updates to bring back to life, more to come soon#1
Conversation
- fixes for sprocket attrs - detach logic from cmd files for cleanliness - restructure for proper packaging (WIP)
MLEBot/task_sprocket.py
Outdated
| known_guild = next((x for x in self.bot.guild_ids if str(x['id']) == interaction.guild.id.__str__()), None) | ||
| if not known_guild: | ||
| return None | ||
| return next((x for x in self.data['sprocket_teams'] if x['Franchise'].upper() == known_guild['team'].upper()), |
There was a problem hiding this comment.
You use this pattern often, so it's worth noting that it's horrifically inefficient. Just for clarity:
(x for x in data['dataset_key'] if x['prop_key'] = thing_were_searching_for)
is the same as
for x in data['dataset_key']:
if x['prop_key'] == thing_were_searching_for:
return x
I.e., all of these methods are using linear search over potentially large datasets to find bits of data. Python (and every other modern language/runtime combo) has a muuuuuch better choice for this type of operation: the hash map, or dictionary.
Whenever the datasets are fetched, it would make a great deal more sense to spend the time to process them into datasets, keyed by whatever you're going to search with later (like 'Franchise', 'discord_id', 'name', 'member_id', etc). Then, all f these methods here on lines 146-205 become lightning fast dictionary lookups, instead of linear searches through arrays.
There was a problem hiding this comment.
Worth noting what I mean by slow and fast: Linear search is O(n), n being the size of the dataset being searched (which in the case of player_stats is many millions). Dictionary lookup is O(1).
There was a problem hiding this comment.
with my latest commit i believe i have mostly solved this issue. there's a few more tweaks i'm sure i could make in the future, but this should do for now
When pulled into main, will distribute to Roxanne.
|
sorry, idiot with git. getting used to it still |
addition of unit-tests that i would like another dev to start on another branch.
…increase later but wanted to save time for now).
2 participants
this gets the bot booting and performing the same functions as before.
now, to bring more robustness / features.