refactor: update project

- add documentation CMake build script. re-organise document layout for future changes.
- move LIBCMO_EXPORT to BMap and rename it to BMAP_EXPORT because only BMap need to use this macro.
- fully refactor VTEncoding to make it more like Python
	- Now language name is platform independent.
	- Hide implementation detail as possible as I can.
	- Language mapping are still work in progress.
- add code gen for new added universal encoding feature to generate language name mapping in Windows and Iconv respectively.
- remove old code of CMake build script.
- update VTUtils for new requirement.
	- remove useless functions.
	- create LibCmo specific custom exception classes.
This commit is contained in:
2024-08-16 22:07:23 +08:00
parent afa06339b2
commit f870d4dde3
20 changed files with 1013 additions and 726 deletions

View File

@ -1,85 +1,24 @@
#pragma once
#include "VTUtils.hpp"
#include <string>
#include <filesystem>
#if defined(LIBCMO_OS_WIN32)
#include <Windows.h>
#include <fileapi.h>
// disable annoy macro at the same time
#undef GetObject
#undef GetClassName
#undef LoadImage
#undef GetTempPath
#else
#include <iconv.h>
#endif
#include <string_view>
namespace LibCmo::EncodingHelper {
#pragma region assist functions
using EncodingToken = void*;
constexpr EncodingToken INVALID_ENCODING_TOKEN = nullptr;
#if defined(LIBCMO_OS_WIN32)
EncodingToken CreateEncodingToken(const std::u8string_view& enc_name);
void DestroyEncodingToken(EncodingToken token);
bool GetWindowsCodePage(const char* u8_encoding_spec, UINT* result);
bool ToOrdinary(const std::u8string_view& src, std::string& dst, EncodingToken token);
bool ToOrdinary(const char8_t* src, std::string& dst, EncodingToken token);
std::string ToOrdinary(const std::u8string_view& src, EncodingToken token);
std::string ToOrdinary(const char8_t* src, EncodingToken token);
bool WcharToChar(const wchar_t* src, std::string& dest, const UINT codepage);
bool WcharToChar(const std::wstring& src, std::string& dest, const UINT codepage);
bool CharToWchar(const char* src, std::wstring& dest, const UINT codepage);
bool CharToWchar(const std::string& src, std::wstring& dest, const UINT codepage);
bool CharToChar(const char* src, std::string& dest, const UINT src_codepage, const UINT dest_codepage);
bool CharToChar(const std::string& src, std::string& dest, const UINT src_codepage, const UINT dest_codepage);
#else
bool CreateIconvDescriptor(const char* enc_from, const char* enc_to, iconv_t& val);
void DestroyIconvDescriptor(iconv_t& val);
bool DoIconv(iconv_t& cd, const std::string& str_from, std::string& str_to);
#endif
#pragma endregion
#pragma region core functions
#if defined(LIBCMO_OS_WIN32)
// Token is the ticket for using encoding functions.
// It should be created by "GenerateEncodingToken" and free by "DestroyEncodingToken".
using ENCODING_TOKEN = UINT*;
constexpr const ENCODING_TOKEN ENCODING_TOKEN_DEFAULT = nullptr;
#else
class IconvPair {
public:
IconvPair();
IconvPair(const IconvPair&) = delete;
IconvPair& operator=(const IconvPair&) = delete;
~IconvPair();
iconv_t FromUtf8;
iconv_t ToUtf8;
};
using ENCODING_TOKEN = IconvPair*;
constexpr const ENCODING_TOKEN ENCODING_TOKEN_DEFAULT = nullptr;
#endif
ENCODING_TOKEN CreateEncodingToken(const std::string& token_string);
void DestroyEncodingToken(const ENCODING_TOKEN& token);
bool GetUtf8VirtoolsName(const std::string& native_name, std::string& u8_name, const ENCODING_TOKEN& token);
bool GetNativeVirtoolsName(const std::string& u8_name, std::string& native_name, const ENCODING_TOKEN& token);
void U8PathToStdPath(std::filesystem::path& stdpath, const char* u8_path);
void StdPathToU8Path(std::string& u8path, std::filesystem::path& stdpath);
FILE* U8FOpen(const char* u8_filepath, const char* u8_mode);
#pragma endregion
bool ToUTF8(const std::string_view& src, std::u8string& dst, EncodingToken token);
bool ToUTF8(const char* src, std::u8string& dst, EncodingToken token);
std::u8string ToUTF8(const std::string_view& src, EncodingToken token);
std::u8string ToUTF8(const char* src, EncodingToken token);
}