Mode updates

[gjtorikian]

Fixes #1219
Fixes #1235

[redchair123]

You are replacing the regex for the shell comment to

regex : "[^\{]#.*$"

I think that's also broken, because

echo 123#456

isn't actually commented (so it would print 123#456). I think you probably want " #.*$"

[gjtorikian]

@nightwing That doesn't work either, what about comments that start right at the beginning of the line (without a space)?

Maybe... .?#.*$ ?

[nightwing]

maybe /(?:^|\s)#.*$/?

[redchair123]

@gjtorikian @nightwing what you really need here is a lookbehind.

@nightwing The only downside is that it captures the whitespace character before the hash as well.

@gjtorikian if you are willing to accept a more complicated solution, how about capturing /(^|\s)(#.*)$/ and grabbing the capture block? To show the difference:

> "abc #def".match(/(^|\s)#.*$/)
[" #def"]
> "abc #def".match(/(?:^|\s)(#.*)$/)
[" #def", "#def"]

[gjtorikian]

I don't see how /(?:^|\s)#.*$/ is capturing the whitespace. ?: is effectively telling the highlighter to ignore these parens. It would only mark the rest up.

[redchair123]

(?:) instructs the regexp engine not to create a separate capture group (meaning you won't get a separate whitespace character in the match) but it will be contained in the complete match.

To explain the difference:

>"abc #def".match(/(?:^|\s)#.*$/)
[" #def"] 

Notice how that whitespace in front of # was captured. That's because the entire match includes the non-captured groups. To show an example involving multiple non-capturing groups:
> "foo#bar".match(/(?:^|[a-z])#(?:b|$)/)
["o#b"]
> "#bar".match(/(?:^|[a-z])#(?:b|$)/)
["#b"]
> "#".match(/(?:^|[a-z])#(?:b|$)/)
["#"]
> "foo#".match(/(?:^|[a-z])#(?:b|$)/)
["o#"]

When using lookaheads, you have to select for the part you want:

>"abc #def".match(/(?:^|\s)(#.*)$/)
[" #def", "#def"]

and use the capture (index 1).

[gjtorikian]

Ok, I got it and made the appropriate change/fix.

[nightwing]

Looks good! Merging.