Ponysearch/searx/engines/bing_news.py

103 lines
3.3 KiB
Python
Raw Normal View History

"""
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
"""
2014-09-01 14:38:59 +02:00
2014-03-04 13:10:04 +01:00
from urllib import urlencode
from cgi import escape
from lxml import html
2014-09-01 14:38:59 +02:00
from datetime import datetime, timedelta
from dateutil import parser
import re
from searx.engines.xpath import extract_text
2014-03-04 13:10:04 +01:00
2014-09-01 14:38:59 +02:00
# engine dependent config
2014-03-04 13:10:04 +01:00
categories = ['news']
paging = True
language_support = True
2014-09-01 14:38:59 +02:00
# search-url
base_url = 'https://www.bing.com/'
search_string = 'news/search?{query}&first={offset}'
2014-03-04 13:10:04 +01:00
2014-09-02 17:13:44 +02:00
2014-09-01 14:38:59 +02:00
# do search-request
2014-03-04 13:10:04 +01:00
def request(query, params):
offset = (params['pageno'] - 1) * 10 + 1
2014-09-01 14:38:59 +02:00
2014-03-04 13:10:04 +01:00
if params['language'] == 'all':
language = 'en-US'
else:
language = params['language'].replace('_', '-')
2014-09-01 14:38:59 +02:00
2014-03-04 13:10:04 +01:00
search_path = search_string.format(
query=urlencode({'q': query, 'setmkt': language}),
offset=offset)
params['cookies']['_FP'] = "ui=en-US"
2014-09-01 14:38:59 +02:00
2014-03-04 13:10:04 +01:00
params['url'] = base_url + search_path
2014-03-04 13:10:04 +01:00
return params
2014-09-01 14:38:59 +02:00
# get response from search-request
2014-03-04 13:10:04 +01:00
def response(resp):
results = []
2014-09-01 14:38:59 +02:00
2014-03-04 13:10:04 +01:00
dom = html.fromstring(resp.content)
2014-09-01 14:38:59 +02:00
# parse results
for result in dom.xpath('//div[@class="sn_r"]'):
link = result.xpath('.//div[@class="newstitle"]/a')[0]
2014-03-04 13:10:04 +01:00
url = link.attrib.get('href')
title = extract_text(link)
contentXPath = result.xpath('.//div[@class="sn_txt"]/div//span[@class="sn_snip"]')
content = escape(extract_text(contentXPath))
2014-09-01 14:38:59 +02:00
# parse publishedDate
2014-12-16 17:26:16 +01:00
publishedDateXPath = result.xpath('.//div[@class="sn_txt"]/div'
2015-05-27 12:08:50 +02:00
'//div[contains(@class,"sn_ST")]'
'//span[contains(@class,"sn_tm")]')
publishedDate = escape(extract_text(publishedDateXPath))
2014-03-04 13:10:04 +01:00
2014-09-01 14:38:59 +02:00
if re.match("^[0-9]+ minute(s|) ago$", publishedDate):
timeNumbers = re.findall(r'\d+', publishedDate)
publishedDate = datetime.now() - timedelta(minutes=int(timeNumbers[0]))
2014-09-01 14:38:59 +02:00
elif re.match("^[0-9]+ hour(s|) ago$", 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$", publishedDate):
2014-09-01 14:38:59 +02:00
timeNumbers = re.findall(r'\d+', publishedDate)
publishedDate = datetime.now()\
- timedelta(hours=int(timeNumbers[0]))\
- timedelta(minutes=int(timeNumbers[1]))
elif re.match("^[0-9]+ day(s|) ago$", publishedDate):
timeNumbers = re.findall(r'\d+', publishedDate)
publishedDate = datetime.now() - timedelta(days=int(timeNumbers[0]))
2014-09-01 14:38:59 +02:00
else:
try:
publishedDate = parser.parse(publishedDate, dayfirst=False)
except TypeError:
publishedDate = datetime.now()
except ValueError:
publishedDate = datetime.now()
2014-09-01 14:38:59 +02:00
# append result
results.append({'url': url,
'title': title,
2014-09-01 14:38:59 +02:00
'publishedDate': publishedDate,
'content': content})
# return results
2014-03-04 13:10:04 +01:00
return results