Closed james-strauss-uwa closed 1 year ago
I know most of these graphs are used for various tests, perhaps we could work out how to programmatically generate these cases instead, so that they are correct by default, rather than maintaining them manually which could definitely be tedious.
I like this idea. We could use the JSON schema to (partially?) generate the JSON. Maybe using something like this: https://github.com/ghandic/jsf
I'll write up a ticket and assign to myself, hopefully I can get to this in the next week or so.
Sorry, I've already added a ticket: https://icrar.atlassian.net/browse/LIU-325
I assigned it to myself, but feel free to take it if you have time. I won't tackle it immediately
I just noticed this out-of-date PR and synced it with master.
@awicenec Could you take a look and see if this is still useful, or if it is even required?
Will merge and close this once PR220 has been accepted
This has now been merged through PR liu-338 into master, will thus close this one.
Added a CI task to validate logical graphs in the repository against the logical graph JSON schema. The CI task fails if any graphs are invalid.
Any feedback would be appreciated.
The consequence of this is that we'd have to keep all the graphs up-to-date whenever we change the schema. It feels nice to have all the graphs in the most current format, but is it necessary?
I added tools/checkGraph.py to do the JSON validation of each file. But we may be able to replace this with a command line JSON validator (pip install json-spec ?). In that case we'd just have to recognise and skip the physical graphs, or validate them separately using the appropriate schema.