[Architecture]: Capture high-level as-is class diagram

[joshbruce]

Marked version: 0.3.17

Markdown flavor: n/a

Proposal type: other

What pain point are you perceiving?

1. The code is difficult to read per 🚧[Experimenting]: Refactoring marked js🚧 #1102.
2. Feature deprecation strategy #1114 comment from @UziTech recommends "best way" (sorry, knee-jerk reaction whenever someone says "best way" 😆) related to a possible to-be architecture.
3. Related to vision alignment: [Vision alignment]: Comparative benchmarking - low-level #965
4. And more...wasn't planning to spend this long writing this... 😝

#1089, #1096, #1097

What solution are you suggesting?

Capture at least a basic (no members or functions necessary) class diagram (like??) model illustrating Marked's current architecture.

Think @intcreator should get first refusal since he seems most passionate about needing an architecture (and architect) to be able to get up to speed in understanding the code easier. I know you said you've only had a couple year's experience with JS but that'll be fine, that's why we have review.

Let us know?? (No harm, no foul.)

[joshbruce]

Huh? Fascinating.

You can tag two projects now.

Projects is a new-ish feature in GitHub...not sure it's always been able to do that.

Don't think you can do two milestones though.

[snowyu]

Use typescript make the Classes and helper functions export more clear and useful.
And make plugins more easy. It is very hard to write plugin for marked currently.
I have to use the kramed to do a simple task with a little patch, see the hacked code front-matter-markdown, The kramed still need to reconstruct too.

Tokenizer, Parser, Render Classes and userful helper functions.

here is a good starter: https://github.com/alexjoverm/typescript-library-starter

[UziTech]

If you are looking for marked written in TypeScript you can look at the fork marked-ts

[snowyu]

I have read it just now, but it is not pretty code. The RegExp passed everywhere.

[snowyu]

The best way should be inherited from a common tokenizer/Lexer, such as Moo Tokenizer Generator or PEG.js.