In Zebra we currently have this index: <index_subfields tag="952" subfields="u"> <target_index>uri:u</target_index> </index_subfields> It seems to be connected only to items.uri however. The generic index name would suggest a broader range. Or perhaps, allow a broader range? We still have one other specific uri indexed: <index_subfields tag="024" subfields="a" condition="marc:subfield[@code='2' and text()='uri']"> <target_index>Identifier-other:u</target_index> </index_subfields> Two popular MARC fields for URIs, 520$u and 856$u are not indexed in a specific index. How about adding them to uri ? Since we do now have uri bound to items, do we need a specific subindex (like item_uri) for that one? Or the other way around, given that uri is already taken, why not add generic_uri or any other (better) new index name? And still elaborating here, a urx search does not really work in Koha. Several special characters in URLs seem to return nothing, since C4/Search does not like them probably. (While the search works in yaz-client.) Should we make it possible to search those just like other text fields too? If I am looking for a record containing a URL like sub.domain.country, why not look for a word in the string like sub or domain?
I might be wrong, but I do not see the uri as a specific index type in Elastic?