gitea
/
Koha-clms
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Bug 41358: Add migration script for legacy circulation logs 

This patch adds an out-of-band migration script to convert legacy
CIRCULATION ISSUE action logs from the old itemnumber-only format
to the new consistent JSON format.

The script:
- Finds all CIRCULATION ISSUE logs where info is just an itemnumber
- Converts them to JSON format with the standard structure
- Attempts to recover issue_id and branchcode from old_issues/issues
- Runs in dry-run mode by default (--commit flag to actually update)
- Processes in batches to handle large databases safely
- Includes verbose mode for detailed output

Usage:
  perl misc/maintenance/migrate_action_logs_to_json.pl --help
  perl misc/maintenance/migrate_action_logs_to_json.pl --verbose
  perl misc/maintenance/migrate_action_logs_to_json.pl --commit

This is designed as an out-of-band maintenance script rather than
an atomic update because:
1. May be expensive on large databases with many log entries
2. Migration is optional - display code handles both formats
3. Can be run during off-peak hours at admin's discretion

Sponsored-by: OpenFifth <https://openfifth.co.uk/>
Signed-off-by: David Nind <david@davidnind.com>
Signed-off-by: Lisette Scheer <lisette@bywatersolutions.com>
Signed-off-by: Lucas Gass <lucas@bywatersolutions.com>
(cherry picked from commit 314df725d5)
Signed-off-by: Baptiste Wojtkowski <baptiste.wojtkowski@biblibre.com>
This commit is contained in:
Martin Renvoize 2025-12-18 14:50:57 +00:00 committed by Baptiste Wojtkowski
parent a12b2cd77d
commit 37438f85f4
No known key found for this signature in database
GPG Key ID: 47A8A13E10BC0D40
1 changed files with 257 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

257
misc/maintenance/migrate_action_logs_to_json.pl Executable file
View File

@ -0,0 +1,257 @@
#!/usr/bin/perl
# Copyright (C) 2025 OpenFifth
#
# This file is part of Koha.
#
# Koha is free software; you can redistribute it and/or modify it
# under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 3 of the License, or
# (at your option) any later version.
#
# Koha is distributed in the hope that it will be useful, but
# WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Koha; if not, see <https://www.gnu.org/licenses>.
use Modern::Perl;
use Getopt::Long qw( GetOptions );
use Pod::Usage qw( pod2usage );
use JSON ();
use Koha::Script;
use C4::Context;
use Koha::ActionLogs;
use Koha::Checkouts;
use Koha::Old::Checkouts;
=head1 NAME
migrate_action_logs_to_json.pl - Migrate old circulation log entries to JSON format
=head1 SYNOPSIS
migrate_action_logs_to_json.pl [ -c | --commit ] [ -v | --verbose ] [ --help ]
Options:
--help or -h Brief usage message
--commit or -c Actually update the database (default is dry run)
--verbose or -v Print detailed information about each conversion
--batch-size Number of records to process per batch (default: 1000)
=head1 DESCRIPTION
This script migrates old CIRCULATION ISSUE action log entries from the legacy
itemnumber-only format to the new consistent JSON format. This is necessary
after the fix for Bug 41358 which ensures all circulation logs are stored in
JSON format for consistent reporting.
The script will:
1. Find all CIRCULATION ISSUE log entries where info is just an itemnumber (not JSON)
2. Convert them to JSON format with the structure:
{
"issue": <issue_id if available>,
"branchcode": <branchcode if available>,
"itemnumber": <itemnumber>,
"confirmations": [],
"forced": []
}
By default, the script runs in dry-run mode. Use --commit to actually update the database.
=head1 WARNING
This script may take a long time to run on large databases with many action_logs entries.
It's recommended to run it during off-peak hours.
=cut
my $help = 0;
my $commit = 0;
my $verbose = 0;
my $batch_size = 1000;
GetOptions(
'h|help' => \$help,
'c|commit' => \$commit,
'v|verbose' => \$verbose,
'batch-size=i' => \$batch_size,
) || pod2usage(1);
if ($help) {
pod2usage(1);
}
my $dbh = C4::Context->dbh;
print "Starting migration of CIRCULATION ISSUE logs to JSON format...\n";
print $commit
? "COMMIT MODE - Changes will be saved\n"
: "DRY RUN MODE - No changes will be saved (use --commit to save)\n";
print "\n";
# First, count how many entries need conversion
my $count_sql = q(
SELECT COUNT(*)
FROM action_logs
WHERE module = 'CIRCULATION'
AND action = 'ISSUE'
AND info IS NOT NULL
AND info NOT LIKE '{%'
AND info REGEXP '^[0-9]+$'
);
my ($total_count) = $dbh->selectrow_array($count_sql);
print "Found $total_count log entries to migrate\n";
print "\n";
if ( $total_count == 0 ) {
print "No entries need migration. Exiting.\n";
exit 0;
}
# Process in batches using keyset pagination by action_id.
# Using OFFSET/LIMIT here is unsafe in commit mode: each UPDATE removes the
# row from the matching set, shifting subsequent OFFSETs and silently skipping
# records. Paging by "action_id > last_seen" is stable under mutation and works
# identically in dry-run mode.
my $last_action_id = 0;
my $processed = 0;
my $converted = 0;
my $errors = 0;
while (1) {
my $select_sql = q(
SELECT action_id, info, object, timestamp
FROM action_logs
WHERE module = 'CIRCULATION'
AND action = 'ISSUE'
AND info IS NOT NULL
AND info NOT LIKE '{%'
AND info REGEXP '^[0-9]+$'
AND action_id > ?
ORDER BY action_id
LIMIT ?
);
my $sth = $dbh->prepare($select_sql);
$sth->execute( $last_action_id, $batch_size );
my $batch_count = 0;
while ( my $row = $sth->fetchrow_hashref ) {
my $action_id = $row->{action_id};
my $info = $row->{info};
my $borrowernumber = $row->{object};
my $timestamp = $row->{timestamp};
$last_action_id = $action_id;
# The info should be an itemnumber
my $itemnumber = $info;
# Try to find the corresponding issue to get issue_id and branchcode
my $issue_id;
my $branchcode;
# First check old_issues (most likely location for old logs)
my $old_checkout = Koha::Old::Checkouts->search(
{
itemnumber => $itemnumber,
borrowernumber => $borrowernumber,
},
{
order_by => { -desc => 'returndate' },
rows => 1
}
)->next;
if ($old_checkout) {
$issue_id = $old_checkout->issue_id;
$branchcode = $old_checkout->branchcode;
} else {
# Try current issues table (unlikely but possible)
my $current_issue = Koha::Checkouts->search(
{
itemnumber => $itemnumber,
borrowernumber => $borrowernumber,
}
)->next;
if ($current_issue) {
$issue_id = $current_issue->issue_id;
$branchcode = $current_issue->branchcode;
}
}
# Build the JSON structure
my $json_data = {
issue => $issue_id,
branchcode => $branchcode,
itemnumber => $itemnumber,
confirmations => [],
forced => []
};
# Match the pretty/canonical encoding used by AddIssue so the
# converted entries are byte-identical to freshly logged ones.
my $json_info = JSON->new->pretty(1)->canonical(1)->encode($json_data);
if ($verbose) {
print "Converting action_id $action_id:\n";
print " Old: $info\n";
print " New: " . ( $json_info =~ s/\n/ /gr ) . "\n";
print " issue_id: " . ( $issue_id // 'NULL' ) . "\n";
print " branchcode: " . ( $branchcode // 'NULL' ) . "\n";
print "\n";
}
if ($commit) {
my $update_sql = q(
UPDATE action_logs
SET info = ?
WHERE action_id = ?
);
my $update_sth = $dbh->prepare($update_sql);
if ( $update_sth->execute( $json_info, $action_id ) ) {
$converted++;
} else {
print STDERR "ERROR: Failed to update action_id $action_id: " . $dbh->errstr . "\n";
$errors++;
}
} else {
$converted++;
}
$batch_count++;
}
last unless $batch_count;
$processed += $batch_count;
unless ($verbose) {
print "Processed $processed / $total_count records...\r";
}
}
print "\n";
print "Migration complete!\n";
print " Converted: $converted\n";
print " Errors: $errors\n";
print "\n";
if ( !$commit && $converted > 0 ) {
print "This was a DRY RUN. Use --commit to actually update the database.\n";
}
=head1 AUTHOR
Koha Development Team
=cut