parse_qs
2013-10-19 21:12:18 +00:00
from json import loads
from lxml import html
2013-10-19 20:19:14 +00:00
2014-09-01 13:10:05 +00:00
# engine dependent config
2013-10-19 20:19:31 +00:00
categories = ['images']
2014-09-01 13:10:05 +00:00
paging = True
2015-02-08 21:15:25 +00:00
safesearch = True
2013-10-19 20:19:14 +00:00
search_url = 'https://www.google.com/search'\
'?{query}'\
'&tbm=isch'\
'&ijn=1'\
'&start={offset}'
2014-01-20 01:31:20 +00:00
2013-10-19 20:19:14 +00:00
2014-09-01 13:10:05 +00:00
# do search-request
2013-10-19 20:19:14 +00:00
def request(query, params):
offset = (params['pageno'] - 1) * 100
2015-02-08 21:15:25 +00:00
2014-01-30 00:21:33 +00:00
params['url'] = search_url.format(query=urlencode({'q': query}),
2015-02-08 21:15:25 +00:00
offset=offset,
safesearch=safesearch)
2014-09-01 13:10:05 +00:00
if safesearch and params['safesearch']:
params['url'] += '&' + urlencode({'safe': 'active'})
2013-10-19 20:19:14 +00:00
return params
2014-01-20 01:31:20 +00:00
2014-09-01 13:10:05 +00:00
# get response from search-request
2013-10-19 20:19:14 +00:00
def response(resp):
results = []
2014-09-01 13:10:05 +00:00
dom = html.fromstring(resp.text)
2014-09-01 13:10:05 +00:00
# parse results
for result in dom.xpath('//div[@data-ved]'):
metadata = loads(result.xpath('./div[@class="rg_meta"]/text()')[0])
thumbnail_src = metadata['tu']
2014-09-01 13:10:05 +00:00
# http to https
thumbnail_src = thumbnail_src.replace("http://", "https://")
2014-09-01 13:10:05 +00:00
# append result
2016-04-07 06:04:36 +00:00
results.append({'url': metadata['ru'],
'title': metadata['pt'],
'content': metadata['s'],
2016-04-07 06:04:36 +00:00
'thumbnail_src': thumbnail_src,
'img_src': metadata['ou'],
2014-01-20 01:31:20 +00:00
'template': 'images.html'})
2014-09-01 13:10:05 +00:00
# return results
2013-10-19 20:19:14 +00:00
return results