[Koha] Corrupted record

[Beda Szukics]

Henri-Damien LAURENT schrieb:
> Frederic Demians a écrit :
>   
>>> When I do a "rebuild_zebra.pl -b" I allways get the following error:
>>>
>>> problem with :577 : :38: parser error : Premature end of data in tag
>>> subfield line 38
>>> <subfield code="6">SCH
>>> ^
>>> :38: parser error : Premature end of data in tag datafield line 33
>>> <subfield code="6">SCH
>>> ^
>>> :38: parser error : Premature end of data in tag record line 2
>>> <subfield code="6">SCH
>>> ^
>>> at /usr/lib/perl5/XML/LibXML/SAX/Parser.pm line 31
>>>
>>> <?xml version="1.0" encoding="UTF-8"?>
>>> <record
>>> xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
>>> xsi:schemaLocation="http://www.loc.gov/MARC21/slim
>>> http://www.loc.gov/ standards/marcxml/schema/MARC21slim.xsd"
>>> xmlns="http://www.loc.gov/MARC21/slim">
>>>
>>> <leader>00521nam a2200133 4500</leader>
>>> <controlfield tag="008">
>>> 960105s1970||||||||||||||||||||||||||||ger</controlfield>
>>> <datafield tag="245" ind1="0" ind2="4">
>>> <subfield code="a">Die schönsten Helden- und Rittersagen des
>>> Mittelalters /</subfield>
>>> <subfield code="c">[Für die Jugend bearb. von] Gerhard Aick ; Mit
>>>     
>>>       
>>   
>>     
> That record seems to be in UTF8 and MARC21.
> Leader has correct value on leader 8th position.
> But the characters has been double encoded.
> Maybe the import of this record was not properly done.
>
> Do you know its origin ?
>   

This particular record was part of an import I did some time ago. Its 
origin is my old "handmade" lybrary management. I used a german program 
called FAUST to manage the biblios of the library in question. I used 
its printing facilities to create the marc-records. Those record were in 
ISO-8859-1. I did a second step importing them in OO and then exporting 
the file using the filter off OO to create an UTF-8 file. This one I 
imported in MarcEdit, did some corrections, compiled it and then 
imported them in koha.

Everything went fine so far. Now there are about 5000 records in my 
database and the only one giving trouble is the one above. The problem 
seems to be the call number with in this case is "SCHÖ" (I usually 
avoided to use non-ASCII-caracters in call-numbers, this one seems to be 
the only exception). Somehow the record can't be deleted or edited in 
any way using the usual methods of koha. So it seems that a need an 
unusual way to get rid of that Ö or of the whole record so I could enter 
a corrected version of it.

(The somehow corrupted way of representing the data above seems to be 
the result of my copiing and pasting the files from a linux to a windows 
system).

The record wich I imported is (un-compiled with MarcEdit):

=LDR  00438nam  2200109   45e0
=008  \\960105s1970||||||||||||||||||||||||||||ger
=245  04$aDie schönsten Helden- und Rittersagen des Mittelalters /$c[Für 
die Jugend bearb. von] Gerhard Aick ; Mit vielen Zeichnungen von Erich Hölle
=260  \\$aWien :$bTosa Verlag,$c[ca. 1970]
=300  \\$a508 S. :$bIll. ;$c23 cm
=700  1\$aAick, Gerhard
=700  1\$aHölle, Erich
=942  \\$cBUCH$kSCHÖ
=952  \\$p160$oSCHÖ$aSB$bSB$yBUCH

Hope that gives some hints

Thank you                                    Beda