German MARC record issues
Hi, I do have a similar problem: I use the US Library of Congress and the Canada National Library as my primary sources of Z39.50 records. In both cases the format is MARC21. The English book notices coming from both libraries are OK but the French books from the Canadian Library get their French accented letters converted to two letters (é -> Âe). Michel Virard, Montreal mvirard@symtec.ca
I'd wait to see what other folks say first, but here's what I'd try. I can't speak for Canada but the records coming from LC should all be UTF-8 encoded (Leader byte 9 is set to 'a'). Take a look at this page: http://wiki.koha.org/doku.php?id=encodingscratchpad AFAIK, Koha does not handle Unicode when first set up. Follow some of the instructions here to get your environment and database properly set up. You will have to rebuildnonmarc after you're done (or even export your records and bulkmarcimport them again). Even then you may need to make some changes to your templates to have a default encoding of utf8. Check the archives, I think there was a message about changing templates for this recently. I don't know if this would effect your version of Koha but recently there was a character encoding problem with a Perl module that's been fixed. You might get some input on whether it would be advisable to upgrade all relevant Perl modules. Well, I hope this gives you some leads. --Jason On 5/16/07, Michel Virard <mvirard@symtec.ca> wrote:
Hi,
I do have a similar problem: I use the US Library of Congress and the Canada National Library as my primary sources of Z39.50 records. In both cases the format is MARC21. The English book notices coming from both libraries are OK but the French books from the Canadian Library get their French accented letters converted to two letters (é -> Âe).
Michel Virard, Montreal mvirard@symtec.ca _______________________________________________ Koha mailing list Koha@lists.katipo.co.nz http://lists.katipo.co.nz/mailman/listinfo/koha
Jason Ronallo wrote:
I'd wait to see what other folks say first, but here's what I'd try. I can't speak for Canada but the records coming from LC should all be UTF-8 encoded (Leader byte 9 is set to 'a'). Take a look at this page: http://wiki.koha.org/doku.php?id=encodingscratchpad We preprocess our MARC records (using a Java program I wrote) to make sure we have some details correct (ISBN, DDC that matches the one on our book,...). This also handles the conversion from MARC-8 to UTF-8.
Our main problem with display of non-ASCII characters is finding fonts that actually contain the characters. We have some Arabic characters in records (not too difficult) and some records with odd accented characters from Nigerian languages (particularly Yoruba and Hausa). For example, this following letter should be a k with a hook at the top: ƙ Lots of good advice on that Koha wiki page, thanks. Kevin -- Kevin O'Rourke ICT Coordinator, National Teachers' Institute, Kaduna, Nigeria 062 316972
Hi! On 5/17/07, Michel Virard <mvirard@symtec.ca> wrote:
Hi,
I do have a similar problem: I use the US Library of Congress and the Canada National Library as my primary sources of Z39.50 records. In both cases the format is MARC21. The English book notices coming from both libraries are OK but the French books from the Canadian Library get their French accented letters converted to two letters (é -> Âe).
From the Koha Intranet interface go to the Parameters->System Preferences
To have unicode support in Koha you need to do a couple of things. When creating the tables in mysql you need to set the encoding to UTF8. page. The select the Intranet tab. There is a setting for "Template encoding". What is this value in your installation? It should be set to utf-8. This may be the default in the latest versions. Regards, Rado
participants (4)
-
Jason Ronallo -
Kevin O'Rourke -
Michel Virard -
Radoslav Kolev