Releases: okube-ai/laktory
Releases · okube-ai/laktory
v0.6.4 (2025-01-24)
v0.6.3 (2025-01-22)
v0.6.2 (2025-01-22)
What’s new
Added
- Support for Terraform
moved
option (resource rename)
Fixed
- Model variables not carried overs to all environments
Full Changelog
v0.6.1 (2025-01-20)
What’s new
Updated
- Reduced import time
- Optimized tests to load spark only when required
- Optimized performance during stack export in IaC backend native format
Full Changelog
v0.6.0 (2025-01-18)
What’s new
Added
MemoryDataSource
support for reading dict or list of data [#337]- Support for
!use
tag in YAML files to directly inject content of other file - Support for
!update
tag in YAML files to use another file to update the content of a dictionary - Support for
!extend
tag in YAML files to use another file to extend (append) more items to a list
Updated
- Laktory variables to support python expressions [#335]
- Laktory variables to support complex types such as dictionaries and lists
- All Laktory model fields to allow
str
type for receiving a variable or expression - Reference to external YAML file path can use variables injection[#335]
Breaking Change
- Removed support for ${include.*} variables in YAML files.
- Reference to external YAML files is now relative to the calling file instead of being relative to the stack entry point.
Full Changelog
v0.5.13 (2025-01-07)
What’s new
Added
- Support for Python 3.12
- Support for Python 3.13
- Contribution guidelines
- Support for external PR
Updated
uv
as the recommended package manager- Formatting and linting with ruff (instead of black)
- Added ruff formatting and linting as a pre-commit
- Added pytest fixtures to run tests only when required environment variables are available
Full Changelog
v0.5.12 (2024-12-30)
v0.5.11 (2024-12-30)
v0.5.10 (2024-12-29)
What’s new
Added
- Support for view creation from pipeline node
--version
andversion
CLI commands- Support for
for_each_task
in Databricks job resource - Support for external libraries in pipelines
Fixed
- Variables supports referencing environment variables and other variables
Updated
- Replaced
dataframe_backend
propagation with dynamic parent lookup - Introduced
PipelineChild
internal class to manage child/parent relationship - Added laktory package as a task cluster dependency when Databricks Job is used as a pipeline orchestrator
Breaking changes
- Renamed
dataframe_type
todataframe_backend
- Renamed pipeline orchestrator from
"DLT"
to"DATABRICKS_DLT"
- Renamed pipeline databricks job and dlt resource names. May cause a re-deployment.
- Moved pipeline
notebook_path
underdatabricks_job
attribute.
Full Changelog