We are having trouble with non-Latin characters in Query Collections output.
Symptom
We use Query Collections to produce MARC records. Our records occasionally contain 245 fields that are blank, or have only a couple of characters, and that seems to be connected to the presence of 880 tags for the alternate display of certain scripts. Our example records have a non-Latin character (such as the Greek character for delta Δ) present in the 245 and in one of the 500 notes. Each of these fields has a linked 880. In the output from Query Collections, the 245 is almost entirely empty (only contains <>) and the 500 is not even present. All of the data is only in the linked 880 fields.
Applies to
- WorldShare Record Manager
- WorldShare Collection Manager
Resolution
When a field contains non-Latin characters, OCLC automatically changes that field into a linking 880 field. It is then expected that the user will provide a 245 field with all Latin characters. If the Latin 245 is not provided, then the placeholder <> is supplied.
This is a result of OCLC implementing MARC Model A to accommodate records in multiple scripts where the Latin is entered in the regular fields while the non-Latin is entered in the linking 880.
The WorldShare Record Manager workflow for creating the 245 field correctly is, after entering the field with the vernacular, a user can enter a new field below the vernacular field with the same tag, and it will automatically be linked to the vernacular field above it, on the next operation (reformat, validate, add, replace, etc.) The user will then see those two fields linked in the Record Manager user interface. Once the record correctly has a Latin 245 field, then Collection Manager will be able to output it in the MARC record output as expected.