Packages: [Markdown] Not closing fenced code block with invalid JavaScript code

[evandrocoan]

````
Bug no markdown;

class Research extends React.Component {...})

deve-se

````

[FichteFoll]

Works for me. Make sure you're using the correct syntax.

[evandrocoan]

I think it is because my JavaScript syntax a ReactJS is from https://packagecontrol.io/packages/JSCustom package.

[Thom1729]

I'll take a look later and see if I can reproduce.

On Wed, Oct 24, 2018, 12:59 evandrocoan [e-mail] wrote:

I think it is because my JavaScript syntax a ReactJS is from
https://packagecontrol.io/packages/JSCustom package.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
https://github.com/sublimehq/Packages/issues/1750#issuecomment-432741655,
or mute the thread
https://github.com/notifications/unsubscribe-auth/AQLEI5DFAzsSxyqS1zqd8jlNK0xiRjaCks5uoJyBgaJpZM4X3Fc3
.

[evandrocoan]

Just one more thing. I am using the latest version on this master branch, but I think I have added a customization to my Markdown fork https://github.com/evandroforks/Packages/blob/master/Markdown/Markdown.sublime-syntax because on FichteFoll picture, there is not highlighting to the embedded JS code.

[deathaxe]

The issue is caused by popping off from the statements context if a stray brace/bracket/paren is detected and statements being included in the main context.

https://github.com/sublimehq/Packages/blob/14a35f641f348b4ee0ead558e5ee474f7255b50f/JavaScript/JavaScript.sublime-syntax#L288

As removing the pop: true does not break any existing test, I wonder why it exists.

Another solution was to not directly include statements in main, but add its body without the pop: true

[evandrocoan]

I think @Thom1729 could help. After running investigate:

That line seems to be added by him earlier this year on the commit:

1. 08/02/2018 19:39 JavaScript - Improved handling of unexpected close punctuation c581495e33d3135dc3b0bdada2b123bb84cf9317

[FichteFoll]

See #1755.

[Thom1729]

Should be an easy fix. I'll try to take a look this weekend.

[evandrocoan]

If @Thom1729 can "fix" the JavaScript syntax, then, I think both fixes (pull #1755 and @deathaxe suggestion for Javascript) are welcome because JavaScript should behave better when included by other syntaxes and Markdown would behave better when including some syntax with the same problem.

[Thom1729]

I'm not sure there's a general solution on the Markdown side. What we want to do is wrap the embedded context in a guard, but I think that would have to be done manually for each embed.

[FichteFoll]

I was just referring to the discussion in #1755, not to the solution, though that might've been misleading. Since I liked the comment by @deathaxe mentioning how the proper fix would be in the JavaScript syntax, I assumed it would be clear that I support that solution instead of the one proposed in the PR.

My bad if this caused confusion.

[deathaxe]

Sorry, didn't count the votes. :-D