What's Changed

• Move testhelpers out of internal. by @copybara-service in #184

Full Changelog: v0.1.6...v0.1.7

Notable Changes:

• Go module and repository name updates to github.com/google/bulk_fhir_tools
• Added orchestration workflows and readme
• BCDA V1 API support removed
• Improved error handling

What's Changed

• Replace all manually written GCS server implementations with the canonical implementation by @copybara-service in #86
• Improve error handling when polling for job status by @copybara-service in #89
• Replace some BCDA references with more generalized terms. by @copybara-service in #91
• Create a fake cloud logging server for testing. by @copybara-service in #103
• Define a common metric interface and add an OpenCensus implementation. by @copybara-service in #112
• Support a slightly broader set of X-Progress header values. According to the bulk fhir specification [1] the format is at the server's discretion so this will support pulling out a percentage number if it is mentioned anywhere in the header value string. by @copybara-service in #115
• Fix test_server_test to work on windows by @copybara-service in #116
• Move metric results into its own file. by @copybara-service in #117
• Upgrade golang.org/x/text by @suyashkumar in #118
• Export types and fields in the OAuth token exchange code so that other implementations can be created outside the package by @copybara-service in #119
• Split metrics_test into counter_test and latency_test. by @copybara-service in #123
• Only log metric results of local implementation. by @copybara-service in #126
• Add fhir-resource-counter, process-url-time, operation-outcome-counter and fhir-rectify-counter metrics to bulk fhir fetch. by @copybara-service in #127
• Add small warning about defaulting to Patient endpoint when no Group ID is set. by @copybara-service in #133
• Include link to the logs in Pantheon if using GCP logging. by @copybara-service in #137
• Remove BCDA V1 support from code. by @copybara-service in #139
• Add logs and monitoring documentation. by @copybara-service in #147
• Add orchestration README with instructions for setup. by @copybara-service in #151
• Create helper functions to download Synthea FHIR and convert it to ndjson. by @copybara-service in #153
• Add support for different aggregation types in Counter metric. by @copybara-service in #155
• Add documentation on the different FHIR Store upload options. by @copybara-service in #163
• Fix logged duration of GCS Import. by @copybara-service in #173
• Improve error handling for FHIR Store batch upload option. by @copybara-service in #174

Full Changelog: v0.1.5...v0.1.6

(Notable changes description coming soon).

Full Changelog: v0.1.4...v0.1.5

Notable Changes

• Support for importing FHIR NDJSONs to GCS and triggering GCS-based import to FHIR store (this can be nicer with quotas).
• Support for various under-the-hood generalization updates to work with importing data from arbitrary bulk fhir servers, beyond BCDA.
• Support for batched updates to FHIR store via the executeBundle API.

Full Changelog: v0.1.3...v0.1.4

Notable changes

• Ability to write out resources with upload errors to disk for downstream handling #23
• Early support for uploading to FHIR store in batches
• Support for testing on multiple operating systems
• Addition of an example BigQuery notebook showing how to use user defined functions on FHIR data. #29
• Support for retrying 404 errors which BCDA has produced transiently recently #32
• internal/testhelpers additions and updates to improve test infrastructure.
• various internal updates and changes.

Full Changelog: v0.1.2...v0.1.3

• 6 hours timeout when waiting for BCDA to finish (we will make this configurable soon)
• The ability to provide a -bcdaJobID=1234 flag to bcda_fetch, which will download the data from the provided job ID if possible.
• If you set -outputPrefix="" (or just don't set it), then bcda_fetch will upload to FHIR store without writing any of the NDJSONs out to disk.

Full Changelog: v0.1.1...v0.1.2

In this release, the primary functional changes are:

• #13: GetData and reauthorization are tried multiple times if BCDA returns unauthorized. This is because in practice we have started to see that these calls take a couple of retries on the BCDA server to proceed without the unauthorized error.
• #12: An internal update to how the concurrent counters work to prevent the possibility of a negative WaitGroup counter.

Full Changelog: v0.1.0...v0.1.1

This is the initial release for medical_claims_tools. This is still an early stage for this project and we are kicking the tires, but core functionality to instrument BCDA imports should be there. Feedback is welcome.