Whitespace afer token

[anderay]

If we change Hello example this way
Person: 'person' name=ID 'age' age=INT;
then following example is perfectly valid

    person personone
    age 11person persontwo age 22
    Hello persontwo!

any possibility (more generic, parser level) other than "validation rule that ensures this based on the CST" mentioned in #1447?

[msujew]

Hey @anderay,

parser level

I don't think it's solvable on the parser level.

more generic

There are definitely ways to solve this generically. You can generically iterate through the $cstNode starting from the root and assert that after any INT token, there's some space until the next token (assuming the next token is some sort of word-like token).

[ydaveluy]

Hey @anderay ,

You can add a word boundary marker \b at the end of your INT rule , something like:

terminal INT:/[0-9]+\b/;

edit: and before:

terminal INT:/\b[0-9]+\b/;

[msujew]

I've thought of that as well, but that's a very heavy handed solution, as it will lead to lexer errors (which likely lead to parsing errors as a consequence, which then results in an incomplete AST). Custom validation errors are usually a much better solution. TypeScript will do something similar: