## Validator Changes
* Validate of experimental property on bindings
* Add -no-experimental-content flag
* Allow JSON named extensions to be structure types other than logical
* Process relative URLs properly in base when generating snapshots
## Other code changes
* Fix rendering path for value set references for kindling
* refactor error handling in ProfileUtilities
* Fix for NPE processing packages
***NO_CI***
## Validator Changes
* Fix expression for con-3 properly (fix validation problem on some condition resources)
* Fix FHIRPath bug using wrong type on simple elements when checking FHIRPath types
* FHIRPath: Allow _ in constant names (per FHIRPath spec)
* Fix value set rendering creating wrong references
* Fix bug processing value set includes / excludes that are just value sets (no system value)
* Alter processing of unknown code systems per discussion at ,https://chat.fhir.org/#narrow/stream/179252-IG-creation/topic/Don't.20error.20when.20you.20can't.20find.20code.20system and implement unknown-codesystems-cause-errors
* Improve message for when elements are out of order in profile differentials
## Other code changes
* fix problem where profile rendering had spurious 'slices for' nodes everywhere
* Update SQL-On-FHIR implementation for latest cases, and clone test cases to general test care repository
* Fix problem generating value set spreadsheets
* fix concurrent modification error processing language translations
* Check for null fetcher processing ConceptMaps (#1728)
***NO_CI***
## Validator Changes
* Security Fix: Move all instantiation of transformerFactory to XMLUtils and set ACCESS_EXTERNAL flags automatically (slightly improvement to output message, but closes possible attack vector)
* Handle 308 redirects when fetching packages
* Fix NPE in ValueSetValidator
* Fix for NPE in IGLoader
* fix version conversion problem for ConceptMap (4<->5): relationship comment getting lost for noMap entries
* Add check for multiple WG extensions (HL7 context)
## Other code changes
* Suppress spurious logging statement
* Add Australian CDA name prefix & uri to mapping transform code
* Add new messages for tx.fhir.org
* Remove narrative from snapshot tests
* Improve complex extension rendering when rendering by profile
* Rewrite processing of map statements when generating snapshots to fix known bugs
* Updates to Capability Statement rendering (and minor Operation Definition rendering improvement)
* Fix wrong reference to CDA classes for unscoped class names
* fix rendering issue for R4 relationship codes in ConceptMap
* rework decimal lowBoundary() and highBoundary() after discussion on Zulip, and add extensive testing
* Fix existing test case for fixed up boundary handling in FHIRPath
* Add CodeSystem fetching for selected code systems for VSAC import
* Drop to saxon 11.6 for now
* fix issue with json unicode whitespace enscaping
* get Java generator running again (though no code regenerated)
* don't escape unicode whitespace in json files
* Fix NPE in questionnaire renderer
* Hack workaround for R4 issue with ProcessPriority
* Rewrite concurrency management for file system package cache
***NO_CI***
## Validator Changes
* Improved handling of display validation for various languages
* fix xml:space message after discussion on Zulip
## Other code changes
* ValueSet rendering improvements
* Add isUnknownLocalReferencesNotLinks to RenderingContext
* Fix up conceptMap rendering (remove spurious summary table)
* Add R6 ballot 2 to FHIRVersion enum in R5
* rendering fixes for implementers
* improve rendering of bindings with no value sets
***NO_CI***
## Validator Changes
* Improve performance of supplement processing
* Add support for validating xml:space
* Fix NPE validating maps
* Fix duplicate removal code that caused missed WARNING messages (#1698)
* Fix interaction between terminology cache and validation location (#1700)
* Better concept map validation
* Improved base64 parsing error
* Add support for unknown code system version validating concept maps
* Fix bug using wrong profile when profiling recursive properties
## Other code changes
* Return null for all toCode enum params that are null
* fix bugs adding publishing WG in element mode
* Fix equality method that caused missed WARNING messages
* fix rendering issues
* rendering fixes for R6 build
* don't throw error for wrong rendering mode
* Document rendering fixes - subject and section
* more work on forcing valid html generation
* more hyperlink validation improvements
* further fixes to link validation
* fix list inside paragraph
* fix actor capabilities in wrong place
* Support missing datatypes
* Fill out missing anchors
* Fix wrong handling of contained resources)
* #1678 - add div to children when exploring Narrative
***NO_CI***
## Validator Changes
* No Changes
## Other code changes
* Naming System fix for THO
* Fix issue with rendering OperationOutcomes that don't have 'detail'
* Fix problem with value set anchor generation when rendering
***NO_CI***
## Validator Changes
* work around THO 6.0.0 problem (hack, to be reversed later)
* Improve MeasureReport validation for checking subject count
* Fix for R2B Resource.id cardinality problem
* Fix for validator using wrong property for list determination when parsing json
## Other code changes
* no changes
***NO_CI***
## Validator Changes
* Add clearer message for errors around unable to validate because unknown code system
* Fix problem with fhirVersion cardinality validating IG resources
## WHO I18n Project
* Remove Questionnaire.item.linkId from translatable content
## Other code changes
* Further revisions to rendering framework for id uniqueness, main spec rendering, and various reported issues
* Improve doco for R4/R4B renderers
* Introduce backwards compatible constructor in ValidationEngineBuilder and fix useEcosystem
* Use v2 of PublishCodeCoverageResults (#1665)
* #1583 do not serialize choice group + add testcase
***NO_CI***
## Validator Changes
* Add support for ND-JSON. Note that the validator is not expected to perform well on large ND-JSON files (or any other file type)
* Significant change: the validator was not checking for the end of the input when parsing a JSON resource finished. It will now start giving errors when JSON continues once the object is complete
* Add support for the create object syntax in FML when validating FML
* Improved error message when supplement url used instead of code system URL
## WHO Translation Project
* Add complete dutch translations (Thanks Alexander Henket)
## Other code changes
* Fix various rendering problems leading to non-unique html anchors
* Fix for unrendered data types
***NO_CI***
## Validator Changes
* Change how count is calculated when expanding value sets
* Fix value set expansion bugs
* Fix for fhirpath issue on command line (#1650)
* Add IG dependency validator
* work around bad r4 extension definitions
* don't reload different sub-version of extensions pack
## Other code changes
* Rework rendering library from ground up
* Merge Fix: Fix split logic of canonical into url and version in several places (#1663)
* Suppress spurious message when code system is unknown
* don't raise needless and wrong exceptions about extension definitions when rendering
* fix duplicate link creation
* Stop recursive rendering crash
* Obligation rendering improvements
* Handle extra profiles on resources in an IG when converting between versions
* Fixed issue with actor title not rendering in obligations
* Test and adaption for dateOp #1655
* Use assertj instead of hamcrest (#1662)
* More transifex work
* Add support for local packages during testing
***NO_CI***
* Fix split logic of canonical into url and version in several places
Removed
- org.hl7.fhir.r4b.renderers.DataRenderer#versionFromCanonical
- org.hl7.fhir.r4b.renderers.DataRenderer#systemFromCanonical
- org.hl7.fhir.r5.renderers.DataRenderer#versionFromCanonical
- org.hl7.fhir.r5.renderers.DataRenderer#systemFromCanonical
The logic of these methods is wrong, the functionality of extracting
the version and the system URL have been swapped.
Therefore, all usages of aforementioned methods were replaced by the
use of org.hl7.fhir.utilities.CanonicalPair. This not only fixes the
problem, but also reduces duplicated functionality.
* Remove dead code from merge
* Fix spacing issues in rendering
* Empty commit, make-pipeline-go
---------
Co-authored-by: Max Bureck <max.bureck@fokus.fraunhofer.de>
## Validator Changes
* Find value sets on tx server if not tx.fhir.org
* Do not send Content-Type header with GET requests for tx servers
* Fix npe validating code system
* Support discriminator by position
* Don't check type characteristics for unknown types
## WHO I18n Project
* Fix typos in phrases, and fix up handling of apostrophes in messages without parameters
* Test and fix for single quote escapes in I18n
* Create and test transifex configuration
## Rendering Changes
* Fix contact rendering to work properly
* Fix issue resolving contained resource rendering DiagnosticReports etc
* Handle case where Contact.value has extensions instead of a string value
* Render Parameterised types
* Fix bug with LOINC Code row showing wrongly in Profile Details view
* Partial implementation of type parameters
* Fixed rendering of actor-specific obligations, added elementIds to obligation narrative rendering
* Corrected ObligationsRenderer to handle multiple actors and multiple codes. Also got obligations with elements to render properly (which means knowing whether you're on a table page or definitions page, what tab you're on, and whether the element for the obligation is in-scope for that tab (so you know whether to hyperlink or not). Had to make links on the tables point to definitions because table anchors are not unique.
## Other code changes
* split out Base.canHavePrimitiveValue() from Base.hasPrimitiveValue()
* Refactor how observation profiles are registered in order to generalise management of extra profiles
* Use validation policy to implement Observation profile rules
* Rework validation policy management in standalone validator
* Try deploying javadoc with SNAPSHOT
* Add utility to unpack the spec.internals file
* v2 in FHIR tests
* Fix bug loading language pack NPMs
***NO_CI***
## Validator Changes
* Find value sets on tx server if not tx.fhir.org
* Do not send Content-Type header with GET requests for tx servers
* Fix npe validating code system
* Support discriminator by position
* Don't check type characteristics for unknown types
## WHO I18n Project
* Fix typos in phrases, and fix up handling of apostrophes in messages without parameters
* Test and fix for single quote escapes in I18n
* Create and test transifex configuration
## Rendering Changes
* Fix contact rendering to work properly
* Fix issue resolving contained resource rendering DiagnosticReports etc
* Handle case where Contact.value has extensions instead of a string value
* Render Parameterised types
* Fix bug with LOINC Code row showing wrongly in Profile Details view
* Partial implementation of type parameters
* Fixed rendering of actor-specific obligations, added elementIds to obligation narrative rendering
* Corrected ObligationsRenderer to handle multiple actors and multiple codes. Also got obligations with elements to render properly (which means knowing whether you're on a table page or definitions page, what tab you're on, and whether the element for the obligation is in-scope for that tab (so you know whether to hyperlink or not). Had to make links on the tables point to definitions because table anchors are not unique.
## Other code changes
* split out Base.canHavePrimitiveValue() from Base.hasPrimitiveValue()
* Refactor how observation profiles are registered in order to generalise management of extra profiles
* Use validation policy to implement Observation profile rules
* Rework validation policy management in standalone validator
* Try deploying javadoc with SNAPSHOT
* Add utility to unpack the spec.internals file
* v2 in FHIR tests
* Fix bug loading language pack NPMs
***NO_CI***
markiantorno
Grahame Grieve
Grahame Grieve
Rob Eastwood
dotasek
dotasek