jsheunis
commented
2 days ago With https://github.com/psychoinformatics-de/shacl-vue/commit/043439ee01f8cc52f472937f5d6474fe5d0d3a6c we now have the database graph being loaded on application startup. This is then made available to form editor components (a work in progress). The next step is to allow the database to be written to, and to be exported after edits have been made.
It is conceivable that the application is not only supplied with shacl for generating the user interface, which is then used for metadata entry (as happens currently), but that the app also has to interact with an existing set of triples in a scope wider than the user's browser session. A simple use case would be when a user wants to add the authors for a specific dataset, and the interface might want to present the user with previously entered authors in the same consortium (for example). This would mean that other users have entered other authors for datasets in the same consortium (scope of identification), and the current user might want to select from them, so they should be made available e.g. via some request to an external database.
For this reason it would be good to factor out some functionality that allows reading from and writing to a database of sorts. An example of a specific implementation would be reading the triples from a text file available via http request. The more tricky part would be updates, because it would be sensible to require some form of authentication for that.