supposedly / code-jam-1 Goto Github PK

On human wikipedia, randomness can be reasonably achieved by visiting Special:RandomInCategory/<category name>. We'll have to check and see how this is usable from within the API. Presumably it'd deal with the hardcoded.categories thing I had to add.

-- unless there's, potentially, a better way than RandomInCategory for us to get a random snek?

One of the most-important parts of the info fetching is where Wikipedia's HTML-styled text is converted to a Markdown representation of itself, so that Discord can display everything properly. Only problem is that this is... hard.

I poked around github for a bit yesterday and found html2text, which is nice and works okay for some pages (see bot.get('black mamba')), but it screws up... a lot... on some others (see bot.get('pythonidae'), and that's not even the worst). Following it around with re.sub() and str.replace() doesn't sound like too Pythonic (nor logical) an endeavor.

The two ways around this I can think of, while still fulfilling the end goal ofc of converting HTML-encoded text markup to Discord-markdown-formatted text, are (1) finding a better Python markdown parser and (2) going full regex from start to end.

The second would work swimmingly once finally implemented, but (speaking from experience!) it'd be somewhat tricky to debug and get working and it might sap up a few precious hours of our work time.
The first would be wonderful, but I don't think it's likely. Haven't found anything myself yet, at least.

Thoughts? Maybe other methods of parsing?