[Koha-cvs] koha/misc/migration_tools bulkmarcimport.pl [rel_2_2]

[Thomas D]

CVSROOT:	/sources/koha
Module name:	koha
Branch:		rel_2_2
Changes by:	Thomas D <thd>	06/09/01 17:08:56

Modified files:
	misc/migration_tools: bulkmarcimport.pl 

Log message:
	For MARC 21, instead of deleting the  whole subfield when a character does not
	translate properly from MARC8 into UTF-8, only the problem characters are
	deleted.

CVSWeb URLs:
http://cvs.savannah.gnu.org/viewcvs/koha/misc/migration_tools/bulkmarcimport.pl?cvsroot=koha&only_with_tag=rel_2_2&r1=1.1.2.4&r2=1.1.2.5

Patches:
Index: bulkmarcimport.pl
===================================================================
RCS file: /sources/koha/koha/misc/migration_tools/bulkmarcimport.pl,v
retrieving revision 1.1.2.4
retrieving revision 1.1.2.5
diff -u -b -r1.1.2.4 -r1.1.2.5
--- bulkmarcimport.pl	26 Apr 2006 11:48:46 -0000	1.1.2.4
+++ bulkmarcimport.pl	1 Sep 2006 17:08:56 -0000	1.1.2.5
@@ -12,6 +12,17 @@
 use MARC::Record;
 use MARC::Batch;
 use MARC::Charset;
+
+# According to kados, an undocumented feature of setting MARC::Charset to 
+# ignore_errors(1) is that errors are not ignored.  Instead of deleting the 
+# whole subfield when a character does not translate properly from MARC8 into 
+# UTF-8, just the problem characters are deleted.  This should solve at least 
+# some of the fixme problems for fMARC8ToUTF8().
+# 
+# Problems remain if there are MARC 21 records where 000/09 is set incorrectly. 
+# -- thd.
+MARC::Charset->ignore_errors(1);
+
 use C4::Context;
 use C4::Biblio;
 use Time::HiRes qw(gettimeofday);