NEWS

hubAdmin 1.9.0

Fixed bug that caused unhelpful validation errors when using custom task IDs (e.g., reference_date) as round ID variables with inconsistent values across model tasks in schema versions v4.0.0+.
Added support for additional properties via ... argument in create_round() and updated both create_round() and create_target_metadata_item() to support schema v6.0.0 additional_metadata field. In v6.0.0+, additional properties are wrapped in an additional_metadata field. For earlier versions: create_round() adds properties directly to the round object (note: while the schema has always allowed additional properties in rounds, the programmatic functionality was previously missing); create_target_metadata_item() adds properties directly for v5.1.0+ or ignores them with a message for earlier versions.
has_target_data_config() has been moved to and is now re-exported from hubUtils for backward compatibility (#139).
Added dynamic validation check for target-data.json that ensures when date_col references a task ID, that task ID must be a Date type task ID (i.e., contains ISO 8601 date values) (#145).

Added basic tests for v6.0.0 schema (#127).
Added support for target-data.json config validations (#132).
Added has_target_data_config() function to check for presence of target-data.json config file (#132).

update create_target_metadata_item()function to add optional additional properties, to match schema version 5.1.0. (#108)

validate_config() now detects the presence of duplicate property names (#98).
validate_config() now checks that the target_id property in each target_metadata item matches the value specified in target_keys (#102). Note this check is only performed if target_keys is not NULL and is does not contain more than one element itself.

Use options to set schema_version and branch arguments in download_tasks_schema() and the create_*() family of functions for creating config files programmatically. This allows for setting the schema version and branch globally for the session (#85).
Make validation of round_id patterns explicit. This ties in with schema version v4.0.1 where the pattern the round_id property must match if round_id_from_variable is false is now specified as a regular expression in the schema. This check is also now implemented dynamically on values of the round_id variable if round_id_from_variable is true when validating tasks.json config files. Checks on round_id patterns are also now implemented in create_round() when creating rounds programmatically.
As of schema version v5.0.0, only a single target_keys element is allowed when creating target metadata items programmatically (#89).

Support v4.0.0 schema configuration of output types and output type IDs when creating config files programmatically. Specifically, whether an output type is required or not is specified via the is_required logical property whereas the output_type_id values as provided through the required property only (#53). In addition, output_type_id required value is now encoded as NULL instead of NA (#72).
Programmatically created higher level config elements now have a branch attribute that can be used to create and validate objects against a schema which is still in development and has not been released to the main branch yet.
Add dynamic checks to ensure derived_task_ids match valid task ID names to validation of task.json config files (#69).

Add derived_task_ids property to create_config() (for specifying hub level derived task IDs) and create_round() (for specifying round level derived task IDs) (#52). This adds compatibility for the upcoming schemas v4.0.0 release.

Add as_config() function to convert a list representation of a tasks.json config file to a <config> class object (#42). Useful when wanting to programmatically manipulate the contents of a tasks.json config file.
Add append_round() function to append one or more <round> class objects to the rounds property of a <config> class object (#42).
Add schema_autobox() function that uses the schema to "box" length one vectors in a <config> class object that should be arrays in JSON format (#44). This transformation is now applied by default when writing a <config> class object to a JSON file using write_json() but can be deactivated using the autobox argument. The transformation is also applied to any properties that should be arrays covered by additionalProperties in the schema (e.g. custom task IDs).
Add get_array_schema_paths() utility function for extracting paths to potential array properties in a JSON schema and export previously internal download_tasks_schema().
Add link to development version documentation to README.

Add output_type_id_datatype argument to create_config(). This allows for the specification of the data type of the output_type_id column of model output data through schema property output_type_id_datatype, introduced in v3.0.1 version of the hubverse schema (#41)

Add write_config() functions to write objects of class <config> to JSON files (#3)

New feature: Add ci_validate_hub_config() as a non-interactive function that works with GitHub to produce a validation report on Continous Integration (#37)

Update create_output_type_cdf() to accommodate less restrictive output_type_id checks introduced in schema version v3.0.1 (#29).
URL for hubdocs updated (#27)
Output of validate_config() and validate_hub_config() are now classed so the summary of their contents is printed nicely to the screen, reducing the amount of screen space needed to report success or failure (#35)

Breaking changes: Support schema v3.0.0 specification of sample output type IDs which are now specified through a output_type_id_params object instead of output_type_id. The main breaking change is in create_output_type_sample() which now takes arguments incompatible with previous schema versions and returns an object with an output_type_id_params object instead of output_type_id. Additional but back-compatible dynamic validation checks on sample output types have been added to validate_config().

Introduce validation check that ensures no all null task IDs exist in tasks.json.

Allow task ID create_task_id() arguments required and optional to both be set to NULL, facilitating the encoding of NA task IDs in modeling tasks where no value is expected for a given task ID.

Initial package release resulting from split of hubUtils package. See hubUtils NEWS.md for details including previous release notes.