Thanks Terry.
The records are extracted from our ILS (SirsiDynix's Symphony) using a custom report that uses a list of Item IDs (i.e. barcodes) as input. One MARC record is output for each barcode, with a 955 field containing the call number ($a) and barcode ($b).
We've extracted and sent tens of thousands of records this way. This record is anomalous and we can't determine why. In other words, where is that Hex code BF coming from? Still a mystery.
We're resubmitting the record, having manually updated the Leader and recompiled, but we'd still like to figure out why our system is seemingly corrupting this data (and seemingly only this data) and (again, seemingly) only on this record.
Jeff
----- Original Message -----
From: "Terry Reese" <[log in to unmask]>
To: [log in to unmask]
Sent: Thursday, February 6, 2014 1:23:12 PM
Subject: Re: [MARCEDIT-L] Bad Leader
So -- I'm not sure where you are getting your data, but various ILS systems take the undefined bytes and reset them. In the example you provided, the mnemonic {BF} is representative of the Hex code BF -- not a printable character. If you wanted to fix that -- you could remove the entire {BF} mnemonic with the desired byte and recompile.
I'm not sure that's all that's going on here (given the difference in the record sizes) -- but I couldn't tell without seeing the raw datafile.
--tr
-----Original Message-----
From: MarcEdit support in technical and instructional matters [mailto:[log in to unmask]] On Behalf Of Jeff Edmunds
Sent: Thursday, February 6, 2014 12:57 PM
To: [log in to unmask]
Subject: [MARCEDIT-L] Bad Leader
All,
Anyone else submitting records to HathiTrust/Google via Zephir? We recently had a record rejected (OCLC # 56437757, for what that's worth) because of a bad Leader. The Zephir diagnostic looks like this:
bad line: <marc:record><marc:leader>01380&#xbf;amÆ°a2200361 a 450Ì</marc:leader> yaz_marc_read_xml failed
When we opened the record in MarcEdit, we saw that LDR Position 05 was not a single character, but {BF} (whatever that means). We tried manually changing this string to "c" and re-compiling the record into MARC, but this didn't seem to work. Then we tried overwriting the copy of the record in our catalog with the OCLC version, reconverting it to XML, and resending it to the HathiTrust folks.
The record was again flagged as having a bad Leader, but this time the diagnostic looks slightly different:
bad line: <marc:record><marc:leader>01750&#xbf;amÆ°a2200433 a 450Ì</marc:leader> yaz_marc_read_xml failed
Does anyone have any experience with odd characters somehow making their way into Leaders?
Thanks,
Jeff
________________________________________________________________________
This message comes to you via MARCEDIT-L, a Listserv(R) list for technical and instructional support in MarcEdit. If you wish to communicate directly with the list owners, write to [log in to unmask] To unsubscribe, send a message "SIGNOFF MARCEDIT-L" to [log in to unmask]
________________________________________________________________________
This message comes to you via MARCEDIT-L, a Listserv(R) list for technical and instructional support in MarcEdit. If you wish to communicate directly with the list owners, write to [log in to unmask] To unsubscribe, send a message "SIGNOFF MARCEDIT-L" to [log in to unmask]
________________________________________________________________________
This message comes to you via MARCEDIT-L, a Listserv(R) list for technical and instructional support in MarcEdit. If you wish to communicate directly with the list owners, write to [log in to unmask] To unsubscribe, send a message "SIGNOFF MARCEDIT-L" to [log in to unmask]
|