Ignore MyST syntax labels

[stevepiercy]

Check for existing issues

• [X] Completed

Describe the feature

In my accept.txt, I have a regular expression that I expect to match labels in MyST markup. I also have proper nouns, such as Plone and Volto.

accept.txt

Plone
Volto
^(.*-label)=$

Here is an example label.

(contributing-volto-sign-and-return-the-plone-contributor-agreement-label)=

Here is my regular expression demo:

https://regex101.com/r/MqBJoC/1

When proper nouns appear lower-cased in the labels, Vale returns errors.

 36:15   error       Use 'Volto' instead of          Vale.Terms             
                     'volto'.                                               
 36:41   error       Use 'Plone' instead of          Vale.Terms             
                     'plone'.                                         

If I remove the proper nouns from accept.txt, then the terms Volto and Plone are flagged as misspellings everywhere except in the labels. I also get the exact same result when I also remove the regular expression from accept.txt.

It seems that my only option is to use proper casing in labels. I would like to avoid the work of changing all of our MyST labels. Is this possible?

[jdkato]

This is an issue because MyST is not one of Vale's supported formats. As a result, your options are limited (most of Vale's scope-related features will not work) but you could do this:

\sPlone\s
\sVolto\s

This will generally only flag instances that occur in a "prose-like" context. In any case, I'd say that your expectations when using Vale with unsupported formats should be fairly low when it comes to ignoring markup.

[stevepiercy]

Thanks for the suggestion. In accept.txt, I tried replacing:

Plone
Volto

with

\sPlone\s
\sVolto\s

but that generated over 600 additional errors. Here's example source and result:

- `mywebsite.com-volto` starts the Node.js process that's responsible for Volto {term}`server-side rendering`.

 57:75  error  Did you really mean 'Volto'?  Vale.Spelling 

It appears that this kind of regular expression is not supported in our custom accept.txt. It supports Plone[a-z]{4} and accepts Plonetest.

Aha! This worked:

-{0,1}plone-{0,1}
-{0,1}vlone-{0,1}

There's more than one way to skin a regular expression!

I am still concerned that neither of our earlier regexes worked. Should I keep this issue open or close it?

[jdkato]

but that generated over 600 additional errors. Here's example source and result ...

Ah, it likely "solved" the issue with Vale.Terms but caused another one with Vale.Spelling.

I would say that what we're doing here is generally a misuse of vocabulary files: they're designed to work with supported markup formats, not workaround false positives caused by unrecognized ones.

The real solution here would be to add support for MyST, but I can't put a timetable on that.

[stevepiercy]

Thanks for the explanation. I'm happy to help bring MyST support to Vale to the best of my ability. A cursory look at support for reStructuredText as an example to adapt for MyST was over my head.