Archives of the TeradataForum

Message Posted: Fri, 31 Mar 2006 @ 13:58:27 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

Many times. DataParcel contains the record in raw bytes as it came from the source system. There may be some additional fields on the front (lengths and such). If the record came from a mainframe, the data will be in EBCDIC. The Perl script below will display the record in character and hex form.

Not sure what you mean here, but it would be best to correct the source data and/or the FastLoad script and rerun the job. See the FastLoad Reference, section "Handling FastLoad Errors".

     #!/usr/bin/perl
     # Take a file of printable hex digits, such as those shown in a
     # Teradata load error table, and print them in character with
     # hex vertical.
     # Options:
     #     -wN  Width at which to break the lines; default=78.
     #     -a   Show plain ASCII characters (up to '~') only. Default is
     #          to show all characters up to x'FF' except controls.
     #     -e   Assume data is EBCDIC.
     use Getopt::Std;
     getopts("w:ae");
     $WIDTH = $opt_w || 78;
     $ASCII_ONLY = $opt_a;
     $EBCDIC = $opt_e;
     if ($EBCDIC) {
        $TRAN_TBL =
       ' ' x 64 .
       '           .<(+|&         !$*); -/         ,%_>?         `:#@\'="'.
       ' abcdefghi       jklmnopqr       ~stuvwxyz   [               ]  '.
       '{ABCDEFGHI      }JKLMNOPQR      \\ STUVWXYZ      0123456789      ';
     # Printable ASCII characters
     } elsif ($ASCII_ONLY) {
        $TRAN_TBL = (' 'x32) .
          ' !"#$%&\'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNO' .
          'PQRSTUVWXYZ[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~ ' . (' 'x128);
     } else {
        $TRAN_TBL = (' 'x32);
        for ($c = 32; $c < 127; $c++) { $TRAN_TBL .= chr($c); }
        $TRAN_TBL .= ' 'x34;
        for ($c = 161; $c < 256; $c++) { $TRAN_TBL .= chr($c); }
     }
     # "Zone" and "numeric" portions
     $ZONES = '0'x16 . '1'x16 . '2'x16 . '3'x16 . '4'x16 . '5'x16 .
      '6'x16 . '7'x16 . '8'x16 . '9'x16 . 'A'x16 .
      'B'x16 . 'C'x16 . 'D'x16 . 'E'x16 . 'F'x16;
     $NUMS = '0123456789ABCDEF'x16;
     # Dashes
     $DASHES = '-'x256;
     while (<>) {
        chomp;
        $chars = pack("H*", $_);
        print_line($chars);
     }
     # Print a line in EBCDIC|ASCII and hex with line breaks and such.
     sub print_line {
      my $line = shift;
      print fcs_xlate($line, $TRAN_TBL), "\n";
      print fcs_xlate($line, $ZONES), "\n";
      print fcs_xlate($line, $NUMS), "\n";
      print fcs_xlate($line, $DASHES), "\n";
     }
     # Full Collating Sequence translation
     sub fcs_xlate {
      my ($in, $table) = @_;
      my ($i, $result);
      $result = '';
      for ($i=0; $i < length($in); $i++) {
         $result .= substr($table, ord(substr($in, $i,1)), 1);
      }
      return $result;
     }

	<Prev	Next>		<<First	<Prev	Next>	Last>>