Course of Raku / Regexes and grammars / Grammars / Creating grammars / Exercises / Parse a time

Solution: Parse a time

Here is a possible solution to the task.

Code

grammar Time {
    token TOP    { <hour> ':' <minute> ':' <second> }
    token hour   { \d ** 2 }
    token minute { \d ** 2 }
    token second { \d ** 2 }
}

say Time.parse('09:30:45')<second>;

🦋 You can find the source code in the file parse-a-time.raku.

Output

｢45｣

Comments

1. TOP describes the whole time as three tokens separated by colons. Each part is \d ** 2 — exactly two digits — rather than an open-ended \d+, so a malformed value like 09:30:456 is rejected. (If a part may be unpadded, such as the hour in 9:30:45, write \d ** 1..2 to allow one or two digits.)

2. After parsing, each token is a named capture, so the second is read as <second> on the match. Adding a third part is just one more token and one more colon — the grammar scales naturally.

Course navigation

← Parse a time   |   Tokens and rules →