rlucioni / craigbot Goto Github PK

Public transportation stops are currently the only supported POIs. Add support for extending this with additional POIs (e.g., grocery stores, gyms).

An infinite loop with a configurable sleep time is currently used. Cron would be cleaner. For reference, see http://stackoverflow.com/questions/37458287/how-to-run-a-cron-job-inside-a-docker-container.

Zillow might be a better data source than Craigslist.

Listings without images aren't interesting. There should be an option to exclude them.

Users should be notified when the bot can't do its job anymore. You can tell if an IP is blocked by making a request to Craigslist and checking if the status code is 403.

>>> import requests
>>> r = requests.get('https://www.craigslist.org')
>>> r.status_code
403

For example, we may want to hear about listings whose location contains the word "harvard" unless the location also contains the words "allston" or "longwood."

https://boston.craigslist.org/search/gbs/gms

http://vignette4.wikia.nocookie.net/southpark/images/c/c2/Craig-tucker.png

Will probably want to use icon_url. See Slack docs for more.