#!/usr/bin/perl

=begin metadata

Name: strings
Description: extract strings
Author: Nathan Scott Thompson, quimby at city-net dot com
License:

=end metadata

=cut


# Copyright 1999 Nathan Scott Thompson <quimby at city-net dot com>

=head1 NAME

strings - extract strings

=head1 SYNOPSIS

	strings [-afo] [-n length] [-t {d|o|x}] [file ...]

=head1 DESCRIPTION

C<strings> prints strings gleaned from the given files
(or from standard input.)  By default strings of less than 4 characters
are ignored.  A string is considered to be any sequence of graphical ASCII
characters (plus space and tab).

Options:

  -f            Print the file name with each string.

  -n length     Set the minimum length of strings to print; 4 by default.

  -o            Print each string with the octal offset in the file.
                (Same as -to).

  -t {d|o|x}    Print each string with the offset in the file;
                specify decimal, octal or hexadecimal respectively.

The -a option is included for compatibility with older versions but does
nothing.

=head1 BUGS

  Things are seldom what they seem,
  Skim milk masquerades as cream;
  Highlows pass as patent leathers;
  Jackdaws strut in peacock's feathers.

=cut

$usage = "usage: strings [-afo] [-n length] [-t {d|o|x}] [file ...]\n";

use vars qw($opt_a $opt_f $opt_o $opt_n $opt_t);
use Getopt::Std;
getopts( 'afon:t:' ) or die $usage;

$opt_n ||= 4;
$opt_t ||= 'o' if $opt_o;
if ( $opt_t )
{
    $opt_t =~ /[dox]/ or die $usage;
    $offset_format = "\%07$opt_t ";
}

# Consider all graphic characters plus space and tab to be printable.
# Escape all punctuation characters out of paranoia.

$punctuation = join '\\', split //, q/`~!@#$%^&*()-+={}|[]\:";'<>?,.\//;
$printable = '\w \t' . $punctuation;
$chunksize = 4096; # whatever

for my $filename ( @ARGV )
{
    open( IN, '<', $filename ) or die "Can't open $filename: $!\n";
    binmode IN;
    $offset = 0;

    while ( $_ or read( IN, $_, $chunksize ) )
    {
        $offset += length($1) if s/^([^$printable]+)//o;
        $string = '';

	do {
            $string .= $1 if s/^([$printable]+)//o;
        } until ( $_ or ! read( IN, $_, $chunksize ) );

        if ( length($string) >= $opt_n )
        {
            print $filename, ':' if $opt_f;
            printf $offset_format, $offset if $opt_t;
            print $string, "\n";
        }
        $offset += length($string);
    }
    close IN;
}