Running noark5-validator/src/importExtraction.php on Debian and MySQL

[Petter Reinholdtsen]

[Ole Aamot]
> I think it may be possible since it appears to me that Thomas Sødring
> has already written PHP code for importing XML in
> noark5-validator/src/importExtraction.php from
> https://raw.githubusercontent.com/KDRS-SA/noark5-validator/master/src/resources/test-uttrekk/uttrekk1/n5uttrekk/arkivuttrekk.xml

Cool.

Adding the ability to importe existing data sets will make nikita useful
for many new use cases. :)

> Can you describe these tables and their purpose in Noark5?

If you browse to
<URL: http://nikita.hioa.no:8092/noark5v4/hateoas-api/arkivstruktur >
(and sakarkiv) using the API browser available from
<URL: http://nikita.hioa.no:3000/ > and the spec available from
<URL: https://samdok.com/leveranser/leveranser-2016/noark5-tjenestegrensesnitt-v-1-0-beta/ >
you should get quite far.

Here is my understanding of the API end point names:

>> > +-------------------------------+
>> > | Tables_in_noark5              |
>> > +-------------------------------+
>> > | basic_record                  | arkivstruktur/basisregistrering
>> > | basic_record_keyword          | arkivstruktur/basisregistrering.noekkelord
>> > | case_file                     | sakarkiv/saksmappe
>> > | classification_system         | metadata/klassifikasjonssystem
>> > | correspondence_part           | arkivstruktur/basisregistrering.korrespondansepart
>> > | document_description          | arkivstruktur/dokumentbeskrivelse
>> > | document_object               | arkivstruktur/dokumentobjekt
>> > | file                          | arkivstruktur/mappe
>> > | file_keyword                  | arkivstruktur/mappe.noekkelord
>> > | fonds                         | arkivstruktur/arkiv
>> > | fonds_creator                 | arkivstruktur/arkivskaper
>> > | keyword                       | metadata/noekkelord
>> > | record                        | arkivstruktur/registrering
>> > | record_correspondence_part    | arkivstruktur/registrering.korrespondansepart
>> > | registry_entry                | sakarkiv/journalpost
>> > | series                        | arkivstruktur/arkivdel
>> > | series_classfication_system   | metadata/klassifikasjonssystem
>> > | series_storage_location       | arkivstruktur/arkivdel.oppbevaringssted
>> > +-------------------------------+

Written from memory, so I might have gotten some of them wrong.

There are some base/subclass relations:

  mappe -> saksmappe
  registrering -> basisregistrering -> journalpost

There are also some parent/child container relationships:

 arkivskaper -> arkiv -> arkivdel -> mappe -> registrering ->
 dokumentbeskrivelse -> dokumentobjekt -> {selve filen}

Note, it is possible to skip some of these steps in the chain.  For
example, I believe this is valid too:

 arkiv -> arkivdel -> registrering -> dokumentbeskrivelse -> dokumentobjekt

ie. dropping arkivskaper and/or mappe.

I hope this make it more clear.

> Is the Nikita Noark5 Core storing the data in the MySQL database, in
> memory (Java) or both?

It is using a database agnostic framework, so it can work with many
databases.  The demo setup is the memory database, but both PostgreSQL
and MySQL should work too.

-- 
Happy hacking
Petter Reinholdtsen