[Rules] Improving removal of comments

[tonhuisman]

Rules where quite rigorously stripped from comments, interfering with url's and // when used in strings.
Features:

• Don't cut off url's like http://myserver.local:8080/some/path after the colon
• Don't cut off quoted strings that contain //, the slashes are ignored when between matching quotes

[uzi18]

I see here regex usefull

[TD-er]

But is parsing regex fast?
At least the required library for it isn't smaller, but it could be used for more parts in the code, so the extra bytes for it may be worth it as other code could be simplified.
However I don't think it will be faster.

[uzi18]

@TD-er check first if it is not already included (by any libs?). Regex is always slower.
Fought it is feature to remove comments by one click ;)
As coders know code without comments are headache

[TD-er]

I'm not that fluent in regex, so not sure if reading a regex will be done in less time than my internal grey mass C++ parser needs :)

[uzi18]

There are some pages to test it over internet.

[tonhuisman]

While I have quite some experience with regular expressions, this one is a bit complex, because of the 3 quotes that can be used to delimit a value, and have to match start/end to apply, within where the // is to be ignored. Having a dual character comment start isn't helping also.
Just for giggles I'll try to create one Gave up after 10 minutes, way too complex to match the start/end quotes.
My go-to site for validating (and explaining) a regex is https://regex101.com

[uzi18]

@tonhuisman maybe you could change these quotes temporary with one known just to find position of last // with comment

[tonhuisman]

@uzi18 I expect to have nailed it with current code, wouldn't want to complicate it further, and replacing quotes isn't going to help finding the comment-start quicker or easier, IMHO.