Update docs if you want: now works with my Atom package

[ghost]

I just incorporated jadelint into my linting package (that was previously using the jade compiler) https://atom.io/packages/linter-jade

[rrdelaney]

If you have any specific feature requests I'd be glad to implement them!

[rrdelaney]

The linter now catches compile errors. I mean, it always did but I just had some silly semantics in there that cause the linter to throw errors.

The error reported is in the normal format of {name, level, filename, line}, so I hope this helps!

[ghost]

Nice, if it's in the same format as the other errors they should be consumed automatically. And I think I can get ride of the jade compiler with that change :)

The other nice things to have in general would be:

• [ ] column numbers
• [ ] return the rule that caused an error (preferably in err.code)
• [ ] why are 'ignore' rules being processed? It seems like they'd just slow down the linter. I'm filtering them out at the moment

Thanks for your efforts!

[edit] And random thing I just noticed -- I went to RPI :smiley:

[rrdelaney]

RPI thing is cool :+1:

• So column number in general would be pretty difficult, and would require some serious reworking of the core linter.
• I can return the code on compile errors really easily.

And I process ignore rules because

• The linter is fast, most rules are just simple checks so I'm not worried about slowing it down really
• You get the whole view of the lint results, regardless of the configuration

I'll do the error code thing now!

[rrdelaney]

So normal errors are formatted

{
    "name": "...",
    "level": "...",
    "filename": "...",
    "line": 0
}

but now compile errors have a special code attribute that returns the jade compiler code (eg JADE:ELSE_NO_IF)