Note: These instructions are for Bepress Digital Commons, but many steps will apply to any harvest pipe.
Basic Information
Useful Orbis Cascade instructions to start with: https://www.orbiscascade.org/blog/9/?bid=153
Important to know before you begin:
- Go in order: source, scope, norm rules, then pipe last.
- Deploy after every step, and Deploy All before running the pipe for the first time.
- Records will not be discoverable in OneSearch until indexing is run for the entire installation. The pipe is for our local instance, but the records are still added to a single installation-level database, which is all indexed at once.
- View harvested records in the PNX viewer at any time after the pipe finishes. Search using [source code*] in Record ID. Adding date/time to Created after or Updated after will make the search go much better (it's super slow otherwise). Put date AND time (e.g. 05/03/2017 01:01) or it won't work.
...
Add the scope to any view where you want the IR records to be in your search results. The One Ex Libris trainer defined the scopes you build in the Views area as "scope buckets." They're called scopes, but they're really containers of one or more scopes. You can add your IR as a separate scope bucket so it will be searched in a separate tab/dropdown, or add it as a value to an existing scope bucket so the results are just automatically included with whatever else it's searching.
Separate scope bucket:
Add to existing scope bucket:
If you added it to an existing scope bucket, you're done. If not, you need to choose which tab to put it in, click Edit, and then Add a New Scope:
Norm Rules
Duplicate from Generic Dublin Core. Start with the Orbis Cascade recommended changes. For an example of modified norm rules (for local additions and to adapt to Qualified Dublin Core) see here: Example of DC Norm Rules Changes for Institutional Repository Harvest
Test records: Load a harvested file, or do this quick-and-dirty method:
...