I am not sure I really got this.
Because if I did:
- The error messages in your orig posting were not the exact error messages.
- The original posted code was not the actual code producing the errors
- And the sample MARC records, supplied to demonstrate the errors, were
actually OK
Sometimes
Mystery solved. It was indeed bad leaders.
This fixes it: substr($line,20,8)=~s/45.{3,5}1/451/;
Bits 22 and 23 of the leader were corrupt. (Anne Highsmith, our SysAdmin,
enlightened me. They are used by some record vendors for record source
info.)
In case you run into mysterious
