Last active
December 8, 2018 04:12
-
-
Save sornars/581c940c002605766dba4a74bf8a738c to your computer and use it in GitHub Desktop.
Script for archiving PMs from therewillbe.games
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import getpass | |
import requests | |
from lxml import html | |
username = input('username: ') | |
password = getpass.getpass() | |
session = requests.Session() | |
# Authorize session | |
forum = session.get('https://therewillbe.games/forum/recent') | |
forum_root = html.fromstring(forum.content) | |
csrf_form = forum_root.xpath('//form[@action="/forum" and @name="login"]/input')[-1] | |
csrf_token = csrf_form.attrib['name'] | |
csrf_token_value = csrf_form.attrib['value'] | |
login = session.post('https://therewillbe.games/forum', | |
data={'view': 'user', 'task': 'login', csrf_token: csrf_token_value, 'username': username, | |
'password': password, 'remember': '1', 'submit': 'Login'}) | |
# Scrape inbox | |
inbox = session.get('https://therewillbe.games/site-tools/trash-mail?task=inbox') | |
inbox_root = html.fromstring(inbox.content) | |
show_all_link = inbox_root.xpath('//a[text()="show all"]')[0].attrib['href'] | |
full_inbox = session.get(show_all_link) | |
full_inbox_root = html.fromstring(full_inbox.content) | |
message_links = [a.attrib['href'] for a in full_inbox_root.xpath( | |
'//form[@name="messages"]//td/a[contains(@href, "task=show") and contains(@href, "messageid=")]')] | |
archive_links = [a.attrib['href'] for a in full_inbox_root.xpath( | |
'//form[@name="messages"]//td/a[contains(@href, "task=archive") and contains(@href, "messageid=")]')] | |
# Capture message content | |
with open('./therewillbegames_messages.txt', 'w', newline='') as o: | |
for ml in message_links: | |
message = session.get(ml) | |
message_root = html.fromstring(message.content) | |
from_div = message_root.xpath('//div[@class="uddeim-messagefrom"]')[0] | |
from_profile = from_div.xpath('//a[contains(@href, "userprofile")]')[0].attrib['href'] | |
from_username = from_profile.split('/')[-1] | |
sent_time_string = list(from_div.itertext())[-1] | |
body_div = message_root.xpath('//div[@class="uddeim-messagebody"]')[0] | |
message = body_div.text_content() | |
logged_message = """MESSAGE BEGIN: {sent_time_string} - {from_username} - {from_profile} | |
{message} | |
MESSAGE END | |
""".format(sent_time_string=sent_time_string, from_username=from_username, from_profile=from_profile, | |
message=message) | |
print('Saving PM from: {from_username} sent at {sent_time_string}'.format(from_username=from_username, | |
sent_time_string=sent_time_string)) | |
o.write(logged_message) | |
# Archive messages | |
for al in archive_links: | |
print('Archiving {al}'.format(al)) | |
session.get(al) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment