Course of Raku / Regexes and grammars / Grammars / Creating grammars

Parsing with parse

To run a grammar, call its .parse method with the string to analyse:

grammar Pair {
    token TOP   { <key> '=' <value> }
    token key   { \w+ }
    token value { \w+ }
}

my $m = Pair.parse('x=5');
say $m<key>;   # ｢x｣
say $m<value>; # ｢5｣

.parse returns a match object when the grammar matches the entire string, and Nil when it does not. Because the whole input must match, there is no need for ^ and $ anchors — .parse adds that requirement for you:

say Pair.parse('x=5').defined; # True
say Pair.parse('x=').defined;  # False

The match object works just like the ones you met earlier. Each token used in the grammar becomes a named capture, so $m<key> and $m<value> give the matched parts. The tokens nest, so a grammar builds a small tree of matches — the subject of a later section.

Course navigation

← The TOP rule   |   Quiz — Creating grammars →

💪 Or jump directly to the exercises in this section.