6 |
## |
## |
7 |
## ---------------------------------------------------------------------------------------- |
## ---------------------------------------------------------------------------------------- |
8 |
## $Log$ |
## $Log$ |
9 |
|
## Revision 1.4 2004/05/06 12:53:07 jonen |
10 |
|
## + added use of File::RsycP::digest |
11 |
|
## |
12 |
|
## Revision 1.3 2003/05/13 08:19:00 joko |
13 |
|
## switched to crc32 |
14 |
|
## |
15 |
|
## Revision 1.2 2003/02/11 09:53:07 joko |
16 |
|
## + metadata-structure-change, fixed some code here |
17 |
|
## |
18 |
## Revision 1.1 2003/02/09 05:10:13 joko |
## Revision 1.1 2003/02/09 05:10:13 joko |
19 |
## + initial commit |
## + initial commit |
20 |
## |
## |
21 |
## ---------------------------------------------------------------------------------------- |
## ---------------------------------------------------------------------------------------- |
22 |
|
|
23 |
|
|
24 |
|
=pod |
25 |
|
|
26 |
|
=head1 Todo |
27 |
|
|
28 |
|
o Data::Transfer::Sync::Compare::Slot using Compare::Struct |
29 |
|
|
30 |
|
=cut |
31 |
|
|
32 |
package Data::Transfer::Sync::Compare::Checksum; |
package Data::Transfer::Sync::Compare::Checksum; |
33 |
|
|
34 |
use strict; |
use strict; |
37 |
use mixin::with qw( Data::Transfer::Sync ); |
use mixin::with qw( Data::Transfer::Sync ); |
38 |
|
|
39 |
|
|
|
# - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - main |
|
|
|
|
40 |
use Data::Dumper; |
use Data::Dumper; |
|
use Digest::MD5 qw(md5 md5_hex md5_base64); |
|
41 |
|
|
42 |
|
# TODO: Load these appropriatly at runtime. |
43 |
|
use Digest::MD5 qw( md5 md5_hex md5_base64 ); |
44 |
|
use String::CRC32; |
45 |
|
use File::RsyncP::Digest; |
46 |
|
|
47 |
# get logger instance |
# get logger instance |
48 |
my $logger = Log::Dispatch::Config->instance; |
my $logger = Log::Dispatch::Config->instance; |
49 |
|
|
50 |
|
|
51 |
|
# Maybe refactor to shortcuts::checksum? |
52 |
sub _calcChecksum { |
sub _calcChecksum { |
53 |
|
|
54 |
my $self = shift; |
my $self = shift; |
69 |
#$logger->dump( __PACKAGE__ . ": " . $dump ); |
#$logger->dump( __PACKAGE__ . ": " . $dump ); |
70 |
|
|
71 |
# calculate checksum from dump |
# calculate checksum from dump |
72 |
# note: the 32-bit integer hash from DBI seems |
|
73 |
# to generate duplicates with small payloads already in ranges of hundreds of items/rows!!! |
# 1. md5-based fingerprint, base64 encoded (from Digest::MD5) |
74 |
# try to avoid to use it or try to use it only for payloads greater than, hmmm, let's say 30 chars? |
#$self->{node}->{$descent}->{checksum} = md5_base64($dump) . '=='; |
75 |
# (we had about 15 chars average per item (row)) |
|
76 |
|
# 2. 32-bit integer "hash" value (maybe faster?) (from DBI) |
77 |
# md5-based fingerprint, base64 encoded (from Digest::MD5) |
# Note: The 32-bit integer hash from DBI seems to generate duplicates |
78 |
$self->{node}->{$descent}->{checksum} = md5_base64($dump) . '=='; |
# with small payloads already in ranges of hundreds of items/rows!!! |
79 |
# 32-bit integer "hash" value (maybe faster?) (from DBI) |
# Try to avoid it or try to use it only for payloads greater than, hmmm, let's say 30 chars? |
80 |
|
# (we had about 15 chars average per item (row)) |
81 |
|
# Possible (generic) solution: Just generate checksum, if length(checksum(payload)) < payload |
82 |
#$self->{node}->{$descent}->{checksum} = DBI::hash($dump, 1); |
#$self->{node}->{$descent}->{checksum} = DBI::hash($dump, 1); |
83 |
|
|
84 |
|
# 3. good ol' crc32??? |
85 |
|
#$self->{node}->{$descent}->{checksum} = crc32($dump); |
86 |
|
|
87 |
|
# 4. File::RsyncP::Digest - Perl interface to rsync message digest algorithms |
88 |
|
my $rsDigest = new File::RsyncP::Digest; |
89 |
|
$rsDigest->add($dump); |
90 |
|
my $digest = $rsDigest->digest(); |
91 |
|
$self->{node}->{$descent}->{checksum} = unpack("H*", $digest); |
92 |
|
|
93 |
|
# 5. some more modern Digest::SHA1 or similar? |
94 |
|
|
95 |
# signal good |
# signal good |
96 |
return 1; |
return 1; |
111 |
# trace |
# trace |
112 |
#print "desc: $descent", "\n"; |
#print "desc: $descent", "\n"; |
113 |
#print Dumper($self); |
#print Dumper($self); |
114 |
|
#print Dumper($self->{meta}->{$descent}); |
115 |
#exit; |
#exit; |
116 |
|
|
117 |
# get checksum for current entry |
# get checksum for current entry |
118 |
# TODO: don't have the checksum column/property hardcoded as "cs" here, make this configurable somehow |
# TODO: don't have the checksum column/property hardcoded as "cs" here, make this configurable somehow |
119 |
if ($self->{options}->{$descent}->{storage}->{isChecksumAuthority}) { |
if ($self->{meta}->{$descent}->{isChecksumAuthority}) { |
120 |
$self->_calcChecksum($descent); |
$self->_calcChecksum($descent); |
121 |
} else { |
} else { |
122 |
$self->{node}->{$descent}->{checksum} = $self->{node}->{$descent}->{payload}->{cs}; |
$self->{node}->{$descent}->{checksum} = $self->{node}->{$descent}->{payload}->{cs}; |
129 |
|
|
130 |
|
|
131 |
1; |
1; |
132 |
|
__END__ |