Internationalization : redesign the architecture to handle metadatablocks

[JayanthyChengan]

Have to figure out, how internationalization will be handled for the files under scripts/api/data/metadatablocks, so that new dynamic content, like domain metadata fields, can be added to the application and available in all languages.

[pameyer]

@JayanthyChengan It may make sense to have internationalization handled separately, but I'd be curious to have your input on how it might fit into something like #4451.

[4tikhonov]

I've managed to get all metadata fields translated in French

I can share my solution if the same functionality isn't coming with new release.

[JayanthyChengan]

Hi @4tikhonov : Thanks.
I was able to get the output and I committed the code to
https://github.com/scholarsportal/dataverse/tree/4684-metadatablock-internationalization

Will do PR soon after committing french metadata files.

[4tikhonov]

Hi @JayanthyChengan, thanks, I see, it's good idea to have a separate folder for translated properties. I'm looking forward to see it in the master branch, we'll use it for DataverseEU as well.

[4tikhonov]

Hi @4tikhonov : Thanks.
I was able to get the output and I committed the code to
https://github.com/scholarsportal/dataverse/tree/4684-metadatablock-internationalization

Will do PR soon after committing french metadata files.

@JayanthyChengan, we've introduced ISO codes in DataverseEU for all languages we're supporting, can you please also test this properties with current release?
https://github.com/IQSS/dataverse-docker/tree/master/dataversedock/dataverse-property-files
fr-CA here is coming from Scholarsportal, other translations were done by our community.

Also we modified
https://github.com/IQSS/dataverse-docker/blob/master/dataversedock/lang.templates/metadataFragment.xhtml to get translated properties for metadata fields. I don't know if it will be incorporated in the master branch soon so just in the case.

[scolapasta]

@JayanthyChengan thoughts on removing columns from db? It's not ideal to have duplicate info. (alternatively, the code could expect to use the db if the primary language and only use the bundle otherwis; but that seems like possible overkill in logic)

[4tikhonov]

@scolapasta, the ideal solution for DataverseEU is to use controlled vocabularies as properties delivered as microservice with language parameter. We're connecting Dataverse to ELLST multilingual thesaurus now (13 languages, distributed as RDF) to get similar functionality.

[JayanthyChengan]

@scolapasta - PR #5290 is the latest one with all changes.

[pdurbin]

@JayanthyChengan thanks for all work on this! Shouldn't your pull request include a change to the rst file for http://guides.dataverse.org/en/4.9.4/admin/metadatacustomization.html to explain to creators of custom metadata blocks that they must create not only a TSV file but also a bundle file? Once your pull request is merged, I mean.

[JayanthyChengan]

@pdurbin @scolapasta , updated doc/sphinx-guides/source/admin/metadatacustomization.rst in the PR #5290 , please review it.

[pdurbin]

@JayanthyChengan I didn't make html but the info you added to about to metadatacustomization.rst looks great! I made a small tweak in a561055 to remove some references to this issue. I put this issue back in code review. Thanks!

[JayanthyChengan]

@kcondon please let me know, if I can answer any of your questions. Thanks

[kcondon]

@JayanthyChengan Thanks, I have been occupied with other issues but hope to get to this shortly.

[kcondon]

@JayanthyChengan This looks good and hope to merge shortly but ran into a slight doc issue with unicode char when building pdf format. I'm checking with the team to see whether this is something we need to worry about.

For what it's worth, here is the doc build error for pdf. html works fine:

! Package inputenc Error: Unicode char ṭ (U+1E6D)
(inputenc) not set up for use with LaTeX.

See the inputenc package documentation for explanation.
Type H for immediate help.
...

l.3758 ...age.marathi_(marathi)=Marathi (Marāṭ
hī)
?
! Emergency stop.
...

l.3758 ...age.marathi_(marathi)=Marathi (Marāṭ

[JayanthyChengan]

@kcondon replaced with unicode character in this commit
86e5b7c

does this help in pdf generation error?

[kcondon]

@JayanthyChengan Yes, that did it. Thanks! Merged.