From 1f7c4ab7f6c5b65d6569a1981b04c1d2ac33b713 Mon Sep 17 00:00:00 2001 From: archshift Date: Mon, 10 Nov 2014 14:36:32 -0800 Subject: Use std::u16string for conversion between UTF-8 and UTF-16, FS:USER functions --- src/common/string_util.cpp | 158 +++++++++++++++++++++++++++++++-------------- src/common/string_util.h | 8 ++- 2 files changed, 115 insertions(+), 51 deletions(-) (limited to 'src/common') diff --git a/src/common/string_util.cpp b/src/common/string_util.cpp index 61f0939c..54943d30 100644 --- a/src/common/string_util.cpp +++ b/src/common/string_util.cpp @@ -9,6 +9,7 @@ #ifdef _WIN32 #include + #include #else #include #endif @@ -411,7 +412,19 @@ std::string UriEncode(const std::string & sSrc) #ifdef _WIN32 -std::string UTF16ToUTF8(const std::wstring& input) +std::string UTF16ToUTF8(const std::u16string& input) +{ + std::wstring_convert, char16_t> convert; + return convert.to_bytes(input); +} + +std::u16string UTF8ToUTF16(const std::string& input) +{ + std::wstring_convert, char16_t> convert; + return convert.from_bytes(input); +} + +static std::string UTF16ToUTF8(const std::wstring& input) { auto const size = WideCharToMultiByte(CP_UTF8, 0, input.data(), input.size(), nullptr, 0, nullptr, nullptr); @@ -424,7 +437,7 @@ std::string UTF16ToUTF8(const std::wstring& input) return output; } -std::wstring CPToUTF16(u32 code_page, const std::string& input) +static std::wstring CPToUTF16(u32 code_page, const std::string& input) { auto const size = MultiByteToWideChar(code_page, 0, input.data(), input.size(), nullptr, 0); @@ -437,7 +450,7 @@ std::wstring CPToUTF16(u32 code_page, const std::string& input) return output; } -std::wstring UTF8ToUTF16(const std::string& input) +std::wstring UTF8ToUTF16W(const std::string &input) { return CPToUTF16(CP_UTF8, input); } @@ -455,61 +468,123 @@ std::string CP1252ToUTF8(const std::string& input) #else template -std::string CodeToUTF8(const char* fromcode, const std::basic_string& input) +static std::string CodeToUTF8(const char* fromcode, const std::basic_string& input) { std::string result; iconv_t const conv_desc = iconv_open("UTF-8", fromcode); - if ((iconv_t)-1 == conv_desc) + if ((iconv_t)(-1) == conv_desc) { ERROR_LOG(COMMON, "Iconv initialization failure [%s]: %s", fromcode, strerror(errno)); + iconv_close(conv_desc); + return {}; } - else - { - size_t const in_bytes = sizeof(T) * input.size(); - size_t const out_buffer_size = 4 * in_bytes; - std::string out_buffer; - out_buffer.resize(out_buffer_size); + const size_t in_bytes = sizeof(T) * input.size(); + // Multiply by 4, which is the max number of bytes to encode a codepoint + const size_t out_buffer_size = 4 * in_bytes; - auto src_buffer = &input[0]; - size_t src_bytes = in_bytes; - auto dst_buffer = &out_buffer[0]; - size_t dst_bytes = out_buffer.size(); + std::string out_buffer; + out_buffer.resize(out_buffer_size); - while (src_bytes != 0) - { - size_t const iconv_result = iconv(conv_desc, (char**)(&src_buffer), &src_bytes, - &dst_buffer, &dst_bytes); + auto src_buffer = &input[0]; + size_t src_bytes = in_bytes; + auto dst_buffer = &out_buffer[0]; + size_t dst_bytes = out_buffer.size(); - if ((size_t)-1 == iconv_result) + while (0 != src_bytes) + { + size_t const iconv_result = iconv(conv_desc, (char**)(&src_buffer), &src_bytes, + &dst_buffer, &dst_bytes); + + if (static_cast(-1) == iconv_result) + { + if (EILSEQ == errno || EINVAL == errno) { - if (EILSEQ == errno || EINVAL == errno) + // Try to skip the bad character + if (0 != src_bytes) { - // Try to skip the bad character - if (src_bytes != 0) - { - --src_bytes; - ++src_buffer; - } - } - else - { - ERROR_LOG(COMMON, "iconv failure [%s]: %s", fromcode, strerror(errno)); - break; + --src_bytes; + ++src_buffer; } } + else + { + ERROR_LOG(COMMON, "iconv failure [%s]: %s", fromcode, strerror(errno)); + break; + } } + } - out_buffer.resize(out_buffer_size - dst_bytes); - out_buffer.swap(result); - + out_buffer.resize(out_buffer_size - dst_bytes); + out_buffer.swap(result); + + iconv_close(conv_desc); + + return result; +} + +std::u16string UTF8ToUTF16(const std::string& input) +{ + std::u16string result; + + iconv_t const conv_desc = iconv_open("UTF-16", "UTF-8"); + if ((iconv_t)(-1) == conv_desc) + { + ERROR_LOG(COMMON, "Iconv initialization failure [UTF-8]: %s", strerror(errno)); iconv_close(conv_desc); + return {}; } + + const size_t in_bytes = sizeof(char) * input.size(); + // Multiply by 4, which is the max number of bytes to encode a codepoint + const size_t out_buffer_size = 4 * sizeof(char16_t) * in_bytes; + + std::u16string out_buffer; + out_buffer.resize(out_buffer_size); + + char* src_buffer = const_cast(&input[0]); + size_t src_bytes = in_bytes; + char* dst_buffer = (char*)(&out_buffer[0]); + size_t dst_bytes = out_buffer.size(); + + while (0 != src_bytes) + { + size_t const iconv_result = iconv(conv_desc, &src_buffer, &src_bytes, + &dst_buffer, &dst_bytes); + + if (static_cast(-1) == iconv_result) + { + if (EILSEQ == errno || EINVAL == errno) + { + // Try to skip the bad character + if (0 != src_bytes) + { + --src_bytes; + ++src_buffer; + } + } + else + { + ERROR_LOG(COMMON, "iconv failure [UTF-8]: %s", strerror(errno)); + break; + } + } + } + + out_buffer.resize(out_buffer_size - dst_bytes); + out_buffer.swap(result); + + iconv_close(conv_desc); return result; } +std::string UTF16ToUTF8(const std::u16string& input) +{ + return CodeToUTF8("UTF-16", input); +} + std::string CP1252ToUTF8(const std::string& input) { //return CodeToUTF8("CP1252//TRANSLIT", input); @@ -523,19 +598,6 @@ std::string SHIFTJISToUTF8(const std::string& input) return CodeToUTF8("SJIS", input); } -std::string UTF16ToUTF8(const std::wstring& input) -{ - std::string result = - // CodeToUTF8("UCS-2", input); - // CodeToUTF8("UCS-2LE", input); - // CodeToUTF8("UTF-16", input); - CodeToUTF8("UTF-16LE", input); - - // TODO: why is this needed? - result.erase(std::remove(result.begin(), result.end(), 0x00), result.end()); - return result; -} - #endif } diff --git a/src/common/string_util.h b/src/common/string_util.h index a41ccc69..787a5663 100644 --- a/src/common/string_util.h +++ b/src/common/string_util.h @@ -89,20 +89,22 @@ std::string ReplaceAll(std::string result, const std::string& src, const std::st std::string UriDecode(const std::string & sSrc); std::string UriEncode(const std::string & sSrc); +std::string UTF16ToUTF8(const std::u16string& input); +std::u16string UTF8ToUTF16(const std::string& input); + std::string CP1252ToUTF8(const std::string& str); std::string SHIFTJISToUTF8(const std::string& str); -std::string UTF16ToUTF8(const std::wstring& str); #ifdef _WIN32 -std::wstring UTF8ToUTF16(const std::string& str); +std::wstring UTF8ToUTF16W(const std::string& str); #ifdef _UNICODE inline std::string TStrToUTF8(const std::wstring& str) { return UTF16ToUTF8(str); } inline std::wstring UTF8ToTStr(const std::string& str) -{ return UTF8ToUTF16(str); } +{ return UTF8ToUTF16W(str); } #else inline std::string TStrToUTF8(const std::string& str) { return str; } -- cgit v1.2.3