Manifest generation?

[pdiveris]

Hi,

I am sorry if this sounds like an idiotic question - I am struggling to get my head round this whole IIIF.

I've installed the jar with TurboJPEG etc and I get nice images with zoom as below

The JPEGs are on the file system and my config is like this:

- pattern: ^(.*)$
  substitutions:
    - 'file:/home/pete/17n/$1.jpg'

- pattern: ^manifest/(.*)$
  substitutions:
    - 'file:/home/pete/17n/$1.json

• How do I get manifests generated?
• How do I get collections to work?

At the moment both throw a 404!

My intended use is to expose OCRed images and provide search (with SOLR I guess) and highligting (so expect more questions!)

[morpheus-87]

Hi, thank you for your question, at the moment manifest/collection generation is not supported by hymir itself, see #18. That means, you need to already have generated the manifests and collections and define the location in the corresponding config file.

[jbaiter]

Or, as suggested in #18, you can implement your own PresentationRepository and hook it into the app :-)

[pdiveris]

Hi Matthias,

Thank you for your quick and prompt (and honest ;o) response. I've had a look at #18, I will give it a go.

It's been a very long time since I last coded Java, so my approach and attempt should be at least "interesting." Here when I say "interesting" I use to term the way my mum used to describe an embarasing gallery/modern art event.

I am sorry to hijack this ticket for a totally different set of questions - I was drawn into your software in my search for an embedded "book" or "document" viewer, bit a rich one which would be semantic, would allow searches, can support the notion of collections etc. My original visual ideal has been nothing more or less than Google Books and Amazon's Look Inside. In order to implement this it seems to me inevitable that I'll have to struggle with Solr, Lucene etc. This is on top of an existing stack I am using which includes Elasticsearch and ArangoDB. So questions will arise for example about the inevitability of having to use Solr - when in fact ArangoDB should be fine in the fist place - alas, ArangoDB doesn't support NPL (stemming for example) n Greek, nor does it support a plugin model for NPL

I think I am going to start work on it NP, Solr, Highligh and Sdarch API) and create the appropriate tickets, and generally start bothering you but if you;e got anything you might want t point me please do - and have a nice weekend!

Petros

Johannes Baiter wrote on 20/07/2018 14:53:

Or, as suggested in #18 https://github.com/dbmdz/iiif-server-hymir/issues/18, you can implement your own |PresentationRepository| and hook it into the app :-)

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/dbmdz/iiif-server-hymir/issues/19#issuecomment-406607239, or mute the thread https://github.com/notifications/unsubscribe-auth/AA-ZpZxzq8_T-ixP4RRWW73CLx4dt4Ewks5uIeDQgaJpZM4VX3-Z.

-- Petros Diveris

Open Source dev Manchester

h: +44 (0) 0161 249 0536 m: +44 (0) 7771 608 617

[morpheus-87]

Hi Petros,

if I understood right and you would like to have something similar to Google Books, then you should have a look at our Bookshelf :-)

If you wish to use Elasticsearch and ArangoDB as backend, you can plug in custom implementations of these two repositories: https://github.com/dbmdz/iiif-bookshelf-webapp/blob/master/src/main/java/de/digitalcollections/iiif/bookshelf/backend/api/repository/IiifManifestSummaryRepository.java and https://github.com/dbmdz/iiif-bookshelf-webapp/blob/master/src/main/java/de/digitalcollections/iiif/bookshelf/backend/api/repository/IiifManifestSummarySearchRepository.java.

Do you already have a platform running in production?

[pdiveris]

Hi

I only saw your post about Bookshelf now, by accident. This is probably what I need. I do not have a production environment running as yet. I will check Bookshelf and let you know asap. You probably saved me a lot of time, will let you know!