spark
spark copied to clipboard
$everything is not implemented
I think that's the same reason as #236
It's actually much worse than not implemented.
Background: using the provided docker images (sparkfhir/spark:r4-latest)
{
"resourceType": "CapabilityStatement",
"version": "1.5",
"name": "Spark",
"experimental": true,
"date": "2021-01-20",
"publisher": "Kufu",
"description": "This FHIR SERVER is a reference Implementation server built in C# on HL7.Fhir.Core (nuget) by Furore and others",
"kind": "capability",
"fhirVersion": "4.0.0",
"format": [
"xml",
"json"
]
...
}
The capability statement does include the $everything operation. If it's not supported, it should not be there:
{
"name": "Fetch Patient Record",
"definition": "http://localhost:5555/fhir/OperationDefinition/Patient-everything"
}
The result of $everything operation is incorrect in few aspects. When calling curl http://localhost:5555/fhir/Patient/ID/\$everything:
- It doesn't matter if Patient with
IDexists - The returned results include Patients which ids start with the regular expression
^ID - The
IDis not escaped for regular expression and thereforecurl http://localhost:5555/fhir/Patient/./\$everythingreturns all the patients - The result has no other resources besides unrelated patients