#!/usr/bin/env perl

# words - list all words (a-zA-Z0-9 only) in the input, translated to lowercase
#
# $Id$

# -x for extended filtering (i.e., fewer words)

use warnings;
use strict;

use Getopt::Std;

my %opt;
getopts( 'hux:', \%opt );

$opt{h} and HELP_MESSAGE();

my $x = '\W';

sub HELP_MESSAGE
{
  print STDERR <<ZZ;

Usage: $0 [-x pattern] [file1 [file2 [...]]

to print the words in the input, translated to lowercase.

Options:

  -u

    Unbuffered output: print the output after every line.

  -x pattern

    The Perl regular expression to match text that isn't part of words.
    The default value is $x (everything that is no ASCII letter,
    digit, or underscore).

Examples:

   To exclude digits and underscores from words, use

      -x  '[^a-zA-Z]'

    or

      -x '[\\W^\\d_]'

   To make words everything that isn't whitespace, use

      -x  ''\\s'

ZZ
  exit(0);
}

$| = !!$opt{u};    # if -u, flush after every line

defined $opt{x} and $x = $opt{x};

my $rx;
eval { $rx = qr($x) }
 or die "invalid pattern for -x: $opt{x}\n";

while (<>)
{
  chomp;
  map { length and print "\L$_\n" } split /$rx/;
}