Store string literals in their unescaped form about effekt HOT 6 CLOSED

I propose we use \u<DIGIT>+ for Unicode character literals and likewise \u<DIGIT>+ inside strings, simply because then we can use StringContext.processEscapes for the latter and don't have to take care of it ourselves

Sounds fine to me, at least as a first step, especially if we can reuse StringContext.processEscapes 👍

Given time, we could actually do the super-correct solution of parsing/lexing the escapes ourselves and then storing the escaped version directly (so without any need to call StringContext.processEscapes at all), but I'd be very happy if we could postpone that for later in order to prioritise the merge of #495 :)

from effekt.

Some things to consider:

• What to do about the ANSI escape sequences we currently have here?
• Related to that, I find it weird that we have character literal syntax of \u{001b} but then also have "\u001b[30m". We should decide on a common notation here.

from effekt.

The difference is due to the different handling of escapes in backends. The right thing to do is exactly to have one format and then translate it when rendering out in backends.

from effekt.

I propose we use \u<DIGIT>+ for Unicode character literals and likewise \u<DIGIT>+ inside strings, simply because then we can use StringContext.processEscapes for the latter and don't have to take care of it ourselves

from effekt.

@jiribenes you had some opinion on the syntax of escapes, didn't you?

from effekt.

Resolved by #495, tested with repro from #494 (comment) 👍

from effekt.