This question is not directly related to Koha itself, but to preparing MARC records for import. Some of the records we've been downloading from Libraries and Archives Canada contain a "5||" (two pipe characters) tag with information about translation. As far as I can tell this is not valid MARC. I developed a little Java program to pre-process records before importing into Koha, using the marc4j library to read MARC. This library refuses to read records containing non-numeric tags, causing us problems. Can anyone recommend any tools for ensuring that a MARC record contains only valid MARC? I could use MarcEdit to 'break' the files, edit out the bad tags and then re'make' them but this is a bit complicated, time-consuming and error-prone. -- Kevin O'Rourke ICT Coordinator, National Teachers' Institute, Kaduna, Nigeria 062 316972