updates to koha3 nozebra search indexing for MARC21
Hi koha-devel people, Ive been updating the marc index mapping for koha3-nozebra. The French team did a nice job on the UNIMARC stuff and ive had a go at padding out the currently pretty minimal MARC21 mapping. FYI: the *'s are wildcards, so, subject = '600*' means all 600 level subfields are used, when doing a 'subject' search any obvious errors or suggestions? thd, how does this look to your sharp arnd eagle-like gaze? ------------------------------------------------------------------------ ------------------------------------------------- 'title' => '210a,222a,240a,243a,245a,245b,246a,246b,247a,247b,250a, 250b,440a', 'author' => '100a,110a,245c,600a,610a,611a,700a710a,711a,800a,810a, 811a', 'isbn' => '020a', 'issn' => '022a', 'lccn' => '010a', 'biblionumber => '999c', 'itemtype' => '942c', 'publisher' => '260b', 'date' => '260c', 'note' => '500a, 501a,504a,505a,508a,511a,518a,520a,521a,522a,524a, 526a,530a,533a,538a,541a,546a,555a,556a,562a,563a,583a,585a,582a', 'subject' => '600*,610*,611*,630*,650*,651*,653*,654*,655*,662*,690*', 'dewey' => '082', 'bc' => '952p', 'host-item' => '952a,952c' ------------------------------------------------------------------------ ------------------------------------------------- Cheers, Mason -- Mason James mason.james@liblime.com Programmer www.liblime.com LibLime +64 21 542 117
Hi Mason, what about for 'author' adding tags and subfiled codes: 111a, 130a, 611a, 630a and 830a ? Irma _____ From: koha-devel-bounces+irma=bigpond.net.au@nongnu.org [mailto:koha-devel-bounces+irma=bigpond.net.au@nongnu.org] On Behalf Of Mason James Sent: 11 December 2007 08:34 To: koha-devel@nongnu.org Subject: [Koha-devel] updates to koha3 nozebra search indexing for MARC21 Hi koha-devel people, Ive been updating the marc index mapping for koha3-nozebra. The French team did a nice job on the UNIMARC stuff and ive had a go at padding out the currently pretty minimal MARC21 mapping. FYI: the *'s are wildcards, so, subject = '600*' means all 600 level subfields are used, when doing a 'subject' search any obvious errors or suggestions? thd, how does this look to your sharp arnd eagle-like gaze? ---------------------------------------------------------------------------- --------------------------------------------- 'title' => '210a,222a,240a,243a,245a,245b,246a,246b,247a,247b,250a,250b,440a', 'author' => '100a,110a,245c,600a,610a,611a,700a710a,711a,800a,810a,811a', 'isbn' => '020a', 'issn' => '022a', 'lccn' => '010a', 'biblionumber => '999c', 'itemtype' => '942c', 'publisher' => '260b', 'date' => '260c', 'note' => '500a, 501a,504a,505a,508a,511a,518a,520a,521a,522a,524a,526a,530a,533a,538a,541a,5 46a,555a,556a,562a,563a,583a,585a,582a', 'subject' => '600*,610*,611*,630*,650*,651*,653*,654*,655*,662*,690*', 'dewey' => '082', 'bc' => '952p', 'host-item' => '952a,952c' ---------------------------------------------------------------------------- --------------------------------------------- Cheers, Mason -- Mason James mason.james@liblime.com Programmer www.liblime.com LibLime +64 21 542 117
I think 111a should be added for 'author' I don't think 130a,830a are appropriate--those are title data. I'd like to see all 6XX pulled out of the 'author' index. A book *about* Shakespeare isn't the same as a book *by* Shakespeare, so I don't think the 600a,610a,611a should be in the author index. I'm wondering about subsequent subfields beyond 'a' for many of these--the 100 needs 'b' and 'c' and 'd' to be useful. Similarly 111 would ideally need 'c', 'd', 'n' . But maybe I'm not understanding the question or how this all works. Spencer On Dec 11, 2007 7:53 AM, Irma <irma@bigpond.net.au> wrote:
Hi Mason,
what about for 'author' adding tags and subfiled codes: 111a, 130a, 611a, 630a and 830a ?
Irma
________________________________ From: koha-devel-bounces+irma=bigpond.net.au@nongnu.org [mailto:koha-devel-bounces+irma=bigpond.net.au@nongnu.org] On Behalf Of Mason James Sent: 11 December 2007 08:34 To: koha-devel@nongnu.org Subject: [Koha-devel] updates to koha3 nozebra search indexing for MARC21
Hi koha-devel people,
Ive been updating the marc index mapping for koha3-nozebra.
The French team did a nice job on the UNIMARC stuff and ive had a go at padding out the currently pretty minimal MARC21 mapping.
FYI: the *'s are wildcards, so, subject = '600*' means all 600 level subfields are used, when doing a 'subject' search
any obvious errors or suggestions? thd, how does this look to your sharp arnd eagle-like gaze?
------------------------------------------------------------------------------------------------------------------------- 'title' => '210a,222a,240a,243a,245a,245b,246a,246b,247a,247b,250a,250b,440a', 'author' => '100a,110a,245c,600a,610a,611a,700a710a,711a,800a,810a,811a', 'isbn' => '020a', 'issn' => '022a', 'lccn' => '010a', 'biblionumber => '999c', 'itemtype' => '942c', 'publisher' => '260b', 'date' => '260c', 'note' => '500a, 501a,504a,505a,508a,511a,518a,520a,521a,522a,524a,526a,530a,533a,538a,541a,546a,555a,556a,562a,563a,583a,585a,582a', 'subject' => '600*,610*,611*,630*,650*,651*,653*,654*,655*,662*,690*', 'dewey' => '082', 'bc' => '952p', 'host-item' => '952a,952c' -------------------------------------------------------------------------------------------------------------------------
Cheers, Mason
-- Mason James mason.james@liblime.com Programmer www.liblime.com LibLime +64 21 542 117
_______________________________________________ Koha-devel mailing list Koha-devel@nongnu.org http://lists.nongnu.org/mailman/listinfo/koha-devel
-- Spencer M. Anspach, Library Systems Analyst/Programmer Library Information Technology, Indiana University Library W501 phone: (812) 856-5318 Bloomington, IN 47405 fax: (812) 856-4979 sanspach@indiana.edu pager: (812) 335-7403
On 2007/12/12, at 2:46 AM, Spencer Anspach wrote:
I think 111a should be added for 'author'
I don't think 130a,830a are appropriate--those are title data.
I'd like to see all 6XX pulled out of the 'author' index. A book *about* Shakespeare isn't the same as a book *by* Shakespeare, so I don't think the 600a,610a,611a should be in the author index.
I'm wondering about subsequent subfields beyond 'a' for many of these--the 100 needs 'b' and 'c' and 'd' to be useful. Similarly 111 would ideally need 'c', 'd', 'n' . But maybe I'm not understanding the question or how this all works.
Spencer
Thanks for the replyies Irma and Spencer, Exactly the good advice i was hoping for :) Aaah, the subtle nuances of marc... So, I've updated the indexes with both of your suggestions. I'm sure in the next few weeks they can be refined, but its now at nice and usable state. 'title' => '130a,210a,222a,240a,243a,245a,245b,246a,246b,247a,247b, 250a,250b,440a,830a', 'author' => '100a,110a,111a,245c,700a,710a,711a,800a,810a,811a', 'isbn' => '020a', 'issn' => '022a', 'lccn' => '010a', 'biblionumber => '999c', 'itemtype' => '942c', 'publisher' => '260b', 'date' => '260c', 'note' => '500a, 501a,504a,505a,508a,511a,518a,520a,521a,522a,524a, 526a,530a,533a,538a,541a,546a,555a,556a,562a,563a,583a,585a,582a', 'subject' => '600*,610*,611*,630*,650*,651*,653*,654*,655*,662*,690*', 'dewey' => '082', 'bc' => '952p', 'callnum' => '952o', 'host-item' => '952a,952c' Cheers, Mason -- Mason James mason.james@liblime.com Programmer www.liblime.com LibLime +64 21 542 117
Mason James a écrit :
So, I've updated the indexes with both of your suggestions. I'm sure in the next few weeks they can be refined, but its now at nice and usable state.
'host-item' => '952a,952c'
don't hesitate to add 'an' => '6009,6109,6119...' if you want to have the authorities working well too. NoZebra can work with authorities and without this line, one won't be able to jump from biblio to authority or count how many biblios uses an authority. -- Paul POULAIN BibLibre SARL Expert en Logiciels Libres pour l'info-doc Tel : 04 91 31 45 19
On 12/10/07, Mason James <mason.james@liblime.com> wrote:
Hi koha-devel people,
Ive been updating the marc index mapping for koha3-nozebra.
The French team did a nice job on the UNIMARC stuff and ive had a go at padding out the currently pretty minimal MARC21 mapping.
FYI: the *'s are wildcards, so, subject = '600*' means all 600 level subfields are used, when doing a 'subject' search
any obvious errors or suggestions?
You might also want to add 'mc' => '942c' so that a NoZebra adv search on itemtype only does not fail incorrectly. Regards -- Chris Nighswonger Faculty Member Network & Systems Director Foundations Bible College & Seminary www.foundations.edu www.fbcradio.org
participants (5)
-
Chris Nighswonger -
Irma -
Mason James -
Paul POULAIN -
Spencer Anspach