http://bugs.koha.org/cgi-bin/bugzilla/show_bug.cgi?id=1567 Summary: MARC record parsing from and to XML Product: Koha Version: rel_3_0 Platform: PC URL: http://support.tamil.fr/tako_test.pl.tgz OS/Version: Linux - Debian Status: NEW Severity: blocker Priority: P3 Component: MARC Import AssignedTo: galen.charlton@liblime.com ReportedBy: frederic@tamil.fr QAContact: koha-bugs@nongnu.org On a Debian/Lenny, parsing of XML marc records encoded in UTF-8 doesn't work properly depending of SAX parser used : PurePerl, Expat or LibXML. (1) Records are converted into XML string before being written in bibliotems.marcxml. A function of MARC::File::XML us used in Biblio.pm: $record->as_xml_record($encoding). With Expat or LibXML SAX Parser, encoding is suppressed: <?xml version="1.0" encoding=""?> Consequently, records can't be read back from this biblioitems.marcxml. This is a blocker bug. (2) Records are created from XML string (biblioitems.marcxml). In Biblio.pm: $record = eval {MARC::Record::new_from_xml( $marcxml, "utf8", C4::Context->preference('marcflavour'))}; This function doesn't work properly with Expat or LibXML. UTF-8 extended characters (é for example) are converted in their ISO-8859-1 counterpart. Perl script attached allow to test those bugs. ------- You are receiving this mail because: ------- You are the QA contact for the bug, or are watching the QA contact.