Reordered stylesheets and addressed #281

[richardofsussex]

Put templates in tei2rdf.xsl and crm.xsl into order by element name and
CIDOC-CRM class/property respectively.

Addressed issue #281 (initial part).

If others are happy with this approach, I will look at extending the coverage, particularly of properties within entities.

[peterstadler]

Thanks for putting in so much effort!
But may I kindly ask that you try not to reorder (or reindent) code blocks because it makes it so hard to spot the relevant changes. I see that one of your issues is about reordering, so please don't mix it (in one commit) with other changes.

Second, on line 53 of rdf/crm.xsl you are skipping the further processing in mode rdf2. Could you elaborate a little bit on the rational (or am I missing something)?

[richardofsussex]

OK, so what would you like me to do, now that we are where we are?  Undo my reorderings and just present a solution to #281? The 'further processing' may have been helpful with the original design, but once one entity (a place) is declared as a property of another (an organisation), the post-processing completely screws things up by removing the place from its correct contextual position and putting it at the top level within the RDF.  It took me the best part of a day to work out what was going on - I assumed it was my lack of familiarity with XSLT 2.0! Richard

…

On 08/12/2017 09:09, Peter Stadler wrote: Thanks for putting in so much effort! But may I kindly ask that you try not to reorder (or reindent) code blocks because it makes it so hard to spot the relevant changes. I see that one of your issues is about reordering, so please don't mix it (in one commit) with other changes. Second, on line 53 of |rdf/crm.xsl| you are skipping the further processing in mode |rdf2|. Could you elaborate a little bit on the rational (or am I missing something)? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#301 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ACGGbnDAB9ScyL-XC-TisxToRV72XSjmks5s-PzFgaJpZM4Q5p96>.

-- *Richard Light*

[peterstadler]

sorry for the delay …

Admittedly, I'm not so adept with the CIDOC CRM that I could judge the changes appropriately so I hope others will chime in who's CIDOC-foo is better. Just a few notes:

Within the Stylesheets repo we have some test files and their expected results (after processing with the Stylesheets) to prevent regressions. In this case, it's Test/test12.xml which is transformed by profiles/default/rdf/to.xsl. The expected result file is expected-results/test12.rdf. (You can run the whole tests by entering make test from the root directory of the repo but will probably need to install some dependencies.)

After testing your changes and comparing with the expected results I see that residences like

<P74_has_current_or_former_residence rdf:resource="http://www.example.com/place/copenhagen http://www.example.com/place/denmark"/>

are changed into

<P74_has_current_or_former_residence>
   <E53_Place rdf:about="http://www.example.com/place/copenhagen">
      <P2_has_type rdf:resource="http://www.tei-c.org/type/place/settlement"/>
      <P87_is_identified_by>
         <E48_Place_Name rdf:about="http://www.example.com/placename/copenhagen">
            <rdf:value>Copenhagen</rdf:value>
         </E48_Place_Name>
      </P87_is_identified_by>
   </E53_Place>
   <E53_Place rdf:about="http://www.example.com/place/denmark">
      <P2_has_type rdf:resource="http://www.tei-c.org/type/place/country"/>
      <P87_is_identified_by>
         <E48_Place_Name rdf:about="http://www.example.com/placename/denmark">
            <rdf:value>Denmark</rdf:value>
         </E48_Place_Name>
      </P87_is_identified_by>
   </E53_Place>
</P74_has_current_or_former_residence>

I don't know whether this is syntactically nor semantically valid? Help wanted here …

Second, the test file is missing an <org>, so I added the one from #281 and got it transformed to

<E74_Group xmlns="http://purl.org/NET/crm-owl#"
           rdf:about="http://clas-lgpn2.classics.ox.ac.uk/id//org/c06.a40b0b7b-ce3b-45eb-82f9-006bb7deb712">
   <P131_is_identified_by>
      <E82_Actor_Appellation rdf:about="http://clas-lgpn2.classics.ox.ac.uk/id//orgname/d4e449">
         <rdf:value>National Library of Greece</rdf:value>
      </E82_Actor_Appellation>
   </P131_is_identified_by>
   <P76_has_contact_point>
      <p>before E53</p>
      <E53_Place rdf:about="http://clas-lgpn2.classics.ox.ac.uk/id//place/c06.55309772-00d6-43e1-9744-daffc24bf1d8">
         <P87_is_identified_by>
            <E47_Place_Spatial_Coordinates rdf:about="http://clas-lgpn2.classics.ox.ac.uk/id//placecoords/c06.4b819b9f-32d3-4744-a9d3-4fc45bb6f2e9">
               <rdf:value>23.733249999999998 37.98116666666667</rdf:value>
            </E47_Place_Spatial_Coordinates>
         </P87_is_identified_by>
         <P87_is_identified_by>
            <E48_Place_Name rdf:about="http://clas-lgpn2.classics.ox.ac.uk/id//placename/c06.55309772-00d6-43e1-9744-daffc24bf1d8">
               <rdf:value>Athens Greece</rdf:value>
            </E48_Place_Name>
         </P87_is_identified_by>
      </E53_Place>
      <p>after E53</p>
   </P76_has_contact_point>
</E74_Group>

Here, you are turning the <place> into a <P76_has_contact_point> rather than a <P74_has_current_or_former_residence> as proposed by #281? Additionally, are these <p>before E53</p> and <p>after E53</p> really necessary?
Help wanted here, as well, for I do not know what the desired output should look like …

[hcayless]

Council has reviewed this. @scstanley7 and I don't think that P76_has_contact_point is a safe assumption for org/place, nor that it's range should be a place. Otherwise, this is probably ok.

[sabineseifert]

North-American subgroup at VF2F on April 13:
Action on @ebeshero: reach out to Ontologies SIG and request their advice as this is ancient PR