VuFind was initially designed with the MARC bibliographic record format in mind, though additional formats are supported through the use of Record Drivers starting with version 1.0. For general information on MARC, see Understanding MARC Bibliographic from the Library of Congress. The Code4Lib Working with MARC page also provides some useful tools.

VuFind comes packaged with the SolrMarc tool for importing MARC records. Follow these steps to take advantage of it.

Before you can load the records into VuFind, you need to get them out of your Integrated Library System (ILS). If you are just testing VuFind, you can also download sample records from sources listed lower on this page.

Every ILS has a different procedure for exporting records, and detailing all of them is beyond the scope of this document. Check your ILS documentation or talk to your vendor if you need help. You can also check the MARC Export Notes page to see if there are notes specific to your ILS; please consider adding to the page if you have knowledge to share. If you still need help, you can always ask on the mailing lists on the Support page – the VuFind community is always happy to help when it can.

Keep these notes in mind to ensure that your records can be imported without any problems:

• Export your records in binary (ISO2709) MARC format, not human-readable ASCII. If for some reason you cannot export the records in binary form, you can use a tool like yaz-marcdump from the YAZ toolkit to convert one MARC format to another.
• Make sure your resulting file has a ”.mrc” extension. Most versions of SolrMarc require this extension, so it is a good practice to use it just to be on the safe side.
• Each exported record must contain a unique identifier so that VuFind can tell it apart from the others. We recommend including your ILS's bibliographic record ID in the exported data for this purpose; you may need to add a special configuration option to your ILS's exporter to make this happen. VuFind's importer expects to find the unique ID in the 001 field, but you can customize this by editing the marc.properties file (for more details, see Customizing Import Mappings).

To begin an import, follow the platform-specific instructions listed below. This may take hours or days for very large data sets!

Switch to your VuFind installation directory and run:

./import-marc.sh your_records_file.mrc

Note: In versions of VuFind prior to 1.0RC2, import-marc.sh was named import.sh.

Switch to your VuFind installation directory and run:

import-marc.bat your_records_file.mrc

Note: prior to VuFind 1.0RC2, import-marc.bat was not available, and it was necessary to run SolrMarc manually:

Java -Xms256m -Xmx256m -Dsolr.core.name=biblio -Dsolrmarc.path=C:/vufind/import -Dsolr.path=C:/vufind/solr -Dmarc.path=c:/vufind/import/catdump.mrc -jar c:/vufind/import/dist/MarcImporter.jar c:/vufind/import/import.properties

(thanks to mike_beccaria)

If the imported records do not show up in VuFind immediately, you will have to restart the program as described here.

For improved performance (and, if applicable, correct spellchecker behavior), it is a good idea to optimize your Solr index after you import records.

See the SolrMarc page for more details on how you can customize the behavior of the import process to meet your needs.

This section is for listing sources of binary MARC records helpful for testing purposes if you want to try VuFind without using your own records:

• Annenberg Media: MARC Records
• Internet Archive: Library Catalog Records (many large files; some may require tweaking!)

• Automation - Notes on automating VuFind, including how to regularly load the latest records.
• Re-indexing - How to clear out your index if you want to start over.

• VuFind Import Strategies