From 3443056e15d7f2cc36e137a5ac50580375f3e1e4 Mon Sep 17 00:00:00 2001
From: Jonathan Druart <jonathan.druart@biblibre.com>
Date: Wed, 5 Nov 2014 16:09:04 +0100
Subject: [PATCH] Bug 8218: qa followup

This patch
- rename _entity_clean as _clean_ampersand
- rename the script to sanitize_records.pl
- add a --fix-ampersand switch (the only one FOR NOW, enabled by
  default) so it is obvious what the script does.
- make POD and usage reflect this changes.

Signed-off-by: Tomas Cohen Arazi <tomascohen@gmail.com>
---
 C4/Charset.pm                                  |  8 ++++----
 ...sanitize_records.pl => sanitize_records.pl} | 18 ++++++++++++++----
 2 files changed, 18 insertions(+), 8 deletions(-)
 rename misc/maintenance/{batch_sanitize_records.pl => sanitize_records.pl} (90%)

diff --git a/C4/Charset.pm b/C4/Charset.pm
index e56422fced..e43404bce2 100644
--- a/C4/Charset.pm
+++ b/C4/Charset.pm
@@ -429,7 +429,7 @@ sub nsb_clean {
 SanitizeRecord($marcrecord);
 
 Sanitize a record
-This routine is called in the maintenance script misc/maintenance/batch_sanitize_records.pl.
+This routine is called in the maintenance script misc/maintenance/sanitize_records.pl.
 It cleans any string with '&amp;amp;...', replacing it by '&'
 
 =cut
@@ -444,7 +444,7 @@ sub SanitizeRecord {
     foreach my $field ( $record->fields() ) {
         if ( $field->is_control_field() ) {
             my $value           = $field->data();
-            my $sanitized_value = _entity_clean($value);
+            my $sanitized_value = _clean_ampersand($value);
             $record_modified = 1 if $sanitized_value ne $value;
             $field->update($sanitized_value);
         }
@@ -456,7 +456,7 @@ sub SanitizeRecord {
                   if $url_field eq $field->tag()
                       and $url_subfield eq $subfield->[0];
                 my $value           = $subfield->[1];
-                my $sanitized_value = _entity_clean($value);
+                my $sanitized_value = _clean_ampersand($value);
                 push @new_subfields, $subfield->[0] => $sanitized_value;
                 $record_modified = 1 if $sanitized_value ne $value;
             }
@@ -481,7 +481,7 @@ sub SanitizeRecord {
     return $record, $record_modified;
 }
 
-sub _entity_clean {
+sub _clean_ampersand {
     my ($string) = @_;
     $string =~ s/(&)(amp;)+/$1/g;
     return $string;
diff --git a/misc/maintenance/batch_sanitize_records.pl b/misc/maintenance/sanitize_records.pl
similarity index 90%
rename from misc/maintenance/batch_sanitize_records.pl
rename to misc/maintenance/sanitize_records.pl
index b164ec225e..b9caa4bfb7 100755
--- a/misc/maintenance/batch_sanitize_records.pl
+++ b/misc/maintenance/sanitize_records.pl
@@ -26,7 +26,7 @@ use Getopt::Long;
 use Pod::Usage;
 
 my ( $help, $verbose, $confirm, $biblionumbers, $reindex, $filename,
-    $auto_search );
+    $auto_search, $fix_ampersand );
 my $result = GetOptions(
     'h|help'          => \$help,
     'v|verbose'       => \$verbose,
@@ -35,8 +35,13 @@ my $result = GetOptions(
     'reindex'         => \$reindex,
     'f|filename:s'    => \$filename,
     'auto-search'     => \$auto_search,
+    'fix-ampersand'   => \$fix_ampersand,
 ) || pod2usage(1);
 
+# This script only fix ampersand at the moment.
+# It is enabled by default.
+$fix_ampersand = 1;
+
 if ($help) {
     pod2usage(0);
 }
@@ -154,13 +159,13 @@ sub biblios_to_sanitize {
 
 =head1 NAME
 
-batch_sanitize_biblios - This script sanitizes a biblio, replacing '&amp;amp;amp;etc.' with '&amp;' in it.
+sanitize_records - This script sanitizes a record.
 
 =head1 SYNOPSIS
 
-batch_sanitize_biblios.pl [-h|--help] [-v|--verbose] [-c|--confirm] [--biblionumbers=BIBLIONUMBER_LIST] [-f|--filename=FILENAME] [--auto-search] [--reindex]
+sanitize_records.pl [-h|--help] [-v|--verbose] [-c|--confirm] [--biblionumbers=BIBLIONUMBER_LIST] [-f|--filename=FILENAME] [--auto-search] [--reindex] [--fix-ampersand]
 
-Replace '&amp;' by '&' in a record. You can either give some biblionumbers or a file with biblionumbers or ask for an auto-search.
+You can either give some biblionumbers or a file with biblionumbers or ask for an auto-search.
 
 =head1 OPTIONS
 
@@ -193,6 +198,11 @@ Give a biblionumber list using a filename. One biblionumber by line or separate
 
 Automatically search records containing "&amp;" in biblioitems.marcxml or in the specified fields.
 
+=item B<--fix-ampersand>
+
+Replace '&amp;' by '&' in the records.
+Replace '&amp;amp;amp;etc.' with '&amp;' in the records.
+
 =item B<--reindex>
 
 Reindex the modified records.
-- 
2.39.5