Requested feature (biblio title order)
Hi, In 2.2, when you order a search by title, stopwords at the beginning are still used in the ordering. "A simple software" will be ordered before "The last minute". In France (& I think in other countries it's the same), we are supposed to have "A" and "The" ignored at the beginning of the title having "(A) simple software" after "(the) last minute". To add this feature in Koha we could : * add a field in biblio.title. Call it biblio.titlenostopword (anybody a suggestion for a better name ;-) ?) * in Biblio.pm, in OLDnewbiblio / OLDmodbiblio => put the title with stopwords removed in this new field * don't order on biblio.title but on the new field. Quite simple to do, a day of work not more. Do we think it's an important feature that could be in 2.2.x or not ? I vote yes. the risk of unstability is really low, for a nice feature. -- Paul POULAIN Consultant indépendant en logiciels libres responsable francophone de koha (SIGB libre http://www.koha-fr.org)
Paul POULAIN said:
In 2.2, when you order a search by title, stopwords at the beginning are still used in the ordering. <snip> Do we think it's an important feature that could be in 2.2.x or not ?
I think this is a _very_ important feature to add. Months ago, we made a quick (and crude) change to the 2.0.0 opac software for NPL, so that titles would display in the 'correct' order -- "The Big Bang" and "A Big Bang" display next to each other in a title-order list. This is the expected behavior for an English-language OPAC. We'll probably make the same change to the 2.2.0 opac software before NPL starts using it in production. Stephen -- Stephen Hedges Skemotah Solutions, USA www.skemotah.com -- shedges@skemotah.com
Stephen Hedges a écrit :
Paul POULAIN said:
In 2.2, when you order a search by title, stopwords at the beginning are still used in the ordering. <snip> Do we think it's an important feature that could be in 2.2.x or not ? I think this is a _very_ important feature to add. Months ago, we made a quick (and crude) change to the 2.0.0 opac software for NPL, so that titles would display in the 'correct' order -- "The Big Bang" and "A Big Bang" display next to each other in a title-order list. This is the expected behavior for an English-language OPAC. We'll probably make the same change to the 2.2.0 opac software before NPL starts using it in production. As you want, but it should be in official release in a month or two ;-)
-- Paul POULAIN Consultant indépendant en logiciels libres responsable francophone de koha (SIGB libre http://www.koha-fr.org)
In article <41E78307.8060006@free.fr> you write:
In 2.2, when you order a search by title, stopwords at the beginning are still used in the ordering. "A simple software" will be ordered before "The last minute". In France (& I think in other countries it's the same), we are supposed to have "A" and "The" ignored at the beginning of the title having "(A) simple software" after "(the) last minute".
The "pedantically correct" way to do this would be to encode the number of characters to skip in the second indicator field in the MARC record. In traditional MARC-based systems, this second indicator field is used to control how many characters to skip when sorting. The advantage of this is that sometimes words like "A" and "The" should not be ignored (e.g. if the library has multilingual collections and the title in question happens to be in a language where "the" is a real and very important word). The disadvantage of this is firstly of course Koha does not support indicators, and most likely the semantics of such skipping may need to be more concretely defined (e.g., what to do when the title begins with non-ASCII Unicode characters) before it can be useful, and lastly this is much more work :-( Regards, Ambrose -- Ambrose LI Cheuk-Wing <a.c.li@ieee.org> http://ada.dhs.org/~acli/
Excellent point, Ambrose! And it seems like it shouldn't really be that hard, since that subfield indicator is stored in the marc_subfield_table as the second character in the tag_indicator column. Once you have a list of biblionumbers returned by Search.pm, you could check them against that column to see if any leading characters in the title need to be skipped. This solution has the great advantage of leaving control of the title order in the hands of the catalogers. Stephen Ambrose LI said:
In article <41E78307.8060006@free.fr> you write:
In 2.2, when you order a search by title, stopwords at the beginning are still used in the ordering. "A simple software" will be ordered before "The last minute". In France (& I think in other countries it's the same), we are supposed to have "A" and "The" ignored at the beginning of the title having "(A) simple software" after "(the) last minute".
The "pedantically correct" way to do this would be to encode the number of characters to skip in the second indicator field in the MARC record. In traditional MARC-based systems, this second indicator field is used to control how many characters to skip when sorting.
The advantage of this is that sometimes words like "A" and "The" should not be ignored (e.g. if the library has multilingual collections and the title in question happens to be in a language where "the" is a real and very important word).
The disadvantage of this is firstly of course Koha does not support indicators, and most likely the semantics of such skipping may need to be more concretely defined (e.g., what to do when the title begins with non-ASCII Unicode characters) before it can be useful, and lastly this is much more work :-(
Regards, Ambrose
-- Ambrose LI Cheuk-Wing <a.c.li@ieee.org>
Stephen Hedges a écrit :
Excellent point, Ambrose! And it seems like it shouldn't really be that hard, since that subfield indicator is stored in the marc_subfield_table as the second character in the tag_indicator column. Once you have a list of biblionumbers returned by Search.pm, you could check them against that column to see if any leading characters in the title need to be skipped.
Yes, but it has the default of being MARC=ON specific, and requires a cataloguer that speaks MARC really good. I think i'll try to find a solution that calculates the "title for order" automatically, unless the indicator has been typed by the librarian, in which case Koha will use it.
This solution has the great advantage of leaving control of the title order in the hands of the catalogers.
Stephen
Ambrose LI said:
In article <41E78307.8060006@free.fr> you write:
In 2.2, when you order a search by title, stopwords at the beginning are still used in the ordering. "A simple software" will be ordered before "The last minute". In France (& I think in other countries it's the same), we are supposed to have "A" and "The" ignored at the beginning of the title having "(A) simple software" after "(the) last minute".
The "pedantically correct" way to do this would be to encode the number of characters to skip in the second indicator field in the MARC record. In traditional MARC-based systems, this second indicator field is used to control how many characters to skip when sorting.
The advantage of this is that sometimes words like "A" and "The" should not be ignored (e.g. if the library has multilingual collections and the title in question happens to be in a language where "the" is a real and very important word).
The disadvantage of this is firstly of course Koha does not support indicators, and most likely the semantics of such skipping may need to be more concretely defined (e.g., what to do when the title begins with non-ASCII Unicode characters) before it can be useful, and lastly this is much more work :-(
-- Paul POULAIN Consultant indépendant en logiciels libres responsable francophone de koha (SIGB libre http://www.koha-fr.org)
Hi, Paul - Paul POULAIN said:
Stephen Hedges a écrit :
Excellent point, Ambrose! And it seems like it shouldn't really be that hard, since that subfield indicator is stored in the marc_subfield_table as the second character in the tag_indicator column. Once you have a list of biblionumbers returned by Search.pm, you could check them against that column to see if any leading characters in the title need to be skipped.
Yes, but it has the default of being MARC=ON specific, and requires a cataloguer that speaks MARC really good. I think i'll try to find a solution that calculates the "title for order" automatically, unless the indicator has been typed by the librarian, in which case Koha will use it.
I don't think very many catalogers do original cataloging, when you can get MARC records from vendors, from Z39.50, from OCLC, etc. And those that do their own cataloging are good enough to handle the subfield indicators with no problem. At least that's the case in the US. What you're proposing to do is solve a problem that MARC(21) has already solved. Stephen -- Stephen Hedges Skemotah Solutions, USA www.skemotah.com -- shedges@skemotah.com
participants (3)
-
Ambrose LI -
Paul POULAIN -
Stephen Hedges