https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=32906

Phil Ringnalda <[email protected]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[email protected]

--- Comment #2 from Phil Ringnalda <[email protected]> ---
I do see an architectural bug, but it's a big stretch to say that this bug is
about it.

Both do support UNIMARC, the editor because the poorly-named
GetMandatoryFieldZ3950 maps 200a to 'authorpersonal', 210a to 'authorcorp', and
230a to 'uniformtitle', the details display because the authfields_mapping
added by attachment 146380 in bug 29173 maps 200a to 'authorpersonal', 210a to
'authorcorp' and 'authormeetingcon', 230a to 'uniformtitle', and 250a to
'subject'.

Wait, those are rather different mappings! And indeed, if you Replace via
Z39.50 from the details display for a UNIMARC corporate name, the existing
value is prefilled for both Author (corporate) and Author (meeting /
conference) which will fail with most if not all servers, and you only get
prefill for a topical subject record if you replace from the details display,
not the editor (probably because the patch for UNIMARC in
GetMandatoryFieldZ3950 was written by a different author than the original
feature patch, and the UNIMARC person realized that the other person didn't
actually make subject work and just skipped it, not realizing that when I
changed topic to subject to make MARC21 work six years later, I wouldn't know
to also add in 250a for UNIMARC).

So, yes, it might be better if there was an exported
GetAuthorityZ3950SRUMapping that both the editor and details used, but it's a
deep rabbithole. The right thing to prefill actually depends on heading, MARC
flavor, and target server (from what I use, 130a should go in Title (uniform)
if I'm searching LoC, but Title (any) if I'm searching OCLC). And given that
SRU can map any index or no index to a particular input, and you can be
searching one or more servers, really making mapping work requires some weird
ranked-choice prefill pref on Z39.50/SRU server definitions.

Not convinced that the non-rabbithole "just have a single exported GetMapping
that both can use rather than having two separate mappings" would be worth the
trouble. And don't even think about making it worthwhile by doing the same
thing for biblios, you absolutely Do. Not. Want. to look at what Replace via
Z39.50 from a bib detail display does.

-- 
You are receiving this mail because:
You are watching all bug changes.
You are the assignee for the bug.
_______________________________________________
Koha-bugs mailing list
[email protected]
https://lists.koha-community.org/cgi-bin/mailman/listinfo/koha-bugs
website : http://www.koha-community.org/
git : http://git.koha-community.org/
bugs : http://bugs.koha-community.org/

Reply via email to