Re: [Koha-devel] Using MARC::Record to insert MARC data into MySQL tables

[paul POULAIN]

Larry Currie wrote:

> I hoping that within a couple of days we will have the item data for our
> collections loaded into the 945, 946, and 949 fields of our MARC
> records, and then we would like to run a Perl script which will go
> through each of our MARC records and extract and insert into the Koha
> MySQL tables biblio, biblioitems, and items all of the data needed to
> run version 1.9.0.  Has anyone developed such a script at this point or
> give me an idea of what the commands would look like to accomplish this?
>
> Larry Currie
>  
C4/Biblio.pm is your friend in this matter :

=item (oldbibnum,$oldbibitemnum) = 
NEWnewbibilio($dbh,$MARCRecord,$oldbiblio,$oldbiblioitem);
creates a new biblio from a MARC::Record. The 3rd and 4th parameter are 
hashes and may be ignored. If only 2 params are passed to the sub, the 
old-db hashes
are builded from the MARC::Record. If they are passed, they are used.

=item NEWnewitem($dbh, $record,$bibid);
adds an item in the db.

Some explanations :
* load your MARC data with a MARC::Record
* open a db handler
* call NEWnewbiblio($dbh,$MARCrecord) to add a biblio.

IF you plan to load biblio AND item in the same MARC Record :
* remember that all item fields MUST be in the same tag (995 proposed)
* split the MARC record in 2 parts : one with biblio and one with item
* call NEWnewbiblio and NEWnewitem

hope it will work fine.
If you write such a script, please submit it in cvs, it should be useful 
for 2.0 version (migrating from another ILS)

Comments from Biblio.pm
========================================================================
move from 1.2 to 1.4 version :
1.2 and previous version uses a specific API to manage biblios. This API 
uses old-DB style parameters.
In the 1.4 version, we want to do 2 differents things :
- keep populating the old-DB, that has a LOT less datas than MARC
- populate the MARC-DB
To populate the DBs we have 2 differents sources :
- the standard acquisition system (through book sellers), that does'nt 
use MARC data
- the MARC acquisition system, that uses MARC data.

Thus, we have 2 differents cases :
- with the standard acquisition system, we have non MARC data and want 
to populate old-DB and MARC-DB, knowing it's an incomplete MARC-record
- with the MARC acquisition system, we have MARC datas, and want to 
loose nothing in MARC-DB. So, we can't store datas in old-DB, then copy 
in MARC-DB. we MUST have an API for true MARC data, that populate 
MARC-DB then old-DB

That's why we need 4 subs :
all I<subs beginning by MARC> manage only MARC tables. They manage 
MARC-DB with MARC::Record parameters
all I<subs beginning by OLD> manage only OLD-DB tables. They manage 
old-DB with old-DB parameters
all I<subs beginning by NEW> manage both OLD-DB and MARC tables. They 
use MARC::Record as parameters. it's the API that MUST be used in MARC 
acquisition system
all I<subs beginning by seomething else> are the old-style API. They use 
old-DB as parameter, then call internally the OLD and MARC subs.

- NEW and old-style API should be used in koha to manage biblio
- MARCsubs are divided in 2 parts :
* some of them manage MARC parameters. They are heavily used in koha.
* some of them manage MARC biblio : they are mostly used by NEW and 
old-style subs.
- OLD are used internally only

all subs requires/use $dbh as 1st parameter.

--
Paul POULAIN
Consultant indépendant en logiciels libres
responsable francophone de koha (SIGB libre http://www.koha-fr.org)