mirror of
https://github.com/searxng/searxng
synced 2024-01-01 18:24:07 +00:00
87 lines
2.8 KiB
Python
87 lines
2.8 KiB
Python
## Bing (News)
|
|
#
|
|
# @website https://www.bing.com/news
|
|
# @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, content, publishedDate
|
|
|
|
from urllib import urlencode
|
|
from cgi import escape
|
|
from lxml import html
|
|
from datetime import datetime, timedelta
|
|
from dateutil import parser
|
|
import re
|
|
|
|
# engine dependent config
|
|
categories = ['news']
|
|
paging = True
|
|
language_support = True
|
|
|
|
# search-url
|
|
base_url = 'https://www.bing.com/'
|
|
search_string = 'news/search?{query}&first={offset}'
|
|
|
|
# do search-request
|
|
def request(query, params):
|
|
offset = (params['pageno'] - 1) * 10 + 1
|
|
|
|
if params['language'] == 'all':
|
|
language = 'en-US'
|
|
else:
|
|
language = params['language'].replace('_', '-')
|
|
|
|
search_path = search_string.format(
|
|
query=urlencode({'q': query, 'setmkt': language}),
|
|
offset=offset)
|
|
|
|
params['cookies']['SRCHHPGUSR'] = \
|
|
'NEWWND=0&NRSLT=-1&SRCHLANG=' + language.split('-')[0]
|
|
|
|
params['url'] = base_url + search_path
|
|
return params
|
|
|
|
|
|
# get response from search-request
|
|
def response(resp):
|
|
results = []
|
|
|
|
dom = html.fromstring(resp.content)
|
|
|
|
# parse results
|
|
for result in dom.xpath('//div[@class="sn_r"]'):
|
|
link = result.xpath('.//div[@class="newstitle"]/a')[0]
|
|
url = link.attrib.get('href')
|
|
title = ' '.join(link.xpath('.//text()'))
|
|
content = escape(' '.join(result.xpath('.//div[@class="sn_txt"]/div//span[@class="sn_snip"]//text()')))
|
|
|
|
# parse publishedDate
|
|
publishedDate = escape(' '.join(result.xpath('.//div[@class="sn_txt"]/div//span[@class="sn_ST"]//span[@class="sn_tm"]//text()')))
|
|
|
|
if re.match("^[0-9]+ minute(s|) ago$", publishedDate):
|
|
timeNumbers = re.findall(r'\d+', publishedDate)
|
|
publishedDate = datetime.now()\
|
|
- timedelta(minutes=int(timeNumbers[0]))
|
|
elif re.match("^[0-9]+ hour(s|) ago$", publishedDate):
|
|
timeNumbers = re.findall(r'\d+', publishedDate)
|
|
publishedDate = datetime.now()\
|
|
- timedelta(hours=int(timeNumbers[0]))
|
|
elif 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:
|
|
publishedDate = parser.parse(publishedDate)
|
|
|
|
# append result
|
|
results.append({'url': url,
|
|
'title': title,
|
|
'publishedDate': publishedDate,
|
|
'content': content})
|
|
|
|
# return results
|
|
return results
|