access and quality UX on results page

[berger-n]

currently we have these access/quality indications:

• restricted by tbrc:

• sealed/fair use (with possible IA link):

• temporarily restricted:

• low quality:

---

now the question is how to display it using this new indicator: https://github.com/buda-base/git-to-dbs/blob/elasticsearch/src/main/java/io/bdrc/gittodbs/ESUtils.java#L820-L831

    // rank_features must be strictly positive (non-zero)
    // etext access:
    //   1: no access
    //   2: search only
    //   3: open access

    // scans access:
    //   1: no access
    //   2: extract only
    //   3: IA
    //   4: open access through IIIF
    //   5: open access on BDRC

[berger-n]

added the scans/etext access filters: https://library-dev.bdrc.io/osearch/search?q=bka%27%20%27gyur

[berger-n]

regarding quality, I think we can have a widget but I'll need more category names like:

• 4: computer input I guess?
• 3: excellent
• 2: very good
• 1-0.66: good
• 0.66-0.33: average
• 0.33-0: poor?

https://library-dev.bdrc.io/osearch/search

[eroux]

thanks! For etexts I think we need stronger standards:

• 1.0-0.95 : very good OCR
• 0.95-0.8: usable OCR
• 0.8-0: poor quality OCR

[eroux]

and let's not make an aggregate with scans_quality, there's almost no data for it (yet)

[berger-n]

now with etext quality ranges: https://library-dev.bdrc.io/osearch/search?etext_quality%5B0%5D=2.99-3.01&sortBy=firstScanSyncDate_desc (note that there are a few scans_quality already available/displayed, should I remove them for now? everything is ready to use, we just have to update the ranges/labels here and here)

---

BTW I think I've fixed the issue with sorting not returning the same results/returning too many results: (this used to return like ~1000 results) https://library-dev.bdrc.io/osearch/search?q=%27ba%27%20bro&sortBy=firstScanSyncDate_desc

[eroux]

wonderful, thanks! you can leave the scans quality, that's fine, the most important is that they don't appear in the filters