Update search logic - AND between facets, OR within a facet

[AntonPetrov]

This will make RNAcentral search consistent with Kayak, Amazon and others.

Burkov: Time estimate is: ??? (ballpark of 10-20 days)

[AntonPetrov]

We need to change how faceted queries are constructed. Instead of adding facets to the query parameter, they need to be specified as a facets parameter. In this example, both tRNA and miRNA are found: http://www.ebi.ac.uk/ebisearch/ws/rest/rnacentral?query=rna&facets=rna_type:miRNA,rna_type:tRNA

The facets need to be added to the URL, but not shown in the search box.

Example EBI search where both human and mouse miRBase hits are retained: https://www.ebi.ac.uk/ebisearch/search.ebi?db=rnacentral&query=mirbase&facets=TAXONOMY:9606,TAXONOMY:10090

[blakesweeney]

Just ran into this issue so I'm adding some comments about the searches I tried. If I do sequences in gorrilla I don't see any that are also in human I should find some examples, because I know those species are highly related. If I click on E. coli I should at least see some tRNA's as there are structures in PDB with tRNA's that have the same sequence.

So what I think the search does: a sequence found in hgnc and found in GENCODE, so I would expect the queries to work like: a sequence found in human and found in E. coli. However this isn't the case since the search operates at the level of species specific URS's and not a generic URS (but returning results at that level is correct).