* Start work on validation rules interceptor
* Work on interceptor
* Add tests
* Work on validation interceptor
* Work on interceptor
* Test fix
* Documentation tweaks
* 2229 - Remove support for the upload-igpack CLI command.
* 2229 - Revert changes made to the ValidateCommand class since it only takes an igpack as a parameter and therefore, should NOT have been removed.
* 2229 - Remove the igpack option from the Validate command.
* Update hapi-fhir-docs/src/main/resources/ca/uhn/hapi/fhir/changelog/5_3_0/2229-remove-upload-igpack-command.yaml
Co-authored-by: James Agnew <jamesagnew@gmail.com>
Co-authored-by: Kevin Dougan <kevindougan@kevins-mbp.lan>
Co-authored-by: James Agnew <jamesagnew@gmail.com>
* Partitioning rework
* Work on partition improvements
* Partition updates
* Work on partitiong
* Test fixes
* Add docs
* Add changelog
* Resolve FIXME
* Test fixes
* Test fixes
* Test fixes
* Compile fix
* Fix compile error
* Test fix
* Test fixes
* init rev
* fix build
* Tweaked the POM config settings and also added some dependency exclusions (commented out for now).
* More re-factoring of the CQL Unit Tests.
* Removed a LogMessages.html file and a minor Unit Test change.
* Unit Tests.
* added debug logging to troubleshoot the dao that has no name
* added debug logging to troubleshoot the dao that has no name
* workaround to get past null dao resourceName issue
* fix jsons to get test to pass. Test still fails with library id problem
* gitignore
* gitignore
* test passes! Woohoo!
* undo troubleshooting logging
* added timer and logging.
* added asserts and time multiple evaluations and measure the average
* readme
* adding explanations
* added more explanatory notes
* measure 2 patients
* move pom to use cqf snapshot
* roughed out cache
* roughed out cache
* Added code to VersionChangeCache class.
* added tests
* added polling test
* wrote init version
* wrote init version
* optimized versioncache
* worked on getting tests to pass
* redesigned interfaces
* all tests pass
* fixmes
* fixmes
* rename param
* Added Unit Tests.
* javadoc
* Fixed the 2-Patient Unit Test.
* More Unit Test work.
* make ResourceVersionMap immutable
* Fixed a Unit Test that was failing intermittently by adding a new way to refresh the cache.
* Use a new method called doRefreshAllCaches(0) to force a refresh and have all Listeners called immediately.
* Cleaned up IVersionChangeListenerRegistry interface to make methods more clear and resolved all Unit Tests.
* disabled tests
* disabled tests
* removed unused test method
* fixed refresh logic and added asserts
* moved cache so it can be used by searchparamregistry
* Updated the Cql Unit Tests to be properly configured for Dstu3 or R4.
* started rewriting SearchParamRegistryImpl to use new cache
added init method to listener interface
* added fixmes
* adding tests
* tests pass
* added tests
* Fixed the way CqlProviderFactory Autowires Beans so it can work with both Dstu3 and R4 contexts.
* moar tests
* fix test
* work tests
* reverting unneccessary refactors
* undo unneccessary import changes to reduce MR size
* undo unneccessary import changes to reduce MR size
* Unit Test fixes...more to come...
* add unregister
* fix tests
* Changed ResourceVersionCache to use a Map of Maps.
* searchparam test
* test passes
* resolved fixme
* fixmies
* strengthen test asserts
* More Unit Test changes and added some FIMXME items.
* changed from long to changeresult
* renamed VersionChange -> ResourceChange
* fixed delete bug
* organize imports
* fix test
* add update test
* add test reset function
* fix stack overflow
* fix startup race condition (might still be intermittent)
* found the problem. delete doesn't work because we can't look up the deleted resource to find out what its name is
* fixed regression
* abandoned idea of incrementally updating searchparam registry. Rebuilding every time--it doesn't change that often.
* fix test
* begin with failing test
* test passes
* fixmes and javadoc
* fix test
* fixme
* fix test
* whack-a-mole. Either subs pass or cql passes. Something's fishy with the FhirContext
* fix subscription test initialization
* fix method name
* Re-factored the CqlProvider Unit Tests.
* changed ResourceChange API
* add interface
* add interface
* fix test
* add schedule test
* add doc
* init rev
* FIXME
* modify FhirContext change
* change fhirContext.getResourceTypes to lazy load
* converted subscriptions
* converted subscriptions
* begin with failing test
* test passes
* fix test
* test coverage
* test coverage
* test coverage
* test coverage
* good coverage now
* pre-review cleanup. I think I found a bug.
* moved cache into listener entry
tests pass with fixmes
* fix test
* fix test
* fix test
* fixme
* FIXMEs
* merge cache and registry
* method reorg
* javadoc
* javadoc done. all FIXMEs resolved.
* change log
* changes needed by cdr
* spring config cleanup
* james feedback
* james feedback
* might not work. Try moving resourcechangeconfig into searchparam config
* merge ResourceChangeListenerRegistryConfig.java into SearchParamConfig
* fix test
* fix SubscriptionLoader
* fix SubscriptionLoader
* create ResourceVersionMap from resources
* added cache handle interface
* fix test
* javadoc
* fix test
* fix test
* James feedback: clone searchparametermap
* fix startup
* fix test
* fix test
* fix intermittent
* pre-review cleanup
* FIXME
* final FIXME yay!
* Address a couple of my own reviw comments
Co-authored-by: Kevin Dougan <kevin.dougan@smilecdr.com>
Co-authored-by: jamesagnew <jamesagnew@gmail.com>
* switched to adding annotations. Just did a test with Token, but the sql looks clean.
* switched to adding annotations. Just did a test with Token, but the sql looks clean.
* added the other six index links
* trying different annotations
* Start testing
* Update resources on package install
* Add changelog
* Join rework
* CLean up SQL builder
* Add docs
* SP rework
* Join work
* Work on params
* Work on refactor
* Work on chains
* Work on joins
* Rework queries
* Work on queries
* Many more tests passing
* Refs test
* Work on sorting
* Work on tests
* More joins work
* Work on tests
* Work on queries
* Tests passing
* More test fixes
* Test fixes
* Work on SQL
* Tests passing
* Add some tests
* Add some tests
* License headers
* Use entity manager to get datasourcd
* One more fix
* Model cleanup
* Ongoing work
* Fixes
* Fixes
* Work on joins
* Ongoing fixes
* Merge conflict
* Cleanup
* clean up unused fields
* Work on join
* COmpile fix
* Rework querying
* Update hapi-fhir-jpaserver-base/src/main/java/ca/uhn/fhir/jpa/config/HibernateDialectProvider.java
Co-authored-by: Ken Stevens <khstevens@gmail.com>
* Address review comments
* Resolve fixmes
* Test fix
* Test fixes
* Test fix
Co-authored-by: Ken Stevens <khstevens@gmail.com>
* Improve synchronous searching by providing offset & limit support
Add support for offset querying which leverages paging at the query level
Add configuration for search default page size and search maximum page size
If using offset, always use synchronous searching to avoid extra database insert/update etc.
When using offset, only calculate count if it's wanted
Validate params closer to provider and handle size if search returned "all" (or last)
* Review and test fixes
Comment _offset as nonstandard parameter
Make synchronous search always count the total value (for now)
Fix issue with FulltextSearchSvcImpl mutating param map
Dirty fix for BaseJpaTest (not sure how to fix the including resources issue)
* Remove temporary count querying fix for synchronous loads
* Fix offsetting of everything operations, do not drop zero offset
* Fix jpa test default and maximum page size, add some fixmes to tests before resolved
* Ignore one failing test, fix others
* Fix Dereferenced variable may be null
* Fix everything paging in R4 by adding DISTINCT if synchronous load is used
Also fix assertion of size when hitting fetchSizeDefaultMaximum
* Add documentation about offset annotation and paging
Co-authored-by: James Agnew <jamesagnew@gmail.com>
* Add test
* Fix
* Add changelog
* Test fixes
* Test fix
* Test fixes
* Resolve fixme
* Test fixes
* Remove debug logs
* Try to get CI tests passing again