🛠️ Remove tag_invoke entirely

CMakeLists.txt

@@ -28,8 +28,7 @@
 #
 # ============================================================================>
 
-cmake_minimum_required(VERSION 3.21.0)
-cmake_policy(VERSION 3.21)
+cmake_minimum_required(VERSION 3.28.0)
 
 # # Project kickstart
 # Includes a bunch of basic flags and utilities shared across projects
@@ -223,9 +222,9 @@ if (ZTD_TEXT_EXAMPLES OR ZTD_TEXT_BENCHMARKS)
 			simdutf
 			GIT_REPOSITORY https://github.com/simdutf/simdutf
 			GIT_SHALLOW    ON
-			GIT_TAG        master)
+			GIT_TAG        master
+			EXCLUDE_FROM_ALL)
 		FetchContent_MakeAvailable(simdutf)
-		set_property(DIRECTORY "${simdutf_SOURCE_DIR}" PROPERTY EXCLUDE_FROM_ALL YES)
 	endfunction()
 	simdutf_dependency_jail()
 
@@ -240,13 +239,26 @@ if (ZTD_TEXT_EXAMPLES OR ZTD_TEXT_BENCHMARKS)
 			boost.text
 			GIT_REPOSITORY https://github.com/ThePhD/text
 			GIT_SHALLOW    ON
-			GIT_TAG        master)
+			GIT_TAG        master
+			EXCLUDE_FROM_ALL)
 		FetchContent_MakeAvailable(boost.text)
-		set_property(DIRECTORY "${boost.text_SOURCE_DIR}" PROPERTY EXCLUDE_FROM_ALL YES)
 	endfunction()
 	if (ZTD_TEXT_BOOST.TEXT)
 		boost_text_dependency_jail()
 	endif()
+	if (ZTD_TEXT_BENCHMARKS)
+		function(unicode_lipsum_dependency_jail)
+			# unicode_lipsum check
+			FetchContent_Declare(
+				unicode_lipsum
+				GIT_REPOSITORY https://github.com/lemire/unicode_lipsum
+				GIT_SHALLOW    ON
+				GIT_TAG        main
+				EXCLUDE_FROM_ALL)
+			FetchContent_MakeAvailable(unicode_lipsum)
+		endfunction()
+		unicode_lipsum_dependency_jail()
+	endif()
 endif()
 
 if(ZTD_TEXT_TESTS)

benchmarks/CMakeLists.txt

@@ -36,6 +36,7 @@ function (google_benchmark_dependency_jail)
 		google_benchmark
 		GIT_REPOSITORY https://github.com/google/benchmark.git
 		GIT_TAG main
+		EXCLUDE_FROM_ALL
 	)
 	# Benchmark variables, internal setup
 	set(BENCHMARK_ENABLE_EXCEPTIONS ON)
@@ -55,7 +56,6 @@ function (google_benchmark_dependency_jail)
 		set(HAVE_STD_REGEX ON)
 	endif()
 	FetchContent_MakeAvailable(google_benchmark)
-	set_property(DIRECTORY "${google_benchmark_SOURCE_DIR}" PROPERTY EXCLUDE_FROM_ALL YES)
 endfunction()
 google_benchmark_dependency_jail()
 

benchmarks/barrier/CMakeLists.txt

@@ -27,7 +27,7 @@
 #
 # =========================================================================== #
 
-# # Benchmarks
+# # Benchmarks data, compiled behind a shared data barrier
 file(GLOB_RECURSE ztd.text.benchmarks.barrier.sources
 	LIST_DIRECTORIES FALSE
 	CONFIGURE_DEPENDS

benchmarks/conversion_speed/CMakeLists.txt

@@ -94,8 +94,8 @@ function (generate_converion_speed_benchmark_targets name data_name title)
 			ZTD_TEXT_BENCHMARKS_CONVERSION_SPEED_CUNEICODE_SINGLE_FROM_BULK_BENCHMARKS=$<IF:$<BOOL:${ZTD_TEXT_BENCHMARKS_INTERNALS}>,1,0>
 			ZTD_TEXT_BENCHMARKS_CONVERSION_SPEED_CUNEICODE_BASIC_UNCHECKED_BENCHMARKS=$<IF:$<BOOL:${ZTD_TEXT_BENCHMARKS_INTERNALS}>,1,0>
 			ZTD_TEXT_BENCHMARKS_CONVERSION_SPEED_CUNEICODE_SINGLE_FROM_BULK_UNCHECKED_BENCHMARKS=$<IF:$<BOOL:${ZTD_TEXT_BENCHMARKS_INTERNALS}>,1,0>
-			ZTD_TEXT_BENCHMARKS_CONVERSION_SPEED_ICU_BENCHMARKS=1
-			ZTD_TEXT_BENCHMARKS_CONVERSION_SPEED_ICONV_BENCHMARKS=1
+			ZTD_TEXT_BENCHMARKS_CONVERSION_SPEED_ICU_BENCHMARKS=$<IF:$<BOOL:${ZTD_TEXT_ICU}>,1,0>
+			ZTD_TEXT_BENCHMARKS_CONVERSION_SPEED_ICONV_BENCHMARKS=$<IF:$<BOOL:${ZTD_TEXT_LIBICONV}>,1,0>
 			ZTD_TEXT_BENCHMARKS_CONVERSION_SPEED_ENCODING_C_BENCHMARKS=1
 			ZTD_TEXT_BENCHMARKS_CONVERSION_SPEED_CTRE_BENCHMARKS=1
 			ZTD_TEXT_BENCHMARKS_CONVERSION_SPEED_SIMDUTF_BENCHMARKS=1

benchmarks/conversion_speed/source/cuneicode.cpp

@@ -58,7 +58,7 @@
 			const ztd_char##FROM_N##_t* input = input_data.data();                                                  \
 			size_t output_size                = output_data.size();                                                 \
 			ztd_char##TO_N##_t* output        = output_data.data();                                                 \
-			cnc_mcerr err = cnc_c##FROM_N##sntoc##TO_N##sn(&output_size, &output, &input_size, &input, &cstate);    \
+			cnc_mcerr err = cnc_c##FROM_N##snrtoc##TO_N##sn(&output_size, &output, &input_size, &input, &cstate);   \
 			if (err != cnc_mcerr_ok) {                                                                              \
 				result = false;                                                                                    \
 			}                                                                                                       \

benchmarks/conversion_speed/source/standard_c++.cpp

@@ -41,6 +41,7 @@
 #include <ztd/idk/charN_t.hpp>
 #include <ztd/idk/detail/unicode.hpp>
 #include <ztd/idk/endian.hpp>
+#include <ztd/idk/mbstate_t.hpp>
 
 #include <barrier/barrier.h>
 
@@ -66,7 +67,7 @@ static void utf16_to_utf32_well_formed_standard_cpp(benchmark::State& state) {
 	to_char_t* output                   = output_data.data();
 	to_char_t* const output_last        = output_data.data() + output_data.size();
 	for (auto _ : state) {
-		std::mbstate_t state {};
+		ztd_mbstate_t state {};
 		const char* char_input      = reinterpret_cast<const char*>(input);
 		const char* char_input_last = reinterpret_cast<const char*>(input_last);
 		const char* char_input_next;
@@ -110,7 +111,7 @@ static void utf32_to_utf16_well_formed_standard_cpp(benchmark::State& state) {
 	to_char_t* output                   = output_data.data();
 	to_char_t* const output_last        = output_data.data() + output_data.size();
 	for (auto _ : state) {
-		std::mbstate_t state {};
+		ztd_mbstate_t state {};
 		char* char_output      = reinterpret_cast<char*>(output);
 		char* char_output_last = reinterpret_cast<char*>(output_last);
 		const from_char_t* input_next;
@@ -155,7 +156,7 @@ static void utf32_to_utf8_well_formed_standard_cpp(benchmark::State& state) {
 	to_char_t* output                   = output_data.data();
 	to_char_t* const output_last        = output_data.data() + output_data.size();
 	for (auto _ : state) {
-		std::mbstate_t state {};
+		ztd_mbstate_t state {};
 		char* char_output      = reinterpret_cast<char*>(output);
 		char* char_output_last = reinterpret_cast<char*>(output_last);
 		const from_char_t* input_next;
@@ -201,7 +202,7 @@ static void utf8_to_utf32_well_formed_standard_cpp(benchmark::State& state) {
 	to_char_t* output                   = output_data.data();
 	to_char_t* const output_last        = output_data.data() + output_data.size();
 	for (auto _ : state) {
-		std::mbstate_t state {};
+		ztd_mbstate_t state {};
 		const char* char_input      = reinterpret_cast<const char*>(input);
 		const char* char_input_last = reinterpret_cast<const char*>(input_last);
 		const char* char_input_next;
@@ -246,7 +247,7 @@ static void utf16_to_utf8_well_formed_standard_cpp(benchmark::State& state) {
 	to_char_t* output                   = output_data.data();
 	to_char_t* const output_last        = output_data.data() + output_data.size();
 	for (auto _ : state) {
-		std::mbstate_t state {};
+		ztd_mbstate_t state {};
 		char* char_output      = reinterpret_cast<char*>(output);
 		char* char_output_last = reinterpret_cast<char*>(output_last);
 		const from_char_t* input_next;
@@ -292,7 +293,7 @@ static void utf8_to_utf16_well_formed_standard_cpp(benchmark::State& state) {
 	to_char_t* output                   = output_data.data();
 	to_char_t* const output_last        = output_data.data() + output_data.size();
 	for (auto _ : state) {
-		std::mbstate_t state {};
+		ztd_mbstate_t state {};
 		const char* char_input      = reinterpret_cast<const char*>(input);
 		const char* char_input_last = reinterpret_cast<const char*>(input_last);
 		const char* char_input_next;

benchmarks/conversion_speed/source/standard_c++.init.cpp

@@ -41,6 +41,7 @@
 #include <ztd/idk/charN_t.hpp>
 #include <ztd/idk/detail/unicode.hpp>
 #include <ztd/idk/endian.hpp>
+#include <ztd/idk/mbstate_t.hpp>
 
 #include <barrier/barrier.h>
 
@@ -65,7 +66,7 @@ static void utf16_to_utf32_init_well_formed_standard_cpp(benchmark::State& state
 		std::locale current_loc = {};
 		auto codecvt_ptr        = new std::codecvt_utf16<char32_t, codecvt_max_code_utfs, codecvt_target_mode>();
 		current_loc             = std::locale(current_loc, codecvt_ptr);
-		std::mbstate_t state {};
+		std::ztd_mbstate_t state {};
 		const char* char_input      = reinterpret_cast<const char*>(input);
 		const char* char_input_last = reinterpret_cast<const char*>(input_last);
 		const char* char_input_next;
@@ -108,7 +109,7 @@ static void utf32_to_utf16_init_well_formed_standard_cpp(benchmark::State& state
 		std::locale current_loc = {};
 		auto codecvt_ptr        = new std::codecvt_utf16<char32_t, codecvt_max_code_utfs, codecvt_target_mode>();
 		current_loc             = std::locale(current_loc, codecvt_ptr);
-		std::mbstate_t state {};
+		ztd_mbstate_t state {};
 		char* char_output      = reinterpret_cast<char*>(output);
 		char* char_output_last = reinterpret_cast<char*>(output_last);
 		const from_char_t* input_next;
@@ -153,7 +154,7 @@ static void utf32_to_utf8_init_well_formed_standard_cpp(benchmark::State& state)
 		std::locale current_loc = {};
 		auto codecvt_ptr        = new std::codecvt_utf8<char32_t, codecvt_max_code_utfs, codecvt_target_mode>();
 		current_loc             = std::locale(current_loc, codecvt_ptr);
-		std::mbstate_t state {};
+		ztd_mbstate_t state {};
 		char* char_output      = reinterpret_cast<char*>(output);
 		char* char_output_last = reinterpret_cast<char*>(output_last);
 		const from_char_t* input_next;
@@ -199,7 +200,7 @@ static void utf8_to_utf32_init_well_formed_standard_cpp(benchmark::State& state)
 		std::locale current_loc = {};
 		auto codecvt_ptr        = new std::codecvt_utf8<char32_t, codecvt_max_code_utfs, codecvt_target_mode>();
 		current_loc             = std::locale(current_loc, codecvt_ptr);
-		std::mbstate_t state {};
+		ztd_mbstate_t state {};
 		const char* char_input      = reinterpret_cast<const char*>(input);
 		const char* char_input_last = reinterpret_cast<const char*>(input_last);
 		const char* char_input_next;
@@ -244,7 +245,7 @@ static void utf16_to_utf8_init_well_formed_standard_cpp(benchmark::State& state)
 		std::locale current_loc = {};
 		auto codecvt_ptr        = new std::codecvt_utf8_utf16<char16_t, codecvt_max_code_utfs, codecvt_target_mode>();
 		current_loc             = std::locale(current_loc, codecvt_ptr);
-		std::mbstate_t state {};
+		ztd_mbstate_t state {};
 		char* char_output      = reinterpret_cast<char*>(output);
 		char* char_output_last = reinterpret_cast<char*>(output_last);
 		const from_char_t* input_next;
@@ -290,7 +291,7 @@ static void utf8_to_utf16_init_well_formed_standard_cpp(benchmark::State& state)
 		std::locale current_loc = {};
 		auto codecvt_ptr        = new std::codecvt_utf8_utf16<char16_t, codecvt_max_code_utfs, codecvt_target_mode>();
 		current_loc             = std::locale(current_loc, codecvt_ptr);
-		std::mbstate_t state {};
+		ztd_mbstate_t state {};
 		const char* char_input      = reinterpret_cast<const char*>(input);
 		const char* char_input_last = reinterpret_cast<const char*>(input_last);
 		const char* char_input_next;

benchmarks/conversion_speed/source/standard_c.cpp

@@ -39,6 +39,7 @@
 
 #include <ztd/idk/c_span.h>
 #include <ztd/idk/charN_t.hpp>
+#include <ztd/idk/mbstate_t.hpp>
 #include <barrier/barrier.h>
 
 #if ZTD_IS_ON(ZTD_CUCHAR)
@@ -55,8 +56,8 @@ static void utf16_to_utf32_well_formed_standard_c(benchmark::State& state) {
 	std::vector<to_char_t> output_data(c_span_char32_t_size(u32_data));
 	bool result = true;
 	for (auto _ : state) {
-		mbstate_t from_state {};
-		mbstate_t to_state {};
+		ztd_mbstate_t from_state {};
+		ztd_mbstate_t to_state {};
 		char intermediate_data[MB_LEN_MAX * 2];
 		to_char_t* output                   = output_data.data();
 		to_char_t* output_last              = output_data.data() + output_data.size();
@@ -149,8 +150,8 @@ static void utf32_to_utf16_well_formed_standard_c(benchmark::State& state) {
 	std::vector<to_char_t> output_data(c_span_char16_t_size(u16_data));
 	bool result = true;
 	for (auto _ : state) {
-		mbstate_t from_state {};
-		mbstate_t to_state {};
+		ztd_mbstate_t from_state {};
+		ztd_mbstate_t to_state {};
 		char intermediate_data[MB_LEN_MAX * 2];
 		to_char_t* output                   = output_data.data();
 		to_char_t* output_last              = output_data.data() + output_data.size();
@@ -244,7 +245,7 @@ static void utf32_to_utf8_well_formed_standard_c(benchmark::State& state) {
 	std::vector<to_char_t> output_data(c_span_char8_t_size(u8_data));
 	bool result = true;
 	for (auto _ : state) {
-		mbstate_t from_state {};
+		ztd_mbstate_t from_state {};
 		to_char_t* output                   = output_data.data();
 		to_char_t* output_last              = output_data.data() + output_data.size();
 		const from_char_t* input            = input_data.data();
@@ -286,7 +287,7 @@ static void utf8_to_utf32_well_formed_standard_c(benchmark::State& state) {
 	std::vector<to_char_t> output_data(c_span_char32_t_size(u32_data));
 	bool result = true;
 	for (auto _ : state) {
-		mbstate_t to_state {};
+		ztd_mbstate_t to_state {};
 		const from_char_t* input            = input_data.data();
 		const from_char_t* const input_last = input_data.data() + input_data.size();
 		to_char_t* output                   = output_data.data();
@@ -362,7 +363,7 @@ static void utf16_to_utf8_well_formed_standard_c(benchmark::State& state) {
 	std::vector<to_char_t> output_data(c_span_char8_t_size(u8_data));
 	bool result = true;
 	for (auto _ : state) {
-		mbstate_t from_state {};
+		ztd_mbstate_t from_state {};
 		to_char_t* output                   = output_data.data();
 		to_char_t* output_last              = output_data.data() + output_data.size();
 		const from_char_t* input            = input_data.data();
@@ -404,7 +405,7 @@ static void utf8_to_utf16_well_formed_standard_c(benchmark::State& state) {
 	std::vector<to_char_t> output_data(c_span_char16_t_size(u16_data));
 	bool result = true;
 	for (auto _ : state) {
-		mbstate_t to_state {};
+		ztd_mbstate_t to_state {};
 		const from_char_t* input            = input_data.data();
 		const from_char_t* const input_last = input_data.data() + input_data.size();
 		to_char_t* output                   = output_data.data();

documentation/CMakeLists.txt

@@ -30,7 +30,11 @@
 
 # Dependencies
 find_package(Doxygen REQUIRED)
+if (NOT ZTD_TEXT_DOCUMENTATION_NO_SPHINX)
+find_package(Sphinx REQUIRED)
+else()
 find_package(Sphinx QUIET)
+endif()
 
 # For the install target and related directories
 include(GNUInstallDirs)

documentation/source/conf.py

@@ -68,7 +68,7 @@
 rst_prolog = """
 .. |ub| replace:: ☢️☢️Undefined Behavior☢️☢️
 
-.. |specializations_okay_different_types| replace:: User Specializations: ✔️ Okay! You can add other types to this classification by specializing the class template. Your specialization must have a type definition named ``type`` (as in, ``using type = ...;`` or `typedef ... type;``) inside of the class specialization that is ``public:``\ ly accessible. Note that specializing any type not explicitly marked with this notice is |ub|.
+.. |specializations_okay_different_types| replace:: User Specializations: ✔️ Okay! You can add other types to this classification by specializing the class template. Your specialization must have a type definition named ``type`` (as in, ``using type = ...;`` or `typedef ... type;``) inside of the class specialization that is ``public:``-ly accessible. Note that specializing any type not explicitly marked with this notice is |ub|.
 
 .. |specializations_okay_true_false_type| replace:: User Specializations: ✔️ Okay! You can add other types to this classification by specializing the class template to a definition that derives from ``std::true_type``, or turn it off explicitly by having a definition that derives from ``std::false_type``. Note that specializing any type not explicitly marked with this notice is |ub|.
 

examples/documentation/snippets/source/runtime_locale_encoding.cpp

@@ -34,6 +34,8 @@
 // This example doesn't work on Apple/libc++ because they don't have
 // standard C or C++ headers.
 
+#include <ztd/idk/mbstate_t.hpp>
+
 #include <cstddef>
 #if defined(__has_include) && __has_include(<cuchar>)
 #include <cuchar>
@@ -50,7 +52,6 @@ extern "C" {
 #define WIN32_LEAN_AND_MEAN 1
 #include <windows.h>
 #else
-
 #endif
 
 class runtime_locale {
@@ -63,23 +64,31 @@ class runtime_locale {
 	inline static constexpr std::size_t max_code_units  = MB_LEN_MAX;
 
 	struct decode_state {
-		std::mbstate_t c_stdlib_state;
+		ztd_mbstate_t c_stdlib_state;
 
 		decode_state() noexcept : c_stdlib_state() {
 			// properly set for mbrtoc32 state
 			code_point ghost_ouput[2] {};
 			UCHAR_ACCESS mbrtoc32(ghost_ouput, "\0", 1, &c_stdlib_state);
 		}
+
+		bool is_complete() const noexcept {
+			return UCHAR_ACCESS mbsinit(&c_stdlib_state) != 0;
+		}
 	};
 
 	struct encode_state {
-		std::mbstate_t c_stdlib_state;
+		ztd_mbstate_t c_stdlib_state;
 
 		encode_state() noexcept : c_stdlib_state() {
 			// properly set for c32rtomb state
 			code_unit ghost_ouput[MB_LEN_MAX] {};
 			UCHAR_ACCESS c32rtomb(ghost_ouput, U'\0', &c_stdlib_state);
 		}
+
+		bool is_complete() const noexcept {
+			return UCHAR_ACCESS mbsinit(&c_stdlib_state) != 0;
+		}
 	};
 
 	bool contains_unicode_encoding() const noexcept {

include/ztd/text/code_point.hpp

@@ -42,15 +42,26 @@
 #include <ztd/prologue.hpp>
 
 //////
-/// @file code point abstractions
+/// @file code_point.hpp
+///
+/// @brief code point abstractions
+
+//////
+/// @namespace ztd
+///
+/// @brief The core library namespace.
+
+//////
+/// @namespace ztd::text
+///
+/// @brief The text library namespace.
 
 namespace ztd { namespace text {
 	ZTD_TEXT_INLINE_ABI_NAMESPACE_OPEN_I_
 
 	//////
 	/// @addtogroup ztd_text_properties Property and Trait Helpers
 	/// @{
-	/////
 
 	//////
 	/// @brief Retrieves the code point type for the given type.

include/ztd/text/decode_view.hpp

@@ -213,7 +213,7 @@ namespace ztd { namespace text {
 	/// @brief The reconstruct extension point for rebuilding an encoding view from its iterator and sentinel
 	/// type.
 	template <typename _Encoding, typename _Range, typename _ErrorHandler, typename _State>
-	constexpr decode_view<_Encoding, _Range, _ErrorHandler, _State> tag_invoke(ztd::tag_t<ranges::reconstruct>,
+	constexpr decode_view<_Encoding, _Range, _ErrorHandler, _State> reconstruct(
 		::std::in_place_type_t<decode_view<_Encoding, _Range, _ErrorHandler, _State>>,
 		typename decode_view<_Encoding, _Range, _ErrorHandler, _State>::iterator __it,
 		typename decode_view<_Encoding, _Range, _ErrorHandler, _State>::sentinel) noexcept(::std::