Smarter handling of null objects in FLAT format

[freshehr]

Background

When integrating data from an external source, an error is generated if a mandatory attribute in the target object is not completed , even if all of the other attributes are empty and the object is optional

For example if the target json is

    "generic_laboratory_report/laboratory_test_result:0/specimen:0/specimen_type|value": "A value",
    "generic_laboratory_report/laboratory_test_result:0/specimen:0/specimen_type|code": "some_code",
    "generic_laboratory_report/laboratory_test_result:0/specimen:0/specimen_type|terminology": "some_terminology",

and the incoming specimen_type datapoint is null or absent....

In the Better CDR if you simply make each of the |value, |code and |terminology attributes an empty string, the whole row will be ignored by the CDR, which is correct, as the Specimen type element is optional.

EhrBase is a more complicated since it enforces the openEHR constraint which says that a CodedText object must have all 3 attributes populated, and ignores the fact that the overall element is itself optional.

Enhancement

Detect if an imported datatype object is completely null, after population with empty strings or null (for numerics), do not import that element and only report an exception if the parent element/attribute is mandatory.

Discussion

No response

Further information

No response