stop indexing collection if they aren't public #1322
Conversation
dnoneill
force-pushed
the
3840-collection
branch
from
18e571e to
6520c7a
Compare
dnoneill
force-pushed
the
3840-collection
branch
from
6520c7a to
238233b
Compare
|
@dbranchini is the designed solution to not to link to the collection, or to not show any collection information at all? I think this code change only accomplishes the latter. |
|
@jcoyne I think that's more of a question for @andrewjbtw. |
Are you referring to the collection information being in the MODS? I don't know how we can prevent that from appearing. My understanding of the scope of this issue is to prevent displaying links to collection pages that don't exist. |
| get_value(public_xml_doc.xpath('/publicObject/identityMetadata/objectLabel')) | ||
| end | ||
|
|
||
| def collection_in_searchworks? |
There was a problem hiding this comment.
I think I've been tripped up by this method name and/or the context for how this gets used. Maybe it's just the method name, but I wonder if there's a more explicit/clearer way to implement this (kind of confusing in the first place) requirement.
From your description, it sounds like we're trying to suppress the id and title from the collection and collection_with_title fields?
Instead of tweaking the underlying data for collection objects (which.. feels like it might be useful in some contexts.. debugging if nothing else?), what if PublicXmlRecord had a released_collections accessor that we used anywhere we needed it?
There was a problem hiding this comment.
That works. The upside of this is we don't have to make any changes to the code in searchworks or the solr index. I have also left collection_with_title to still index with this it just returns an empty value of -|- which I thought would be good for debugging. Basically it would indicate that this has a collection attached but it isn't public. Probably not the best indicator so I can update the code, but I think that means we would need to update the solr index along with Searchworks.
From what I can tell this is working on sw. It will not fix the issue with currently indexed items but it will stop this issue from re-occuring. I have been using https://searchworks-preview-stage.stanford.edu/view/bh250xv2418 + https://argo-stage.stanford.edu/view/druid:qr592gj5093 to test that this works. It looks like you can release/withdraw a collection without its objects and the items will get reindex. Be warned, it can take a while to reindex.
Closes sul-dlss/SearchWorks#3840
If a argo collection has not been released to Searchworks, the solr index with not provide the id in the collection field. It still provides collection_with_title which becomes -|- which I thought would be good to keep to let us know that there is a non public collection associated with the record but Searchworks doesn't display those non viewable collections.