@jwang1919 so I think we might be at a point where we should just contribute a fix upstream to PrettyCSS. We’ve done so before and the process was pretty painless—the library is well-factored and it’s easy to add tests and such.

I would lean in that direction because PrettyCSS itself has a full tokenizer and lexer built in, so adding support for a given syntactic construct is much more robust and expressive than trying to work around a parser error using regular expressions. (I admit to being a little lost looking at the regexps you introduce here, and I consider myself reasonably competent with regexps).

WDYT?

@outoftime Yeah that sounds like a better option to me as well.
I'll revisit this branch once I submit the PR and it gets accepted in the PrettyCSS library.

The regexp was a difficult come up with, but not too hard to interpret.
It looks for any strings that has a comma followed by another comma with any number of blank spaces in between.

, // look for comma
( // start lookahead assertion
?= // positive assertion, only match for these things
\s* // zero to many blank spaces
, // a comma
) // close lookahead assertion

@jwang1919 gotcha. yeah I’ve only worked with lookaheads a couple times, that’s what threw me off.

FWIW, we don’t necessarily need to wait for PrettyCSS to integrate your patch—we could temporarily target https://github.com/popcodeorg/PrettyCSS (which I think I did in the past while waiting for PRs to get merged)

Okay, made the changes and submitted PRs to both the original branch and the popcode fork. Once those are dealt with, I can change this branch's code.

@outoftime When possible, please review again. I have pushed the fix onto the PrettyCSS library (the maintainer merged it pretty quickly) and I've updated the PrettyCSS library.