Lucy::Analysis::RegexTokenizer(3) Split a string into tokens.

SYNOPSIS


my $whitespace_tokenizer
= Lucy::Analysis::RegexTokenizer->new( pattern => '\S+' );
# or...
my $word_char_tokenizer
= Lucy::Analysis::RegexTokenizer->new( pattern => '\w+' );
# or...
my $apostrophising_tokenizer = Lucy::Analysis::RegexTokenizer->new;
# Then... once you have a tokenizer, put it into a PolyAnalyzer:
my $polyanalyzer = Lucy::Analysis::PolyAnalyzer->new(
analyzers => [ $case_folder, $word_char_tokenizer, $stemmer ], );

DESCRIPTION

Generically, ``tokenizing'' is a process of breaking up a string into an array of ``tokens''. For instance, the string ``three blind mice'' might be tokenized into ``three'', ``blind'', ``mice''.

Lucy::Analysis::RegexTokenizer decides where it should break up the text based on a regular expression compiled from a supplied "pattern" matching one token. If our source string is...

    "Eats, Shoots and Leaves."

... then a ``whitespace tokenizer'' with a "pattern" of "\\S+" produces...

    Eats,
    Shoots
    and
    Leaves.

... while a ``word character tokenizer'' with a "pattern" of "\\w+" produces...

    Eats
    Shoots
    and
    Leaves

... the difference being that the word character tokenizer skips over punctuation as well as whitespace when determining token boundaries.

CONSTRUCTORS

new( [labeled params] )

    my $word_char_tokenizer = Lucy::Analysis::RegexTokenizer->new(
        pattern => '\w+',    # required
    );
  • pattern - A string specifying a Perl-syntax regular expression which should match one token. The default value is "\w+(?:[\x{2019}']\w+)*", which matches ``it's'' as well as ``it'' and ``O'Henry's'' as well as ``Henry''.

INHERITANCE

Lucy::Analysis::RegexTokenizer isa Lucy::Analysis::Analyzer isa Lucy::Object::Obj.