searxng/searx/engines/bing_images.py

99 lines
2.6 KiB
Python
Raw Normal View History

"""
Bing (Images)
@website https://www.bing.com/images
@provide-api yes (http://datamarket.azure.com/dataset/bing/search),
max. 5000 query/month
@using-api no (because of query limit)
@results HTML (using search portal)
@stable no (HTML can change)
@parse url, title, img_src
@todo currently there are up to 35 images receive per page,
because bing does not parse count=10.
limited response to 10 images
"""
2014-09-01 12:39:18 +00:00
from urllib import urlencode
from lxml import html
from yaml import load
import re
# engine dependent config
categories = ['images']
paging = True
safesearch = True
2014-09-01 12:39:18 +00:00
# search-url
base_url = 'https://www.bing.com/'
search_string = 'images/search?{query}&count=10&first={offset}'
2015-04-26 17:23:06 +00:00
thumb_url = "http://ts1.mm.bing.net/th?id={ihk}" # no https, bad certificate
2014-09-01 12:39:18 +00:00
# safesearch definitions
safesearch_types = {2: 'STRICT',
1: 'DEMOTE',
0: 'OFF'}
2014-09-02 15:13:44 +00:00
2015-02-08 21:01:24 +00:00
2014-09-01 12:39:18 +00:00
# do search-request
def request(query, params):
offset = (params['pageno'] - 1) * 10 + 1
# required for cookie
2015-01-29 19:15:52 +00:00
if params['language'] == 'all':
language = 'en-US'
else:
language = params['language'].replace('_', '-')
2014-09-01 12:39:18 +00:00
search_path = search_string.format(
query=urlencode({'q': query}),
offset=offset)
params['cookies']['SRCHHPGUSR'] = \
'NEWWND=0&NRSLT=-1&SRCHLANG=' + language.split('-')[0] +\
'&ADLT=' + safesearch_types.get(params['safesearch'], 'DEMOTE')
2014-09-01 12:39:18 +00:00
params['url'] = base_url + search_path
return params
# get response from search-request
def response(resp):
results = []
dom = html.fromstring(resp.content)
# init regex for yaml-parsing
p = re.compile('({|,)([a-z]+):(")')
2014-09-01 12:39:18 +00:00
# parse results
for result in dom.xpath('//div[@class="dg_u"]'):
link = result.xpath('./a')[0]
# parse yaml-data (it is required to add a space, to make it parsable)
yaml_data = load(p.sub(r'\1\2: \3', link.attrib.get('m')))
2014-09-01 12:39:18 +00:00
title = link.attrib.get('t1')
ihk = link.attrib.get('ihk')
2015-01-17 18:24:35 +00:00
# url = 'http://' + link.attrib.get('t3')
2014-09-01 12:39:18 +00:00
url = yaml_data.get('surl')
img_src = yaml_data.get('imgurl')
# append result
results.append({'template': 'images.html',
'url': url,
'title': title,
'content': '',
'thumbnail_src': thumb_url.format(ihk=ihk),
2014-09-01 12:39:18 +00:00
'img_src': img_src})
# TODO stop parsing if 10 images are found
if len(results) >= 10:
break
# return results
return results