Ponysearch/searx/engines/bing_images.py

107 lines
3 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 14:39:18 +02:00
from urllib import urlencode
from lxml import html
2016-02-01 21:40:28 +01:00
from json import loads
2014-09-01 14:39:18 +02:00
import re
# engine dependent config
categories = ['images']
paging = True
safesearch = True
2016-10-30 20:58:34 +01:00
time_range_support = True
2014-09-01 14:39:18 +02:00
# search-url
base_url = 'https://www.bing.com/'
search_string = 'images/search?{query}&count=10&first={offset}'
2016-10-30 20:58:34 +01:00
time_range_string = '&qft=+filterui:age-lt{interval}'
thumb_url = "https://www.bing.com/th?id={ihk}"
2016-10-30 20:58:34 +01:00
time_range_dict = {'day': '1440',
'week': '10080',
'month': '43200',
'year': '525600'}
2014-09-01 14:39:18 +02:00
# safesearch definitions
safesearch_types = {2: 'STRICT',
1: 'DEMOTE',
0: 'OFF'}
2014-09-02 17:13:44 +02:00
2015-02-08 22:01:24 +01:00
2016-02-01 21:40:28 +01:00
_quote_keys_regex = re.compile('({|,)([a-z][a-z0-9]*):(")', re.I | re.U)
2014-09-01 14:39:18 +02:00
# do search-request
def request(query, params):
offset = (params['pageno'] - 1) * 10 + 1
# required for cookie
2015-01-29 20:15:52 +01:00
if params['language'] == 'all':
language = 'en-US'
else:
language = params['language'].replace('_', '-')
2014-09-01 14:39:18 +02:00
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 14:39:18 +02:00
params['url'] = base_url + search_path
2016-10-30 20:58:34 +01:00
if params['time_range'] in time_range_dict:
params['url'] += time_range_string.format(interval=time_range_dict[params['time_range']])
2014-09-01 14:39:18 +02:00
return params
# get response from search-request
def response(resp):
results = []
dom = html.fromstring(resp.text)
2014-09-01 14:39:18 +02:00
# parse results
2016-03-26 02:07:52 +01:00
for result in dom.xpath('//div[@class="dg_u"]/div'):
2014-09-01 14:39:18 +02:00
link = result.xpath('./a')[0]
2016-02-01 21:40:28 +01:00
# parse json-data (it is required to add a space, to make it parsable)
json_data = loads(_quote_keys_regex.sub(r'\1"\2": \3', link.attrib.get('m')))
2014-09-01 14:39:18 +02:00
title = link.attrib.get('t1')
ihk = link.attrib.get('ihk')
2015-01-17 19:24:35 +01:00
# url = 'http://' + link.attrib.get('t3')
2016-02-01 21:40:28 +01:00
url = json_data.get('surl')
img_src = json_data.get('imgurl')
2014-09-01 14:39:18 +02:00
# append result
results.append({'template': 'images.html',
'url': url,
'title': title,
'content': '',
'thumbnail_src': thumb_url.format(ihk=ihk),
2014-09-01 14:39:18 +02:00
'img_src': img_src})
# TODO stop parsing if 10 images are found
if len(results) >= 10:
break
# return results
return results