When parsing JSON documents with duplicate fields, e.g.
during ingestion Analytics currently rejects such documents and reports that the documents cannot be parsed. In other parts of the platform such objects are accepted, either because
- duplicate fields are removed (e.g. in the query service) or because
- there is no validation of such JSON objects.
Instead of rejecting such documents, we could consider accepting them (and possibly "correcting" them by removing additional fields) and report a warning that describes the modification.