For every file in GTFS we currently define which element of the header we use as identification of the line.
For example stop_id is defined as a key for stops.txt.
If, for a file, one of the keys is missing in the header, we just stop the whole diff, i.e. we have no result.
For example if stops.txt has ho stop_id column, we don't produce any diff.
Obviously stops.txt without stop_id is an invalid file, so it would be OK to not have a diff of that file, but we should still get the diff of the other files if they are valid.
This will probably require a change of the diff schema.
Possibly add a missingKey for the status for that file. We hould consider having a details field in the schema to explain the problem (e.g. missing stop_id key in base dataset)
For every file in GTFS we currently define which element of the header we use as identification of the line.
For example
stop_idis defined as a key forstops.txt.If, for a file, one of the keys is missing in the header, we just stop the whole diff, i.e. we have no result.
For example if stops.txt has ho stop_id column, we don't produce any diff.
Obviously stops.txt without stop_id is an invalid file, so it would be OK to not have a diff of that file, but we should still get the diff of the other files if they are valid.
This will probably require a change of the diff schema.
Possibly add a
missingKeyfor the status for that file. We hould consider having adetailsfield in the schema to explain the problem (e.g. missing stop_id key in base dataset)