Trying to create a lexer for kind-of-classic 80's BASIC

[main Alexey]

RegEx engine in the lexer is not the same as RegEx engine in the syntax parser. wiki:

=Regular expressions=

Lexer parser uses EControl regex engine. You use this regex syntax only in the "Lexer Properties" dialog in SynWrite, not in the CudaText normal usage. EControl regex has custom features:

* class \A: begin of the document
* class \Z: end of the document
* class \l: Unicode word-char except the underscore char
* class \L: inversion to \l
* lookahead/lookbehind can find match of variable length
* modifier (?r): \w catches all Unicode letters too
* modifier (?g): greedy

CudaText search/replace uses [https://regex.sorokin.engineer/en/lates ... sions.html TRegExpr engine]

[main Alexey]

as alternative way to make the lexer:
post your wishes to the lexer (my version, which I attached). after one iteration from me, post next wishes.