[Koha] Authority main heading indexing for UNIMARC

[Kathy Hines]

Please delete me from this mailing list.
Kathy

-----Original Message-----
From: Koha [mailto:koha-bounces at lists.katipo.co.nz] On Behalf Of Pedro Amorim
Sent: Monday, August 22, 2016 1:11 PM
To: Koha list
Subject: [Koha] Authority main heading indexing for UNIMARC

Hello all,

Koha 3.22 from packages UNIMARC install on Debian.
This is not a fresh install but all the configuration (and indexing) files have not been modified.
Everything was seemly working fine up until the point we started using authority records.

This is a follow-up on a previous mail I sent:
https://lists.katipo.co.nz/public/koha/2016-August/045991.html

I have narrowed the problem down to the indexing (or lack of) of the Heading:p index for authorities.

Say I have this authority (NP) record:
200 a$Amorim $bPedro $f1991-1992

Conducting an authority search from the staff interface with search term:
"Amorim Pedro 1991-1992"
With exact match on main heading, should result on the above record. But it doesn't.

Instead, an exact search for each of the above subfields alone ('Amorim', 'Pedro' or '1991-1992') will all return the same expected correct record.

As if the authority indexing is indexing all the subfields, but as words instead of a whole phrase.

Additional info:
I've stumbled upon the GRS-1/DOM terms but from what I understand, after Koha 3.18 it's the DOM indexing that's used and GRS-1 is deprecated. Also, all config files point to DOM by default.

------

I conducted test searches on (UNIMARC) Koha implementations for this incorrect behaviour and oddly enough, it's quite common.

The UNIMARC demo test Biblibre implements this search as phrase correctly, searching for all the subfields in the heading will result in the correct
record:
http://intranet-demo.biblibre.com/

However, Tamil doesn't. It behaves the same as my installation, not interpreting phrase but only words:
https://kpro.tamil.fr/

------

Additional yaz info results:

Z> f @and @attr 1=authtype NP @attr 1=Heading @attr 4=1 @attr 6=3 "Amorim"
Sent searchRequest.
Received SearchResponse.
Search was a success.
Number of hits: 1, setno 1
SearchResult-1: term=NP cnt=1, term=Amorim cnt=1 records returned: 0
Elapsed: 0.005944
Z> f @and @attr 1=authtype NP @attr 1=Heading @attr 4=1 @attr 6=3 
Z> "Amorim
Pedro 1991-1992"
Sent searchRequest.
Received SearchResponse.
Search was a success.
Number of hits: 0, setno 2
SearchResult-1: term=NP cnt=1, term=Amorim Pedro 1991-1992 cnt=0 records returned: 0
Elapsed: 0.000615
Z> scan @attr 1=Personal-name @attr 4=1 @attr 6=3 "amorim"
query: @attr 1=Personal-name @attr 4=1 @attr 6=3 "amorim"
Received ScanResponse
6 entries, position=1
Scan returned code 5
* Amorim (1)
Pedro (1)
Elapsed: 0.037127

The yaz results make it obvious the indexes are wrong and only indexing words, I think.

-----------------------

Sorry for long post, wanted to provide as much as useful info as possible.
Any help would be greatly appreciated.

Thank you,

Pedro Amorim
_______________________________________________
Koha mailing list  http://koha-community.org Koha at lists.katipo.co.nz https://lists.katipo.co.nz/mailman/listinfo/koha