Bug 24675

Summary: No update of the list of MARC21 subfields in C4/Heading/MARC21.pm, Generated authorities are missing subfields
Product: Koha Reporter: Bouzid Fergani <bouzid.fergani>
Component: CatalogingAssignee: Bouzid Fergani <bouzid.fergani>
Status: Failed QA --- QA Contact: Testopia <testopia>
Severity: normal    
Priority: P5 - low CC: 1joynelson, dcook, katrin.fischer, m.de.rooy, martin.renvoize, nick, phil
Version: unspecified   
Hardware: All   
OS: All   
See Also: https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=20391
Change sponsored?: --- Patch complexity: ---
Documentation contact: Documentation submission:
Text to go in the release notes:
Version(s) released in:
Attachments: Bug24675: Update MARC21 config for bib_heading_fields
Bug24675: Update MARC21 config for bib_heading_fields
Bug 24675: Update MARC21 config for bib_heading_fields
Bug 24675: Update MARC21 config for bib_heading_fields

Description Bouzid Fergani 2020-02-17 17:31:24 UTC

    
Comment 1 Bouzid Fergani 2020-02-17 18:58:44 UTC Comment hidden (obsolete)
Comment 2 Bouzid Fergani 2020-02-17 19:06:11 UTC
Created attachment 99125 [details] [review]
Bug24675: Update MARC21 config for bib_heading_fields

To reproduce:
      - Enable the sysprefs
            AutoCreateAuthorities
            BiblioAddsAuthorities
            CatalogModuleRelink
      - Edit or open any biblio record
      - fill all subfields of the field 100
         In my case , I have $a, $d, $e, $q
      - Save and continue editing
      - Check the $9 of the field 100 filled for the authid(auto create of authority)
      - Open the detail of the new authority created using authid ($9)
      - In the new authority, $e is not filled
This problem produce, because in C4/Heading/MARC21.pm, the list of subfields for each field is not upated since 2011.
The source of new list of subfields is : library of congress (https://www.loc.gov/marc/bibliographic)
When applying the patch, all subfields is reported correctly in the new authority.
NB: All the subfields not written in the file C4/Heading/MARC21.pm have the same problem.
Comment 3 Bouzid Fergani 2020-02-17 20:09:50 UTC
Created attachment 99130 [details] [review]
Bug 24675: Update MARC21 config for bib_heading_fields

To reproduce:
      - Enable the sysprefs
            AutoCreateAuthorities
            BiblioAddsAuthorities
            CatalogModuleRelink
      - Edit or open any biblio record
      - fill all subfields of the field 100
         In my case , I have $a, $d, $e, $q
      - Save and continue editing
      - Check the $9 of the field 100 filled for the authid(auto create of authority)
      - Open the detail of the new authority created using authid ($9)
      - In the new authority, $e is not filled
This problem produce, because in C4/Heading/MARC21.pm, the list of subfields for each field is not upated since 2011.
The source of new list of subfields is : library of congress (https://www.loc.gov/marc/bibliographic)
When applying the patch, all subfields is reported correctly in the new authority.
NB: All the subfields not written in the file C4/Heading/MARC21.pm have the same problem.
Comment 4 Martin Renvoize 2020-02-24 12:31:52 UTC
Created attachment 99508 [details] [review]
Bug 24675: Update MARC21 config for bib_heading_fields

To reproduce:
      - Enable the sysprefs
            AutoCreateAuthorities
            BiblioAddsAuthorities
            CatalogModuleRelink
      - Edit or open any biblio record
      - fill all subfields of the field 100
         In my case , I have $a, $d, $e, $q
      - Save and continue editing
      - Check the $9 of the field 100 filled for the authid(auto create of authority)
      - Open the detail of the new authority created using authid ($9)
      - In the new authority, $e is not filled
This problem produce, because in C4/Heading/MARC21.pm, the list of subfields for each field is not upated since 2011.
The source of new list of subfields is : library of congress (https://www.loc.gov/marc/bibliographic)
When applying the patch, all subfields is reported correctly in the new authority.
NB: All the subfields not written in the file C4/Heading/MARC21.pm have the same problem.

Signed-off-by: Martin Renvoize <martin.renvoize@ptfs-europe.com>
Comment 5 Nick Clemens (kidclamp) 2020-03-04 10:57:46 UTC
The reference here is to the MARC bibliographic records - $e is not a subfield that is supposed to be recorded in authorities - that would end up creating a different record for each relator terms for the same individual.

Check here, under the 'LC Guidelines' link near the bottom - LC has a list of subfields not to use, including e

We might need an update here, but I think we need to get an authoritative list with some cataloger input before moving it ahead
Comment 6 Katrin Fischer 2020-04-04 22:38:40 UTC
(In reply to Nick Clemens from comment #5)
> The reference here is to the MARC bibliographic records - $e is not a
> subfield that is supposed to be recorded in authorities - that would end up
> creating a different record for each relator terms for the same individual.
> 
> Check here, under the 'LC Guidelines' link near the bottom - LC has a list
> of subfields not to use, including e
> 
> We might need an update here, but I think we need to get an authoritative
> list with some cataloger input before moving it ahead

Nick is correct, $e, $4 and maybe others should not be part of the generated authority record as they can be differerent in every record using the same authority.
Comment 7 Phil Ringnalda 2020-04-08 06:39:52 UTC
The actual situation with $e is even weirder, because although LC chooses not to use it, $e is actually defined for authority 100 tags in http://www.loc.gov/marc/authority/ad100.html but it isn't defined for what you would expect. 

Instead, as documented in http://www.loc.gov/marc/authority/adx00.html it's the relator between the $a and the $t in a name/title heading. So if you were cataloging according to MARC but not according to LC, you could have an "800 1# $aThunkitup, John,$ecreator$tOthers write John's characters series", and when Koha created an authority record from that and dropped the $ecreator it would be wrong.

Making that work for 700, where $a$e means the relationship between the $a and the work the bib record describes, but $a$e$t means the relationship between the $a and the $t, would require a rewrite of how Koha decides what subfields are appropriate, to make them dependent on the existence of others, and that horrible ambiguity is probably why LC decided that it wasn't a good idea so they wouldn't use it.