[Koha-devel] New Koha 2.2 search....

[Ernesto Silva]

Hi,
   I've said before we've developed a different search in Koha 2.2, as 
our Koha installation has many modifications I'm not able to send you
a 2.2 working version with the new features.

Anyway, I can send you our pages, templates and modules so "somebody" 
can make a integration-testing with actual Koha versions and evaluate.

I'm attaching a ZIP file with all our new staff, please feel free to use 
it, test it, and finally comment on it.

I don't have any other files, so if someone wants to give it a try and 
needs any assistance please contact me (address@hidden) or Jean Pierre 
Ducassou (address@hidden) so we can guide you.

Other data:
- all texts, help and comments are in Spanish, please contact us if you 
need any help or assistance.

- we use MySQL 5.0.18, I think that any version with sub-queries support 
should work fine.

- all templates belongs to the Intranet section, when the system is 
fully functional we'll make some for the OPAC.

- There's a help file with an explanation of the grammar used by the 
librarians, the grammar which we convert to MySQL 5 queries/subqueries.

- In the grammar the slowest operation is the negation of something, we 
have operators like "contains" and "nocontains" and that works fine, but 
if you use the "not" operator the queries gets slow. We are trying to 
find an elegant solution, but this will take some days

- We used YAPP to parse the grammar and generate a perl module, the .yp 
file with the grammar source is also included.

Finally, I know this lacks a lot of information, but please, first look 
at it and then contact us.

Best regards,
Ernesto Silva.






********************************************************************
Hi,
   here at Universidad ORT Uruguay we've been using koha since may'05.
Andres Tarallo and I were the primary developers, Andres is still with
us but working on another proyect and we've hired Jean Pierre Ducassou
to continue our Koha developement. And... no Paul, Jean Pierre is not
french, but his ancients were.

Anyway, as our Koha 2.2 installation has it's own flavour and we were
aiming to improve the search system we've made some work during 2006 but
the developer we hired has gone away... pity.

We've been reviewing 2006 search code (with Jean Pierre) wich was based
on MySQL 5 using sub-queries and Full Text Indexing. We've got a 95%
working search system, we're giving it the final tweaks.

We were wondering if anyone is interested in it, we can contribute with
the code, explanations and basic documentation.

The basics of the systems are:
- our own query language, simple but yet powerful.
- a parser to convert the query language expressions to sql expressions.
- a new table, similar to marc_subfield_table, but with a subfieldvalue
field of blob type, and a nsubfieldvalue field of float type.
- full text indexing.

In our developement server (2 x PIII of 800 MHz) the new search system
works very fast with 52000 bibles and 75000 items.

When the system is fully functional (in 2 or 3 days I hope) we will
begin the migration off the rest of our Koha installation to MySQL 5,
and when this is ready we'll put into the production server, I hope we
can achieve this in October.

Sorry about my poor english and best regards,

--
Ernesto Silva.
Coordinador de Desarrollo Web y Sistemas Abiertos
Centro de Procesamiento de Datos
Universidad ORT Uruguay.
E-mail: address@hidden
Tel: (+598-2) 902-1505 ext. 206
Fax: (+598-2) 900-2952



_______________________________________________
Koha-devel mailing list
address@hidden
http://lists.nongnu.org/mailman/listinfo/koha-devel

ORT_KohaSearch.zip
Description: Zip archive