update bing engines and fix bing_news

This commit is contained in:
Thomas Pointhuber 2014-09-01 14:38:59 +02:00
parent 55dfb305a0
commit 4b1e0423a0
2 changed files with 92 additions and 23 deletions

View File

@ -1,48 +1,81 @@
## Bing (Web)
#
# @website https://www.bing.com
# @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
#
# @todo publishedDate
from urllib import urlencode
from cgi import escape
from lxml import html
base_url = 'http://www.bing.com/'
search_string = 'search?{query}&first={offset}'
# engine dependent config
categories = ['general']
paging = True
language_support = True
# search-url
base_url = 'https://www.bing.com/'
search_string = '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]
#if params['category'] == 'images':
# params['url'] = base_url + 'images/' + search_path
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="sa_cc"]'):
link = result.xpath('.//h3/a')[0]
url = link.attrib.get('href')
title = ' '.join(link.xpath('.//text()'))
content = escape(' '.join(result.xpath('.//p//text()')))
results.append({'url': url, 'title': title, 'content': content})
# append result
results.append({'url': url,
'title': title,
'content': content})
# return results if something is found
if results:
return results
# parse results again if nothing is found yet
for result in dom.xpath('//li[@class="b_algo"]'):
link = result.xpath('.//h2/a')[0]
url = link.attrib.get('href')
title = ' '.join(link.xpath('.//text()'))
content = escape(' '.join(result.xpath('.//p//text()')))
results.append({'url': url, 'title': title, 'content': content})
# append result
results.append({'url': url,
'title': title,
'content': content})
# return results
return results

View File

@ -1,50 +1,86 @@
## 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']
base_url = 'http://www.bing.com/'
search_string = 'news/search?{query}&first={offset}'
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]
#if params['category'] == 'images':
# params['url'] = base_url + 'images/' + search_path
params['url'] = base_url + search_path
return params
# get response from search-request
def response(resp):
results = []
dom = html.fromstring(resp.content)
for result in dom.xpath('//div[@class="sa_cc"]'):
link = result.xpath('.//h3/a')[0]
# 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('.//p//text()')))
results.append({'url': url, 'title': title, 'content': content})
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 results:
return results
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)
for result in dom.xpath('//li[@class="b_algo"]'):
link = result.xpath('.//h2/a')[0]
url = link.attrib.get('href')
title = ' '.join(link.xpath('.//text()'))
content = escape(' '.join(result.xpath('.//p//text()')))
results.append({'url': url, 'title': title, 'content': content})
# append result
results.append({'url': url,
'title': title,
'publishedDate': publishedDate,
'content': content})
# return results
return results