Course of Raku / Regexes and grammars / Grammars / The parse tree, make and made / Exercises / Reverse the words

Solution: Reverse the words

Here is a possible solution to the task.

Code

grammar Phrase {
    token TOP  { <word> [ ' ' <word> ]* { make $<word>».made.join(' ') } }
    token word { \w+ { make $/.flip } }
}

say Phrase.parse('hello world').made;

🦋 You can find the source code in the file reverse-the-words.raku.

Output

olleh dlrow

Comments

  1. Each word token makes its own reversed text: its inline block runs whenever a word matches, and make $/.flip stores the word — $/ is the current match — spelled backwards.

  2. TOP matches the words with <word> [ ' ' <word> ]* — one word, then any number of “space then word” — and combines them. $<word> is the list of every word match; ».made pulls out the reversed text each one stored, and .join(' ') rebuilds the phrase with spaces.

  3. That “item, then item, then item…” shape is common enough to have a shorthand: the % separator. Writing <word>+ % ' ' means “one or more <word>, separated by a space”, and matches exactly the same phrases as <word> [ ' ' <word> ]* — just more compactly.

  4. So make and made work at two levels here: the small pieces make their values, and the whole is made from them. Reading .made on the result gives olleh dlrow.

Course navigation

← Reverse the words   |   Action classes →