From 0b7d58968b37dc9d36a90d3fb439891a57b81a18 Mon Sep 17 00:00:00 2001 From: "Matthias P. Braendli" Date: Wed, 17 Jan 2024 09:48:56 +0100 Subject: Update ClockTAI and RemoteControl --- lib/charset/utf8/core.h | 275 ++++++++++++++++++++++++++++++++++++++---------- 1 file changed, 219 insertions(+), 56 deletions(-) (limited to 'lib/charset/utf8/core.h') diff --git a/lib/charset/utf8/core.h b/lib/charset/utf8/core.h index 693d388..4494c53 100644 --- a/lib/charset/utf8/core.h +++ b/lib/charset/utf8/core.h @@ -29,15 +29,42 @@ DEALINGS IN THE SOFTWARE. #define UTF8_FOR_CPP_CORE_H_2675DCD0_9480_4c0c_B92A_CC14C027B731 #include +#include +#include + +// Determine the C++ standard version. +// If the user defines UTF_CPP_CPLUSPLUS, use that. +// Otherwise, trust the unreliable predefined macro __cplusplus + +#if !defined UTF_CPP_CPLUSPLUS + #define UTF_CPP_CPLUSPLUS __cplusplus +#endif + +#if UTF_CPP_CPLUSPLUS >= 201103L // C++ 11 or later + #define UTF_CPP_OVERRIDE override + #define UTF_CPP_NOEXCEPT noexcept +#else // C++ 98/03 + #define UTF_CPP_OVERRIDE + #define UTF_CPP_NOEXCEPT throw() +#endif // C++ 11 or later + namespace utf8 { - // The typedefs for 8-bit, 16-bit and 32-bit unsigned integers - // You may need to change them to match your system. - // These typedefs have the same names as ones from cstdint, or boost/cstdint - typedef unsigned char uint8_t; - typedef unsigned short uint16_t; - typedef unsigned int uint32_t; +// The typedefs for 8-bit, 16-bit and 32-bit code units +#if UTF_CPP_CPLUSPLUS >= 201103L // C++ 11 or later + #if UTF_CPP_CPLUSPLUS >= 202002L // C++ 20 or later + typedef char8_t utfchar8_t; + #else // C++ 11/14/17 + typedef unsigned char utfchar8_t; + #endif + typedef char16_t utfchar16_t; + typedef char32_t utfchar32_t; +#else // C++ 98/03 + typedef unsigned char utfchar8_t; + typedef unsigned short utfchar16_t; + typedef unsigned int utfchar32_t; +#endif // C++ 11 or later // Helper code - not intended to be directly called by the library users. May be changed at any time namespace internal @@ -45,61 +72,62 @@ namespace internal // Unicode constants // Leading (high) surrogates: 0xd800 - 0xdbff // Trailing (low) surrogates: 0xdc00 - 0xdfff - const uint16_t LEAD_SURROGATE_MIN = 0xd800u; - const uint16_t LEAD_SURROGATE_MAX = 0xdbffu; - const uint16_t TRAIL_SURROGATE_MIN = 0xdc00u; - const uint16_t TRAIL_SURROGATE_MAX = 0xdfffu; - const uint16_t LEAD_OFFSET = LEAD_SURROGATE_MIN - (0x10000 >> 10); - const uint32_t SURROGATE_OFFSET = 0x10000u - (LEAD_SURROGATE_MIN << 10) - TRAIL_SURROGATE_MIN; + const utfchar16_t LEAD_SURROGATE_MIN = 0xd800u; + const utfchar16_t LEAD_SURROGATE_MAX = 0xdbffu; + const utfchar16_t TRAIL_SURROGATE_MIN = 0xdc00u; + const utfchar16_t TRAIL_SURROGATE_MAX = 0xdfffu; + const utfchar16_t LEAD_OFFSET = 0xd7c0u; // LEAD_SURROGATE_MIN - (0x10000 >> 10) + const utfchar32_t SURROGATE_OFFSET = 0xfca02400u; // 0x10000u - (LEAD_SURROGATE_MIN << 10) - TRAIL_SURROGATE_MIN // Maximum valid value for a Unicode code point - const uint32_t CODE_POINT_MAX = 0x0010ffffu; + const utfchar32_t CODE_POINT_MAX = 0x0010ffffu; template - inline uint8_t mask8(octet_type oc) + inline utfchar8_t mask8(octet_type oc) { - return static_cast(0xff & oc); + return static_cast(0xff & oc); } template - inline uint16_t mask16(u16_type oc) + inline utfchar16_t mask16(u16_type oc) { - return static_cast(0xffff & oc); + return static_cast(0xffff & oc); } + template inline bool is_trail(octet_type oc) { return ((utf8::internal::mask8(oc) >> 6) == 0x2); } - template - inline bool is_lead_surrogate(u16 cp) + inline bool is_lead_surrogate(utfchar32_t cp) { return (cp >= LEAD_SURROGATE_MIN && cp <= LEAD_SURROGATE_MAX); } - template - inline bool is_trail_surrogate(u16 cp) + inline bool is_trail_surrogate(utfchar32_t cp) { return (cp >= TRAIL_SURROGATE_MIN && cp <= TRAIL_SURROGATE_MAX); } - template - inline bool is_surrogate(u16 cp) + inline bool is_surrogate(utfchar32_t cp) { return (cp >= LEAD_SURROGATE_MIN && cp <= TRAIL_SURROGATE_MAX); } - template - inline bool is_code_point_valid(u32 cp) + inline bool is_code_point_valid(utfchar32_t cp) { return (cp <= CODE_POINT_MAX && !utf8::internal::is_surrogate(cp)); } + inline bool is_in_bmp(utfchar32_t cp) + { + return cp < utfchar32_t(0x10000); + } + template - inline typename std::iterator_traits::difference_type - sequence_length(octet_iterator lead_it) + int sequence_length(octet_iterator lead_it) { - uint8_t lead = utf8::internal::mask8(*lead_it); + const utfchar8_t lead = utf8::internal::mask8(*lead_it); if (lead < 0x80) return 1; else if ((lead >> 5) == 0x6) @@ -112,8 +140,7 @@ namespace internal return 0; } - template - inline bool is_overlong_sequence(uint32_t cp, octet_difference_type length) + inline bool is_overlong_sequence(utfchar32_t cp, int length) { if (cp < 0x80) { if (length != 1) @@ -127,7 +154,6 @@ namespace internal if (length != 3) return true; } - return false; } @@ -135,22 +161,22 @@ namespace internal /// Helper for get_sequence_x template - utf_error increase_safely(octet_iterator& it, octet_iterator end) + utf_error increase_safely(octet_iterator& it, const octet_iterator end) { if (++it == end) return NOT_ENOUGH_ROOM; if (!utf8::internal::is_trail(*it)) return INCOMPLETE_SEQUENCE; - + return UTF8_OK; } - #define UTF8_CPP_INCREASE_AND_RETURN_ON_ERROR(IT, END) {utf_error ret = increase_safely(IT, END); if (ret != UTF8_OK) return ret;} + #define UTF8_CPP_INCREASE_AND_RETURN_ON_ERROR(IT, END) {utf_error ret = increase_safely(IT, END); if (ret != UTF8_OK) return ret;} /// get_sequence_x functions decode utf-8 sequences of the length x template - utf_error get_sequence_1(octet_iterator& it, octet_iterator end, uint32_t& code_point) + utf_error get_sequence_1(octet_iterator& it, octet_iterator end, utfchar32_t& code_point) { if (it == end) return NOT_ENOUGH_ROOM; @@ -161,11 +187,11 @@ namespace internal } template - utf_error get_sequence_2(octet_iterator& it, octet_iterator end, uint32_t& code_point) + utf_error get_sequence_2(octet_iterator& it, octet_iterator end, utfchar32_t& code_point) { if (it == end) return NOT_ENOUGH_ROOM; - + code_point = utf8::internal::mask8(*it); UTF8_CPP_INCREASE_AND_RETURN_ON_ERROR(it, end) @@ -176,7 +202,7 @@ namespace internal } template - utf_error get_sequence_3(octet_iterator& it, octet_iterator end, uint32_t& code_point) + utf_error get_sequence_3(octet_iterator& it, octet_iterator end, utfchar32_t& code_point) { if (it == end) return NOT_ENOUGH_ROOM; @@ -195,7 +221,7 @@ namespace internal } template - utf_error get_sequence_4(octet_iterator& it, octet_iterator end, uint32_t& code_point) + utf_error get_sequence_4(octet_iterator& it, octet_iterator end, utfchar32_t& code_point) { if (it == end) return NOT_ENOUGH_ROOM; @@ -220,21 +246,23 @@ namespace internal #undef UTF8_CPP_INCREASE_AND_RETURN_ON_ERROR template - utf_error validate_next(octet_iterator& it, octet_iterator end, uint32_t& code_point) + utf_error validate_next(octet_iterator& it, octet_iterator end, utfchar32_t& code_point) { + if (it == end) + return NOT_ENOUGH_ROOM; + // Save the original value of it so we can go back in case of failure // Of course, it does not make much sense with i.e. stream iterators octet_iterator original_it = it; - uint32_t cp = 0; + utfchar32_t cp = 0; // Determine the sequence length based on the lead octet - typedef typename std::iterator_traits::difference_type octet_difference_type; - const octet_difference_type length = utf8::internal::sequence_length(it); + const int length = utf8::internal::sequence_length(it); // Get trail octets and calculate the code point utf_error err = UTF8_OK; switch (length) { - case 0: + case 0: return INVALID_LEAD; case 1: err = utf8::internal::get_sequence_1(it, end, cp); @@ -273,16 +301,133 @@ namespace internal template inline utf_error validate_next(octet_iterator& it, octet_iterator end) { - uint32_t ignored; + utfchar32_t ignored; return utf8::internal::validate_next(it, end, ignored); } + template + utf_error validate_next16(word_iterator& it, word_iterator end, utfchar32_t& code_point) + { + if (it == end) + return NOT_ENOUGH_ROOM; + // Save the original value of it so we can go back in case of failure + // Of course, it does not make much sense with i.e. stream iterators + word_iterator original_it = it; + + utf_error err = UTF8_OK; + + const utfchar16_t first_word = *it++; + if (!is_surrogate(first_word)) { + code_point = first_word; + return UTF8_OK; + } + else { + if (it == end) + err = NOT_ENOUGH_ROOM; + else if (is_lead_surrogate(first_word)) { + const utfchar16_t second_word = *it++; + if (is_trail_surrogate(second_word)) { + code_point = (first_word << 10) + second_word + SURROGATE_OFFSET; + return UTF8_OK; + } else + err = INCOMPLETE_SEQUENCE; + + } else { + err = INVALID_LEAD; + } + } + // error branch + it = original_it; + return err; + } + + // Internal implementation of both checked and unchecked append() function + // This function will be invoked by the overloads below, as they will know + // the octet_type. + template + octet_iterator append(utfchar32_t cp, octet_iterator result) { + if (cp < 0x80) // one octet + *(result++) = static_cast(cp); + else if (cp < 0x800) { // two octets + *(result++) = static_cast((cp >> 6) | 0xc0); + *(result++) = static_cast((cp & 0x3f) | 0x80); + } + else if (cp < 0x10000) { // three octets + *(result++) = static_cast((cp >> 12) | 0xe0); + *(result++) = static_cast(((cp >> 6) & 0x3f) | 0x80); + *(result++) = static_cast((cp & 0x3f) | 0x80); + } + else { // four octets + *(result++) = static_cast((cp >> 18) | 0xf0); + *(result++) = static_cast(((cp >> 12) & 0x3f)| 0x80); + *(result++) = static_cast(((cp >> 6) & 0x3f) | 0x80); + *(result++) = static_cast((cp & 0x3f) | 0x80); + } + return result; + } + + // One of the following overloads will be invoked from the API calls + + // A simple (but dangerous) case: the caller appends byte(s) to a char array + inline char* append(utfchar32_t cp, char* result) { + return append(cp, result); + } + + // Hopefully, most common case: the caller uses back_inserter + // i.e. append(cp, std::back_inserter(str)); + template + std::back_insert_iterator append + (utfchar32_t cp, std::back_insert_iterator result) { + return append, + typename container_type::value_type>(cp, result); + } + + // The caller uses some other kind of output operator - not covered above + // Note that in this case we are not able to determine octet_type + // so we assume it's utfchar8_t; that can cause a conversion warning if we are wrong. + template + octet_iterator append(utfchar32_t cp, octet_iterator result) { + return append(cp, result); + } + + // Internal implementation of both checked and unchecked append16() function + // This function will be invoked by the overloads below, as they will know + // the word_type. + template + word_iterator append16(utfchar32_t cp, word_iterator result) { + if (is_in_bmp(cp)) + *(result++) = static_cast(cp); + else { + // Code points from the supplementary planes are encoded via surrogate pairs + *(result++) = static_cast(LEAD_OFFSET + (cp >> 10)); + *(result++) = static_cast(TRAIL_SURROGATE_MIN + (cp & 0x3FF)); + } + return result; + } + + // Hopefully, most common case: the caller uses back_inserter + // i.e. append16(cp, std::back_inserter(str)); + template + std::back_insert_iterator append16 + (utfchar32_t cp, std::back_insert_iterator result) { + return append16, + typename container_type::value_type>(cp, result); + } + + // The caller uses some other kind of output operator - not covered above + // Note that in this case we are not able to determine word_type + // so we assume it's utfchar16_t; that can cause a conversion warning if we are wrong. + template + word_iterator append16(utfchar32_t cp, word_iterator result) { + return append16(cp, result); + } + } // namespace internal /// The library API - functions intended to be called by the users // Byte order mark - const uint8_t bom[] = {0xef, 0xbb, 0xbf}; + const utfchar8_t bom[] = {0xef, 0xbb, 0xbf}; template octet_iterator find_invalid(octet_iterator start, octet_iterator end) @@ -296,12 +441,36 @@ namespace internal return result; } + inline const char* find_invalid(const char* str) + { + const char* end = str + std::strlen(str); + return find_invalid(str, end); + } + + inline std::size_t find_invalid(const std::string& s) + { + std::string::const_iterator invalid = find_invalid(s.begin(), s.end()); + return (invalid == s.end()) ? std::string::npos : static_cast(invalid - s.begin()); + } + template inline bool is_valid(octet_iterator start, octet_iterator end) { return (utf8::find_invalid(start, end) == end); } + inline bool is_valid(const char* str) + { + return (*(utf8::find_invalid(str)) == '\0'); + } + + inline bool is_valid(const std::string& s) + { + return is_valid(s.begin(), s.end()); + } + + + template inline bool starts_with_bom (octet_iterator it, octet_iterator end) { @@ -311,17 +480,11 @@ namespace internal ((it != end) && (utf8::internal::mask8(*it)) == bom[2]) ); } - - //Deprecated in release 2.3 - template - inline bool is_bom (octet_iterator it) + + inline bool starts_with_bom(const std::string& s) { - return ( - (utf8::internal::mask8(*it++)) == bom[0] && - (utf8::internal::mask8(*it++)) == bom[1] && - (utf8::internal::mask8(*it)) == bom[2] - ); - } + return starts_with_bom(s.begin(), s.end()); + } } // namespace utf8 #endif // header guard -- cgit v1.2.3