Benvenuto Fabio,
20030730d1968----km-y0itay0103----ba
but, after the import from Koha backend interface, it becomes
20030730d1968----km-y0frey50 ----ba
(as you can see, cataloguing language is changed from "ita" to "fre", character sets from "01" and "03" to "50", and the whole string is 35 characters long instead of 36).
During the import process, your biblio records are converted into UTF-8 encoding. That's the reason why 01 is transformed into 50. This is required. Otherwise, biblio records parsing inside Koha would fail. The transformation of 'ita' into 'fre' is not required. Koha could be modified to leave this 100 tag part unchanged. Take a look at C4/Charset.pm module: if you fix that you should submit a patch to the community. For tag length truncation, I have no idea... -- Frédéric DEMIANS http://www.tamil.fr/u/fdemians.html