A character class matches one character from a constrained set. The word character class, written \w, matches letters, numbers, and underscore, making it useful for recognizing word-like clumps without accepting every possible character. section 11.3.3 section 11.7.1
Scanning word characters
Ruby can scan each word-like run with /\w+/: "one two three".scan(/\w+/) produces "one", "two", and "three". The + quantifier expands the one-character class into one-or-more consecutive matches. section 11.7.1 section 11.8
Use character-class syntax when the allowed characters need to be controlled:
- Write an explicit class such as [dr] to match either d or r, but only one character. section 11.3.3
- Use ranges such as [A-Fa-f0-9] for hexadecimal digits, or negate the class with a leading caret to match anything else. section 11.3.3
- Prefer \d for digits and \w for word characters when their predefined sets express the intended constraint. section 11.3.3
>> "testing 1 2 3 testing 4 5 6".scan(/\d/)
=> ["1", "2", "3", "4", "5", "6"]
A runnable scan example shows \d matching individual digits and demonstrates the same single-character behavior used by \w. section 11.7.1
>> str = "This is an epsilon: \u03a3"
=> "This is an epsilon: Σ"
>> str.match(/\p{Greek}/)
=> #<MatchData "Σ">
A Unicode example shows how property-based classes can match a Greek character when a narrower shorthand does not express the desired character set. section 11.3.3
