Generators Release - 1.10.0 (documentation in progress)
22/07/2025
We’re excited to introduce a new version of our generators with a range of powerful enhancements and improvements:
- Pre- & Post-Template Customizations: Gain unprecedented flexibility by customizing the generated code before and after template execution.
- Smarter Change Detection for Data Store: Data change tracking configurable either as column or hash comparison.
- Customizable Data Quality Rules for Dimensional Generators: Define tailored rules for your data quality needs.
- Parent-Child Relationships in Dimensional Models: Now supports complex hierarchies, including mixed SCD configurations.
- New Model Object: Mart Flatten Hierarchy (Spark only)
- Enhanced PowerShell-based deployments to release your code to the target platform.
- Drop Table Commands Now Optional: Can be toggled in the project settings.
How to update: To upgrade to the latest generator version, open a project and a branch (or create one), and update the project via the project settings.
Databricks Generator – Key Changes
Changes in replacement_config.json
- Delta Logging: The variable loadcontrol#loadcontrol#log_table_name is now replaced by (when Delta Logging is enabled):
- execution_log_table_name
- deployment_log_table_name
- New Section: deployment
- Allows setting a timeout for notebooks
- Defines the location of the Jupyter notebooks
Deployment Mechanism
- Split Deployment Files
- Deployment is now split into multiple notebooks
- New project-level property:
- Notebook Deployment Objects Limit (default: 100)
- The *_Deployment.ipynb notebook no longer performs the deployment itself. Instead, it triggers the individual split deployment notebooks
- It can be executed:
- From its respective subfolder
- Or from the project root folder using an input variable to specify the subfolder location
- New Root-Level Deployment Notebook
- Orchestrates the overall deployment process
- Deploys model objects
- Also deploys logging tables if Delta Logging is enabled
- Renamed Logging Deployment File
- The notebook in the LoadControl folder has been renamed (now lowercase).
Execution Enhancements
- Improved Error Handling
- The multi-threaded loader no longer returns a 0/1 result
- Instead, it throws an exception in case of errors
- Better Logging Messages
- If the Jupyter loader notebook is not found, the log now includes a clear hint to verify the file location
Snowflake Generator – Changes
- Uppercase Naming: Object and column names are now generated in UPPERCASE by default
- Simplified Variable Definitions in Loaders
- Switched to CREATE OR REPLACE Statements: Instead of separate DROP and CREATE commands
- Table and View Comments Generated by Default: Can be disabled via the project setting: Target Object Comment Type = None
- Controls whether DROP TABLE statements are generated (default: enabled)
- BG_RowHash Attribute is Now NOT NULL
- PowerShell Script deploy.ps1 Now Accepts Parameters:
- ProfileName
- ArtifactInputPath
- Help
If you have any questions or feedback, feel free to reach out. Happy deploying with the latest release of biGENIUS-X.