search

sul-dlss / exhibits

Stanford University Libraries online exhibits showcase
https://exhibits.stanford.edu
Other
19 stars 7 forks source link

Questions re exhibits and MD import from Rumsey IIIF manifests #2034

Open caaster opened 3 years ago

caaster commented 3 years ago

Hi Cathy,

We have another IIIF question: previously, when importing items into Spotlight from davidrumsey.com via IIIF, the images always loaded so I considered their imports successful.

However, it's true they import with limited metadata (just title and attribution), as seen in the following example: https://exhibits.stanford.edu/islamicworld/catalog/1c152adf83e8ff3e9bfb739a74396599

Here is this example's IIIF manifest URL from davidrumsey.com: https://www.davidrumsey.com/luna/servlet/iiif/m/RUMSEY~8~1~301244~90072207/manifest

I've just noticed that there appears to be more metadata in the IIIF manifest that could be imported than what eventually shows up in Spotlight. If this also appears to be the case to you, can it be fixed?

For reference, here's an item from the Smithsonian imported via IIIF, with more metadata: https://exhibits.stanford.edu/islamicworld/catalog/1b0fe9ff720a307afc67ed856956bf95

Since we use davidrumsey.com images frequently in our exhibitions, we figure it may be worth taking a look at how the metadata imports from davidrumsey.com IIIF URLs. Please let me know if anything's unclear, or if I can provide more information.

Thanks, Rueiyun

Rueiyun Wang Center Services Support Specialist David Rumsey Map Center, Stanford University rueiw@stanford.edu

jkeck commented 3 years ago

Just in case it helps anybody looking into this, I believe the issue at hand is the metadata described in the manifest is canvas level and not at the manifest level.

The IIIF importing code will create custom metadata fields in the exhibit for the manifest level metadata (as well as a few other things like attribution, license, etc) and add the relevant data from the manifest, however it does not currently make any attempt to drill into the metadata for all the canvases and roll them up into object/manifest level metadata.

cbeer commented 3 years ago

I'm not sure what it would even mean to roll up canvas-level metadata onto the object itself, because typical use of canvas metadata is for things like "Page 1", "Page 2", "Page 3", physical dimensions, etc.