move linked_data_platforms to directory

[briesenberg07]

• [x] Copy RDF/XML to new location
• [x] Pre-process RDF/XML
  • [x] Increment version number in RDF/XML as revision; dct:hasFormat values changing
  • [x] Remove dct:hasFormat triples
• [x] run main.py
  • [x] change $rdf2rdfa_stylesheet to xsl/rdf2htmlrdfa-plusdc.xsl *
  • [x] check new dct:hasFormat values
  • [x] check triples in new serializations against old **
  • [ ] change $rdf2rdfa_stylesheet back to xsl/rdf2htmlrdfa.xsl *
• [x] Update DataCite XML
  • [x] Remove version number
  • [x] Update per https://github.com/uwlib-cams/uwlswd/issues/51
  • [x] Update per https://github.com/uwlib-cams/uwlswd/issues/50
• [x] Update DOI at DataCite Fabrica
  • [x] Upload XML
  • [x] Update URL
• [x] Confirm DOI dereferences to new location
• [x] Delete data at old location

* See Run main.py to serialize data ** Only checked number of triples in the graph

[briesenberg07]

20230919_main.py_error.txt

@cspayne sharing terminal error output I suspect the use of both / and \ in the filepath to directory for processing?

[briesenberg07]

Oh wait, I'm looking at the following:

https://uwlib-cams.github.io/gitrepos_uwlswd/uwlswd_vocabs/linked_data_platforms\linked_data_platforms.rdf does not look like a valid URI, trying to serialize this will break.

That seems like a URL string which combines some of the path to the target directory in my computer's file structure with a base URL portion?

[cspayne]

Looking at this now!

[cspayne]

@briesenberg07 - will you try replacing line 36 in main.py (where uri_path is set) with:

uri_path = "https://uwlib-cams.github.io/" + file_path_noext.replace("../", "").replace("\\", "/")

@cspayne I'll try it now!

[briesenberg07]

Hmmm, stymied! I did two things:

• Per L35 comment "assumes top-level directory for file is parallel to uwlswd directory", moved directories on my computer (see * below)
• Changed L36 as above

So I'm mystified by the new error message...

20230919_main.py_error_02.txt

...where it says:

https://uwlib-cams.github.io/gitrepos_uwlswd/uwlswd_vocabs/linked_data_platforms\linked_data_platforms.ttl does not look like a valid URI, trying to serialize this will break.

I don't know where gitrepos_uwlswd/ is coming from at this point, as that directory no longer exists on my machine...

---

* BMR directory structure now/following change

gitrepos/
   uwlswd/
   uwlswd_vocabs/

[cspayne]

@briesenberg07 Did this uri potentially end up in the rdf/xml during your first serialization attempt?

[briesenberg07]

Thanks for the above Cypress--this had indeed happened. I ran the script successfully! The expected files were output in the expected place!

But checking dct:hasFormat values:

Input (excerpt):

...
    <void:feature rdf:resource="http://www.w3.org/ns/formats/Turtle"/>
    <dct:hasFormat rdf:resource="https://uwlib-cams.github.io/uwlswd_vocabs/linked_data_platforms.nt"/>
    <dct:hasFormat rdf:resource="https://uwlib-cams.github.io/uwlswd_vocabs/linked_data_platforms.ttl"/>
    <dct:hasFormat rdf:resource="https://uwlib-cams.github.io/uwlswd_vocabs/linked_data_platforms.jsonld"/>
    <dct:hasFormat rdf:resource="https://uwlib-cams.github.io/uwlswd_vocabs/linked_data_platforms.html"/>
    <ldproc:containsEntity rdf:resource="https://doi.org/10.6069/uwlib.55.b.1#Platform"/>
...

Output (excerpt):

...
    <void:feature rdf:resource="http://www.w3.org/ns/formats/Turtle"/>
    <dcterms:hasFormat rdf:resource="https://uwlib-cams.github.io/uwlswd_vocabs/linked_data_platforms.nt"/>
    <dcterms:hasFormat rdf:resource="https://uwlib-cams.github.io/uwlswd_vocabs/linked_data_platforms.ttl"/>
    <dcterms:hasFormat rdf:resource="https://uwlib-cams.github.io/uwlswd_vocabs/linked_data_platforms.jsonld"/>
    <dcterms:hasFormat rdf:resource="https://uwlib-cams.github.io/uwlswd_vocabs/linked_data_platforms.html"/>
    <dcterms:hasFormat rdf:resource="https://uwlib-cams.github.io/uwlswd_vocabs/linked_data_platforms/linked_data_platforms.ttl"/>
    <dcterms:hasFormat rdf:resource="https://uwlib-cams.github.io/uwlswd_vocabs/linked_data_platforms/linked_data_platforms.nt"/>
    <dcterms:hasFormat rdf:resource="https://uwlib-cams.github.io/uwlswd_vocabs/linked_data_platforms/linked_data_platforms.jsonld"/>
    <dcterms:hasFormat rdf:resource="https://uwlib-cams.github.io/uwlswd_vocabs/linked_data_platforms/linked_data_platforms.html"/>
    <ldproc:containsEntity rdf:resource="https://doi.org/10.6069/uwlib.55.b.1#Platform"/>
...

[cspayne]

Yay!

The repeat dct:hasFormat output is because of the location change - if you remove or update the dct:hasFormat triples before serialization, then the script will add the correct ones.

It is not set to remove any dct:hasFormat triples that are already in the rdf/xml so that in circumstances where a resource does have additional formats, those are not lost. This does mean that when the uri is changing, there is the additional step of removing the old dct:hasFormat triples.

[briesenberg07]

@cspayne did your DOI URL update take effect yet? I checked this vocab this morning and DOI dereferenced to new URL (after making DataCite Fabrica updates on Tuesday). So, it takes... something less than 2 days to take effect?

[cspayne]

@cspayne did your DOI URL update take effect yet? I checked this vocab this morning and DOI dereferenced to new URL (after making DataCite Fabrica updates on Tuesday). So, it takes... something less than 2 days to take effect?

Yes! I had to clear the cache in my browser and since then it's been pretty instantaneous.