Feature Request: consider splitting dictionary in multiple ones

[ccoVeille]

I like how @Jason3S (hi Jason 👋) structured the cspell project

https://github.com/streetsidesoftware/cspell-dicts

Here I'm not asking for a distinct repository as cspell does, but at least to split a dictionary that went too big

[hippietrail]

Regarding our recent chat on Discord, here's the thread where I posted some related thoughts a few days ago: #473

[Jason3S]

@hippietrail, @elijah-potter

Cool project!

[ccoVeille]

Here is a thread about "should we accept min as a word"

Here we are reaching the logic cspell project faced and why they split their dictionaries, and why I suggested the idea for this project too

A developer would like to have min, while a non-technical writer won't.

Please take a look at how cspell-dicts is organized, and how they can be enabled, disabled in cspell.

Also, cspell-dicts are almost all (maybe all?) with MIT license, so they could be reused by project like yours

Originally posted by @ccoVeille in #596 (comment)

So the idea could be not only to split the dictionary, but also consider external source of data to feed harper dictionaries

[hippietrail]

Here is a thread about "should we accept min as a word"

Here we are reaching the logic cspell project faced and why they split their dictionaries, and why I suggested the idea for this project too
A developer would like to have min, while a non-technical writer won't.
Please take a look at how cspell-dicts is organized, and how they can be enabled, disabled in cspell.
Also, cspell-dicts are almost all (maybe all?) with MIT license, so they could be reused by project like yours

Originally posted by @ccoVeille in #596 (comment)

So the idea could be not only to split the dictionary, but also consider external source of data to feed harper dictionaries

There are going to be lots of identifiers in comments, and also technical terms for things that are not really English words, or don't have settled spellings, etc. think "hash map" vs "hashmap" vs "HashMap" vs "hash_map" etc. not to mention things like printf, stdout, const, etc.

On the one hand we don't want to mark them all as possible errors, and on the other hand we don't want to pollute the English dictionary with lots of that stuff.

Completely separately but related, any class/object/interface/variable/constant/function etc. name will be referenced in comments. So ideally in the future a way to gather all those from the current file or the project/repo into an additional dynamic dictionary would be something to have. But it may also be worth finding English mistakes inside identifiers too - I know I've contributed fixes for that sort of thing to OSS projects before.