Bashir Sadjad
Bashir Sadjad
Currently we require the user to tell us what FHIR resource types need to be fetched/monitored. We should add a feature for fetching "everything" or at least every type in...
We will keep using this for pure unit-test related PRs until we reach 80% coverage; then we will make sure we don't go below it.
It seems to have support for Parquet files too. The code is here and the license is Apache 2.0: https://github.com/IBM/FHIR/tree/master/fhir-bulkimportexport-webapp
When we implemented the support for `ViewDefinition` (#821 and #916) we added relational tables as the only supported output (#822). We should extend this functionality and add an option for...
Our current flags/options handling is a mess; some come from config files, some from command line (which uses two different ways). We should consolidate all of these and probably add...
I have been trying to debug an e2e failure that I cannot explain (context: PR #254) hence tried steps described [here](https://github.com/GoogleCloudPlatform/openmrs-fhir-analytics/blob/master/doc/developers.md#running-end-to-end-tests-locally) to run the tests locally. Running the cloud-build tool...
There are at least two aspects of the build that might be improved by using Gradle: - Speed, in particular incremental build speed is supposed to be much faster with...
Flink is the recommended runner for production grade local runs. So at least one of the e2e scenarios should use Flink. This is to make sure that if anything breaks...
This is particularly important when running e2e tests when thousands of Spark log lines can make it hard to find real issues. In general, we should not show these logs...
As part of our performance analysis with HAPI as the source, we need to load synthetic data into HAPI. To make future tests easier, it would be nice to create...