6 |
## |
## |
7 |
## ---------------------------------------------------------------------------------------- |
## ---------------------------------------------------------------------------------------- |
8 |
## $Log$ |
## $Log$ |
9 |
|
## Revision 1.3 2003/05/13 08:19:00 joko |
10 |
|
## switched to crc32 |
11 |
|
## |
12 |
|
## Revision 1.2 2003/02/11 09:53:07 joko |
13 |
|
## + metadata-structure-change, fixed some code here |
14 |
|
## |
15 |
## Revision 1.1 2003/02/09 05:10:13 joko |
## Revision 1.1 2003/02/09 05:10:13 joko |
16 |
## + initial commit |
## + initial commit |
17 |
## |
## |
18 |
## ---------------------------------------------------------------------------------------- |
## ---------------------------------------------------------------------------------------- |
19 |
|
|
20 |
|
|
21 |
|
=pod |
22 |
|
|
23 |
|
=head1 Todo |
24 |
|
|
25 |
|
o Data::Transfer::Sync::Compare::Slot using Compare::Struct |
26 |
|
|
27 |
|
=cut |
28 |
|
|
29 |
package Data::Transfer::Sync::Compare::Checksum; |
package Data::Transfer::Sync::Compare::Checksum; |
30 |
|
|
31 |
use strict; |
use strict; |
34 |
use mixin::with qw( Data::Transfer::Sync ); |
use mixin::with qw( Data::Transfer::Sync ); |
35 |
|
|
36 |
|
|
|
# - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - main |
|
|
|
|
37 |
use Data::Dumper; |
use Data::Dumper; |
|
use Digest::MD5 qw(md5 md5_hex md5_base64); |
|
38 |
|
|
39 |
|
# TODO: Load these appropriatly at runtime. |
40 |
|
use Digest::MD5 qw( md5 md5_hex md5_base64 ); |
41 |
|
use String::CRC32; |
42 |
|
|
43 |
# get logger instance |
# get logger instance |
44 |
my $logger = Log::Dispatch::Config->instance; |
my $logger = Log::Dispatch::Config->instance; |
45 |
|
|
46 |
|
|
47 |
|
# Maybe refactor to shortcuts::checksum? |
48 |
sub _calcChecksum { |
sub _calcChecksum { |
49 |
|
|
50 |
my $self = shift; |
my $self = shift; |
65 |
#$logger->dump( __PACKAGE__ . ": " . $dump ); |
#$logger->dump( __PACKAGE__ . ": " . $dump ); |
66 |
|
|
67 |
# calculate checksum from dump |
# calculate checksum from dump |
68 |
# note: the 32-bit integer hash from DBI seems |
|
69 |
# to generate duplicates with small payloads already in ranges of hundreds of items/rows!!! |
# 1. md5-based fingerprint, base64 encoded (from Digest::MD5) |
70 |
# try to avoid to use it or try to use it only for payloads greater than, hmmm, let's say 30 chars? |
#$self->{node}->{$descent}->{checksum} = md5_base64($dump) . '=='; |
71 |
# (we had about 15 chars average per item (row)) |
|
72 |
|
# 2. 32-bit integer "hash" value (maybe faster?) (from DBI) |
73 |
# md5-based fingerprint, base64 encoded (from Digest::MD5) |
# Note: The 32-bit integer hash from DBI seems to generate duplicates |
74 |
$self->{node}->{$descent}->{checksum} = md5_base64($dump) . '=='; |
# with small payloads already in ranges of hundreds of items/rows!!! |
75 |
# 32-bit integer "hash" value (maybe faster?) (from DBI) |
# Try to avoid it or try to use it only for payloads greater than, hmmm, let's say 30 chars? |
76 |
|
# (we had about 15 chars average per item (row)) |
77 |
|
# Possible (generic) solution: Just generate checksum, if length(checksum(payload)) < payload |
78 |
#$self->{node}->{$descent}->{checksum} = DBI::hash($dump, 1); |
#$self->{node}->{$descent}->{checksum} = DBI::hash($dump, 1); |
79 |
|
|
80 |
|
# 3. good ol' crc32??? |
81 |
|
$self->{node}->{$descent}->{checksum} = crc32($dump); |
82 |
|
|
83 |
|
# 4. some more modern Digest::SHA1 or similar? |
84 |
|
|
85 |
# signal good |
# signal good |
86 |
return 1; |
return 1; |
101 |
# trace |
# trace |
102 |
#print "desc: $descent", "\n"; |
#print "desc: $descent", "\n"; |
103 |
#print Dumper($self); |
#print Dumper($self); |
104 |
|
#print Dumper($self->{meta}->{$descent}); |
105 |
#exit; |
#exit; |
106 |
|
|
107 |
# get checksum for current entry |
# get checksum for current entry |
108 |
# TODO: don't have the checksum column/property hardcoded as "cs" here, make this configurable somehow |
# TODO: don't have the checksum column/property hardcoded as "cs" here, make this configurable somehow |
109 |
if ($self->{options}->{$descent}->{storage}->{isChecksumAuthority}) { |
if ($self->{meta}->{$descent}->{isChecksumAuthority}) { |
110 |
$self->_calcChecksum($descent); |
$self->_calcChecksum($descent); |
111 |
} else { |
} else { |
112 |
$self->{node}->{$descent}->{checksum} = $self->{node}->{$descent}->{payload}->{cs}; |
$self->{node}->{$descent}->{checksum} = $self->{node}->{$descent}->{payload}->{cs}; |
119 |
|
|
120 |
|
|
121 |
1; |
1; |
122 |
|
__END__ |