### GET /quaternion/as-matrix

Source: https://smath.slendi.dev/structsmath_1_1Quaternion.html

Converts the current quaternion instance into a 4x4 homogeneous rotation matrix.

```APIDOC
## GET /quaternion/as-matrix

### Description
Converts the quaternion representation into a 4x4 rotation matrix suitable for linear algebra transformations.

### Method
GET

### Endpoint
/quaternion/as-matrix

### Parameters
#### Path Parameters
- **id** (string) - Required - The unique identifier of the quaternion instance.

### Response
#### Success Response (200)
- **matrix** (Mat<4, 4, T>) - The 4x4 rotation matrix.

#### Response Example
{
  "matrix": [
    [1, 0, 0, 0],
    [0, 1, 0, 0],
    [0, 0, 1, 0],
    [0, 0, 0, 1]
  ]
}
```

--------------------------------

### Define C++ class hierarchy for Doxygen graph generation

Source: https://smath.slendi.dev/graph_legend.html

A collection of C++ class definitions demonstrating various inheritance types, template usage, and documentation states. These examples are used by Doxygen to generate visual inheritance and collaboration graphs.

```cpp
/*! Invisible class because of truncation */
class Invisible { };
/*! Truncated class, inheritance relation is hidden */
class Truncated : public Invisible { };
/* Class not documented with doxygen comments */
class Undocumented { };
/*! Class that is inherited using public inheritance */
class PublicBase : public Truncated { };
/*! A template class */
template<class T> class Templ { };
/*! Class that is inherited using protected inheritance */
class ProtectedBase { };
/*! Class that is inherited using private inheritance */
class PrivateBase { };
/*! Class that is used by the Inherited class */
class Used { };
/*! Super class that inherits a number of other classes */
class Inherited : public PublicBase,
protected ProtectedBase,
private PrivateBase,
public Undocumented,
public Templ<int>
{
private:
Used *m_usedClass;
};
```

--------------------------------

### Vector Element Accessors

Source: https://smath.slendi.dev/smath_8hpp_source.html

Provides standard access to vector elements using template-based get functions, supporting both mutable and immutable references with compile-time bounds checking.

```cpp
template<size_t I, size_t N, class T> constexpr T &get(Vec<N, T> &v) noexcept
{
	static_assert(I < N);
	return v[I];
}

template<size_t I, size_t N, class T> constexpr T const &get(Vec<N, T> const &v) noexcept
{
	static_assert(I < N);
	return v[I];
}
```

--------------------------------

### Constructor Methods

Source: https://smath.slendi.dev/structsmath_1_1Mat.html

Methods for initializing matrices, including zero matrices, diagonal matrices, and column-based initialization.

```APIDOC
## POST /smath/Mat/construct

### Description
Constructs a new matrix instance. Supports zero initialization, diagonal initialization, or column-based initialization.

### Method
POST

### Endpoint
/smath/Mat/construct

### Parameters
#### Request Body
- **type** (string) - Required - 'zero', 'diagonal', or 'columns'
- **diag** (T) - Optional - Value for diagonal matrix
- **cols** (Vec<R, T>[]) - Optional - Array of column vectors

### Response
#### Success Response (200)
- **matrix** (Object) - The constructed matrix object
```

--------------------------------

### Vector Initialization and Unpacking

Source: https://smath.slendi.dev/smath_8hpp_source.html

Constructors for initializing vectors from scalars or variadic arguments, and utility methods for unpacking vector components into references.

```cpp
constexpr Vec(T const &s) noexcept;
constexpr Vec(Args &&...args) noexcept;
constexpr auto unpack(Args &...args) noexcept -> void;
```

--------------------------------

### Camera and Projection Matrix Generation

Source: https://smath.slendi.dev/smath_8hpp_source.html

Functions for creating view and projection matrices, including orthographic, perspective, and infinite perspective projections, as well as LookAt camera matrices.

```cpp
template<typename T>
[[nodiscard]] inline auto matrix_ortho3d(T const left, T const right, T const bottom, T const top, T const near, T const far, bool const flip_z_axis = true) -> Mat<4, 4, T> {
	Mat<4, 4, T> res {};
	res[0, 0] = 2 / (right - left);
	res[1, 1] = 2 / (top - bottom);
	res[2, 2] = -2 / (far - near);
	res[0, 3] = -(right + left) / (right - left);
	res[1, 3] = -(top + bottom) / (top - bottom);
	res[2, 3] = -(far + near) / (far - near);
	res[3, 3] = 1;
	if (flip_z_axis) { res[2] = -res[2]; }
	return res;
}

template<typename T>
[[nodiscard]] inline auto matrix_look_at(Vec<3, T> const eye, Vec<3, T> const center, Vec<3, T> const up) -> Mat<4, 4, T> {
	auto f = (center - eye).normalized_safe();
	auto s = f.cross(up).normalized_safe();
	auto u = s.cross(f);
	return Mat<4, 4, T> {
		Vec<4, T> { s.x(), u.x(), -f.x(), 0 },
		Vec<4, T> { s.y(), u.y(), -f.y(), 0 },
		Vec<4, T> { s.z(), u.z(), -f.z(), 0 },
		Vec<4, T> { -s.dot(eye), -u.dot(eye), f.dot(eye), 1 },
	};
}
```

--------------------------------

### Matrix Constructors

Source: https://smath.slendi.dev/structsmath_1_1Mat.html

Constructors for initializing matrices as zero-filled, diagonal-filled, or from a set of column vectors. These constructors are constexpr and noexcept for optimal performance.

```cpp
template<std::size_t R, std::size_t C, typename T = float>
Mat<R, C, T> m1; // Zero matrix

Mat<R, C, T> m2(5.0f); // Diagonal matrix with 5.0

Mat<R, C, T> m3(col1, col2, col3); // Initialized from columns
```

--------------------------------

### Constants

Source: https://smath.slendi.dev/structsmath_1_1Mat.html

Static configuration constants for the matrix class.

```APIDOC
## [DATA] EPS_DEFAULT

### Description
The default epsilon value used for floating point comparisons within the matrix class.

### Value
1e-6

### Type
static constexpr T
```

--------------------------------

### Vector Initialization

Source: https://smath.slendi.dev/smath_8hpp_source.html

Basic vector structure definitions for the smath library, supporting zero-initialization and component access.

```cpp
constexpr Vec() noexcept;
```

--------------------------------

### Basic Vector Operations with smath

Source: https://smath.slendi.dev/

Demonstrates the basic usage of smath's Vec class for vector addition, dot product, and normalization. It requires the smath library and C++23 standard.

```cpp
#include <print>
#include <smath/smath.hpp>

int main() {
  using namespace smath;
  Vec3 a{1.0f, 2.0f, 3.0f};
  Vec3 b{3.0f, 2.0f, 1.0f};
  std::println("a + b = {}", a + b);
  std::println("dot(a, b) = {}", a.dot(b));
  std::println("normalized(a) = {}", a.normalized());
}
```

--------------------------------

### POST /quaternion/from-axis-angle

Source: https://smath.slendi.dev/structsmath_1_1Quaternion.html

Constructs a new quaternion instance based on a provided rotation axis and angle.

```APIDOC
## POST /quaternion/from-axis-angle

### Description
Builds a quaternion from a 3D rotation axis and an angle. This is a static factory method used to represent rotations.

### Method
POST

### Endpoint
/quaternion/from-axis-angle

### Parameters
#### Request Body
- **axis** (Vec<3, T>) - Required - The rotation axis vector.
- **angle** (T) - Required - The rotation angle in configured units.

### Request Example
{
  "axis": [0, 1, 0],
  "angle": 1.57
}

### Response
#### Success Response (200)
- **quaternion** (Quaternion) - The resulting rotation quaternion.

#### Response Example
{
  "x": 0.0,
  "y": 0.707,
  "z": 0.0,
  "w": 0.707
}
```

--------------------------------

### Vector Normalization Utilities

Source: https://smath.slendi.dev/smath_8hpp_source.html

Provides safe and standard normalization methods to produce unit vectors. Includes checks for epsilon values to prevent division by zero.

```cpp
constexpr auto normalized_safe(U eps = EPS_DEFAULT) const noexcept -> Vec {
	auto m = magnitude();
	return (m > eps) ? (*this) / m : Vec {};
}

[[nodiscard]] constexpr auto normalized() noexcept -> Vec<N, T>
requires std::is_floating_point_v<T> {
	return (*this) / this->magnitude();
}
```

--------------------------------

### Vec Constructors

Source: https://smath.slendi.dev/structsmath_1_1Vec.html

Documentation for the various constructors of the Vec class.

```APIDOC
## Constructors

### Vec() [1/3]

```cpp
template<std::size_t N, typename T = float>
constexpr smath::Vec< N, T >::Vec() noexcept
```

**Description:** Constructs a zero-initialized vector.

### Vec() [2/3]

```cpp
template<std::size_t N, typename T = float>
constexpr smath::Vec< N, T >::Vec(T const &s) noexcept
```

**Description:** Fills all components with the same scalar value.

**Parameters:**
- **s** (T): Scalar value assigned to every component.

### Vec() [3/3]

```cpp
template<std::size_t N, typename T = float>
template<typename... Args>
requires ((detail::is_scalar_v<Args> || detail::is_Vec_v<Args>) && ...) && (total_extent<Args...>() == N) && (!(sizeof...(Args) == 1 && (detail::is_Vec_v<Args> && ...)))
constexpr smath::Vec< N, T >::Vec(Args &&...args) noexcept
```

**Description:** Constructs a vector from a variadic list of arguments, which can be scalars or other Vec objects, ensuring the total extent matches N.
```

--------------------------------

### Quaternion Transformations

Source: https://smath.slendi.dev/smath_8hpp_source.html

Handles quaternion-based rotations and conversions. Includes methods for Hamilton product calculation and conversion to 4x4 rotation matrices.

```cpp
constexpr auto as_matrix() const noexcept -> Mat< 4, 4, T >;
constexpr auto operator*(Quaternion const &rhs) const noexcept -> Quaternion;
static constexpr auto from_axis_angle(Vec< 3, T > const &axis, T const angle) noexcept -> Quaternion;
```

--------------------------------

### Matrix Equality and Comparison

Source: https://smath.slendi.dev/structsmath_1_1Mat.html

Methods for comparing matrices, including an approximate equality check for floating-point values and standard equality/inequality operators.

```cpp
bool eq = mat.approx_equal(rhs, 1e-5f);
bool is_equal = (mat == rhs);
bool is_not_equal = (mat != rhs);
```

--------------------------------

### POST /quaternion/multiply

Source: https://smath.slendi.dev/structsmath_1_1Quaternion.html

Performs the Hamilton product between two quaternions.

```APIDOC
## POST /quaternion/multiply

### Description
Calculates the Hamilton product of two quaternions, effectively combining two rotations.

### Method
POST

### Endpoint
/quaternion/multiply

### Parameters
#### Request Body
- **lhs** (Quaternion) - Required - The left-hand quaternion.
- **rhs** (Quaternion) - Required - The right-hand quaternion.

### Request Example
{
  "lhs": {"x": 0, "y": 0, "z": 0, "w": 1},
  "rhs": {"x": 1, "y": 0, "z": 0, "w": 0}
}

### Response
#### Success Response (200)
- **result** (Quaternion) - The product of the two quaternions.

#### Response Example
{
  "x": 1, "y": 0, "z": 0, "w": 0
}
```

--------------------------------

### Compile-time Vector Swizzling

Source: https://smath.slendi.dev/smath_8hpp_source.html

Implements GLSL-style swizzling for vectors using template metaprogramming. It validates component characters and bounds at compile-time to ensure type safety.

```cpp
template<detail::FixedString S, std::size_t N, typename T>
requires detail::ValidSwizzle<S, N>
constexpr auto swizzle(Vec<N, T> const &v) -> VecOrScalar<S.size, T>
{
	return detail::swizzle_impl<S>(v, std::make_index_sequence<S.size> {});
}
```

--------------------------------

### Matrix Orthographic Projection (C++)

Source: https://smath.slendi.dev/smath_8hpp_source.html

Builds an orthographic projection matrix for 3D space. It takes parameters defining the boundaries of the view frustum (left, right, bottom, top, near, far) and an optional flag to flip the Z-axis. This is useful for creating 2D-like views within a 3D environment.

```cpp
auto matrix_ortho3d(T const left, T const right, T const bottom, T const top, T const near, T const far, bool const flip_z_axis=true) -> Mat< 4, 4, T >
```

--------------------------------

### Matrix Perspective Projection (C++)

Source: https://smath.slendi.dev/smath_8hpp_source.html

Builds a finite perspective projection matrix. This function takes the field of view angle (fovy), aspect ratio, near clipping plane distance, and far clipping plane distance as input. An optional flag can be used to flip the Z-axis. This is fundamental for creating realistic 3D camera views.

```cpp
auto matrix_perspective(T fovy, T aspect, T znear, T zfar, bool flip_z_axis=false) -> Mat< 4, 4, T >
```

--------------------------------

### smath Interop Conversion

Source: https://smath.slendi.dev/

Shows how to convert values between smath's internal representation and external types using provided adapter functions. This is useful for integrating smath with other libraries.

```cpp
auto v = smath::from_external(external_value);
auto ext = smath::to_external<ExternalType>(smath_value);
```

--------------------------------

### Vector Cross Product and Projection

Source: https://smath.slendi.dev/smath_8hpp_source.html

Advanced geometric operations including 3D cross product and vector projection onto another vector.

```cpp
constexpr auto cross(Vec const &r) const noexcept -> Vec
requires(N == 3) {
	return {
		(*this)[1] * r[2] - (*this)[2] * r[1],
		(*this)[2] * r[0] - (*this)[0] * r[2],
		(*this)[0] * r[1] - (*this)[1] * r[0],
	};
}

constexpr auto project_onto(Vec const &n) const noexcept -> Vec
requires std::is_floating_point_v<T> {
	auto d = this->dot(n);
	auto nn = n.dot(n);
	return (nn ? (d / nn) * n : Vec());
}
```

--------------------------------

### Define smath configuration macros

Source: https://smath.slendi.dev/smath_8hpp.html

Configuration macros used to define library behavior, such as the default unit for angle calculations.

```cpp
#define SMATH_ANGLE_UNIT rad
```

--------------------------------

### Data Unpacking Utilities

Source: https://smath.slendi.dev/namespacesmath.html

Functions for unpacking packed 32-bit integers into 4-component vectors.

```APIDOC
## [FUNCTION] smath::unpack_snorm4x8

### Description
Unpacks a 32-bit signed normalized integer into a 4-component vector.

### Parameters
- **packed** (std::uint32_t) - Required - The packed input value.

### Response
- **return** (Vec<4, T>) - The unpacked vector.

## [FUNCTION] smath::unpack_unorm4x8

### Description
Unpacks a 32-bit unsigned normalized integer into a 4-component vector.

### Parameters
- **packed** (std::uint32_t) - Required - The packed input value.

### Response
- **return** (Vec<4, T>) - The unpacked vector.
```

--------------------------------

### Static constants and conversion logic

Source: https://smath.slendi.dev/structsmath_1_1Vec.html

Defines the default epsilon value for floating-point comparisons and illustrates the internal logic for casting vector types.

```cpp
static constexpr T EPS_DEFAULT = T(1e-6);

// Conversion logic example
for (std::size_t i = 0; i < N; ++i) {
	this->operator[](i) = static_cast<T>(other[i]);
}
```

--------------------------------

### Utility Functions for Type Conversion in C++

Source: https://smath.slendi.dev/smath_8hpp_source.html

Provides utility functions for converting between different numeric types, specifically for normalized and signed normalized 8-bit integers. These functions handle the scaling and casting required for accurate conversions to floating-point types.

```cpp
template<class T> constexpr auto unpack_unorm8(std::uint8_t b) -> T
{
    static_assert(std::is_floating_point_v<T>);
    return static_cast<T>(b) / T(255);
}

template<class T> constexpr auto unpack_snorm8(std::int8_t b) -> T
{
    static_assert(std::is_floating_point_v<T>);
    int i = static_cast<int>(b);
    if (i < -127)
        i = -127;
    if (i > 127)
        i = 127;
    return static_cast<T>(i) / T(127);
}
```

--------------------------------

### Data Packing/Unpacking (C++)

Source: https://smath.slendi.dev/smath_8hpp_source.html

Includes functions for packing and unpacking normalized 4-component vectors into and from a single 32-bit unsigned integer. These are often used for efficient storage or transmission of color or other normalized data.

```cpp
constexpr auto pack_unorm4x8(Vec< 4, T > const &v) -> std::uint32_t
```

```cpp
constexpr auto unpack_unorm4x8(std::uint32_t packed) -> Vec< 4, T >
```

```cpp
constexpr auto pack_snorm4x8(Vec< 4, T > const &v) -> std::uint32_t
```

```cpp
constexpr auto unpack_snorm4x8(std::uint32_t packed) -> Vec< 4, T >
```

--------------------------------

### Vec Constructor: Variadic Arguments

Source: https://smath.slendi.dev/structsmath_1_1Vec.html

Constructs a vector using a variadic number of arguments. These arguments can be scalars or other Vec objects, provided they meet specific type and size constraints. This constructor is designed for flexible initialization.

```cpp
template<std::size_t N, typename T = float>
template<typename... Args>
requires ((detail::is_scalar_v<Args> || detail::is_Vec_v<Args>) && ...) && (total_extent<Args...>() == N) && (!(sizeof...(Args) == 1 && (detail::is_Vec_v<Args> && ...)))
constexpr smath::Vec< N, T >::Vec(Args &&...args) noexcept;
```

--------------------------------

### Interop Adapter (C++)

Source: https://smath.slendi.dev/smath_8hpp_source.html

Defines a template struct `interop_adapter` which facilitates conversion between external types (`Ext`) and `smath` types. It provides `to_smath` and `from_smath` functions for bidirectional type conversion, enabling seamless integration with other libraries or systems.

```cpp
template<class Ext>
struct interop_adapter;

template<class Ext>
constexpr auto from_external(Ext const &ext) -> typename interop_adapter<Ext>::smath_type
{
  return interop_adapter<Ext>::to_smath(ext);
}

template<class Ext, class SMathT>
constexpr auto to_external(SMathT const &value) -> Ext
{
  return interop_adapter<Ext>::from_smath(value);
}
```

--------------------------------

### 3D Projection Matrix Generation (C++)

Source: https://smath.slendi.dev/namespacesmath.html

Functions to generate common 3D projection matrices: orthographic, perspective, and look-at view matrices. These are fundamental for setting up the camera and viewing frustum in 3D rendering pipelines. They support customizable parameters like field of view, aspect ratio, and near/far clipping planes.

```cpp
template<typename T >
auto | matrix_ortho3d (T const left, T const right, T const bottom, T const top, T const near, T const far, bool const flip_z_axis=true) -> Mat< 4, 4, T >
| Builds an orthographic projection matrix.   

template<typename T >
auto | matrix_perspective (T fovy, T aspect, T znear, T zfar, bool flip_z_axis=false) -> Mat< 4, 4, T >
| Builds a finite perspective projection matrix.   

template<typename T >
auto | matrix_look_at (Vec< 3, T > const eye, Vec< 3, T > const center, Vec< 3, T > const up) -> Mat< 4, 4, T >
| Builds a right-handed look-at view matrix.   

template<typename T >
auto | matrix_infinite_perspective (T const fovy, T const aspect, T const znear, bool flip_z_axis=false) -> Mat< 4, 4, T >
| Builds a perspective matrix with an infinite far plane.
```

--------------------------------

### 2D Scaling of a 3x3 Matrix (C++)

Source: https://smath.slendi.dev/smath_8hpp_source.html

Applies a 2D scaling transformation to a 3x3 matrix using a 2D vector for scaling factors.

```cpp
template<typename T>
[[nodiscard]] inline auto scale(Mat<3, 3, T> const &m, Vec<2, T> const &v)
-> Mat<3, 3, T> {
Mat<3, 3, T> res;

res[0] = m[0] * v[0];
res[1] = m[1] * v[1];
res[2] = m[2];

return res;
}
```

--------------------------------

### Matrix Infinite Perspective Projection (C++)

Source: https://smath.slendi.dev/smath_8hpp_source.html

Constructs a perspective projection matrix with an infinite far plane. This is useful in scenarios where the far clipping plane is not a concern or needs to extend indefinitely. It requires the field of view angle (fovy), aspect ratio, and the near clipping plane distance.

```cpp
auto matrix_infinite_perspective(T const fovy, T const aspect, T const znear, bool flip_z_axis=false) -> Mat< 4, 4, T >
```

--------------------------------

### Normalization Packing

Source: https://smath.slendi.dev/smath_8hpp_source.html

Utility functions to pack 4-component floating-point vectors into a 32-bit unsigned integer using unsigned or signed normalization (unorm/snorm).

```cpp
template<class T> requires std::is_floating_point_v<T>
[[nodiscard]] constexpr auto pack_unorm4x8(Vec<4, T> const &v) -> std::uint32_t {
	std::uint32_t r = detail::pack_unorm8(v[0]);
	std::uint32_t g = detail::pack_unorm8(v[1]);
	std::uint32_t b = detail::pack_unorm8(v[2]);
	std::uint32_t a = detail::pack_unorm8(v[3]);
	return (r) | (g << 8) | (b << 16) | (a << 24);
}

template<class T> requires std::is_floating_point_v<T>
[[nodiscard]] constexpr auto pack_snorm4x8(Vec<4, T> const &v) -> std::uint32_t {
	std::uint32_t r = static_cast<std::uint8_t>(detail::pack_snorm8(v[0]));
	std::uint32_t g = static_cast<std::uint8_t>(detail::pack_snorm8(v[1]));
	std::uint32_t b = static_cast<std::uint8_t>(detail::pack_snorm8(v[2]));
	std::uint32_t a = static_cast<std::uint8_t>(detail::pack_snorm8(v[3]));
	return (r) | (g << 8) | (b << 16) | (a << 24);
}
```

--------------------------------

### Affine Transformations (Shear, Translate, Rotate, Scale)

Source: https://smath.slendi.dev/smath_8hpp_source.html

Functions to apply geometric transformations to matrices. These include shearing along axes, translation, rotation, and scaling, typically used in 3D graphics pipelines.

```cpp
template<typename T>
[[nodiscard]] inline auto shear_y(Mat<3, 3, T> const &m, T const v) -> Mat<3, 3, T> {
	Mat<3, 3, T> res { 1 };
	res[0][1] = v;
	return m * res;
}

template<typename T>
[[nodiscard]] inline auto translate(Mat<4, 4, T> const &m, Vec<3, T> const &v) -> Mat<4, 4, T> {
	Mat<4, 4, T> res { m };
	res[3] = m[0] * v[0] + m[1] * v[1] + m[2] * v[2] + m[3];
	return res;
}

template<typename T>
[[nodiscard]] inline auto rotate(Mat<4, 4, T> const &m, T const angle) -> Mat<4, 4, T> {
	Mat<4, 4, T> res;
	T const c { std::cos(angle) };
	T const s { std::sin(angle) };
	res[0] = m[0] * c + m[1] * s;
	res[1] = m[0] * -s + m[1] * c;
	res[2] = m[2];
	res[3] = m[3];
	return res;
}

template<typename T>
[[nodiscard]] inline auto scale(Mat<4, 4, T> const &m, Vec<3, T> const &v) -> Mat<4, 4, T> {
	Mat<4, 4, T> res;
	res[0] = m[0] * v[0];
	res[1] = m[1] * v[1];
	res[2] = m[2] * v[2];
	res[3] = m[3];
	return res;
}
```

--------------------------------

### Identity Matrix Creation (C++)

Source: https://smath.slendi.dev/smath_8hpp_source.html

Creates an identity matrix of the same dimensions as the current matrix type. Requires the matrix to be square (R == C).

```cpp
[[nodiscard]] static constexpr auto identity() noexcept -> Mat<R, C, T>
requires(R == C) {
Mat<R, C, T> m {};
for (std::size_t i = 0; i < R; ++i)
m(i, i) = T(1);
return m;
}
```

--------------------------------

### String Macro Definitions for C++

Source: https://smath.slendi.dev/smath_8hpp_source.html

These macros define stringification utilities in C++. SMATH_STR converts a macro argument into a string literal, and SMATH_XSTR expands its argument before stringifying it. These are useful for debugging and compile-time string manipulation.

```cpp
#define SMATH_STR(x) #x
#define SMATH_XSTR(x) SMATH_STR(x)
```

--------------------------------

### Matrix Structure and Arithmetic

Source: https://smath.slendi.dev/smath_8hpp_source.html

A template-based Matrix class providing column-major storage, element access via operator() and operator[], and standard arithmetic operations including addition, subtraction, and scalar multiplication.

```cpp
template<std::size_t R, std::size_t C, typename T = float>
requires std::is_arithmetic_v<T>
struct Mat : std::array<Vec<R, T>, C>
{
	using Base = std::array<Vec<R, T>, C>;
	constexpr auto operator[](std::size_t const row, std::size_t const column) -> T & { return col(column)[row]; }
	constexpr auto operator[](std::size_t const row, std::size_t const column) const -> T const & { return col(column)[row]; }
	constexpr Mat() noexcept { for (auto &col : *this) col = Vec<R, T> {}; }
	constexpr explicit Mat(T const &diag) noexcept requires(R == C) { for (std::size_t c = 0; c < C; ++c) { (*this)[c] = Vec<R, T> {}; (*this)[c][c] = diag; } }
	constexpr auto col(std::size_t j) noexcept -> Vec<R, T> & { return (*this)[j]; }
	constexpr auto operator()(std::size_t row, std::size_t col) noexcept -> T & { return (*this)[col][row]; }
	constexpr auto operator-() const noexcept -> Mat { Mat r {}; for (std::size_t c = 0; c < C; ++c) r[c] = -(*this)[c]; return r; }
	constexpr auto operator+=(Mat const &rhs) noexcept -> Mat & { for (std::size_t c = 0; c < C; ++c) (*this)[c] += rhs[c]; return *this; }
	constexpr auto operator*=(T const &s) noexcept -> Mat & { for (std::size_t c = 0; c < C; ++c) (*this)[c] *= s; return *this; }
};
```

--------------------------------

### Pack Vector Data

Source: https://smath.slendi.dev/namespacesmath.html

Utility functions to pack 4-component vectors into 32-bit unsigned integers using signed or unsigned normalized formats. Requires floating-point input types.

```cpp
template<class T>
requires std::is_floating_point_v<T>
constexpr std::uint32_t pack_snorm4x8(Vec<4, T> const& v);

template<class T>
requires std::is_floating_point_v<T>
constexpr std::uint32_t pack_unorm4x8(Vec<4, T> const& v);
```

--------------------------------

### Include standard library dependencies for smath

Source: https://smath.slendi.dev/smath_8hpp.html

Essential C++ standard library headers required for the smath library functionality, including support for algorithms, arrays, math functions, and type traits.

```cpp
#include <algorithm>
#include <array>
#include <cmath>
#include <cstddef>
#include <cstdint>
#include <format>
#include <numbers>
#include <optional>
#include <type_traits>
```

--------------------------------

### Vec Class Constructors in C++

Source: https://smath.slendi.dev/smath_8hpp_source.html

Defines constructors for the Vec class, including default, scalar, and variadic template constructors. The variadic constructor handles unpacking scalar values and other Vec instances.

```cpp
template<std::size_t N, typename T = float>
requires std::is_arithmetic_v<T>
struct Vec : std::array<T, N>
{
    // Constructors
    constexpr Vec() noexcept
    {
        for (auto &v : *this)
            v = T(0);
    }
    explicit constexpr Vec(T const &s) noexcept
    {
        for (auto &v : *this)
            v = s;
    }

    template<typename... Args>
    requires((detail::is_scalar_v<Args> || detail::is_Vec_v<Args>) && ...)
    && (total_extent<Args...>() == N)
    && (!(sizeof...(Args) == 1 && (detail::is_Vec_v<Args> && ...)))
    constexpr Vec(Args &&...args) noexcept
    {
        std::size_t i = 0;
        (fill_one(i, std::forward<Args>(args)), ...);
    }
};
```

--------------------------------

### Matrix Transposition (C++)

Source: https://smath.slendi.dev/smath_8hpp_source.html

Returns the transpose of a matrix. The transpose is created by swapping the rows and columns of the original matrix.

```cpp
[[nodiscard]] constexpr auto transposed() const noexcept -> Mat<C, R, T> {
Mat<C, R, T> r {};
for (std::size_t c = 0; c < C; ++c)
for (std::size_t r_idx = 0; r_idx < R; ++r_idx)
r(r_idx, c) = (*this)(c, r_idx);
return r;
}
```

--------------------------------

### Vector Formatting (C++)

Source: https://smath.slendi.dev/smath_8hpp_source.html

Defines a `std::formatter` specialization for `smath::Vec<N, T>`. This allows `smath::Vec` objects to be directly formatted using C++'s `<format>` library, outputting them in a readable `{x, y, z, ...}` format.

```cpp
template<std::size_t N, typename T>
requires std::formattable<T, char>
struct std::formatter<smath::Vec<N, T>> : std::formatter<T>
{
  constexpr auto parse(std::format_parse_context &ctx)
  {
    return std::formatter<T>::parse(ctx);
  }

  template<typename Ctx>
  auto format(smath::Vec<N, T> const &v, Ctx &ctx) const
  {
    auto out = ctx.out();
    *out++ = '{';
    for (std::size_t i = 0; i < N; ++i) {
      if (i) {
        *out++ = ',';
        *out++ = ' ';
      }
      out = std::formatter<T>::format(v[i], ctx);
    }
    *out++ = '}';
    return out;
  }
};
```

--------------------------------

### Identity Matrix Generation

Source: https://smath.slendi.dev/structsmath_1_1Mat.html

Static factory method to generate an identity matrix. Requires the matrix to be square (R == C).

```cpp
auto identity = smath::Mat<3, 3, float>::identity();
```

--------------------------------

### Component Unpacking

Source: https://smath.slendi.dev/structsmath_1_1Vec.html

Utility method to unpack vector components into a variadic list of output references. Useful for structured binding or passing vector data to legacy APIs.

```cpp
template<class... Args>
constexpr auto unpack(Args&... args) -> void;
```

--------------------------------

### Packed Float-to-Uint Conversion (C++)

Source: https://smath.slendi.dev/namespacesmath.html

Functions for packing and unpacking 4-component floating-point vectors into a single 32-bit unsigned integer. `pack_unorm4x8` and `pack_snorm4x8` convert normalized or signed normalized floats to uint8 components, while `unpack_unorm4x8` and `unpack_snorm4x8` perform the reverse operation. Requires `std::is_floating_point_v<T>`.

```cpp
template<class T > 
requires std::is_floating_point_v<T>
constexpr auto | pack_unorm4x8 (Vec< 4, T > const &v) -> std::uint32_t
template<class T > 
requires std::is_floating_point_v<T>
constexpr auto | pack_snorm4x8 (Vec< 4, T > const &v) -> std::uint32_t
template<class T = float> 
requires std::is_floating_point_v<T>
constexpr auto | unpack_unorm4x8 (std::uint32_t packed) -> Vec< 4, T >
template<class T = float> 
requires std::is_floating_point_v<T>
constexpr auto | unpack_snorm4x8 (std::uint32_t packed) -> Vec< 4, T >
```

--------------------------------

### Look-At Matrix (C++)

Source: https://smath.slendi.dev/smath_8hpp_source.html

Builds a right-handed look-at view matrix. This function defines the orientation of a camera in 3D space by specifying its position (eye), the point it is looking at (center), and the direction considered 'up'. This is crucial for setting up camera transformations in 3D rendering.

```cpp
auto matrix_look_at(Vec< 3, T > const eye, Vec< 3, T > const center, Vec< 3, T > const up) -> Mat< 4, 4, T >
```

--------------------------------

### Matrix Inequality Comparison (C++)

Source: https://smath.slendi.dev/smath_8hpp_source.html

Implements the inequality operator (!=) by negating the result of the equality operator (==). This provides a concise way to check if two matrices are not equal.

```cpp
[[nodiscard]] constexpr auto operator!=(Mat const &rhs) const noexcept -> bool {
return !(*this == rhs);
}
```

--------------------------------

### Matrix Arithmetic Operator Overloads

Source: https://smath.slendi.dev/structsmath_1_1Mat.html

Provides friend operator overloads for matrix-scalar and matrix-matrix arithmetic. These allow for intuitive syntax when performing linear algebra operations.

```cpp
template<std::size_t R, std::size_t C, typename T = float>
friend constexpr auto operator*(Mat<R, C, T> lhs, T const &s) -> Mat;

template<std::size_t R, std::size_t C, typename T = float>
friend constexpr auto operator*(T const &s, Mat<R, C, T> rhs) -> Mat;

template<std::size_t R, std::size_t C, typename T = float>
friend constexpr auto operator+(Mat<R, C, T> lhs, Mat<R, C, T> const &rhs) -> Mat;

template<std::size_t R, std::size_t C, typename T = float>
friend constexpr auto operator-(Mat<R, C, T> lhs, Mat<R, C, T> const &rhs) -> Mat;

template<std::size_t R, std::size_t C, typename T = float>
friend constexpr auto operator/(Mat<R, C, T> lhs, T const &s) -> Mat;
```

--------------------------------

### C++ Matrix Class Definition and Operators

Source: https://smath.slendi.dev/structsmath_1_1Mat.html

Defines the `Mat` class template for matrices with R rows, C columns, and type T. It includes constructors for zero and diagonal matrices, member functions for accessing columns and elements, and overloaded operators for arithmetic operations, equality comparison, and approximate equality checks. Dependencies include `Vec` and `std::array`.

```cpp
#include <smath.hpp>

template<std::size_t R, std::size_t C, typename T = float>
class Mat {
public:
    using Base = std::array<Vec<R, T>, C>;

    constexpr Mat() noexcept;
    constexpr Mat(T const &diag) noexcept;

    template<typename... Cols>
    requires (sizeof...(Cols) == C && (std::same_as<std::remove_cvref_t<Cols>, Vec<R, T>> && ...))
    constexpr Mat(Cols const &...cols) noexcept;

    constexpr auto col(std::size_t j) noexcept -> Vec<R, T> &;
    constexpr auto col(std::size_t j) const noexcept -> Vec<R, T> const &;

    constexpr auto operator[](std::size_t const row, std::size_t const column) -> T &;
    constexpr auto operator[](std::size_t const row, std::size_t const column) const -> T const &;
    constexpr auto operator()(std::size_t row, std::size_t col) noexcept -> T &;
    constexpr auto operator()(std::size_t row, std::size_t col) const noexcept -> T const &;

    constexpr auto operator-() const noexcept -> Mat;
    constexpr auto operator+=(Mat const &rhs) noexcept -> Mat &;
    constexpr auto operator-=(Mat const &rhs) noexcept -> Mat &;
    constexpr auto operator*=(T const &s) noexcept -> Mat &;
    constexpr auto operator/=(T const &s) noexcept -> Mat &;

    constexpr auto operator==(Mat const &rhs) const noexcept -> bool;
    constexpr auto operator!=(Mat const &rhs) const noexcept -> bool;

    template<class U = T>
    requires std::is_floating_point_v<U>
    constexpr auto approx_equal(Mat const &rhs, U eps = EPS_DEFAULT) const noexcept -> bool;

    constexpr auto transposed() const noexcept -> Mat<C, R, T>;

    static constexpr auto identity() noexcept -> Mat<R, C, T> requires(R == C);

    static constexpr T EPS_DEFAULT = T(1e-6);
};

constexpr auto operator+(Mat lhs, Mat const &rhs) noexcept -> Mat;
constexpr auto operator-(Mat lhs, Mat const &rhs) noexcept -> Mat;
constexpr auto operator*(Mat lhs, T const &s) noexcept -> Mat;
constexpr auto operator*(T const &s, Mat rhs) noexcept -> Mat;
constexpr auto operator/(Mat lhs, T const &s) noexcept -> Mat;

```

--------------------------------

### Vector Geometric Calculations

Source: https://smath.slendi.dev/smath_8hpp_source.html

Functions to calculate magnitude, length, and dot products. Includes requirements for floating-point types to ensure mathematical correctness.

```cpp
constexpr auto magnitude() const noexcept -> T
requires std::is_floating_point_v<T> {
	T total = 0;
	for (auto const &v : *this)
		total += v * v;
	return std::sqrt(total);
}

[[nodiscard]] constexpr auto dot(Vec<N, T> const &other) const noexcept -> T {
	T res = 0;
	for (std::size_t i = 0; i < N; ++i) {
		res += (*this)[i] * other[i];
	}
	return res;
}
```

--------------------------------

### 2D Transformation Matrices (C++)

Source: https://smath.slendi.dev/namespacesmath.html

Functions for creating and manipulating 3x3 transformation matrices for 2D graphics. This includes functions to apply translations, rotations, scaling, and shearing to existing matrices, as well as functions to create new transformation matrices from scratch.

```cpp
template<typename T >
auto | translate (Mat< 3, 3, T > const &m, Vec< 2, T > const &v) -> Mat< 3, 3, T >
| Applies translation to an existing 3x3 transform.   

template<typename T >
auto | translate (Vec< 2, T > const &v) -> Mat< 3, 3, T >
| Creates a 3x3 translation matrix.   

template<typename T >
auto | rotate (Mat< 3, 3, T > const &m, T const angle) -> Mat< 3, 3, T >
| Applies rotation to an existing 3x3 transform.   

template<typename T >
auto | scale (Mat< 3, 3, T > const &m, Vec< 2, T > const &v) -> Mat< 3, 3, T >
| Applies scaling to an existing 3x3 transform.   

template<typename T >
auto | shear_x (Mat< 3, 3, T > const &m, T const v) -> Mat< 3, 3, T >
| Applies X shear to an existing 3x3 transform.   

template<typename T >
auto | shear_y (Mat< 3, 3, T > const &m, T const v) -> Mat< 3, 3, T >
| Applies Y shear to an existing 3x3 transform.
```

--------------------------------

### Angle Unit Conversion (C++)

Source: https://smath.slendi.dev/smath_8hpp_source.html

Provides utility functions to convert angle values between different units: radians, degrees, and turns. This allows for flexible handling of angular data, accommodating various input formats and internal representations.

```cpp
constexpr auto rad(T value)
```

```cpp
constexpr auto deg(T value)
```

```cpp
constexpr auto turns(T value)
```

--------------------------------

### Create 3x3 Translation Matrix (C++)

Source: https://smath.slendi.dev/smath_8hpp_source.html

Creates a 3x3 translation matrix from a 2D vector. This matrix can then be used to apply translation to other transformations.

```cpp
template<typename T>
[[nodiscard]] inline auto translate(Vec<2, T> const &v) -> Mat<3, 3, T> {
Mat<3, 3, T> res { 1 };
res[2].x() = v.x();
res[2].y() = v.y();
return res;
}
```

--------------------------------

### Geometric Vector Operations in C++

Source: https://smath.slendi.dev/structsmath_1_1Vec.html

Core geometric methods for the smath::Vec class, including cross product, dot product, magnitude, and projection. These methods are marked as constexpr and noexcept for high-performance compile-time evaluation.

```cpp
template<std::size_t N, typename T = float>
constexpr auto cross(Vec<N, T> const& r) const;

template<std::size_t N, typename T = float>
constexpr auto dot(Vec<N, T> const& other) const;

template<std::size_t N, typename T = float>
constexpr auto magnitude() const;

template<std::size_t N, typename T = float>
constexpr auto project_onto(Vec<N, T> const& n) const;
```

--------------------------------

### Vector Geometric Operations

Source: https://smath.slendi.dev/smath_8hpp_source.html

Core geometric functions including magnitude, normalization, dot product, and vector projection. Includes safe variants to handle zero-length vectors.

```cpp
constexpr auto magnitude() const noexcept -> T;
constexpr auto normalized() noexcept -> Vec< N, T >;
constexpr auto normalize_safe(U eps=EPS_DEFAULT) const noexcept -> Vec;
constexpr auto dot(Vec< N, T > const &other) const noexcept -> T;
constexpr auto project_onto(Vec const &n) const noexcept -> Vec;
constexpr auto cross(Vec const &r) const noexcept -> Vec;
```

--------------------------------

### Vec Constructor: Scalar Initialization

Source: https://smath.slendi.dev/structsmath_1_1Vec.html

Constructs a vector of size N with elements of type T, where all components are initialized to the same scalar value. This constructor takes a constant reference to a scalar value.

```cpp
template<std::size_t N, typename T = float>
constexpr smath::Vec< N, T >::Vec(T const &s) noexcept;
```

--------------------------------

### Generate Projection Matrices

Source: https://smath.slendi.dev/namespacesmath.html

Functions for creating standard 3D graphics projection matrices, including perspective and orthographic variants. These return 4x4 matrices based on input parameters like field of view, aspect ratio, and clipping planes.

```cpp
template<typename T>
auto matrix_perspective(T fovy, T aspect, T znear, T zfar, bool flip_z_axis = false) -> Mat<4, 4, T>;

template<typename T>
auto matrix_ortho3d(T left, T right, T bottom, T top, T near, T far, bool flip_z_axis = true) -> Mat<4, 4, T>;
```

--------------------------------

### Matrix Transformations (C++)

Source: https://smath.slendi.dev/smath_8hpp_source.html

Provides functions for applying common transformations to 3x3 and 4x4 matrices, including translation, scaling, rotation, and shearing along the X, Y, and Z axes. These operations are fundamental for manipulating objects in 2D and 3D graphics.

```cpp
auto translate(Mat< 3, 3, T > const &m, Vec< 2, T > const &v) -> Mat< 3, 3, T >
```

```cpp
auto scale(Mat< 3, 3, T > const &m, Vec< 2, T > const &v) -> Mat< 3, 3, T >
```

```cpp
auto rotate(Mat< 3, 3, T > const &m, T const angle) -> Mat< 3, 3, T >
```

```cpp
auto shear_x(Mat< 3, 3, T > const &m, T const v) -> Mat< 3, 3, T >
```

```cpp
auto shear_y(Mat< 3, 3, T > const &m, T const v) -> Mat< 3, 3, T >
```

```cpp
auto shear_z(Mat< 4, 4, T > const &m, T const v) -> Mat< 4, 4, T >
```

--------------------------------

### Quaternion Vector Conversion

Source: https://smath.slendi.dev/structsmath_1_1Quaternion.html

Method to retrieve the vector representation of the quaternion.

```APIDOC
## [GET] smath::Quaternion::vec()

### Description
Returns the quaternion represented as its vector base type.

### Method
GET

### Response
#### Success Response (200)
- **return** (Base) - The vector base type representation of the quaternion.
```