Closed za3k closed 3 years ago
Not really.
In the initial design of the program, I struggled with a simple, flexible way to store 300000 collected trivia questions.
A sqlite database was considered, but then users would have to know something about sqlite in order to edit questions.
So I decided on a stupid format for the files, but then I didn't want a monolithic file that was 300000 lines long, so we broke them up.
The algorithm for question selection picks a random file, then picks a random line from the selected file, which seemed like a good randomization.
But as I've grown up, I think a monolithic file would have been fine. I also think it would make it simpler to have different game modes, like 'sequential' (ie: start at the beginning of the file and walk each question to the end and repeat) or random, but I was learning python and programming at the time so it is what it is.
I've also been considering a python 3 port, just haven't found the time yet for this hobby program. Got a 3d printer and been messing with that for a bit. :)
Oh, and the algorithm I used is very sparing on memory. One should not load the whole file into memory, that's just a waste.
No, the algorithm you've used loads the entire file into memory (lines = fd.read().splitlines()
), which is why I thought you might have split the file into many files.
(P.S. have fun with your 3d printer!)
lol, goes to show you how long it's been since I've read my own code.
Then yeah, ram usage would be a problem then. Wasn't my intention, but like I said, I was a student at the time.
Is there any reason the question data is in multiple files? Is this to reduce RAM usage or something?
Was considering doing a Python3 port (to fix https://github.com/rawsonj/triviabot/issues/78) and saw some unrelated stuff that might be worth cleaning up.