in reply to How to read number of block in binary files?

        open (OUTPUT, ">$fname.tmp");
        open (DATA, "$fname");

[download]

You should always verify that the file opened correctly before trying to use a possibly invalid filehandle. 

        open OUTPUT, '>', "$fname.tmp" or die "Cannot open '$fname.tmp
+' $!";
        open INPUT,  '<', $fname or die "Cannot open '$fname' $!";

[download]

        $offset = 0;
        while (read(DATA, $data, BLOCKLEN, $offset)) {
                my $blockhdr = unpack "H*", substr($data, 0, BLOCKHDR)
+;
                my $blockdt  = unpack "H*", substr($data, BLOCKHDR, BL
+OCKDATA);
                my $cdrdata  = unpack "H*", substr($data, BLOCKHDR+BLO
+CKDATA, CDRLENGTH);
...
                        $offset += BLOCKLEN;
...
                        $offset += BLOCKLEN;
                        $offset -= 286;

[download]

$offset is the position in $data where the data is copied from the file but you are not capturing the actual number of bytes read and just assuming that it is always BLOCKLEN number of bytes.   So each time through the loop you are appending records to the end of $data but you are always reading from the beginning of $data. 

                my $total_record = hextoint(reverse_str(substr($blockd
+t,4)));

[download]

What does reverse_str do that reverse cannot do?   What does hextoint do that unpack or hex cannot do?

Replies are listed 'Best First'.
Re^2: How to read number of block in binary files?
by bh_perl (Monk) on Mar 09, 2009 at 05:51 UTC

    Thanks for your info..

    Yes.. you are right.. I am assuming the numbers of bytes is always BLOCKLEN or 890 bytes.

    But, I am plan to reverse the offset values about 286 bytes back before read the next record. This is because if the block contain 2 record the length is 604 while 3 record is 890 bytes. So that why i set $offset -= 286. is it correct ?

[reply]
      This is because if the block contain 2 record the length is 604 while 3 record is 890 bytes. So that why i set $offset -= 286. is it correct ?

      read( DATA, $data, BLOCKLEN, $offset );

      is the same as saying:

      read( DATA, my $temp, BLOCKLEN );
      substr( $data, $offset ) = $temp;

      and then the next statement:

      my $blockhdr = unpack "H*", substr($data, 0, BLOCKHDR);

      always extracts data from offset 0, at the beginning of $data

      You need to read BLOCKHDR + BLOCKDATA data first, extract the $total_record value from that and then read $total_record * 286 data to get the actual records.

[reply]
[d/l]
[select]

        Hi.. Thank for your info.. that very helpful.. But could you guide me how could i read the binary files until end of file ?..

        This is my latest code:- 

        if ($fname) {
        open (OUTPUT, ">$fname.tmp") or die "Can't open $fname.tmp $!\
+n";
        open (DATA, "<$fname") or die "Can't open $fname $!\n";
        binmode DATA;

        read(DATA, $data, BLOCKHDR);
        $blockhdr = unpack "H*", $data;

        read(DATA, $data, BLOCKDATA);
        $blockdt  = unpack "H*", $data;

        my $t_cdr = hextoint(reverse_str(substr($blockdt,4)));
        my $blocklen = $t_cdr * 286;

        read(DATA, $data, $blocklen);
        my $record = unpack "H*", $data;

        close(DATA);
        close(OUTPUT);
} else {

        [download]
[reply]
[d/l]

In Section Seekers of Perl Wisdom