Ruby/Python: regex parser: group sequences of 'normal' characters

[aibaars]

The Javascript library groups sequences of normal characters, it would be good to make the behaviour of the regex libraries behave as similarly to each other as possible.

[aibaars]

@yoff Could you have a look at this?

[yoff]

Great to see this, thanks for doing it!
I believe that what you are doing, is changing character from being either a simpleCharacter or an escapedCharacter to now being either a maximal simpleCharacter run or an escapedCharacter. I have therefor suggested to rename some of the predicates.

I also wonder if the computation of maximal runs is standard or if a more efficient one might exist, but if the performance looks fine, then we can punt this question.

[yoff]

This (and others like it) looks like an existing bug. I think it is due to this line in simpleCharacter not considering qualifiers longer than one character (such as the multiples notation {n, m}).

[tausbn]

A few comments. Otherwise this looks good to me. 👍

[yoff]

I see from the failed tests that something is off with using simple chars instead of normal chars. I would like to understand what that is, but it does not have to be during this PR.

[aibaars]

I see from the failed tests that something is off with using simple chars instead of normal chars. I would like to understand what that is, but it does not have to be during this PR.

I think simpleCharacter does not exclude characters that are part of escape sequences.

[aibaars]

@yoff @tausbn I think I addressed your comments.

[yoff]

LGTM, is the DCA run from earlier still valid?

[yoff]

I think simpleCharacter does not exclude characters that are part of escape sequences.
Ah, yes, that might be it :-)

[aibaars]

LGTM, is the DCA run from earlier still valid?

Yes, the force-push only fixed a test case.