A web scraping python script to generate an ical calandar file from a list of Facebook events pages. The goal is to retrieve events without having to spend hours on FB. Unfortunately Facebook has won and the majority of events are announced and shared on FB, even when the bands or venues have their own website.
Facebook Graph API no longer supports events, so this script uses selenium. The use of selenium and web scarpping in general is contrary to Facebook's terms of use. If you remain reasonable about the number of pages, the frequency of use, if you don't use selenium in headless mode, tor or other proxy, it should be a okay. If you scarpe thousands of page every 30 minutes in headless mode from a tor gateway, at best the response time will increase until it become unusable or the IP adress will be blacklisted.
For the kinkiest among you, the script supports authentication to access pages reserved for adults. Multiple factor authentification is supported (if you are not in headless mode). If you use the authentication function and play dumb, you risk having your account slowed down, see the number of captcha increase or even get ban. You've been warned, there's no need to come crying to me afterwards.
Facebook uses many browser identification techniques to detect that Firefox is running headless, including SVG and WebGL rendering. So they know that something odd happens when you use the headless mode.
I use this script once a day for about 40 pages and I haven't had any problems for the moment.
Scrap on specific event page:
python3 events-scraper.py -e https://www.facebook.com/somepage/events/
Scrap all the page listed inside a file:
python3 events-scarper.py -f event_velo.txt
Scrap all the page listed inside a file and login with the credential inside credential.txt:
python3 events-scraper.py -f event_velo.txt -c
usage: events-scraper.py [-h] [-e EVENTS [EVENTS ...]] [-f FILE] [-o OUTPUT_FILE] [-c] [-hl] [-q]
Non API public FB event miner
optional arguments:
-h, --help show this help message and exit
-e EVENTS [EVENTS ...], --events EVENTS [EVENTS ...]
List the pages you want to scrape for events
-f FILE, --file FILE file with the list of pages to scrape for events
-o OUTPUT_FILE, --output OUTPUT_FILE
output ical file name
-c, --credentials use credentials from credentials.txt
-hl, --headless run FireFox in headless mode
-q, --quiet silence output
- Selenium
- icalandar
- Tickets url isn't scrapped
- Recurrent events are not supported (Pub quiz every sunday)
- Day saving time was not tested
- URL inside the description of the events are removed for no reason ( selenium getAttribut("text") don't like links)
- need a crash or layout change alert system
- Credential option work only with the classic layout and don't detect the new FB layout