Import records to Goobi Viewer

Hi!
I was looking on the possibility to import records directly to Goobi viewer without starting from Goobi workflow. Is this possible at the current version? if yes, what are the procedures to achieve this?

Thank you and best regards,
Hemed

1 „Gefällt mir“

Hi @hemed

please see chapter 2.7 of the Goobi viewer Indexer documentation for this question:

Best wishes and have a nice weekend,

Jan :slight_smile:

1 „Gefällt mir“

That’s quite asy. We prepare and import files from several workflows without goobi.workflow:

  • Prepare some METS/MODS, LIDO or TEI files.
  • Put your images in the directory beside
  • Copy them to /MYVIEWERPATH/viewer/hotfolder
  • Wait a few seconds and see what happens. (some files should disappear)

if nothing seems to happen:

  • Login into your viewer and check wether your documents are visible.
  • Check your log-fils
1 „Gefällt mir“

Thank you for that @jan. I saw this documentation before and misunderstood the term “Indexing”. I thought this means only getting the records to the Solr index.

Thanks for the procedures, @bbfks! Do you then have a script that convert the metadata from other forms (e.g Excel) to METS/MODS?

Best regards,
Hemed

1 „Gefällt mir“

Well, you could use Goobi Goobi workflow for this :stuck_out_tongue:

You could use the mass import functionallity using the following plugin:

The workflow could be 100% automatic, so simply uploading the Excel file would start it and do everything else you desire automatically (for example copy images, generate derivatives, OCR, generate identifiers, generate METS/MODS, export…)

Cheers,

Jan :slight_smile:

2 „Gefällt mir“

Just to answer your question: yes, I write my own scripts. But it is necesary, if the metadata is delivered in strange files with no standardized formats aso.

I guess, you should look at Jan’s plugin. That’s the documented method, and if you stuck anywhere you can get help here in this community.

Thanks all for the useful information here. @jan , I have a follow up question regarding the export to viewer from Goobi workflow.

When you export a process to viewer, _media folder is copied to viewer. This means media files copied to viewer are technically duplicates if they are on the same shared disc mount with Goobi workflow and it can be costly if you are thinking about backup. Is there away to tell viewer to read images directly from workflow or instead use workflow API to read images?

No. Until now not. And it is even not in our roadmap (yet). There is a quite simple reason for this, too.

Very often, the workflow system is in an internal network while the viewer system is in a DMZ. Often there are various projects managed within one Goobi workflow installation. This can be a typical digitization project where all images are open access on one hand, but maybe you manage a document delivery service or other content with access restrictions in other projects, too.

Goobi workflow can be configured to only transfer the content that is to be published somehow on the Goobi viewer server and the Goobi viewer can deal with access restrictions.

Goobi workflow does not implement this security mechanisms as it only manages the workflow. In an ideal world, you export the master images to a long term preservation system in the end and only stay with small derivatives for correction purposes etc.

By the way: Goobi workflow can ingest into a Fedora repository and the Goobi viewer can read the image data from there. It is just a matter of writing the correct URLs into the METS fileGrps…

1 „Gefällt mir“

Thank you for the clarification, @jan !