- Adding the possibility of explicitly setting a resource type via the (new)
type
property.
- Adding information on
@context
files
- Label is now optional
- Error in the namespace for
Property
(see #16)
- Adding the possibility to define datatypes
- Bug handling...
- Minor improvement on the generated JSON vs language handling (thanks to @pchampin).
- The default value set for
status
was set toreserved
, rather thanstable
.
-
Instead of the (binary)
deprecated
flag this version uses thestatus
value in the yml file whose value can bestable
,reserved
, anddeprecated
. The terms are marked up as such in the final vocabulary, and the HTML output separates these three classes into separate to-level sections.This feature is backward compatible with
deprecated
; older vocabularies that use that flag would still work, but the value ofstatus
has a higher priority. -
Introduction of a
defined_by
term: a URL referring to the formal definition of the term in another document.
- Added more cross-references to the HTML output, to make the end result more useful:
- the class definitions include a further set of information about the properties (in the same vocabulary) that refer to that class either through the range or the domain statements
- the references to classes and properties (e.g., in range or subproperty statements) are now hyperlinks. That is also true if external vocabularies are used with a prefix
- Handled a bug whereby the generated turtle always referred to the same top level vocabulary instead of using the one set in the vocabulary description itself.
- Minor code improvements: separating the schema into its own files, better error handling through
Promise.allSettled
, and also retrofitting some changes required by thedeno
version (which is more strict than the standardtsc
)
- The content of the
comment
field is now considered as an HTML fragment. For most of the vocabularies it does not have any consequence now, because those descriptions/comments are a single paragraph anyway. However, if line feeds or other tricks are used to force some formatting, those should be redone in bona fide HTML. The RDF version of the ontologies use therdf:HTML
data type and therefore keep those structures in the official RDF definition, too.
- Added the (optional) generation of a JSON-LD
@context
file - Added, in the YAML format, the possibility for a
dataset
entry on properties, resulting in the"@container":"@graph"
statement in the generated context
- Better use of the Commander package for the CLI
- Some CSV leftover still persisted in the README.md file...
- Adopted some naming conventions from the airbnb style file.
- Restructured the HTML output: each term in its own subsection instead of into a table (this gives more space for the content)
- Added the possibility to add one or more examples into the YML file
- The YAML content is checked with a built-in JSON Schema before processing further