As a background for those unfamiliar with the need arisen in arXiv is the need
to bundle together several elements of a MARCXML element that's like a history
of the document put in MARC. there will be two fields that should be looked
at. And without a custom index, there would be a lot of false positives to
parse through.
We store this history in 950 $h, 950 $i, and 950 $j. But only two of them are
used. There will be many 950 elements with $h, $i, and $j for each "row" of
update history. The idea is that if
950 $i = FOO
$j = 03/24/2012
950 $i = BAZ
$j = 03/24/2012
And we do a search for FOO 03/24/2012. It would match both records, and I'd
have to do further parsing to really gett the one I wanted. HOwever, I could
create a compound index that would eliminate the need for all the parsing.
------------------------------------------
Peter Halliday
Cornell University Library IT
Repositories Group
[email protected]<mailto:[email protected]>
(Phone:) 607-255-1790
(Cell:) 607-329-6905
On Mar 26, 2012, at 2:41 PM, Peter J. Halliday wrote:
Any ETA on this ticket?
------------------------------------------
Peter Halliday
Cornell University Library IT
Repositories Group
[email protected]<mailto:[email protected]>
(Phone:) 607-255-1790
(Cell:) 607-329-6905
On Mar 26, 2012, at 2:36 PM, Tibor Simko wrote:
On Mon, 26 Mar 2012, Samuele Kaplun wrote:
this was also proposed since long time
As usual, so many things, so little time...
(it's still on Savannah!)
Now part of <http://invenio-software.org/ticket/852>.
Best regards
--
Tibor Simko
