python-ox/ox/web/twitter.py
2024-09-11 22:52:01 +01:00

35 lines
1.3 KiB
Python

# -*- coding: utf-8 -*-
# vi:si:et:sw=4:sts=4:ts=4
import re
from datetime import datetime
from urllib.parse import quote
import lxml.html
import ox
from ox.cache import read_url
def find(query=None, user=None, timeout=60):
if user:
url = 'https://twitter.com/' + quote(user)
else:
url = 'https://twitter.com/search/' + quote(query)
data = ox.cache.read_url(url, timeout=timeout).decode('utf-8')
doc = lxml.html.document_fromstring(data)
tweets = []
for e in doc.xpath("//div[contains(@class, 'original-tweet')]"):
t = lxml.html.tostring(e)
text = e.xpath(".//p[contains(@class, 'js-tweet-text')]")[0]
html = lxml.html.tostring(text, encoding='unicode').strip()
text = ox.decode_html(ox.strip_tags(html)).strip()
user = re.compile('data-name="(.*?)"').findall(t)[0]
user = ox.decode_html(ox.strip_tags(user)).strip()
tweets.append({
'id': re.compile(r'data-tweet-id="(\d+)"').findall(t)[0],
'user-id': re.compile(r'data-user-id="(\d+)"').findall(t)[0],
'name': re.compile(r'data-screen-name="(.*?)"').findall(t)[0],
'time': datetime.fromtimestamp(int(re.compile(r'data-time="(\d+)"').findall(t)[0])),
'user': user,
'text': text,
'html': html,
})
return tweets