Thanks, will give it a shot. I actually ended up doing a substitue
([^\x09-\x0d\x20-\x7e]) on the whole file string and then going
line by line. Almost similar to what you have but ugly.
Thanks.
Sumit
----- Original Message -----
From: <[EMAIL PROTECTED]>
To: "Sumit" <[EMAIL PROTECTED]>
Cc: "active" <[EMAIL PROTECTED]>
Sent: Wednesday, August 20, 2003 1:06 PM
Subject: Re: read binary files
> Hi,
>
> I recently put something together. Hope it's still useful. It even has
> some POD doc ... ;-)
>
> Ekkehard
>
> [snip]
>
> #!perl -w
> use English;
> use Getopt::Std;
> use strict;
>
>
#==========================================================================
> =pod
>
> =item @str strings($file, $maxbytes, $translate, $minchar, $offset)
>
> This method is similar to the GNU version of the strings utility. It
> reads the file $file upto the $maxbytes byte and extracts all strings
> it encounters. A string is here a sequence of at least $minchar bytes
> in the range 0x40 - 0x176 or any umlaut followed by a byte not out of
> those character set. If $translate is set to 1 any umlauts and german
> 'sz' found in IBM encoding are translated to ISO-8859-1 encoding. If
> the $offset parameter is not 0 each element in the returned list
> contains the offset within the file before the string and - tab
> separated - the string found. The offset is given in either decimal
> (d), octal (o) or hexadecimal (x) notation.
>
> =cut
>
#==========================================================================
> sub strings
> {
> my ($file, $maxbytes, $translate, $minchar, $offset) = @ARG;
>
> return unless -r $file;
> return unless $translate =~ /^[01]$/;
> return unless $minchar =~ /^\d+$/;
> return unless $offset =~ /^[0dox]{1}$/;
>
> $maxbytes = 3000 unless defined $maxbytes;
> my $data;
>
> open IN,"<$file" or return;
> binmode IN;
> read IN, $data, $maxbytes or return;
> close IN or return;
>
> # AE OE UE ae oe ue sz
> my $IBM = '\216\231\232\204\224\201\341';
> my $ANSI = '\304\326\334\344\366\374\337';
>
> # What we consider as 'ascii':
> my $ascii="\040-\176$IBM$ANSI";
>
> my @strings;
> while ($data =~ /([$ascii]{$minchar,})[^$ascii]/g)
> {
> if ($offset ne "0")
> {
> push @strings,
> sprintf("%$offset",pos($data)-length($1)-1)."\t$1";
> }
> else
> {
> push @strings, $1;
> }
> }
>
> # Translate IBM to ANSI encoding. We can not use the variables
> # here as the translation table is generated at compile time (see
> # tr documentation) !!!
> if ($translate == 1)
> {
> foreach (@strings)
> { tr [\216\231\232\204\224\201\341]
> [\304\326\334\344\366\374\337]; }
> }
>
> return @strings;
> }
>
>
> our $opt_b = 0;
> our $opt_n = 4;
> our $opt_t = 0;
> our $opt_T = 0;
> our $opt_h;
>
> getopts('b:n:t:Th');
>
> if ($opt_h || @ARGV==0)
> {
> print <<HEND;
>
> Usage: pstrings [-n <minchars>] [-b <maxbytes>] [-t radix] [-T] [-h]
> <filename>
>
> -n at least minchars characters make up the string
> -b at most maxbytes are read from the file if specified.
> Otherwise the entire file is read.
> -t Print the offset within the file before each string. The single
> character argument specifies the radix of the offset: o for octal,
> x for hexadecimal, or d for decimal.
> -T translates IBM encoded umlauts to ISO encoded ones
> -h shows this help
>
> The filename has to be in a NTish form.
>
> HEND
>
> exit 0;
> }
>
> my $file = $ARGV[0];
> $file =~ s|\\|/|g;
>
> die "Canot read from $file: $!" unless -r $file;
>
> $opt_b = -s _ if $opt_b == 0;
>
> my @s = strings($file,$opt_b,$opt_T,$opt_n, $opt_t) or die "No strings:
> $!";
>
> binmode STDOUT;
> print join("\n",@s);
>
> exit 0;
>
>
_______________________________________________
ActivePerl mailing list
[EMAIL PROTECTED]
To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
