Remove cartesian_product and any_check

[zasdfgbnm]

This update removes the cartesian_product function and introduces a more efficient any_check function that utilizes C++20 fold expressions. The new implementation checks if any elements in one or more tuples satisfy a given predicate, improving code clarity and performance.

[zasdfgbnm]

!test

[github-actions]

Review updated until commit 6b6d622

Description

• Remove cartesian_product and any_check utility functions from type_traits.h

• Refactor any_check usages to use C++20 fold expressions with std::apply

• Convert is_dynamic_type trait to variable template is_dynamic_type_v

• Introduce DynamicTypeLike concept for improved type constraints

• Update operator definitions and macros to use modern C++20 features

Changes walkthrough

Relevant files

Enhancement

dynamic_type.h Modernize type checking with C++20 fold expressions            lib/dynamic_type/src/dynamic_type/dynamic_type.h Replace any_check calls with std::apply and fold expressions in can_cast_to, has_square_bracket, has_any_square_bracket Convert is_dynamic_type trait to variable template is_dynamic_type_v Add DynamicTypeLike concept for type constraints Update DEFINE_UNARY_OP, operator*, operator<< to use fold expressions Refactor DEFINE_LEFT_PPMM and DEFINE_RIGHT_PPMM macros with fold expressions Update DEFINE_ASSIGNMENT_OP and has_cross_type_equality to use DynamicTypeLike concept +118/-129 type_traits.h Remove deprecated utility functions                                            lib/dynamic_type/src/dynamic_type/type_traits.h Remove entire cartesian_product function implementation (lines 248-318) Remove entire any_check function implementation (lines 320-355) +0/-96

PR Reviewer Guide

Here are some key observations to aid the review process:

🧪 No relevant tests

⚡ Recommended focus areas for review

Missing Performance Data The PR description claims performance improvements but provides no actual performance metrics, benchmarks, or data to substantiate these claims. This makes it difficult to evaluate the effectiveness of the changes. static constexpr bool can_cast_to = std::apply( [](auto... ts) constexpr { return ( (requires(typename decltype(ts)::type from) { (T)(from); }) || ...); }, type_identities_as_tuple); No Tests for New Implementation The PR removes the cartesian_product and any_check functions but doesn't include tests for the new fold expression implementations. This creates a risk of regressions and makes it difficult to verify correctness of the new approach. template <typename> inline constexpr bool is_dynamic_type_v = false; template <typename Containers, typename... Ts> inline constexpr bool is_dynamic_type_v<DynamicType<Containers, Ts...>> = true; // A DynamicType, allowing cv/ref qualifiers on the deduced type. template <typename T> concept DynamicTypeLike = is_dynamic_type_v<std::remove_cvref_t<T>>; Removed Functionality Without Migration Path The cartesian_product and any_check functions are completely removed from type_traits.h. If these functions were part of the public API or used elsewhere, this could break existing code. The PR should verify these functions are not used externally. static_assert(!belongs_to<int, float, double, long>);

[zasdfgbnm]

!test

[zasdfgbnm]

!test