searxng/searx/engines/twitter.py

83 lines
2.4 KiB
Python
Raw Normal View History

"""
Twitter (Social media)
@website https://twitter.com/
@provide-api yes (https://dev.twitter.com/docs/using-search)
@using-api no
@results HTML (using search portal)
@stable no (HTML can change)
@parse url, title, content
@todo publishedDate
"""
2013-10-20 08:40:02 +00:00
from lxml import html
from datetime import datetime
from searx.engines.xpath import extract_text
2016-11-30 17:43:03 +00:00
from searx.url_utils import urlencode, urljoin
2013-10-20 08:40:02 +00:00
# engine dependent config
2013-10-20 08:40:02 +00:00
categories = ['social media']
language_support = True
2013-10-20 08:40:02 +00:00
# search-url
2013-10-20 08:40:02 +00:00
base_url = 'https://twitter.com/'
search_url = base_url + 'search?'
# specific xpath variables
results_xpath = '//li[@data-item-type="tweet"]'
avatar_xpath = './/img[contains(@class, "avatar")]/@src'
link_xpath = './/small[@class="time"]//a'
2015-05-02 16:20:12 +00:00
title_xpath = './/span[contains(@class, "username")]'
content_xpath = './/p[contains(@class, "tweet-text")]'
timestamp_xpath = './/span[contains(@class,"_timestamp")]'
2014-01-20 01:31:20 +00:00
2013-10-20 08:40:02 +00:00
# do search-request
2013-10-20 08:40:02 +00:00
def request(query, params):
params['url'] = search_url + urlencode({'q': query})
params['cookies']['lang'] = params['language'].split('-')[0]
2013-10-20 08:40:02 +00:00
return params
# get response from search-request
2013-10-20 08:40:02 +00:00
def response(resp):
results = []
2013-10-20 08:40:02 +00:00
dom = html.fromstring(resp.text)
# parse results
for tweet in dom.xpath(results_xpath):
try:
link = tweet.xpath(link_xpath)[0]
content = extract_text(tweet.xpath(content_xpath)[0])
img_src = tweet.xpath(avatar_xpath)[0]
img_src = img_src.replace('_bigger', '_normal')
except Exception:
continue
2013-10-20 08:40:02 +00:00
url = urljoin(base_url, link.attrib.get('href'))
title = extract_text(tweet.xpath(title_xpath))
pubdate = tweet.xpath(timestamp_xpath)
if len(pubdate) > 0:
timestamp = float(pubdate[0].attrib.get('data-time'))
publishedDate = datetime.fromtimestamp(timestamp, None)
# append result
results.append({'url': url,
'title': title,
'content': content,
'img_src': img_src,
'publishedDate': publishedDate})
else:
# append result
results.append({'url': url,
'title': title,
'content': content,
'img_src': img_src})
# return results
2013-10-20 08:40:02 +00:00
return results