You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The CQL engine currently supports certain types of query restriction pushdowns, specifically for date filters. In other words, given the following CQL:
[ Condition ] C where C.onset starts 30 days or less before Today()
the CQL compiler can emit an ELM Retrieve that has that date filter captured in the datePath and dateLow elements.
Ideally, we could support the same type of push-downs for Retrieves by id. In other words, these two statements ought to be equivalent with respect to the ELM that's generated:
[Medication] M where M.id = `123`
[Medication: id in {'123'}]
And the ELM that's generated for the Retrieve ought to have both the id and the idProperty populated in both cases.
Acceptance Criteria:
Compiler has Id pushdown optimization done
Tests for same are implemented
The text was updated successfully, but these errors were encountered:
Note that it's the DataRequirementsProcessor that does these types of optimizations as a post-processing step. There is some date range pushdown in the translator directly, but it's not as sophisticated as what the DataRequirementsProcessor is doing, and should be deprecated in favor of the DataRequirementsProcessor, IMHO.
The CQL engine currently supports certain types of query restriction pushdowns, specifically for date filters. In other words, given the following CQL:
the CQL compiler can emit an ELM Retrieve that has that date filter captured in the
datePath
anddateLow
elements.Ideally, we could support the same type of push-downs for
Retrieve
s by id. In other words, these two statements ought to be equivalent with respect to the ELM that's generated:And the ELM that's generated for the
Retrieve
ought to have both theid
and theidProperty
populated in both cases.Acceptance Criteria:
The text was updated successfully, but these errors were encountered: