#pragma once
#include <string_view>
#include <cstring>
#include <cstdio>
#include <cstdint>
#ifdef __DEBUG
#include <stdexcept>
#endif

namespace asp {
	template<typename T>
	struct Array {
		T* data = nullptr;
		size_t length = 0;
		size_t* refcount = nullptr;

		constexpr Array(void) noexcept = delete;
		constexpr Array(const size_t& size) noexcept : data(new T[size]), length(size), refcount(new size_t(1)) {
#ifdef __DEBUG
			printf("Creating array of size %lu\n", size);
#endif
		}

		constexpr Array(const std::initializer_list<size_t>& init_list) noexcept : data(new T[init_list.size()]), length(init_list.size()), refcount(new size_t(1)) {
#ifdef __DEBUG
			printf("Copying initializer_list of size %lu\n", init_list.size());
#endif
			memcpy(data, init_list.begin(), length * sizeof(T));
		}


		constexpr Array(const Array<T>& other) noexcept : Array<T>(other.length) {
#ifdef __DEBUG
			printf("Copying array of size %lu\n", other.length);
#endif
			memcpy(data, other.data, length * sizeof(T));
		}

		constexpr Array(Array&& other) noexcept {
#ifdef __DEBUG
			printf("Moving array of size %lu\n", other.length);
#endif
			if (refcount != nullptr){
				delete[] data;
				delete refcount;
			}
			data = other.data;
			length = other.length;
			refcount = other.refcount;
			other.data = nullptr;
			other.length = 0;
			other.refcount = nullptr;
		}

		inline ~Array() noexcept {
			if (refcount == nullptr)
				return;
#ifdef __DEBUG
			printf("Destructing array of size %lu and refcount %lu\n", length, *refcount);
#endif

			if (--(*refcount) == 0){
#ifdef __DEBUG
				printf("Freeing array of size %lu\n", length);
#endif
				delete[] data;
				data = nullptr;
				delete refcount;
				refcount = nullptr;
			}
		}

		constexpr Array& operator=(const Array& other) noexcept {
			if (this != &other) {
				if (--(*refcount) == 0) {
					delete[] data;
					delete refcount;
				}
				data = other.data;
				length = other.length;
				refcount = other.refcount;
				++(*refcount);
			}
			return *this;
		}

		constexpr T& operator[](const size_t& index) {
#ifdef __DEBUG
			if (index > length) {
				fprintf(stderr, "Index %ld out of range in Array of length %ld !\n", index, length);
				throw std::out_of_range("Index out of range !");
			}
#endif
			return data[index];
		}

		constexpr const T& operator[](const size_t& index) const {
#ifdef __DEBUG
			if (index > length) {
				fprintf(stderr, "Index %ld out of range in Array of length %ld !\n", index, length);
				throw std::out_of_range("Index out of range !");
			}
#endif
			return data[index];
		}
	};

	template<typename T>
	constexpr T max(const Array<T>& a) noexcept {
		T max_el = a[0];
		for (size_t i = 1; i < a.length; ++i)
			if (a[i] > max_el)
				max_el = a[i];
		return max_el;
	}

	template<typename T>
	constexpr T& min(const Array<T>& a) noexcept {
		T& max_el = a[0];
		for (size_t i = 1; i < a.length; ++i)
			if (a[i] < max_el)
				max_el = a[i];
		return max_el;
	}

	template<typename T, typename F>
	constexpr Array<T>& map(Array<T>& a, const F& fnc) noexcept {
		for (size_t i = 0; i < a.length; ++i)
			a[i] = fnc(i, a[i]);
		return a;
	}

	template<typename T, typename F>
	constexpr void foreach(const Array<T>& a, const F& fnc) noexcept {
		for (size_t i = 0; i < a.length; ++i)
			fnc(i, a[i]);
	}

	inline Array<size_t> range(const size_t& n) noexcept {
		Array<size_t> a(n);
		return map(a, [](const size_t i, const size_t) -> size_t {
			return i;
		});
	}

	template<typename T>
	constexpr void bubble_sort(Array<T>& a) noexcept {
		size_t j = 0;
		for (size_t i = 0; i < a.length; ++i) {
			bool swapped = false;
			for (j = i + 1; j < a.length; ++j)
				if (a[i] > a[j]) {
					std::swap(a[i], a[j]);
					swapped = true;
				}
			if (!swapped)
				break;
		}
	}

	template<typename T>
	inline Array<size_t> bubble_sort_arg(Array<T>& a) noexcept {
		Array<size_t> indices = range(a.length);
		size_t j;
		for (size_t i = 0; i < a.length; ++i) {
			bool swapped = false;
			for (j = i + 1; j < a.length; ++j)
				if (a[i] > a[j]) {
					std::swap(indices[i], indices[j]);
					std::swap(a[i], a[j]);
					swapped = true;
				}
			if (!swapped)
				break;
		}
		return indices;
	}

	template<typename T>
	constexpr size_t qs_partition(Array<T>& a, const size_t& l, const size_t& h) noexcept {
		size_t i = l - 1;
		for (size_t j = l; j <= h; ++j)
			if (a[j] < a[h])
				std::swap(a[++i], a[j]);
		std::swap(a[++i], a[h]);
		return i;
	}

	template<typename T>
	constexpr void quicksort(Array<T>& a, const size_t& l, const size_t& h) noexcept {
		if (l >= h)
			return;

		const size_t p = qs_partition(a, l, h);
		if (p - 1 <= h)
			quicksort(a, l, p - 1);
		quicksort(a, p + 1, h);
	}

	template<typename T>
	constexpr void quicksort(Array<T>& a) noexcept {
		quicksort(a, 0, a.length - 1);
	}

	template<typename T>
	constexpr void quicksort_iter(Array<T>& a, const size_t& l, const size_t& h) noexcept {
		// Create an auxiliary stack

		const size_t total = h - l + 1;
		// push initial values of l and h to stack
		size_t* const stack = new size_t[total]{l, h};

		// initialize top of stack
		size_t top = 1;

		size_t low = l, high = h;

		// Keep popping from stack while is not empty
		while (top <= total) {
			// Pop h and l
			high = stack[top--];
			low = stack[top--];
			if(low >= high)
				break;

			// Set pivot element at its correct position
			// in sorted array
			const size_t p = qs_partition(a, low, high);

			// If there are elements on left side of pivot,
			// then push left side to stack
			if (p - 1 > low && p - 1 < total) {
				stack[++top] = low;
				stack[++top] = p - 1;
			}

			// If there are elements on right side of pivot,
			// then push right side to stack
			if (p + 1 < high) {
				stack[++top] = p + 1;
				stack[++top] = high;
			}
		}

		delete[] stack;
	}

	template<typename T>
	constexpr void quicksort_iter(Array<T>& a) noexcept {
		quicksort_iter(a, 0, a.length - 1);
	}

	template<typename T>
	constexpr size_t qs_arg_partition(Array<T>& a, Array<size_t>& indices, const size_t& l, const size_t& h) noexcept {
		size_t i = l - 1;
		for (size_t j = l; j <= h; ++j)
			if (a[j] < a[h]){
				std::swap(a[++i], a[j]);
				std::swap(indices[i], indices[j]);
			}
		std::swap(indices[++i], indices[h]);
		std::swap(a[i], a[h]);
		return i;
	}

	template<typename T>
	constexpr void quicksort_arg(Array<T>& a, Array<size_t>& indices, const size_t& l, const size_t& h) noexcept {
		if (l >= h)
			return;

		const size_t p = qs_arg_partition(a, indices, l, h);
		if (p - 1 <= h)
			quicksort_arg(a, indices, l, p - 1);
		quicksort_arg(a, indices, p + 1, h);
	}

	template<typename T>
	inline Array<size_t> quicksort_arg(Array<T>& other, const size_t& l, const size_t& h) noexcept {
		Array<size_t> indices = range(other.length);
		quicksort_arg(other, indices, l, h);
		return indices;
	}

	template<typename T>
	inline Array<size_t> quicksort_arg(Array<T>& a) noexcept {
		return quicksort_arg(a, 0, a.length - 1);
	}

	template<typename T>
	inline void quicksort_arg_iter(Array<T>& a, Array<size_t>& indices, const size_t& l, const size_t& h) noexcept {
		// Create an auxiliary stack
		const size_t total = h - l + 1;
		// push initial values of l and h to stack
		size_t* const stack = new size_t[total]{l,h};

		// initialize top of stack
		size_t top = 1, low = l, high = h;

		// Keep popping from stack while is not empty
		while (top <= total) {
			// Pop h and l
			high = stack[top--];
			low = stack[top--];
			if(low >= high)
				break;

			// Set pivot element at its correct position
			// in sorted array
			const size_t p = qs_arg_partition(a, indices, low, high);

			// If there are elements on left side of pivot,
			// then push left side to stack
			if (p - 1 > low && p - 1 < total) {
				stack[++top] = low;
				stack[++top] = p - 1;
			}

			// If there are elements on right side of pivot,
			// then push right side to stack
			if (p + 1 < high) {
				stack[++top] = p + 1;
				stack[++top] = high;
			}
		}

		delete[] stack;
	}

	template<typename T>
	inline Array<size_t> quicksort_arg_iter(Array<T>& a) noexcept {
		Array<size_t> indices = range(a.length);
		quicksort_arg_iter(a, indices, 0, a.length - 1);
		return indices;
	}

	template<typename T>
	struct ArgVal {
		size_t indice;
		T val;

		constexpr ArgVal(void) noexcept = default;
		constexpr ArgVal(const size_t& _i, const T& _v) noexcept : indice(_i), val(_v) {}

		constexpr bool operator>(const ArgVal<T>& other) const noexcept {
			return val > other.val;
		}
		constexpr bool operator<(const ArgVal<T>& other) const noexcept {
			return val < other.val;
		}
		constexpr bool operator>=(const ArgVal<T>& other) const noexcept {
			return val >= other.val;
		}
		constexpr bool operator<=(const ArgVal<T>& other) const noexcept {
			return val <= other.val;
		}
	};

	template<typename T>
	constexpr void merge(Array<T>& a, const size_t& l, const size_t& m, const size_t& r) noexcept {

		Array<T> left_arr(m - l + 1);
		memcpy(&left_arr.data[0], &a[l], left_arr.length * sizeof(T));
		Array<T> right_arr(r - m);
		memcpy(&right_arr.data[0], &a[m + 1], right_arr.length * sizeof(T));

		size_t i_a0 = 0, i_a1 = 0, i = l;

		// Merge the temp arrays back size_to array[left..right]
		for (; i_a0 < left_arr.length && i_a1 < right_arr.length; ++i)
			a[i] = left_arr[i_a0] <= right_arr[i_a1] ? left_arr[i_a0++] : right_arr[i_a1++];
		//
		// Copy the remaining elements of left[], if there are any
		const size_t leftover = left_arr.length - i_a0;
		memcpy(&a[i], &left_arr[i_a0], leftover * sizeof(T));
		// Copy the remaining elements of right[], if there are any
		memcpy(&a[i + leftover], &right_arr[i_a1], (right_arr.length - i_a1) * sizeof(T));
	}

	template<typename T>
	constexpr void mergesort(Array<T>& a, const size_t& l, const size_t& r) noexcept {
		if (l >= r)
			return;

		const size_t m = l + (r - l) / 2;
		mergesort(a, l, m);
		mergesort(a, m + 1, r);
		merge(a, l, m, r);
	}

	template<typename T>
	constexpr void mergesort(Array<T>& a) noexcept {
		mergesort(a, 0, a.length - 1);
	}

	template<typename T>
	inline Array<size_t> mergesort_arg(Array<T>& a) noexcept {

		Array<ArgVal<T>> temp_vals(a.length);
		map(temp_vals, [&a](const size_t& i, const ArgVal<T>&) -> ArgVal<T> {
			return ArgVal<T>(i, a[i]);
		});

		mergesort(temp_vals, 0, a.length - 1);

		Array<size_t> indices(a.length);
		return map(indices, [&temp_vals](const size_t& i, const size_t&) -> size_t {
			return temp_vals[i].indice;
		});
	}

	template<typename T>
	constexpr void insertion_sort(Array<T>& a) noexcept {
		size_t j = 0;
		for(size_t i = 1; i < a.length; ++i) {
			T key = a[i];
			for(j = i; j > 0 && a[j - 1] > key; --j)
				a[j] = a[j - 1];
			a[j] = key;
		}
	}

	template<typename T>
	inline Array<size_t> insertion_argsort(Array<T>& a) noexcept {
		Array<ArgVal<T>> temp_vals(a.length);
		map(temp_vals, [&a](const size_t& i, const ArgVal<T>&) -> ArgVal<T> {
			return ArgVal<T>(i, a[i]);
		});

		insertion_sort(temp_vals);

		Array<size_t> indices(a.length);
		return map(indices, [&temp_vals](const size_t& i, const size_t&) -> size_t {
			return temp_vals[i].indice;
		});
	}

	template<typename T>
	inline void countsort(Array<T>& a, int64_t k, int64_t pos) noexcept {
		Array<T> output(a.length + 1);
		T max = T((a[0] / pos) % k);
		for (size_t i = 1; i < a.length; ++i) {
			const T val = a[i];
			if (((val / pos) % k) > max)
				max = val;
		}

		Array<T> count(max + 1);
		memset(count.data, 0, count.length * sizeof(T));

		for (size_t i = 0; i < a.length; ++i)
			++count[size_t((a[i] / pos) % k)];

		// Prefix sum
		for (size_t i = 1; i < count.length; i++)
			count[i] += count[i - 1];

		for (size_t i = a.length - 1; i > 0; --i) {
			const T val = a[i];
			output[count[size_t((val / pos) % k)] - 1] = val;
			--count[size_t((val / pos) % k)];
		}
		const T val = a[0];
		output[count[size_t((val / pos) % k)] - 1] = val;
		--count[size_t((val / pos) % k)];
		memmove(a.data, output.data, a.length * sizeof(T));
	}

	template<typename T>
	inline void counting_sort(Array<T>& a) noexcept {
		countsort<T>(a, max(a) + 1, 1);
	}

	template<typename T>
	constexpr void radix_sort(Array<T>& a) noexcept {
		constexpr int64_t k = 10;
		T max = a[0];
		for (size_t i = 1; i < a.length; i++)
			if (a[i] > max)
				max = a[i];
		for (int64_t pos = 1; max / pos > 0; pos *= k)
			countsort(a, k, pos);
	}
};