The OAI-PMH Harvester plugin imports records from OAI-PMH data providers.
Some online repositories expose their metadata through the Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH). This plugin makes it possible to harvest that metadata, mapping it to the Omeka data model. The plugin can be used for one-time data transfers, or to keep up-to-date with changes to an online repository.
Currently the plugin is able to import Dublin Core and CDWA Lite metadata. Dublin Core is an internationally recognized standard for describing any resource. Every OAI-PMH data provider should implement this standard. CDWA Lite is a standard for describing works of art and material culture. Very few repositories expose CDWA Lite, but the standard is getting more and more popular.
- Path to PHP-CLI: Path to your server's PHP-CLI command. The PHP version must correspond to normal Omeka requirements. Some web hosts use PHP 4.x for their default PHP-CLI, but many provide an alternative path to a PHP-CLI 5 binary. Check with your web host for more information.
- Memory Limit: Set a memory limit to avoid memory allocation errors during harvesting. We recommend that you choose a high memory limit. Examples include 128M, 1G, and -1. The available options are K (for Kilobytes), M (for Megabytes) and G (for Gigabytes). Anything else assumes bytes. Set to -1 for an infinite limit. Be advised that many web hosts set a maximum memory limit, so this setting may be ignored if it exceeds the maximum allowable limit. Check with your web host for more information.
The data stored by the harvester plugin has changed between versions, so when upgrading, it is necessary to uninstall the old version of the plugin first. This will remove data stored by the harvester, but the harvested items themselves will remain.
To upgrade the plugin:
- Uninstall the old version of the OAI-PMH Harvester plugin from the admin panel.
- Replace the OaipmhHarvester directory with the updated version.
- Install the now-updated OAI-PMH Harvester plugin from the admin panel.
Performing a harvest
- Go to Admin > "OAI-PMH Harvester"
- Enter an OAI-PMH base URL, click "View Sets"
- Select a set and the metadata prefix to harvest, click "Go"
- The harvest process runs in the background and may take a while
- Go to the harvest's "Status" page to check the progress
- If you encounter errors, submit the base URL and status messages to the Omeka forums
Re-harvesting and updating
The harvester includes the ability to make multiple successive harvests from a single repository, keeping in sync with changes to that repository.
After a repository or set has been successfully harvested, a "Re-harvest" button will be added to its entry on the Admin > OAI-PMH Harvester page. Clicking this button will harvest from that repository again using all the same settings, adding new items and updating previously-harvested items as necessary.
Manually specifying the exact same harvest to be run again (same base URL, set, and metadata prefix) will result in the same behavior.
Aborting a harvest
A harvest in progress can be aborted by clicking the "Kill Process" button in the harvest's entry on the Admin > OAI-PMH Harvester page. The harvest will immediately be aborted. Alternatively, the PID of the background process is displayed for currently-running harvests, and this can be used to manually kill or pause the process.
Duplicate items (multiple Omeka items corresponding to the same repository record) can be created if an item in a repository is a member of several OAI-PMH sets. This will also occur if a repository is harvested using more than one metadata prefix. In this case, the duplicate items are idependent, and changes to one will not propagate to the others.
However, the duplicate items, if any, can be accessed from the admin item show page. If an item has duplicates, they will be shown in an infobox on the right-hand side of the page titled "Duplicate Harvested Items."