- 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
- Replace malloc/calloc/free with Standard::AllocateOptimal/Standard::Free for OCCT-consistent memory management
- Modernize constructor declarations and use nullptr instead of 0 for null pointers
- Replace custom swap implementation with std::swap and add noexcept specifications
- Introduces position and dynamic_cast based logic in AddValuesSeparator to decide when to append ", ".
- Adds new GTest suite Standard_Dump_Test.cxx plus an extra Bnd_Box JSON test in TKMath.
- Updates FILES.cmake to include the new test file.
- Replace legacy Standard_Mutex usage across many modules with std::mutex.
- Include <mutex> where needed and remove <Standard_Mutex.hxx> includes.
- Replace Standard_Mutex::Sentry / explicit Lock/Unlock with std::lock_guard or std::unique_lock.
- Convert optional/heap mutex holders to std::unique_ptr<std::mutex> and adapt locking accordingly.
- Simplify several singleton initializations (remove manual double-checked locking where safe).
- Use thread_local for per-thread flags instead of ad-hoc mutex protection.
- Fix BVH_BuildQueue Fetch logic to preserve thread counters and wasBusy handling.
- Remove obsolete TopTools_MutexForShapeProvider sources and update FILES.cmake.
This modernizes mutex usage, reduces dependency on custom mutex types and improves clarity of locking patterns.
- Add noexcept to many small mutating and non-throwing methods (Mirror, Mirrored, SetMirror, SetTranslationPart, etc.) across gp classes.
- Annotate trivial getters, translators and small computations as constexpr where applicable (SetLocation/SetPosition, Area/Length/Radius/Volume, Axis/Location/Position accessors, Translate helpers).
- Update headers and source files to improve noexcept/constexpr conformance for gp geometry and math types, enabling better optimization and stronger exception-safety guarantees.
- Added `noexcept` to transformation and mirroring methods across all 2D geometry classes
- Added `constexpr` to trivial getters, setters, and simple computational methods where applicable
- Removed an unused variable declaration in GeomInt_IntSS_1.cxx
- Delete PLib_Base.hxx / PLib_Base.cxx and remove from PLib/FILES.cmake.
- Convert PLib_JacobiPolynomial and PLib_HermitJacobi from polymorphic Handle-based types to direct value types:
- Remove inheritance from PLib_Base and RTTI macros.
- Replace DEFINE_STANDARD_HANDLE / Handle usage with plain objects/members.
- Make internal PLib_JacobiPolynomial member const where appropriate.
- Make basis evaluation methods const (D0/D1/D2/D3 and internal D0123) and mark trivial accessors noexcept:
- WorkDegree() and NivConstr() now noexcept.
- Update signatures (const correctness) across implementation and headers.
- Update all call sites to new API and value semantics:
- Tests: PLib_HermitJacobi_Test.cxx, PLib_JacobiPolynomial_Test.cxx updated to construct objects by value and call methods without '->'.
- Modeling/approximation & FEM code: AdvApprox_ApproxAFunction, AdvApprox_SimpleApprox, AppDef_LinearCriteria, AppDef_Variational, FEmTool_Curve, FEmTool_ElementsOfRefMatrix, FEmTool_LinearFlexion/Jerk/Tension and related headers updated to accept/use PLib_HermitJacobi by reference/value and call new methods.
- Replace Handle(...) constructions with stack/local objects and adapt calls (ToCoefficients, ReduceDegree, MaxError, AverageError, Points, Weights, etc.).
- Miscellaneous API adjustments to match new declarations (removed Standard_OVERRIDE annotations where not applicable).
- Keep behavior unchanged; this is an API/implementation refactor to remove legacy polymorphic base and improve const-correctness and performance.
- Fixed capitalization of comment opening words to follow standard documentation conventions
- Reformatted multi-line comments to improve readability by reducing excessive line breaks
- Corrected a typo in parameter description
- Remove separate PLib_HermitJacobi.lxx and stop installing it in FILES.cmake.
- Move Hermite matrices and W coefficients into the .cxx translation unit (anonymous namespace) as constexpr/static helpers.
- Replace Handle(PLib_JacobiPolynomial) with a value member; forward WorkDegree() and NivConstr() inline to myJacobi.
- Update ToCoefficients, D0123 and related methods to use the new helper functions and value-style myJacobi API.
- Remove myH and myWCoeff members and adapt matrix/coeff access accordingly.
- Minor naming/variable adjustments for clarity (NivConstr/WorkDegree -> aNivConstr/aWorkDegree).
- Add PLib_JacobiPolynomial_Coeffs.pxx with large constexpr tables and a JacobiCoefficientsCache + GetJacobiCoefficients() fast lookup to serve precomputed TNorm/CofA/CofB/Denom arrays for all constraint levels and degrees.
- Replace several runtime-initialized per-instance arrays/handles with zero-overhead constexpr data; remove myTNorm/myCofA/myCofB/myDenom handles from the class.
- Modernize PLib_JacobiPolynomial implementation:
- Use constructor initializer list and validate inputs there.
- Use constexpr lookup tables for weights/transforms and memcpy/std::fill_n to speed copying into TColStd arrays.
- Replace switch-based pointer selection with indexed arrays for cleaner selection of DB pointers.
- Simplify numerical loops, make offsets/indices const where possible and use clearer variable names.
- Use GetJacobiCoefficients() in D0123 (and callers) to remove on-demand initialization and reduce per-call overhead.
- Change static data types to constexpr double for stronger optimization and clearer intent.
- Various micro-optimizations and safety fixes (avoid NULL, tighten const correctness) to improve performance and reduce runtime allocations.
Make gp basic types and operations more constexpr-friendly and noexcept:
- Mark gp::Resolution as constexpr.
- Add constexpr and/or noexcept to many constructors and small access/modifier methods.
- Convert several functions to constexpr where safe to allow compile-time evaluation.
- Initialize members via initializer lists for several GTrsf/GTrsf2d/Mat/etc constructors.
- Add noexcept to transformation/serialization helpers and small utilities where appropriate.
This enables better compile-time optimizations and safer noexcept semantics across TKMath gp primitives.
Introduce TCollection_AsciiString::EmptyString() and TCollection_ExtendedString::EmptyString()
and update headers.
Replace many local/static empty string instances and default-constructed
temporaries with the shared EmptyString() reference across ApplicationFramework, DataExchange,
Visualization and other modules.
Remove several unused local empty-key/static variables and
adjust default parameters and return paths to use the centralized empty-string accessors.
Add missing #include <TCollection_AsciiString.hxx> to several headers that use/forward TCollection_AsciiString.
- Mark many methods and operators noexcept (iterators, containers, allocators, maps, lists, buffers, UTF/string utilities, etc.)
- Add constexpr to suitable functions and factory methods (matrices, vectors, static sizes) to enable compile-time usage
- Replace non-constexpr memcpy usages in NCollection_Mat3/NCollection_Mat4 with constexpr-friendly initialization and promote static arrays to constexpr
- Make several small API improvements: stronger const-correctness, noexcept on allocation/deallocation operators, delete forbidden copy ctors/assign where appropriate
- Remove obsolete NCollection_TypeDef.hxx and update FILES.cmake
- Miscellaneous small refactors to support modern C++ patterns and better compile-time guarantees
- Converts `BSplCLib::MaxDegree()` to a constexpr function to enable compile-time evaluation
- Replaces the dynamically allocated `BinomAllocator` with a template-based constexpr Pascal's triangle that is computed at compile time
- Add M_SQRT2 macro in multiple sources and replace runtime sqrt/Sqrt(2) calls with the macro
- Add SQRT_3 constexpr in IGESGraph_Color and Bnd_OBB and replace Sqrt(3) usages
- Make BSplSLib_zero a constexpr array
- Replace a few computed sqrt literals with compile-time constants for clarity and consistency
- Updated preprocessor directives to utilize Standard_CPP17_OR_HIGHER instead of direct __cplusplus checks.
- Removed unnecessary preprocessor checks for C++17 in TCollection_AsciiString implementation and tests.
- Enhanced code readability and maintainability by centralizing C++ version checks.
Delete PLib_DoubleJacobiPolynomial sources (cxx/hxx/lxx), remove associated Google Test (PLib_DoubleJacobiPolynomial_Test.cxx)
and update FILES.cmake entries in TKMath/PLib and TKMath/GTests to stop building the removed files.
- Replace std::unordered_map<std::string, Stopwatch> with NCollection_DataMap<TCollection_AsciiString, Stopwatch>
- Update all method signatures to use TCollection_AsciiString instead of std::string
- Replace STL map operations with OCCT equivalents (Bind/Seek/ChangeSeek/IsBound/UnBind/Clear)
Allocate TColStd_Array1OfReal using aDegree = WorkDegree - 2*(NivConstr()+1) in BasisFunctionD0, BasisFunctionDerivatives and StressTests so array sizes match the actual number of basis functions and avoid sizing mismatches.
- Rename internal members mystring/mylength -> myString/myLength across AsciiString
implementation and related files to improve naming consistency.
- Add std::string_view support and efficient overloads:
- constructor from string_view, AssignCat/Copy/Insert/SetValue from string_view
- Cat, operators (+, +=) and comparison helpers accepting string_view
- Token/StartsWith/EndsWith/Search helpers and literal-template overloads
- operator std::string_view() to obtain a non-owning view
- IsEqual/IsSameString/IsDifferent variants for string_view/C-string combinations
- optimized literal templates to avoid runtime strlen for compile-time literals
- Improve memory handling and allocation helpers (allocate/reallocate/deallocate)
and keep padding strategy consistent.
- Add Insert/AssignCat implementations that accept string_view and use memmove/memcpy.
- Fix calls in HAsciiString to use direct method instead of members
- Adjust misc. parameter names and small code-style/clarity improvements.
These changes enable zero-copy interop with std::string_view, reduce redundant C-string
operations for literals, and unify internal naming for maintainability.
- Removal of self-referencing includes where files include themselves
- Elimination of duplicate include statements within the same file
- Cleanup of redundant includes in conditional compilation blocks
- Adding CI validation for validation PRs
Replace the incorrect implementation of Axis Aligned Bounding Boxes transformation to Jim Arvo's efficient AABB transformation algorithm described in his chapter in the Graphics Gems (1990) book.
The algorithm simplifies the number of operations necessary to transform the AABB. Instead of transforming the eight vertices of the original AABB, and then computing the bounding box of the transformed vertices, Arvo's algorithm works by forming both products for each component of the min and max point, and summing the smallest/largest terms, we arrive at the minimal/maximal value.
Used regex in src folder:
^//[\s\n\r\t]*[-=]{2,}[=]*[\n\r]*//[\s\n\r\t]*[a-z]+[\s\n\r\t]*:[\s\n\r\t]*[a-z0-9_:\t\s]*[\s\n\r\t]*//[\s\n\r\t]*[a-z]+[\s\n\r\t]*:[\s\n\r\t]*[a-z0-9_\s\t./-><]{0,22}//[\s\n\r\t]*[-=]{3,}[=]*
- Complete replacement of handle-based arrays (`TColStd_HArray`) with direct `NCollection_Array` instances
- Comprehensive algorithm refactoring using the QL algorithm with Wilkinson shifts for improved numerical stability
- Enhanced documentation with detailed mathematical context and algorithm explanations
- Addition of comprehensive unit tests covering edge cases and numerical stability scenarios
Increase DoubleTab container buffer size to 64 from 16.
Optimize the vector initialisation with 5-10% increasing time.
Before the creating was takes too much branching.
- Complete removal of old Draw Harness test infrastructure for NCollection classes
- Addition of new GTest files testing STL algorithm compatibility (min, max, replace, sort, reverse)
- Migration of OSD_Path and Handle operation tests to GTest format
- Replaces legacy DRAW command-based tests with modern GTest framework
- Removes the `OCC28887` DRAW command and associated test files
- Adds comprehensive unit test coverage for `Standard_ArrayStreamBuffer` functionality
- Complete removal of the `OSD_MAllocHook` class and its related handler implementations
- Removal of the Draw command "mallochook" that provided command-line interface to the memory tracking functionality
- Updates to build configuration to exclude the removed files
- Relaxed mathematical tolerances from extremely tight values (1.0e-15) to more reasonable ones (1.0e-10/1.0e-12) for debug mode compatibility
- Increased iteration limits to allow algorithms more time to converge in debug builds
- Added exception handling with EXPECT_NO_THROW to prevent test crashes in debug mode
- Introduced comprehensive unit tests for the Jacobi polynomial implementation in PLib, covering constructors, edge cases, Gauss integration points, weights, and basis function evaluations.
- Added tests for basic utility functions in PLib, including pole conversion and binomial coefficient calculations.
- Implemented checks for Hermite interpolation and polynomial evaluation with derivatives.
- Enhanced error handling and edge case testing for small and large coefficients.
- Initialised MaxError in PLib_DoubleJacobiPolynomial to ensure consistent behaviour during degree reduction.
- Added missing `<algorithm>` include for C++ standard library functionality
- Suppressed unknown warning options in the macOS build configuration to prevent build failures
- Extensive test coverage for core mathematical algorithms (Newton methods, optimization, linear algebra, root finding)
- Bug fix for proper iteration count tracking in `math_FunctionRoot`
- Bug fix for custom vector bounds handling in `math_SVD`
- Dependency update in `OSD_PerfMeter_Test` by replacing Boolean operations with mathematical computations
- Updates function calls from `sprintf` to `Sprintf` across the entire codebase
- Updates comments and documentation references to reflect the new function name
- Adds format attribute annotations to the `Sprintf` function declaration for better compiler validation
- Updates `.clang-format` to include `OCC_CATCH_SIGNALS`, `DEFINE_STANDARD_RTTIEXT`, and `Draw_Drawable3D_FACTORY` as statement macros
- Adds a GitHub action step to automatically remove empty lines after `Standard_DEPRECATED` macro declarations
- Reformats numerous files by removing blank lines after deprecated method declarations and improving macro formatting
- Standardizes comment formatting by replacing detailed comment blocks with a uniform "//=================================================================================================" separator
- Removes redundant "function:" and "purpose:" fields that often contained minimal or no useful information
- Maintains code functionality while improving visual consistency across the codebase
- Introduces helper functions for consistent padded memory allocation across AsciiString and ExtendedString classes
- Improves documentation for UTF-8 conversion methods with clearer buffer allocation guidance
- Adds comprehensive test coverage for string constructors, memory allocation, and UTF-8 conversion functionality
- Fixed the logical error in `CornerMax()` method by correcting the openness condition checks
- Added comprehensive GTest suite for `Bnd_Box` class to prevent similar regressions
- Minor code formatting improvements in the `DumpJson()` method
