Create widoco mockup

[kara-mela]

Motivation

The possbilities to customize the Ontology Specification created with widoco are very limited. However, it would be great to adapt the style of the generated page (colors, font, ...), but also the displayed content (adding links to PaN Finder and PaN Training to each term).

@paulmillar investigated the possibility to contribute to the widoco repo, but this seems to be very complicated.

Alternatively, we can recreate the page using jekyll pages based on the PaNET output: PaNET_reasoned.owl. The initial effort to generate such a page is relatively high as many information are listed implicitly (only SubclassOf explicitly given, but not SubclassFrom, etc). However, we can customize every aspect of the page ourselves.

Modification

A mockup of the widoco page has been created and is publicly available in parallel to the widoco page; "Browse PaNET (new)" and "Browse PaNET (old)".

Result

Currently, the many of the main features from widoco are already available, however some are still missing (indicated by comments on the page). Especially, the visual webVOWL representation is missing.

Related Issues

contributes to #399

[eilmiv]

Looks good. This nerd sniped me to look into how the json-ld could be formatted more nicely so that the templates don't need to run these queries using for loops. A modified owl2json.py could look like this.

I also tried to change the templates to use this alternative json-ld structure using AI. This worked and approximately halved all the templates in size while being much faster to generate and still having the same result (from what I can tell). I added a pull request with the changes so that it is clear how this could look.

[eilmiv]

I know this file existed in an earlier version, but now it no longer does in the _data directory. Maybe remove the PaNET.json and PaNET_metadata.json generation? Or run these conversions conditionally?

[kara-mela]

You are absolutely right. I do not use the unreasoned owl anymore. And I also figured out that the author information etc from PaNET_metadata are also included in the owl files. So only the conversion of PaNET_reasoned.owl is needed.

[eilmiv]

This works when having an actual newline in instead of \n here.

Suggested change

	{% assign output = include.text | replace: "\n", "<br><br>" %}
	{% comment %}
	The assign command is in principal correct; I tested it with other strings.
	However, this replace does not work. why?
	{% endcomment %}
	{% assign output = include.text | replace: '
	', '<br>' %}

[eilmiv]

A simple approach for styling the other things in the abstract would be like this:

{%- assign output = output | replace: 'https://expands.eu/', '<a href="https://expands.eu/">https://expands.eu/</a>' -%}
{%- assign output = output | replace: 'https://github.com/ExPaNDS-eu/ExPaNDS-experimental-techniques-ontology', '<a href="https://github.com/ExPaNDS-eu/ExPaNDS-experimental-techniques-ontology">https://github.com/ExPaNDS-eu/ExPaNDS-experimental-techniques-ontology</a>' -%}
{%- assign output = output | replace: 'https://doi.org/10.1107/S1600577525005272', '<a href="https://doi.org/10.1107/S1600577525005272">https://doi.org/10.1107/S1600577525005272</a>' -%}
{%- assign output = output | replace: 'https://doi.org/10.5281/zenodo.4806026', '<a href="https://doi.org/10.5281/zenodo.4806026">https://doi.org/10.5281/zenodo.4806026</a>' -%}

{%- assign output = output | replace: '<br>Purpose<br>', '<br><strong>Purpose</strong><br>' -%}
{%- assign output = output | replace: '<br>Design philosophy<br>', '<br><strong>Design philosophy</strong><br>' -%}

[kara-mela]

Wow! Looks great!

[kara-mela]

@eilmiv Thanks for your suggestions. I updated the code following your more simple comments. Maybe the detailed workover you suggested should be part of a dedicated merge request?

[eilmiv]

@eilmiv Thanks for your suggestions. I updated the code following your more simple comments. Maybe the detailed workover you suggested should be part of a dedicated merge request?

Yes, if you want to go in that direction.

[paulmillar]

I don't think it makes sense to commit docs/_data/PaNET_reasoned.json and PaNET_reasoned.owl into git. These should be (auto-)generated.