Конвертер из ЦСЯ Markdown в

[0xGeorgii]

• HTML
• LaTeX
• XML
• PDF?

[0xGeorgii]

@pgmmpk @typiconman Can I mark HTML, LaTeX, XML as 'done' or they require more development?

[pgmmpk]

pdf is irrelevant as it will be done via latex: md = > latex => pdf

[typiconman]

Can we run some tests MD -> XML -> MD and XML -> MD -> XML, to make sure the processes are lossless?

[0xGeorgii]

we have some samples, maybe it makes sense to write some script. however, somebody should verify a result.

[pgmmpk]

MD=>XML needs some work (on specs and implementation). XML => MD is not even started...

…

On Sat, May 11, 2019 at 3:14 AM George Plotnikov ***@***.***> wrote: we have some samples, maybe it makes sense to write some script. however, somebody should verify a result. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#5 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAELA4TCKHEIJWRS5SVCDD3PUZW6RANCNFSM4HMF4WOQ> .

[typiconman]

Wait, I though we already had a converter here.

Also, I think that we now have a lot of duplicated information in cumd and markdown-it-church-slavonic.

[pgmmpk]

Thank you for the reminder - I forgot about existence of Python cumd, shame :/.

The MD=>XML part better be in JS (for the sake of VSCode). Maybe we need to port XML=>MD part to JS as well (so that we can unit-test the roundtrips).

Alternative would be to do the heavy lifting in Python, but integrate Python service into VSCode.

Lets brainstorm this - what the best way?

1. Have JS for MD=>XML=>MD. Pros: easy to put into VSCode and on the web. Cons: can not be directly used from Python (i.e. nltk and other lang modeling tools, most of which are Python)

2. Have Python MD=>XML=>MD. Pros: (see cons above), Cons: painful integration with VSCode

3. Have both. Cons: hard to keep in-sync. One will need to be marked as "reference implementation", while other playing "catch-up".

My initial thoughts:

• doing (3) is impractical as it duplicates the effort
• do not have strong preference between (1) and (2). Python is easier for me personally, but that will require some research how to integrate with VSCode (its definitely possible, just need to learn the plumbing).

[typiconman]

As I understand it, our texts will be stored in MD. On the website, we will display them in HTML. So, we need MD -> HTML (trivial form of MD -> XML) in a method that can be easily automated on a server.

The other application will be our API, where a user can request a text in any format (MD, XML, LaTeX). Again, the conversion needs to be easily automated on a server.

[0xGeorgii]

Conversion MD -> HTML is the MD's purpose. I don't see the reason why it's necessary to use py, what for?

[typiconman]

How do you store text as MD, but display HTML on a webpage for the user?

[0xGeorgii]

@typiconman something like that. there are many libs for that

[0xGeorgii]

Oh, the problem is in custom MD. I see. But there are only ~500 lines of code. Maybe it makes sense to re-write it on JS or TS?