Releases: python-constraint/python-constraint
2.3.1: minor fix after major update
Minor fix to ensure variables less than one are tracked where required for the ExactProdConstraint
and MaxProdConstraint
constraints.
Full Changelog: 2.3.0...2.3.1
2.3.0: major improvements for built-in constraints and parsing
This release brings three exciting improvements:
- Python Constraint now has an entirely new type of built-in constraints, allowing comparisons with variables on both sides (e.g.
"x == y / z"
) to be evaluated much more efficiently. - Constraint rewriting has been extended to include subtraction and division operators.
- Rewriting of restrictions with less-than or greater-than operators now supports floats as well.
In addition:
- Tests have been extended and amended to reflect changes introduced.
- Documentation has been extended and amended to reflect changes introduced.
- Various minor bugs have been resolved.
- Bump requests from 2.32.3 to 2.32.4 in /docs by @dependabot in #99
Full Changelog: 2.2.3...2.3.0
2.2.3: extend wheel binaries to older manylinux versions, updated documentation dependencies
This release adds additional wheel binaries for older manylinux versions.
It also updates the dependencies for the documentation and brings an improved changelog and pyproject.
Full Changelog: 2.2.2...2.2.3
2.2.2: bug fix for sorting issue
This update brings a bug fix for an issue where variable names of incomparable types would raise an issue, as described in #96.
Full Changelog: 2.2.1...2.2.2
2.2.1: update to classifiers and performance benchmarking
This fix update adds Python version classifiers and adjusts the performance benchmarking threshold.
Full Changelog: 2.2.0...2.2.1
2.2.0: overall quality improvements
This release raises the overall quality of python-constraint with a wide variety of improvements, such as:
- Automatic performance benchmarking and validation with real-world cases and automatic reporting (see #93).
- Preparations for free-threading (no GIL) capabilities (see #94).
- Removed NumPy dependency in micro benchmarks, updated dependencies, and optimized test interfaces.
- Changed development status from Beta to Production / Stable.
Full Changelog: 2.1.0...2.2.0
2.1.0: string constraints and parallel solver
This release provides two exciting improvements: string constraints and a parallel solver!
Introducing string constraints, a new interface for addConstraint where users can simply write their constraint(s) as a list of Python-evaluable strings. Writing constraints in the new string format is preferable over functions and lambdas. These strings, even as compound statements, are automatically parsed to faster built-in constraints, are more concise, and do not require constraint solving familiarity by the user to be efficient.
For example, problem.addConstraint(["50 <= x * y < 100"])
is parsed to [MinProdConstraint(50, ["x", "y"]), MaxProdConstraint(100, ["x", "y"])]
.
This feature is in beta and subject to possible change, please provide feedback.
The new Parallel Solver is a solver that is able to parallelize solving for all solutions over the largest domain in both ProcessPool and ThreadPool mode. In ProcessPool mode, string constraints must be used instead of functions or lambdas, as these can't be pickled. This Parallel Solver is currently experimental and unlikely to provide an actual speedup - it is recommended to use the default OptimizedBacktrackingSolver for now.
In addition, Python 3.13 and 3.14 support has been added and dependencies updated.
For more details, see #91.
Full Changelog: 2.0.3...2.1.0
2.0.3: Finding module path at runtime
2.0.2: minor improvements to MinConflictsSolver and packaging
What's Changed
- Add ability to pass custom Random to MinConflictsSolver by @Timmmm in #81
- Include .so compiled cpython libraries to wheel package by @iyanmv in #88
- Minor improvements to changelog
- Updated dependencies
New Contributors
Full Changelog: 2.0.1...2.0.2