highlite: fix #17890 - tokenize Nim escape seq-s

[a-mr]

Before:

After:

[timotheecour]

• pre-existing (maybe?) but what about other escape sequences (refs https://nim-lang.github.io/Nim/manual.html#lexical-analysis-string-literals), eg:

proc main*()=
  ## \x123
  ## `\x123`
  runnableExamples:
    echo "abc\x1234"
    echo "abc\x1"
    echo "abc\uabcd3" # bug with this one
    echo "abc\u{12345}def" # and this one
    echo "abc\e"
main()

is there possibility to reuse code (possibly via a lib/std/private/lexerutils) to avoid duplication with what compiler already does?

(a plausible alternative is to not highlight things inside string literals, but see also #17722 which would specify the language with which to highlight)

supporting \u etc can also be defered to future work, since it's more rare

[a-mr]

I just enabled the existing code by moving case of first \ into surrounding if — Github obviously shows more changes than has been really done.

I can hardly guess why highlite.nim was written with its own separate lexer.

@Araq ?

[Araq]

Lexer: Transforms "\n" into \10. Skips comments and whitespace (more or less).
Highlighter: Highlights "\n". Highlights comments and whitespace.

Different code for different things.