OD2
OD2 copied to clipboard
Blacklight facets sorted A-Z sorts regardless of uppercase/lowercase
Descriptive summary
~If you chose more
from a facet to get the all the facetable values. Sort by A-Z Sort
. All values will be uppercase A-Z. At the end of uppercase will be lowercase A-Z. Facet values should be normalized to uppercase the first letter OR A-Z sort should have A
and a
next to each other.~
Edit: Previously we attempted to sort uppercase before lowercase on the A-Z sort facet modal. This isn't actually possible, and needs to be reverted. A-Z Sort may be sorted with uppercase and lowercase intermingled, preferring correct alphabetical sort.
Expected behavior
Facet A-Z sort is caseless sorted A-Z
Related work
Accessibility Concerns
Values from external controlled vocabs can only be normalized in-system. Can the system be forced to ignore capitalization altogether?
I'm not sure we can do either solution until we look closer, but my thinking was for in-system normalization would be best followed by ignoring capitalization. Assuming we can only ignore and not normalize, would interspersed upper- and lower-case facet values be okay? or would the current sort method be preferable?
A mix of upper- and lowercase might look slightly messy, but it'd be better for browsing than having two alphabetic sequences. Thoughts @kmthorn @petersec ?
I can live with a little messiness; we definitely don't want an A-Z list followed by an a-z list.
The labels in the facets should appear as they come from the vocabularies. We can't force capitalize everything.
The actual value that is indexed and faceted on can be normalized so capitalization doesn't affect display.
This is related to the work to index the faceted value as label$URI
but show just label
QA pass. Noticed that "the" is being used as the first word in the sort
Something is really wrong. The lower-case topic block appears smack in the middle of the upper-case Ws.
@straleyb FYI
This is another one that's got me stumped. Solr itself can't sort case-insensitive and the faceting windowed offsets to retrieve the facets, meaning we can't post-process the sort. Our option would be to have all the facets in lowercase or the first letter capitalized. The original values can still be displayed on the show page though
Integrating the upper and lower case is the primary concern. POSM doesn't care about transforming the uppercase to lowercase or vice-versa.
QA:
- [x] Open facet model using
more >>
button - [x] Model opens w/ option to A-Z Sort
- [x] A-Z sort sorts A-Z
- [x] Ending couple of A-Z sort pages continue to sort correctly
A-Z sort is working, but now displays all the first names/portions of the facet in lower case
@CGillen do you want me to pass it with a new ticket for that issue or send this back to in progress?
We can move this forward. POSM isn't concerned w/ uppercase/lowercase transformation
@jsimic @shieldsb @CGillen I don't think this is acceptable. This looks horrible:
Is there no way to only capitalize labels that are lowercase and retain any existing capitalization?
We can certainly revert for now if you want.
We might be able to do this with a new solr field. It would still mix-up mixed-cased words. Like McDonald
would be above Mca*
, and Mcdonald
would be below Mcc*
When looking at individual letters in the A-Z faceted sort (such as "A") nothing displays. I assume this is related to the rest of alphabetical sorting issues. If you would like a separate ticket to track this please let me know. I think this is related to/impacts https://github.com/OregonDigital/OD2/issues/3023