Course of Raku / Regexes and grammars / Grammars / What is a grammar

The grammar keyword

A grammar groups named regexes together under one name, much as a class groups methods. You declare it with the grammar keyword:

grammar Greeting {
    token TOP  { 'Hello, ' <name> '!' }
    token name { \w+ }
}

Inside the grammar, each named pattern is written with token (a kind of regex you will study in detail shortly). One token, by convention called TOP, is the starting point — it describes the whole thing. Here TOP says “the text is Hello,, then a name, then !”, and name says what a name looks like.

The tokens can refer to one another by name, exactly like the named regexes of the previous topic. TOP uses <name>, which keeps the grammar readable: each rule has one clear job.

To run a grammar against a string, call its .parse method:

my $m = Greeting.parse('Hello, Anna!');
say $m<name>; # ｢Anna｣

The next section looks at TOP and .parse more closely. For now, the idea to take away is simple: a grammar is a named collection of named regexes that together describe a whole piece of structured text.

Course navigation

← Quiz — Named regexes   |   Quiz — What is a grammar →

💪 Or jump directly to the exercises in this section.