#!/usr/bin/perl -w

use strict;

my ( @categories, @record_attributes );
my ( $input, $output );

die usage() unless defined $ARGV[ 0 ] && -e $ARGV[ 0 ];

$input = $ARGV[ 0 ] if ( -e $ARGV[ 0 ] );
if ( defined $ARGV[ 1 ] )
{
  print STDERR "Warning: overwriting '$ARGV[1]'!\n" if ( -e $ARGV[ 1 ]
+ );
  $output = $ARGV[ 1 ];
}

parse_pdb( $input, $output );


sub usage
{
  return "$0 v1.0 - WebLog database parser.\n".
        "(C) 2002 - J.W. Janssen, janwillem.janssen\@lxtreme.nl\n\n".
        "usage: $0 <weblog_hXwl.pdb> [<output file>]\n";
} # usage


sub parse_pdb
{
  my $name = shift;
  my $output = shift;

  # Obtain the necessary header information
  open PDB, "<$name" || die "Couldn't open '$name'";
  binmode PDB;
  my $line;
  read( PDB, $line, 78 );
  my @header = unpack "A32nnN6A4A4NNn", $line;

  # Read & parse the record entry headers...
  my @offsets;

  # The record entry headers start immediatly after the header
  #     of the PDB file (offset 78).
  seek( PDB, 78, 0 );
  # The header gives us the amount of entries in the PDB file.
  for ( 0..$header[ 13 ] )
  {
    read( PDB, $line, 8 );
    # Entry record:
    # name      | # | description
    # ----------+---+----------------------
    # offset    | 0 | the entry file offset
    # attribute | 1 | the upper 4 bits denote the record
    #           |   |  attribute, and the lower 4 the
    #           |   |  category.
    # uniqueID  | 2 | unknown; should be zero.
    my @record_list = unpack "NC4", $line;

    # Category == unknown
    my $category = $record_list[ 1 ] & 0x0F;
    # Attribute == 0x10: secret record bit
    #              0x20: busy bit (in use)
    #              0x40: dirty bit
    #              0x80: delete on next sync
    my $attribute = $record_list[ 1 ] & 0xF0;

    push @offsets, $record_list[ 0 ];
    push @record_attributes, $attribute;
  }
  # Make sure the last offset is the filesize...
  $offsets[ -1 ] = int( -s $name );

  # Open our output channel ...
  if ( defined $output and $output ne "-" )
  {
    open OUT, ">$output" || die "Couldn't open '$output' for writing!"
+;
  }
  else
  {
    open OUT, ">&STDOUT";
  }

  # Converting from Palm date format <-> UNIX format:
  #     number of days between
  #       01/01/1904 and 01/01/1970:           24,107
  #     number of seconds in one day:          86,400 *
  #                                     ---------------
  #     number of seconds between
  #       01/01/1904 and 01/01/1940:    2,082,844,800
  my $diff_secs = 24107*86400;

  print OUT "Last modified on: " . gmtime( $header[ 4 ] - $diff_secs )
+ . "\n";

  # Read as many entries...
  for ( 0..$header[ 13 ] - 1 )
  {
    my $offset = $offsets[ $_ ];
    my $rec_length = $offsets[ $_ + 1 ] - $offset;

    seek( PDB, $offset, 0 );
    read( PDB, $line, $rec_length );

    # Record format for WebLog:
    # name       | # | length  | description
    # -----------+---+---------+------------------------------------
    # record idx | 0 | 2 bytes | the logical entry number.
    # ?          | 1 | 2 bytes | unknown
    # entry date | 2 | 4 bytes | number of seconds since 01/01/1904,
    # entry text | 3 | ? bytes | the actual entry...
    my ( $index, $unknown, $date, $text ) = unpack "nnNa*", $line;
    my $hidden = ( $record_attributes[ $_ ] & 0x10 ) ? 1 : 0;

    print OUT "($index, $hidden) " . gmtime( $date - $diff_secs ) .
                " - " . $text . "\n";
  }

  close OUT;
  close PDB;
} # parse_pdb
[download]