Release v0.17.0 (#496)

* [INTERNAL] Disable downstream `ucx` check
([#495](#495)). The
downstream check for UCX has been disabled due to its current version
being pinned to less than 0.17, rendering the checks unnecessary, and
because the implementation of these checks exposes an issue with
numerous UCX unit tests failing when used in conjunction with the latest
Databricks SDK. As a result, the UCX check has been temporarily
commented out to prevent failures, while other downstream checks, such
as the one for remorph, continue to remain active, allowing for a more
stable testing environment until the necessary updates can be made to
UCX to address these issues and enable the reactivation of its
downstream check.
* Explicitly test/support python 3.13 and 3.14
([#490](#490)). The
acceptance workflow has been enhanced to support Python versions 3.13
and 3.14, in addition to existing versions, by updating environment
variables and dependencies. The workflow now utilizes Hatch version
1.16.5, installed using the `HATCH_VERSION` environment variable, and
the latest versions of `actions/checkout` and `actions/setup-python`,
specifically version 6. The Python version used in the Install Python
step has been updated to 3.14, and the types of pull requests that
trigger this workflow have been modified to exclude `ready_for_review`
events. Additionally, the condition to run the integration job has been
simplified, and environment variables `HATCH_VERBOSE` and
`HATCH_VERSION` have been added to the workflow. These changes aim to
extend the CI/CD pipeline and project metadata to cover Python 3.13 and
3.14, with incidental updates to dependencies, including upgrading Hatch
and GitHub actions, to improve the overall testing and installation
process.
* Fixed dashboard serialisation: `query` is now `queryLines` on upstream
REST API ([#491](#491)).
The `Dataset` class has been enhanced with improved serialization and
deserialization functionality, allowing it to handle compatibility
issues with dashboard APIs. The `from_dict` method now checks for both
`query` and `queryLines` attributes in the input dictionary, using the
`query` attribute directly and joining `queryLines` into a single
string, while setting the query to None if neither attribute is found.
This change enables the method to load datasets from various sources,
including previously saved files, while maintaining compatibility.
Additionally, the class's serialization and deserialization
functionality has been thoroughly tested with the introduction of
several new test methods, which verify its ability to correctly handle
different query formats, including cases where the query is represented
as `queryLines` or "query", and when the query is absent, ensuring
accurate conversion between the object and dictionary representations.
* Limit `sqlglot` to releases earlier than 26.30.0
([#410](#410)). The library
has undergone several updates to improve its functionality and
robustness. The dependency on a key library has been updated to prevent
the use of a version that introduced a breaking change, while also
incorporating type-hinting fixes to accommodate newer versions of a
static type checker. The `Tile` class has been enhanced with various
changes, including input validation, improved position calculation, and
better handling of potential issues, such as missing or invalid values.
Additionally, numerous data classes have been modified to allow for
optional values in various fields, enabling greater flexibility in the
data models and facilitating the representation of incomplete or missing
data. These changes aim to improve the overall reliability and
adaptability of the library, allowing it to handle a wider range of
scenarios and input data.
* Updated the minimum version of `databricks-labs-pytester` to 0.7.2
([#408](#408)). The minimum
required version of a testing library has been updated to 0.7.2, which
resolves an issue caused by a breaking change in the Databricks SDK
version 0.51. This update ensures compatibility and prevents import
failures when using older versions of the testing library with the new
Databricks SDK. As a result, test dependencies must be at version 0.7.2
or later, although this change does not impact normal installs or new
development environments, which will automatically use the latest
version.

---------

Co-authored-by: Andrew Snare <asnare@users.noreply.github.com>

Author: gueniai

CHANGELOG.md

@@ -1,5 +1,13 @@
 # Version changelog
 
+## 0.17.0
+
+* Explicitly test/support python 3.13 and 3.14 ([#490](https://github.com/databrickslabs/lsql/issues/490)). The acceptance workflow has been enhanced to support Python versions 3.13 and 3.14, in addition to existing versions, by updating environment variables and dependencies. The workflow now utilizes Hatch version 1.16.5, installed using the `HATCH_VERSION` environment variable, and the latest versions of `actions/checkout` and `actions/setup-python`, specifically version 6. The Python version used in the Install Python step has been updated to 3.14, and the types of pull requests that trigger this workflow have been modified to exclude `ready_for_review` events. Additionally, the condition to run the integration job has been simplified, and environment variables `HATCH_VERBOSE` and `HATCH_VERSION` have been added to the workflow. These changes aim to extend the CI/CD pipeline and project metadata to cover Python 3.13 and 3.14, with incidental updates to dependencies, including upgrading Hatch and GitHub actions, to improve the overall testing and installation process.
+* Fixed dashboard serialisation: `query` is now `queryLines` on upstream REST API ([#491](https://github.com/databrickslabs/lsql/issues/491)). The `Dataset` class has been enhanced with improved serialization and deserialization functionality, allowing it to handle compatibility issues with dashboard APIs. The `from_dict` method now checks for both `query` and `queryLines` attributes in the input dictionary, using the `query` attribute directly and joining `queryLines` into a single string, while setting the query to None if neither attribute is found. This change enables the method to load datasets from various sources, including previously saved files, while maintaining compatibility. Additionally, the class's serialization and deserialization functionality has been thoroughly tested with the introduction of several new test methods, which verify its ability to correctly handle different query formats, including cases where the query is represented as `queryLines` or "query", and when the query is absent, ensuring accurate conversion between the object and dictionary representations.
+* Limit `sqlglot` to releases earlier than 26.30.0 ([#410](https://github.com/databrickslabs/lsql/issues/410)). The library has undergone several updates to improve its functionality and robustness. The dependency on a key library has been updated to prevent the use of a version that introduced a breaking change, while also incorporating type-hinting fixes to accommodate newer versions of a static type checker. The `Tile` class has been enhanced with various changes, including input validation, improved position calculation, and better handling of potential issues, such as missing or invalid values. Additionally, numerous data classes have been modified to allow for optional values in various fields, enabling greater flexibility in the data models and facilitating the representation of incomplete or missing data. These changes aim to improve the overall reliability and adaptability of the library, allowing it to handle a wider range of scenarios and input data.
+* Updated the minimum version of `databricks-labs-pytester` to 0.7.2 ([#408](https://github.com/databrickslabs/lsql/issues/408)). The minimum required version of a testing library has been updated to 0.7.2, which resolves an issue caused by a breaking change in the Databricks SDK version 0.51. This update ensures compatibility and prevents import failures when using older versions of the testing library with the new Databricks SDK. As a result, test dependencies must be at version 0.7.2 or later, although this change does not impact normal installs or new development environments, which will automatically use the latest version.
+
+
 ## 0.16.0
 
 * Let page name adhere to naming restrictions ([#370](https://github.com/databrickslabs/lsql/issues/370)). In this release, a new method `_clean_resource_name` has been introduced to modify resource names according to the updated naming convention, allowing only alphanumeric characters, hyphens, and underscores. The `as_lakeview` method in the `BaseHandler` class now uses this new method to ensure the `Page` class name adheres to the new restrictions. Furthermore, test files for dashboards have been updated to reflect the change, with a new test function `test_dashboard_metadata_as_lakeview_cleans_page_name` verifying that page names are free of special characters, and an existing test function modified to handle invalid dashboard YAML files. These changes improve consistency, reliability, and adherence to best practices in dashboard naming within the project.

src/databricks/labs/lsql/__about__.py

@@ -1 +1 @@
-__version__ = "0.16.0"
+__version__ = "0.17.0"