Cataloging Home Cataloging Wiki

Libraries' catalog
Journals | E-Matrix
WorldCat NCLive | Beta
Public Department page
Staff server

Department
Annual reports
Directory
Personnel guidelines
Presentations
Production statistics
Tech Services requests

Department sections
Continuing & Electronic Resources
Metadata & Database Quality
Monographs

Cataloging tools
Cataloging agencies
Cutter tables
Design images classification
Format-specific tools
MARC documentation
OPAC tools
Reference sources
Remote OPACs
Special call numbers
Symphony tools
Training

Local procedures
Cataloging guidelines & procedures by topic
Metadata project procedures

News for catalogers
Recent news/events
Technical service e-journals

Open Content Alliance metadata capture procedures

Capture MARC XML metadata from Open Content Alliance Website

• Digitized content available through NCSU Libraries main OCA content page
• For each scanned title splash page, use the HTTP link from the View the book box on the left



• From the HTTP Index page, right click on the file with marc.xml extension; save link to g:\Cataloging\Metadata\Open Content Alliance\pending



• DLI has devised a local script page which automatically harvests OCA metadata and stores it at T:\depts\cataloging\oca\marcxml

Transform OCA XML to MARC21

• From MarcEdit, choose Tools, then Batch Process Records
• Source directory should be g:\Cataloging\Metadata\Open Content Alliance\pending; File Types: (blank); Function: OCAxml=>MARCXML
• Click on Process



• Transformed files will be in g:\Cataloging\Metadata\Open Content Alliance\pending\processed_files
• If resulting MARC XML file has 0 bytes:
• Open corresponding file in g:\Cataloging\Metadata\Open Content Alliance\pending\ in Oxygen XML editor
• Check file for duplicate subfields in 260; fix
• Run MARC Breaker on single files, saving to g:\Cataloging\Metadata\Open Content Alliance\Pending\processed_files to overwrite bad files
• From MarcEdit, run Batch Process Records again, this time to transform MARC XML to MARC



• End result of this process will be separate MARC records for each scanned volume in g:\Cataloging\Metadata\Open Content Alliance\Pending\processed_files\processed_files
• From MarcEdit, choose Tools, then MARCjoin
• Source File: will be the name you wish the end file to have (I use OCAmmdd.mrc, where mm=month, dd=day)
• File(s) to Join: will be the contents of g:\Cataloging\Metadata\Open Content Alliance\Pending\processed_files\processed_files; highlight all files (CTRL A) to grab all



• If you have multi-volume sets to combine before batching to OCLC, open the joined MARC file in Marc Breaker and copy 856s for 2nd, 3rd volume to the record for 1st volume; add $zVol. 1, $Vol. 2, etc. to 856; delete bib records for extra volumes

Import transformed records into WorldCat

• Open OCLC Connexion
• From File open Import Records to import the file into your local save file.
• Enter the path/filename to your saved MARC records, usually g:\Cataloging\Metadata\Open Content Alliance\Pending\OCAmmdd.mrc
• In answer to the question "Delete Original File", say NO



• Within OCLC Connexion, login to WorldCat and open your local save file. Today's records should have "NEW" in "Control #" and today's date in "Date/Time Added".
• Check/edit each record, validate, and UPDATE to utility. Common problems here include:
• Diacritics which don't translate into the MARC XML
• Non-AACR2 punctuation not caught by XSLT
• Capitalization of corporate names, especially in 245, 260, & 4xx
• Use of abbreviated publisher statement in 260$b ("University Press", "The Department", "Gov't Printing Ofc")
• Untraced series that should be traced
• Unacceptable characters in 776 $w (DLC), usually prefixes such as "agr"; simply remove these
• Highlight all records processed for the week and export to your computer

Load OCLC export file into Unicorn

• In Unicorn, open MARCImport wizard, import file onto Unicorn server using filename you will remember (OCAmmdd.mrc)
• From New report menu, Templates, open Load OCA titles, enter file name assigned above on Load tab, File to load.
• From Finished reports, check finished report to ensure records loaded