Stream: bulk data
Topic: includes vs. match in bulk data
Lloyd McKenzie (May 27 2020 at 22:44):
Bulk Data essentially does away with the notion of the search response Bundle in favor of organizing results into a set of resource-specific files. When this happens, we lose the distinction of which data is 'matching' data and which data is simply 'included' data. That's not an issue when the query is formulated such that the resource types for matching resources (the focal resource type) and included resources is disjoint. However, I were - for example - to query on Encounter and ask for the 'part-of' Encounters to be included, I'd have no way of knowing when reading through my Encounter file which Encounters matched my criteria and which were includes. (And if I wrote my query using _has or something like that, I might not even be able to evaluate by looking at the data which were matches and which weren't.)
Is this an issue? Is there value in adding in the 'output' object in the status response an additional element that corresponds to Bundle.search.mode? with the expectation that you'd have separate files per type per mode?
Last updated: Apr 12 2022 at 19:14 UTC
