[Koha-devel] RE: [oss4lib-discuss] MARC -> RDBMS

[Pat Eyler]

On Wed, 11 Dec 2002, Brian Cassidy wrote:

> Hello again,
>
> Thanks to everyone for their quick replies! Here's what I've done since
> yesterday:
>
> I looked into the 'Koha' library software. It looks interesting, but the
> line "It is currently possible to import MARC records, one at a time."
> turned me off of it, due to the fact that I have several repositories
> with thousands of records each.

It sounds like we need to do a better job of documenting something then.
THere are a couple of flavors of bulk-importing scripts at this point.
Other flavors can be derived from the existing scripts at need.  (It's all
a smop, right?)

-pate


>
> I looked into Michael Doran's concerns, but the powers that be want me
> to steer away from a Z39.50 interface. So, that's as far as I got with
> that. :/
>
> After looking at the document
> (http://www.openisis.org/openisis/doc/RdbConv) provided by Ferran Jorba,
> I created a simple DB structure (pardon my ASCII :)
>
> .---------.        .-----------.
> | item    |        | marc_data |
> -----------  ----  -------------
> | item_id |        | item_id   |
> -----------        | data      |
>      |             -------------
>      |
>     /|\
> .-------------.
> | tag_data    |
> ---------------
> | item_id     |
> | tag_order   |
> | tag         |
> | indicator_1 |
> | indicator_2 |
> | subfield    |
> | data        |
> ---------------
>
> Note: marc_data.data is just a 'blob' of the MARC record.
>
> I then used MARC::Record (Hi Ed.) to quickly write a couple perl scripts
> to generate some SQL and run the queries. My database now has 2002
> (MARC) records in a somewhat useable structure. For instance, I can now
> say
>
> SELECT
>       item_id
> FROM
>       tag_data
> WHERE
>       tag = '008' AND
>       data LIKE '%fre__'
>
> to get all of the IDs of the French records.
>
> I'm now at the point where I need to create a data crosswalk (as Art
> mentioned) so I can display the data in a human readable way, as well as
> facilitate easy searching. The Dublin Core crosswalk
> (http://www.loc.gov/marc/marc2dc.html) on the loc.gov website (hat tip:
> Art) seems like a great standard to follow. The problem that I'm now
> facing is that frequently, data in a tag is split up into many
> subfields. For instance:
>
> SELECT
>       item_id,
>       subfield,
>       data
> FROM
>       ust_tag_data
> WHERE
>       tag = '245' -- Title
> ORDER BY
>       item_id
>
> One of my results is:
>
> item_id | subfield | data
> -------------------------------------------------------------
> ...
> 2       | a        | Keeping Alberta competitive :
> 2       | b        | success through workplace learning. --
> ...
>
> Now, it's totally doable to loop through the results (in perl, ASP, etc)
> and concatenate the data together. This COULD, however, get tedious
> depending on the number of fields I need to look at. This is probably
> the wrong forum to ask, but would anyone have any suggestions on how to
> proceed (or perhaps takes some steps back and attack it differently)?
>
> Thanks again for all your help!
>
> -Brian Cassidy (brian at nald dot ca)
>
>
> http://www.gordano.com - Messaging for educators.
>
>
> -------------------------------------------------------
> This sf.net email is sponsored by:
> With Great Power, Comes Great Responsibility
> Learn to use your power at OSDN's High Performance Computing Channel
> http://hpc.devchannel.org/
> _______________________________________________
>
> address@hidden
>
> https://lists.sourceforge.net/lists/listinfo/oss4lib-discuss
> see also http://www.oss4lib.org/
>