Open AndyElliottCRL opened 1 month ago
MarcEdit validation on CRL Catalog file fails with some kind of error like:
field_number: Length appears to be invalid. Reported length is: X. Expected length: Y
where field_number
is the MARC field, usually 008, sometimes 006 or 007.
First record, 008 comes out in MRK (broken by MarcEdit) as:
=008 151002d19581958nr{bsol}ar{bsol}{bsol}{bsol}o{bsol}{bsol}{bsol}{bsol}s0{bsol}{bsol}{bsol}{bsol}0eng{bsol}d
bsol appears to be "Back Solidus" or Reverse Solidus, as an HTML Entity: \
= \
= U+005C
= \
;
DLC:
Backslashes. The backslash “ \” (hexadecimal character value 5C/decimal character value 092) can be used instead of a space or blank (hexadecimal character value 20/decimal character value 032) in an input file. This means that you cannot use the backslash character to represent a real backslash with MARCMaker. To enter a real backslash into a MARC record, use the character string “{bsol}”, an abbreviation for “back solidus”, the official name of the backslash. Note: Not all computers represent the backslash with the same character value. It does not matter what you see on your computer screen, but be sure that the backslash character is the hexadecimal value 5C/decimal value 092. You can usually test this by holding down the “[ALT]” key on your keyboard and typing in the digits “0 9 2” on the numeric keypad to the right of your keyboard. This character value may have been assigned to a special accented letter in your country. Use it as if it were a backslash.
Apparently like this for every record in first 10k. Sample log attached marcedit_validation.log.sample.txt
Here's a zipped sample DOS batch file to validate MRC files. cmarcedit_batch_validator.sample.zip
Error for every record now: 008: Length appears to be invalid. Reported length is: X. Expected length: 40 (always the same Expected length)
Error for about 1 in 10: 007: Length appears to be invalid. Reported length is: X. Expected length: 13
Since (June?) 2024, VuFind is not getting the changes/new records from Folio.
Daily MRK/MRC files of the whole catalog are on a SharePoint thanks to @nflorin . VuFind should be using same APIs are used for that whole catalog harvest (re-verify?).
Restart has not resolved the issue. SOLR index is functioning. Possible causes:
I am going to try split and validate the catalog file.