search

buda-base / public-digital-library

http://library.bdrc.io
5 stars 6 forks source link

handle "AND" in the filters for language #598

Open berger-n opened 3 years ago

berger-n commented 3 years ago

find a way to handle "AND" in the filters for the specific case of languages (ex: Khmer and Pali)

berger-n commented 3 years ago
berger-n commented 3 years ago

having a look at https://library.bdrc.io/search?q=%22dhammapada%22&lg=pi-x-ndia&t=Instance&f=asset,inc,tmp:possibleAccess&f=asset,inc,tmp:catalogOnly

Screenshot_20211108_144339

I'm wondering whether we are really talking of results which some parts are in Pali language and others in Khmer language here? and not results that are in Pali language written in Khmer script?

+ ldspdi-dev crashes for that search in pali: http://ldspdi-dev.bdrc.io/lib/instanceFacetGraph?LG_NAME=inc-x-ndia&L_NAME=%22m%C5%ABla%E1%B9%AD%C4%ABk%C4%81%20p%C4%81%E1%B9%ADh%22 but not ldspdi: http://ldspdi.bdrc.io/lib/instanceFacetGraph?LG_NAME=inc-x-ndia&L_NAME=%22m%C5%ABla%E1%B9%AD%C4%ABk%C4%81%20p%C4%81%E1%B9%ADh%22 needs an update maybe?

eroux commented 3 years ago

the dual languages ("Pali and Khmer") are basically Pali glossed in Khmer (ex: one word of Pali, several words in Khmer to explain it). bdr:PiKhmr is something else entirey (Pali written in Khmer script). Adding that in the query would be quite difficult... We could imagine adding bdr:PiKm for glosses though... not very satisfying in terms of modeling, but at least that would make things easy... let's do that! So I think we can more or less close this, I'll change things in the import

berger-n commented 3 years ago

ok! thanks for the explanation, if you prefer we can do that in the client, something like what we do for available assets (generating a facet info from another one)

eroux commented 3 years ago

hmmm that could be an idea yes... let's try that at some point (but not now)