- Added automated migration scripts for handle syntax, standard types, and macros
- Deprecated legacy `Standard_*` types and macros in favor of native C++ equivalents
- Introduced modern `occ` namespace with template-based type checking helpers
- Enhanced NCollection macros to support variadic arguments for complex template types- Added automated migration scripts for handle syntax, standard types, and macros
- Deprecated legacy `Standard_*` types and macros in favor of native C++ equivalents
- Introduced modern `occ` namespace with template-based type checking helpers
- Enhanced NCollection macros to support variadic arguments for complex template types
- Replaced custom DEFINE_HARRAY1 and DEFINE_HSEQUENCE macros with typedefs to NCollection_HArray1 and NCollection_HSequence for various data types across multiple files.
- Updated header files in the following modules:
- HLRAlgo
- TKShHealing
- TKBRep
- TKG2d
- TKG3d
- TKGeomBase
- TKMeshVS
- TKV3d
- This change improves consistency and reduces the complexity of the codebase by utilizing the standard NCollection templates.
Replaced std::vector with NCollection_Vector and std::unordered_map with NCollection_Array1 to use OCCT's native collection classes.
Avoid hashing to use array of elements with index directly.
Remove bit maps for storing the filling status and use the direct checking for emptiness.
Optimise allocations by using IncAllocator for all NCollection_Vector objects.
math_VectorBase can be used with both stack and heap storage and resized when needed.
This is not often required, but when it is, the Resize method allows to do it efficiently,
preserving existing data and optimizing memory usage.
- Replaced `memcpy`/`memset` calls with direct initialization loops to enable register allocation for small dimensions (1-15)
- Introduced local stack arrays that reduce memory round-trips by writing to output only at the end of computation
- Fused derivative and value update loops to minimize redundant memory reads
- Added six new packages: MathUtils, MathPoly, MathLin, MathOpt, MathRoot, MathSys, and MathInteg
- Implemented modern template-based algorithms with consistent result structures and error handling
- Added comprehensive test files for all new packages
- Refactored Add method in BVH_Box to utilize in-place component-wise operations, reducing temporary vector creation.
- Introduced precomputed reciprocal direction in BVH_Ray for faster ray-box intersection tests.
- Updated RayBoxIntersection methods to leverage the new reciprocal direction for optimal performance.
- Enhanced documentation for clarity on new functionalities and optimizations.
- Added constexpr to constructors and methods in both classes, with special handling for already-normalized inputs
- Refactored coordinate access methods to avoid pointer arithmetic (incompatible with constexpr)
- Implemented a dual-path approach: fast path for normalized inputs (constexpr-compatible) and slow path for runtime normalization
Modernized and refactored the polynomial root-finding implementation
with improved numerical stability and modern C++ practices.
Key changes:
1. Implementation refactoring (math_DirectPolynomialRoots.cxx):
- Extracted helper functions into anonymous namespace for better encapsulation
- Introduced ScaledCoefficients struct for coefficient scaling operations
- Added separate functions for cubic root cases (three real, one real, multiple)
- Replaced deprecated OCCT math functions with std:: equivalents
(std::abs, std::sqrt, std::pow, std::log, std::cos, std::sin, std::atan, std::max)
- Improved code documentation with algorithm references
2. Header modernization (math_DirectPolynomialRoots.hxx):
- Added comprehensive Doxygen documentation for all public methods
- Renamed private members to follow OCCT conventions (myDone, myRoots, etc.)
- Moved inline implementations from .lxx file directly into header
- Removed math_DirectPolynomialRoots.lxx file (merged into .hxx)
3. Test improvements:
- Added test fixture class for math_DirectPolynomialRoots with helper methods
- Extended test coverage with numerical stability tests
- Added regression tests for problematic quartic cases
- Added Geom2dGcc_Circ2d3Tan tests for BUC60622 regression case
4. Minor fixes:
- Removed unused #include <iostream> from test file
- Updated FILES.cmake to remove deleted .lxx file
- Fixed test case expected values in bug28626_2
- Added a default constructor to CSLib_Class2d for creating empty classifiers.
- Implemented move constructor and move assignment operator to optimize resource management.
- Updated IntTools_FClass2d and BRepTopAdaptor_FClass2d to utilize the new move semantics, eliminating unnecessary dynamic memory allocations.
- Replaced deprecated pointer-based storage with NCollection_Sequence for better memory management.
- Removed the obsolete BRepTopAdaptor_SeqOfPtr class to streamline the codebase.
- Added `D0Local`, `D1Local`, `D2Local`, `D3Local` methods to `BSplCLib_Cache` and `BSplSLib_Cache` classes
- Added accessor methods to `Geom_BSplineCurve` and `Geom_BSplineSurface` for direct handle access to internal arrays
- Comprehensive test coverage for both curve and surface cache implementations
Refactored BSplSLib_DataContainer and BSplCLib_DataContainer_T structs:
- Use compile-time constexpr THE_MAX_DEGREE for array sizes
- Remove constructors with runtime initialization overhead
- Add standalone validateBSplineDegree() functions for degree checks
- Eliminate unnecessary zero-initialization of stack buffers
- Replace Standard_Real/Standard_Integer with native double/int types
This improves performance by avoiding runtime memset of large arrays
(~2800 doubles for surfaces, ~200 for curves) on every B-spline evaluation.
- Updated CSLib_Class2d to enhance documentation and improve parameter types for constructors.
- Refined point classification methods with clearer return types and improved comments.
- Modified CSLib_DerivativeStatus to clarify the status of surface derivatives computation.
- Enhanced CSLib_NormalPolyDef for better readability and efficiency in polynomial evaluations.
- Updated CSLib_NormalStatus to provide clearer descriptions of normal computation statuses.
- Updated point classification methods to return clearer result types.
- Enhanced input validation and array management for polygon vertices.
- Improved documentation for methods and parameters.
- Refactored internal methods for consistency and readability.
- Replaced manual absolute value logic (conditional negation) with `std::abs` for clarity
- Added `const` qualifiers to variables that are not modified after initialization
- Moved variable declarations closer to their first use and removed unused variable assignments
- Replaced `Graphic3d_Vec3d` and `Graphic3d_Vec4d` with generic `BVH_VecNt` and `BVH::VectorType<T, 4>::Type`
- Changed `Standard_Integer` and `Standard_Real` to template type `T` and `int` for generic implementation
- Added comprehensive test coverage for float-precision transformations
- Fixed multiple bugs including uninitialized variables (zmin/zmax in IsOut(gp_Lin)), incorrect type usage (Standard_Integer → Standard_Boolean), and improper gap handling
- Optimized performance-critical methods through branchless operations, early exits, and cached computations
- Modernized codebase with C++17 features including constexpr constructors, noexcept specifications, and [[nodiscard]] attributes
- Refactor Bnd_Box and Bnd_Box2d classes to introduce GetXMin, GetXMax, GetYMin, GetYMax, and GetZMin,
GetZMax methods for improved clarity and encapsulation.
- Add Limits struct to represent box limits in both classes.
- Performance optimizations including binary exponentiation for `Powered()`, optimized hash code computation, and fast-path optimizations for common operations
- Code modernization with `noexcept` qualifiers, `constexpr` for compile-time constants, and inline wrapper methods
- Added move constructors and move assignment operators to `math_VectorBase`, `math_Matrix`, and `math_DoubleTab`
- Optimized move operations to avoid unnecessary copying when dimensions match and both objects use heap allocation
- Added comprehensive test coverage for move semantics with both heap-allocated (large) and buffer-allocated (small) objects
- Replace C-style arrays with std::array for internal storage (myCenter and myHSize fields)
- Add constexpr noexcept constructors for all variants (default, gp_XY/gp_XYZ, and std::array overloads)
- Add public getter methods Center() and HSize() for direct read access to internal fields
- Add setter overloads accepting std::array parameters
- Mark appropriate methods as constexpr and noexcept for compile-time evaluation and optimization guarantees
- Clean up boolean comparisons (remove explicit Standard_True/False checks)
Majority of functions now simply call same functions from std namespace.
Functions that duplicate std namespace functionality are declared
deprecated.
Calls of deprecated functions are replaced with std functions calls.
- Enhanced the parameter validation logic in BSplCLib_CacheParams to ensure correct handling of edge cases.
- Added checks for floating point precision when determining if the next knot should be used.
- Improved code readability by restructuring the return conditions.
- Converted validation macros to inline functions for better type safety
- Added `noexcept` specifiers to non-throwing functions for compiler optimization opportunities
- Added `constexpr` to compile-time evaluable functions (comparison operators, leap year calculation)
- Enhanced `Quantity_Color::StringName()` to return "UNDEFINED" instead of throwing exceptions
- Introduced shared time constants header for better maintainability
- Converted all constructor implementations from assignment-based to initializer list-based initialization
- Added `constexpr` and `noexcept` qualifiers to the default constructor where previously missing
- Removed `std::memset` usage in favor of compile-time initialization
- Created BSplCLib_CurveComputation.pxx with template implementations
for 2D/3D curve computation methods
- Updated BSplCLib_1.cxx to use explicit template instantiation for 2D types
(gp_Pnt2d, gp_Vec2d, TColgp_Array1OfPnt2d)
- Updated BSplCLib_3.cxx to use explicit template instantiation for 3D types
(gp_Pnt, gp_Vec, TColgp_Array1OfPnt)
- Removed deprecated BSplCLib_CurveComputation.gxx file
- Deletion of 20 legacy DRAW test files (.tcl format) from tests/bugs/ directories
- Addition of 15 new GTest C++ test files across multiple modules
- Removal of corresponding QA command implementations from TKQADraw
- Replaced macro-based template system with C++ templates for `Bnd_B2` and `Bnd_B3` classes
- Fixed incorrect use of Y-coordinate instead of Z-coordinate in transformation methods
- Added comprehensive GTest test coverage for both 2D and 3D bounding box classes
- Included <mutex> in Interface_Category.cxx to support thread synchronization.
- Added <windows.h> in Standard_StackTrace.cxx for Windows-specific functionality.
- Removed deprecated macro usage example in Standard_Macro.hxx to clean up the code.
- Removed 37 DRAW test scripts from `tests/bugs/` directories
- Added 31 new GTest C++ test files in appropriate `GTests/` directories
- Removed corresponding QAcommands implementations from QABugs source files
- Updated CMake FILES.cmake files to include new test files
- Standardized spacing in comment formatting (removing extra spaces after colons, between words)
- Fixed one typo in a parameter name within a comment
- Translated one French comment to English
- Eliminated redundant member variables (`LowerRowIndex`, `UpperRowIndex`, `LowerColIndex`, `UpperColIndex`) that duplicated information already tracked by the underlying `math_DoubleTab Array`
- Inlined most `math_Matrix` methods into the `.lxx` file for better performance
- Updated all access patterns to use the `Array` member's methods directly
- Added `noexcept` qualifiers to non-throwing methods
- Modernized operator implementations (e.g., `operator-()` now returns `const`)
- Introduced a centralized `normalizeAngle()` function in both ElSLib.cxx and ElCLib.cxx with special handling for values near zero and the 2π seam
- Replaced multiple instances of inline angle normalization code with calls to the new function
- Migrated the OCC24945 bug test from Draw Harness to GTest framework
- Updated expected test values to reflect the improved normalization behavior
- Introduced new static constexpr methods: Computational() and SquareComputational() for machine epsilon precision.
- Enhanced documentation to clarify the purpose and use cases of these methods in numerical computations.
- Emphasized the distinction between machine epsilon and geometric tolerances for better understanding.
- Corrected spelling errors in comments (e.g., "od" to "of", "thr" to "the")
- Removed unnecessary `<br>` HTML tags from comment formatting
- Normalized whitespace in comments for consistency
- Fixed terminology (e.g., "3-d" to "3D")
- Addition of `gp_Dir::D` and `gp_Dir2d::D` enums for standard directions (X, Y, Z, NX, NY, NZ)
- Constexpr/noexcept constructors for geometric primitives (circles, cones, cylinders, etc.)
- Enhanced axis placement classes with enum-based constructors
- Replacement of hardcoded direction values throughout the codebase
- Added `Standard_MACRO_DEPRECATED` macro to mark preprocessor macros as deprecated with compiler-specific warning generation
- Simplified deprecated macros (`Standard_OVERRIDE`, `Standard_DELETE`, `Standard_FALLTHROUGH`, `Standard_NODISCARD`, `Standard_THREADLOCAL`, `Standard_ATOMIC`, `Standard_Noexcept`) to expand directly to their C++11/C++17 equivalents while emitting deprecation warnings
- Removed complex conditional compilation logic that checked for compiler/standard support
