I posted a new update to MarcEdit 5.0 last night relating to how the application handles invalid MARC records. Generally, most MARC tools just spit the records out or stop processing at the offending record. Well, in MarcEdit 4.6 there was a MARC “healer” function that would essentially attempt to evaluate how the record was invalid and then attempt to build data around it. Since I’ve made the default processing engine in MarcEdit 5.0 very strict (much stricter than 4.6), I also wanted to include this healing feature. Last night, I uploaded a new version that included some additional code to handle invalid records. So what does the program handle correctly?
1) Records where the leader and record length don’t match.
2) Tabs, new lines, invalid characters in the MARC record.
3) Record sets translated in unblocked format — i.e., separated by new lines, tabs (believe it or not, I’ve seen this), spaces, etc.
I’m working on adding code to the program that will also handle records were the directory simply has been corrupted beyond repair or is simply missing. I’ve nearly got it finished…so stay tuned.
As always, the download is found at: MarcEdit50_Setup.exe