sgml: upgrade to support UTF-16 and 32

Windows and Apple (NSString) are using UTF-16 wchar_t. All others use
UTF-32. This eliminates need for additional conversion should we need
UTF-32 on Windows&Apple or UTF-16 on others. Mind, UTF-16↔32 is not a
trivial conversion as it may change number of code units.

Signed-off-by: Simon Rozman <simon@rozman.si>
This commit is contained in:
Simon Rozman 2024-02-21 17:33:00 +01:00
parent c6f5a72e70
commit c20d9dbc09
4 changed files with 1867 additions and 1811 deletions

View File

@ -127,7 +127,8 @@ namespace stdex
_In_ int flags = match_default) = 0; _In_ int flags = match_default) = 0;
/// \cond internal /// \cond internal
const wchar_t* next_sgml_cp(_In_ const char* text, _In_ size_t start, _In_ size_t end, _Out_ size_t& chr_end, _Out_ wchar_t(&buf)[3]) template <class T = wchar_t>
const T* next_sgml_cp(_In_ const char* text, _In_ size_t start, _In_ size_t end, _Out_ size_t& chr_end, _Out_ T(&buf)[5])
{ {
if (text[start] == '&') { if (text[start] == '&') {
// Potential entity start // Potential entity start
@ -139,31 +140,9 @@ namespace stdex
} }
if (text[chr_end] == ';') { if (text[chr_end] == ';') {
// Entity end // Entity end
utf32_t buf32[2];
size_t n = chr_end - start - 1; size_t n = chr_end - start - 1;
if (n >= 2 && text[start + 1] == '#') { auto entity_w = utf32_to_wstr(sgml2uni(text + start + 1, n, buf32), buf);
// Numerical entity
utf32_t unicode;
if (text[start + 2] == 'x' || text[start + 2] == 'X')
unicode = static_cast<utf32_t>(strtou32(text + start + 3, n - 2, nullptr, 16));
else
unicode = static_cast<utf32_t>(strtou32(text + start + 2, n - 1, nullptr, 10));
#ifdef _WIN32
if (unicode < 0x10000) {
buf[0] = (wchar_t)unicode;
buf[1] = 0;
}
else {
ucs4_to_surrogate_pair(buf, unicode);
buf[2] = 0;
}
#else
buf[0] = (wchar_t)unicode;
buf[1] = 0;
#endif
chr_end++;
return buf;
}
const wchar_t* entity_w = sgml2uni(text + start + 1, n);
if (entity_w) { if (entity_w) {
chr_end++; chr_end++;
return entity_w; return entity_w;
@ -361,7 +340,7 @@ namespace stdex
m_invert(invert) m_invert(invert)
{ {
_Assume_(chr || !count); _Assume_(chr || !count);
wchar_t buf[3]; wchar_t buf[5];
size_t chr_end; size_t chr_end;
m_chr.assign(count ? next_sgml_cp(chr, 0, count, chr_end, buf) : L""); m_chr.assign(count ? next_sgml_cp(chr, 0, count, chr_end, buf) : L"");
} }
@ -375,7 +354,7 @@ namespace stdex
{ {
_Assume_(text || start >= end); _Assume_(text || start >= end);
if (start < end && text[start]) { if (start < end && text[start]) {
wchar_t buf[3]; wchar_t buf[5];
const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf); const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf);
bool r = ((flags & match_case_insensitive) ? bool r = ((flags & match_case_insensitive) ?
stdex::strnicmp(chr, stdex::strlen(chr), m_chr.data(), m_chr.size(), m_locale) : stdex::strnicmp(chr, stdex::strlen(chr), m_chr.data(), m_chr.size(), m_locale) :
@ -456,7 +435,7 @@ namespace stdex
{ {
_Assume_(text || start >= end); _Assume_(text || start >= end);
if (start < end && text[start]) { if (start < end && text[start]) {
wchar_t buf[3]; wchar_t buf[5];
const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf); const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf);
const wchar_t* chr_end = chr + stdex::strlen(chr); const wchar_t* chr_end = chr + stdex::strlen(chr);
bool r = bool r =
@ -534,7 +513,7 @@ namespace stdex
{ {
_Assume_(text || start >= end); _Assume_(text || start >= end);
if (start < end && text[start]) { if (start < end && text[start]) {
wchar_t buf[3]; wchar_t buf[5];
const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf); const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf);
const wchar_t* chr_end = chr + stdex::strlen(chr); const wchar_t* chr_end = chr + stdex::strlen(chr);
bool r = std::use_facet<std::ctype<wchar_t>>(m_locale).scan_not(std::ctype_base::punct, chr, chr_end) == chr_end; bool r = std::use_facet<std::ctype<wchar_t>>(m_locale).scan_not(std::ctype_base::punct, chr, chr_end) == chr_end;
@ -611,7 +590,7 @@ namespace stdex
{ {
_Assume_(text || start >= end); _Assume_(text || start >= end);
if (start < end && text[start]) { if (start < end && text[start]) {
wchar_t buf[3]; wchar_t buf[5];
const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf); const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf);
const wchar_t* chr_end = chr + stdex::strlen(chr); const wchar_t* chr_end = chr + stdex::strlen(chr);
bool r = bool r =
@ -807,7 +786,7 @@ namespace stdex
{ {
_Assume_(text || start >= end); _Assume_(text || start >= end);
if (start < end && text[start]) { if (start < end && text[start]) {
wchar_t buf[3]; wchar_t buf[5];
const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf); const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf);
const wchar_t* set = m_set.data(); const wchar_t* set = m_set.data();
size_t r = (flags & match_case_insensitive) ? size_t r = (flags & match_case_insensitive) ?
@ -906,7 +885,7 @@ namespace stdex
this->interval.invalidate(); this->interval.invalidate();
return false; return false;
} }
wchar_t buf[3]; wchar_t buf[5];
const wchar_t* chr = next_sgml_cp(text, this->interval.end, end, this->interval.end, buf); const wchar_t* chr = next_sgml_cp(text, this->interval.end, end, this->interval.end, buf);
for (; *chr; ++str, ++chr) { for (; *chr; ++str, ++chr) {
if (!*str || if (!*str ||
@ -2551,7 +2530,7 @@ namespace stdex
{ {
_Assume_(text || start >= end); _Assume_(text || start >= end);
if (start < end && text[start]) { if (start < end && text[start]) {
wchar_t buf[3]; wchar_t buf[5];
const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf); const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf);
const wchar_t* chr_end = chr + stdex::strlen(chr); const wchar_t* chr_end = chr + stdex::strlen(chr);
if (((chr[0] == L'-' || if (((chr[0] == L'-' ||
@ -2865,7 +2844,7 @@ namespace stdex
{ {
_Assume_(text || start >= end); _Assume_(text || start >= end);
if (start < end && text[start]) { if (start < end && text[start]) {
wchar_t buf[3]; wchar_t buf[5];
const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf); const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf);
const wchar_t* chr_end = chr + stdex::strlen(chr); const wchar_t* chr_end = chr + stdex::strlen(chr);
if ((('A' <= chr[0] && chr[0] <= 'Z') || if ((('A' <= chr[0] && chr[0] <= 'Z') ||
@ -3041,7 +3020,7 @@ namespace stdex
{ {
_Assume_(text || start >= end); _Assume_(text || start >= end);
if (start < end && text[start]) { if (start < end && text[start]) {
wchar_t buf[3]; wchar_t buf[5];
const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf); const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf);
const wchar_t* chr_end = chr + stdex::strlen(chr); const wchar_t* chr_end = chr + stdex::strlen(chr);
if (((chr[0] == L'-' || if (((chr[0] == L'-' ||
@ -3143,7 +3122,7 @@ namespace stdex
{ {
_Assume_(text || start >= end); _Assume_(text || start >= end);
if (start < end && text[start]) { if (start < end && text[start]) {
wchar_t buf[3]; wchar_t buf[5];
const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf); const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf);
const wchar_t* chr_end = chr + stdex::strlen(chr); const wchar_t* chr_end = chr + stdex::strlen(chr);
if (((chr[0] == L'-' || if (((chr[0] == L'-' ||
@ -3249,7 +3228,7 @@ namespace stdex
{ {
_Assume_(text || start >= end); _Assume_(text || start >= end);
if (start < end && text[start]) { if (start < end && text[start]) {
wchar_t buf[3]; wchar_t buf[5];
const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf); const wchar_t* chr = next_sgml_cp(text, start, end, this->interval.end, buf);
const wchar_t* chr_end = chr + stdex::strlen(chr); const wchar_t* chr_end = chr + stdex::strlen(chr);
if (((chr[0] == L'/' || if (((chr[0] == L'/' ||

View File

@ -23,11 +23,11 @@ namespace stdex
{ {
/// \cond internal /// \cond internal
template <class T> template <class T>
const wchar_t* sgml2uni(_In_reads_or_z_(count) const T* entity, _In_ size_t count) const utf32_t* sgml2uni(_In_reads_or_z_(count) const T* entity, _In_ size_t count, utf32_t buf[2])
{ {
_Assume_(entity && count); _Assume_(entity && count);
_Assume_(count < 2 || entity[0] != '#'); // No numeric entities
if (count < 2 || entity[0] != '#') {
for (size_t i = 0, j = _countof(sgml_unicode); i < j; ) { for (size_t i = 0, j = _countof(sgml_unicode); i < j; ) {
size_t m = (i + j) / 2; size_t m = (i + j) / 2;
if (sgml_unicode[m].sgml[0] < entity[0]) if (sgml_unicode[m].sgml[0] < entity[0])
@ -49,6 +49,37 @@ namespace stdex
return nullptr; return nullptr;
} }
buf[0] = entity[1] == 'x' || entity[1] == 'X' ?
static_cast<utf32_t>(strtou32(&entity[2], count - 2, nullptr, 16)) :
static_cast<utf32_t>(strtou32(&entity[1], count - 1, nullptr, 10));
buf[1] = 0;
return buf;
}
inline const utf16_t* utf32_to_wstr(_In_opt_z_ const utf32_t* str, utf16_t* buf)
{
if (!str)
return nullptr;
for (size_t i = 0, j = 0;; ++i) {
if (!str[i]) {
buf[j] = 0;
return buf;
}
if (str[i] < 0x10000)
buf[j++] = static_cast<utf16_t>(str[i]);
else {
ucs4_to_surrogate_pair(&buf[j], str[i]);
j += 2;
}
}
}
inline const utf32_t* utf32_to_wstr(_In_opt_z_ const utf32_t* str, utf32_t* buf)
{
_Unreferenced_(buf);
return str;
}
template <class T> template <class T>
const T* sgmlend( const T* sgmlend(
_In_reads_or_z_opt_(count) const T* str, _In_ size_t count) _In_reads_or_z_opt_(count) const T* str, _In_ size_t count)
@ -104,35 +135,11 @@ namespace stdex
for (size_t i = 0; i < count_src && src[i];) { for (size_t i = 0; i < count_src && src[i];) {
if (src[i] == '&') { if (src[i] == '&') {
auto end = sgmlend(src + i + 1, count_src - i - 1); auto end = sgmlend(&src[i + 1], count_src - i - 1);
if (end) { if (end) {
const wchar_t* entity_w; utf32_t chr[2];
wchar_t chr[3];
size_t n = end - src - i - 1; size_t n = end - src - i - 1;
if (n >= 2 && src[i + 1] == '#') { auto entity_w = sgml2uni(&src[i + 1], n, chr);
utf32_t unicode;
if (src[i + 2] == 'x' || src[i + 2] == 'X')
unicode = strtou32(src + i + 3, n - 2, nullptr, 16);
else
unicode = strtou32(src + i + 2, n - 1, nullptr, 10);
#ifdef _WIN32
if (unicode < 0x10000) {
chr[0] = (wchar_t)unicode;
chr[1] = 0;
}
else {
ucs4_to_surrogate_pair(chr, unicode);
chr[2] = 0;
}
#else
chr[0] = (wchar_t)unicode;
chr[1] = 0;
#endif
entity_w = chr;
}
else
entity_w = sgml2uni(src + i + 1, n);
if (entity_w) { if (entity_w) {
i = end - src + 1; i = end - src + 1;
continue; continue;
@ -157,7 +164,7 @@ namespace stdex
} }
/// ///
/// Convert SGML string to Unicode (UTF-16 on Windows) and append to string /// Convert SGML string to Unicode and append to string
/// ///
/// \param[in,out] dst String to append Unicode to /// \param[in,out] dst String to append Unicode to
/// \param[in] src SGML string /// \param[in] src SGML string
@ -166,9 +173,9 @@ namespace stdex
/// \param[in] offset Logical starting offset of source and destination strings. Unused when map parameter is nullptr. /// \param[in] offset Logical starting offset of source and destination strings. Unused when map parameter is nullptr.
/// \param[in,out] map The vector to append index mapping between source and destination string to. /// \param[in,out] map The vector to append index mapping between source and destination string to.
/// ///
template <class T_from, class TR_to = std::char_traits<wchar_t>, class AX_to = std::allocator<wchar_t>> template <class T_to = wchar_t, class T_from, class TR_to = std::char_traits<T_to>, class AX_to = std::allocator<T_to>>
void sgml2strcat( void sgml2strcat(
_Inout_ std::basic_string<wchar_t, TR_to, AX_to>& dst, _Inout_ std::basic_string<T_to, TR_to, AX_to>& dst,
_In_reads_or_z_opt_(count_src) const T_from* src, _In_ size_t count_src, _In_reads_or_z_opt_(count_src) const T_from* src, _In_ size_t count_src,
_In_ int skip = 0, _In_ int skip = 0,
_In_ const mapping<size_t>& offset = mapping<size_t>(0, 0), _In_ const mapping<size_t>& offset = mapping<size_t>(0, 0),
@ -194,49 +201,26 @@ namespace stdex
dst.reserve(dst.size() + count_src); dst.reserve(dst.size() + count_src);
for (size_t i = 0; i < count_src;) { for (size_t i = 0; i < count_src;) {
if (src[i] == '&') { if (src[i] == '&') {
auto end = sgmlend(src + i + 1, count_src - i - 1); auto end = sgmlend(&src[i + 1], count_src - i - 1);
if (end) { if (end) {
const wchar_t* entity_w; utf32_t chr32[2];
wchar_t chr[3]; _Assume_(&src[i + 1] <= end);
_Assume_(src + i + 1 <= end);
size_t n = static_cast<size_t>(end - src) - i - 1; size_t n = static_cast<size_t>(end - src) - i - 1;
if (n >= 2 && src[i + 1] == '#') { T_to chr[5];
utf32_t unicode; auto entity_w = utf32_to_wstr(sgml2uni(&src[i + 1], n, chr32), chr);
if (src[i + 2] == 'x' || src[i + 2] == 'X')
unicode = static_cast<utf32_t>(strtou32(src + i + 3, n - 2, nullptr, 16));
else
unicode = static_cast<utf32_t>(strtou32(src + i + 2, n - 1, nullptr, 10));
#ifdef _WIN32
if (unicode < 0x10000) {
chr[0] = (wchar_t)unicode;
chr[1] = 0;
}
else {
ucs4_to_surrogate_pair(chr, unicode);
chr[2] = 0;
}
#else
chr[0] = (wchar_t)unicode;
chr[1] = 0;
#endif
entity_w = chr;
}
else
entity_w = sgml2uni(src + i + 1, n);
if (entity_w && if (entity_w &&
(skip_quot || (entity_w[0] != L'"')) && (skip_quot || (entity_w[0] != '"')) &&
(skip_apos || (entity_w[0] != L'\'')) && (skip_apos || (entity_w[0] != '\'')) &&
(skip_amp || (entity_w[0] != L'&')) && (skip_amp || (entity_w[0] != '&')) &&
(skip_lt_gt || (entity_w[0] != L'<' && entity_w[0] != L'>')) && (skip_lt_gt || (entity_w[0] != '<' && entity_w[0] != '>')) &&
(skip_bsol || (entity_w[0] != L'\\')) && (skip_bsol || (entity_w[0] != '\\')) &&
(skip_dollar || (entity_w[0] != L'$')) && (skip_dollar || (entity_w[0] != '$')) &&
(skip_percnt || (entity_w[0] != L'%')) && (skip_percnt || (entity_w[0] != '%')) &&
(skip_commat || (entity_w[0] != L'@')) && (skip_commat || (entity_w[0] != '@')) &&
(skip_num || (entity_w[0] != L'#')) && (skip_num || (entity_w[0] != '#')) &&
(skip_lpar_rpar || (entity_w[0] != L'(' && entity_w[0] != L')')) && (skip_lpar_rpar || (entity_w[0] != '(' && entity_w[0] != ')')) &&
(skip_lcub_rcub || (entity_w[0] != L'{' && entity_w[0] != L'}')) && (skip_lcub_rcub || (entity_w[0] != '{' && entity_w[0] != '}')) &&
(skip_lsqb_rsqb || (entity_w[0] != L'[' && entity_w[0] != L']'))) (skip_lsqb_rsqb || (entity_w[0] != '[' && entity_w[0] != ']')))
{ {
if (map) map->push_back(mapping<size_t>(offset.from + i, offset.to + dst.size())); if (map) map->push_back(mapping<size_t>(offset.from + i, offset.to + dst.size()));
dst.append(entity_w); dst.append(entity_w);
@ -252,7 +236,7 @@ namespace stdex
} }
/// ///
/// Convert SGML string to Unicode (UTF-16 on Windows) and append to string /// Convert SGML string to Unicode and append to string
/// ///
/// \param[in,out] dst String to append Unicode to /// \param[in,out] dst String to append Unicode to
/// \param[in] src SGML string /// \param[in] src SGML string
@ -260,9 +244,9 @@ namespace stdex
/// \param[in] offset Logical starting offset of source and destination strings. Unused when map parameter is nullptr. /// \param[in] offset Logical starting offset of source and destination strings. Unused when map parameter is nullptr.
/// \param[in,out] map The vector to append index mapping between source and destination string to. /// \param[in,out] map The vector to append index mapping between source and destination string to.
/// ///
template <class T_from, class TR_to = std::char_traits<wchar_t>, class AX_to = std::allocator<wchar_t>, class TR_from = std::char_traits<T_from>, class AX_from = std::allocator<T_from>> template <class T_to = wchar_t, class T_from, class TR_to = std::char_traits<T_to>, class AX_to = std::allocator<T_to>, class TR_from = std::char_traits<T_from>, class AX_from = std::allocator<T_from>>
void sgml2strcat( void sgml2strcat(
_Inout_ std::basic_string<wchar_t, TR_to, AX_to>& dst, _Inout_ std::basic_string<T_to, TR_to, AX_to>& dst,
_In_ const std::basic_string<T_from, TR_from, AX_from>& src, _In_ const std::basic_string<T_from, TR_from, AX_from>& src,
_In_ int skip = 0, _In_ int skip = 0,
_In_ const mapping<size_t>& offset = mapping<size_t>(0, 0), _In_ const mapping<size_t>& offset = mapping<size_t>(0, 0),
@ -272,7 +256,7 @@ namespace stdex
} }
/// ///
/// Convert SGML string to Unicode (UTF-16 on Windows) and append to string /// Convert SGML string to Unicode and append to string
/// ///
/// \param[in,out] dst String to append Unicode to /// \param[in,out] dst String to append Unicode to
/// \param[in] count_dst Unicode string character count limit. Function throws std::invalid_argument if there is not enough space in Unicode string (including space for zero-terminator). /// \param[in] count_dst Unicode string character count limit. Function throws std::invalid_argument if there is not enough space in Unicode string (including space for zero-terminator).
@ -284,9 +268,9 @@ namespace stdex
/// ///
/// \return Final length of SGML string in code points excluding zero-terminator /// \return Final length of SGML string in code points excluding zero-terminator
/// ///
template <class T_from> template <class T_to = wchar_t, class T_from>
size_t sgml2strcat( size_t sgml2strcat(
_Inout_cap_(count_dst) wchar_t* dst, _In_ size_t count_dst, _Inout_cap_(count_dst) T_to* dst, _In_ size_t count_dst,
_In_reads_or_z_opt_(count_src) const T_from* src, _In_ size_t count_src, _In_reads_or_z_opt_(count_src) const T_from* src, _In_ size_t count_src,
_In_ int skip = 0, _In_ int skip = 0,
_In_ const mapping<size_t>& offset = mapping<size_t>(0, 0), _In_ const mapping<size_t>& offset = mapping<size_t>(0, 0),
@ -314,54 +298,31 @@ namespace stdex
count_src = strnlen(src, count_src); count_src = strnlen(src, count_src);
for (size_t i = 0; i < count_src;) { for (size_t i = 0; i < count_src;) {
if (src[i] == '&') { if (src[i] == '&') {
auto end = sgmlend(src + i + 1, count_src - i - 1); auto end = sgmlend(&src[i + 1], count_src - i - 1);
if (end) { if (end) {
const wchar_t* entity_w; utf32_t chr32[2];
wchar_t chr[3]; T_to chr[5];
size_t n = end - src - i - 1; size_t n = end - src - i - 1;
if (n >= 2 && src[i + 1] == '#') { auto entity_w = utf32_to_wstr(sgml2uni(&src[i + 1], n, chr32), chr);
utf32_t unicode;
if (src[i + 2] == 'x' || src[i + 2] == 'X')
unicode = strtou32(src + i + 3, n - 2, nullptr, 16);
else
unicode = strtou32(src + i + 2, n - 1, nullptr, 10);
#ifdef _WIN32
if (unicode < 0x10000) {
chr[0] = (wchar_t)unicode;
chr[1] = 0;
}
else {
ucs4_to_surrogate_pair(chr, unicode);
chr[2] = 0;
}
#else
chr[0] = (wchar_t)unicode;
chr[1] = 0;
#endif
entity_w = chr;
}
else
entity_w = sgml2uni(src + i + 1, n);
if (entity_w && if (entity_w &&
(skip_quot || (entity_w[0] != L'"')) && (skip_quot || (entity_w[0] != '"')) &&
(skip_apos || (entity_w[0] != L'\'')) && (skip_apos || (entity_w[0] != '\'')) &&
(skip_amp || (entity_w[0] != L'&')) && (skip_amp || (entity_w[0] != '&')) &&
(skip_lt_gt || (entity_w[0] != L'<' && entity_w[0] != L'>')) && (skip_lt_gt || (entity_w[0] != '<' && entity_w[0] != '>')) &&
(skip_bsol || (entity_w[0] != L'\\')) && (skip_bsol || (entity_w[0] != '\\')) &&
(skip_dollar || (entity_w[0] != L'$')) && (skip_dollar || (entity_w[0] != '$')) &&
(skip_percnt || (entity_w[0] != L'%')) && (skip_percnt || (entity_w[0] != '%')) &&
(skip_commat || (entity_w[0] != L'@')) && (skip_commat || (entity_w[0] != '@')) &&
(skip_num || (entity_w[0] != L'#')) && (skip_num || (entity_w[0] != '#')) &&
(skip_lpar_rpar || (entity_w[0] != L'(' && entity_w[0] != L')')) && (skip_lpar_rpar || (entity_w[0] != '(' && entity_w[0] != ')')) &&
(skip_lcub_rcub || (entity_w[0] != L'{' && entity_w[0] != L'}')) && (skip_lcub_rcub || (entity_w[0] != '{' && entity_w[0] != '}')) &&
(skip_lsqb_rsqb || (entity_w[0] != L'[' && entity_w[0] != L']'))) (skip_lsqb_rsqb || (entity_w[0] != '[' && entity_w[0] != ']')))
{ {
if (map) map->push_back(mapping<size_t>(offset.from + i, offset.to + j)); if (map) map->push_back(mapping<size_t>(offset.from + i, offset.to + j));
size_t m = wcslen(entity_w); size_t m = strlen(entity_w);
if (j + m >= count_dst) if (j + m >= count_dst)
throw buffer_overrun; throw buffer_overrun;
memcpy(dst + j, entity_w, m * sizeof(wchar_t)); j += m; memcpy(dst + j, entity_w, m * sizeof(*entity_w)); j += m;
i = end - src + 1; i = end - src + 1;
if (map) map->push_back(mapping<size_t>(offset.from + i, offset.to + j)); if (map) map->push_back(mapping<size_t>(offset.from + i, offset.to + j));
continue; continue;
@ -379,7 +340,7 @@ namespace stdex
} }
/// ///
/// Convert SGML string to Unicode (UTF-16 on Windows) /// Convert SGML string to Unicode
/// ///
/// \param[in,out] dst String to write Unicode to /// \param[in,out] dst String to write Unicode to
/// \param[in] src SGML string /// \param[in] src SGML string
@ -388,9 +349,9 @@ namespace stdex
/// \param[in] offset Logical starting offset of source and destination strings. Unused when map parameter is nullptr. /// \param[in] offset Logical starting offset of source and destination strings. Unused when map parameter is nullptr.
/// \param[in,out] map The vector to write index mapping between source and destination string to. /// \param[in,out] map The vector to write index mapping between source and destination string to.
/// ///
template <class T_from, class TR_to = std::char_traits<wchar_t>, class AX_to = std::allocator<wchar_t>> template <class T_to = wchar_t, class T_from, class TR_to = std::char_traits<T_to>, class AX_to = std::allocator<T_to>>
void sgml2strcpy( void sgml2strcpy(
_Inout_ std::basic_string<wchar_t, TR_to, AX_to>& dst, _Inout_ std::basic_string<T_to, TR_to, AX_to>& dst,
_In_reads_or_z_opt_(count_src) const T_from* src, _In_ size_t count_src, _In_reads_or_z_opt_(count_src) const T_from* src, _In_ size_t count_src,
_In_ int skip = 0, _In_ int skip = 0,
_In_ const mapping<size_t>& offset = mapping<size_t>(0, 0), _In_ const mapping<size_t>& offset = mapping<size_t>(0, 0),
@ -403,7 +364,7 @@ namespace stdex
} }
/// ///
/// Convert SGML string to Unicode (UTF-16 on Windows) /// Convert SGML string to Unicode
/// ///
/// \param[in,out] dst String to write Unicode to /// \param[in,out] dst String to write Unicode to
/// \param[in] src SGML string /// \param[in] src SGML string
@ -411,9 +372,9 @@ namespace stdex
/// \param[in] offset Logical starting offset of source and destination strings. Unused when map parameter is nullptr. /// \param[in] offset Logical starting offset of source and destination strings. Unused when map parameter is nullptr.
/// \param[in,out] map The vector to write index mapping between source and destination string to. /// \param[in,out] map The vector to write index mapping between source and destination string to.
/// ///
template<class T_from, class TR_to = std::char_traits<wchar_t>, class AX_to = std::allocator<wchar_t>, class TR_from = std::char_traits<T_from>, class AX_from = std::allocator<T_from>> template<class T_to = wchar_t, class T_from, class TR_to = std::char_traits<T_to>, class AX_to = std::allocator<T_to>, class TR_from = std::char_traits<T_from>, class AX_from = std::allocator<T_from>>
void sgml2strcpy( void sgml2strcpy(
_Inout_ std::basic_string<wchar_t, TR_to, AX_to>& dst, _Inout_ std::basic_string<T_to, TR_to, AX_to>& dst,
_In_ const std::basic_string<T_from, TR_from, AX_from>& src, _In_ const std::basic_string<T_from, TR_from, AX_from>& src,
_In_ int skip = 0, _In_ int skip = 0,
_In_ const mapping<size_t>& offset = mapping<size_t>(0, 0), _In_ const mapping<size_t>& offset = mapping<size_t>(0, 0),
@ -423,7 +384,7 @@ namespace stdex
} }
/// ///
/// Convert SGML string to Unicode (UTF-16 on Windows) /// Convert SGML string to Unicode
/// ///
/// \param[in,out] dst String to write Unicode to /// \param[in,out] dst String to write Unicode to
/// \param[in] count_dst Unicode string character count limit. Function throws std::invalid_argument if there is not enough space in Unicode string (including space for zero-terminator). /// \param[in] count_dst Unicode string character count limit. Function throws std::invalid_argument if there is not enough space in Unicode string (including space for zero-terminator).
@ -435,9 +396,9 @@ namespace stdex
/// ///
/// \return Final length of SGML string in code points excluding zero-terminator /// \return Final length of SGML string in code points excluding zero-terminator
/// ///
template <class T_from> template <class T_to = wchar_t, class T_from>
size_t sgml2strcpy( size_t sgml2strcpy(
_Inout_cap_(count_dst) wchar_t* dst, _In_ size_t count_dst, _Inout_cap_(count_dst) T_to* dst, _In_ size_t count_dst,
_In_reads_or_z_opt_(count_src) const T_from* src, _In_ size_t count_src, _In_reads_or_z_opt_(count_src) const T_from* src, _In_ size_t count_src,
_In_ int skip = 0, _In_ int skip = 0,
_In_ const mapping<size_t>& offset = mapping<size_t>(0, 0), _In_ const mapping<size_t>& offset = mapping<size_t>(0, 0),
@ -452,7 +413,7 @@ namespace stdex
} }
/// ///
/// Convert SGML string to Unicode string (UTF-16 on Windows) /// Convert SGML string to Unicode string
/// ///
/// \param[in] src SGML string /// \param[in] src SGML string
/// \param[in] count_src SGML string character count limit /// \param[in] count_src SGML string character count limit
@ -462,14 +423,14 @@ namespace stdex
/// ///
/// \return Unicode string /// \return Unicode string
/// ///
template <class T_from> template <class T_to = wchar_t, class T_from, class TR_to = std::char_traits<T_to>, class AX_to = std::allocator<T_to>>
std::wstring sgml2str( std::basic_string<T_to, TR_to, AX_to> sgml2str(
_In_reads_or_z_opt_(count_src) const T_from* src, _In_ size_t count_src, _In_reads_or_z_opt_(count_src) const T_from* src, _In_ size_t count_src,
_In_ int skip = 0, _In_ int skip = 0,
_In_ const mapping<size_t>& offset = mapping<size_t>(0, 0), _In_ const mapping<size_t>& offset = mapping<size_t>(0, 0),
_Inout_opt_ mapping_vector<size_t>* map = nullptr) _Inout_opt_ mapping_vector<size_t>* map = nullptr)
{ {
std::wstring dst; std::basic_string<T_to, TR_to, AX_to> dst;
sgml2strcat(dst, src, count_src, skip, offset, map); sgml2strcat(dst, src, count_src, skip, offset, map);
return dst; return dst;
} }
@ -484,25 +445,61 @@ namespace stdex
/// ///
/// \return Unicode string /// \return Unicode string
/// ///
template <class T_from, class TR_from = std::char_traits<T_from>, class AX_from = std::allocator<T_from>> template <class T_to = wchar_t, class T_from, class TR_to = std::char_traits<T_to>, class AX_to = std::allocator<T_to>, class TR_from = std::char_traits<T_from>, class AX_from = std::allocator<T_from>>
std::wstring sgml2str( std::basic_string<T_to, TR_to, AX_to> sgml2str(
_In_ const std::basic_string<T_from, TR_from, AX_from>& src, _In_ const std::basic_string<T_from, TR_from, AX_from>& src,
_In_ int skip = 0, _In_ int skip = 0,
_In_ const mapping<size_t>& offset = mapping<size_t>(0, 0), _In_ const mapping<size_t>& offset = mapping<size_t>(0, 0),
_Inout_opt_ mapping_vector<size_t>* map = nullptr) _Inout_opt_ mapping_vector<size_t>* map = nullptr)
{ {
return sgml2str(src.data(), src.size(), skip, offset, map); return sgml2str<T_to, T_from, TR_to, AX_to>(src.data(), src.size(), skip, offset, map);
} }
/// \cond internal /// \cond internal
inline const char* chr2sgml(_In_reads_or_z_(count) const wchar_t* entity, _In_ size_t count) inline const char* chr2sgml(_In_reads_or_z_(count) const utf16_t* entity, _In_ size_t count)
{ {
_Assume_(entity && count); _Assume_(entity && count);
const wchar_t e2 = entity[0]; utf32_t e2;
size_t offset;
if (count < 2 || !is_surrogate_pair(entity)) {
e2 = static_cast<utf32_t>(entity[0]);
offset = 1;
}
else {
e2 = surrogate_pair_to_ucs4(entity);
offset = 2;
}
for (size_t i = 0, j = _countof(unicode_sgml); i < j; ) { for (size_t i = 0, j = _countof(unicode_sgml); i < j; ) {
size_t m = (i + j) / 2; size_t m = (i + j) / 2;
wchar_t e1 = sgml_unicode[unicode_sgml[m]].unicode[0]; auto e1 = sgml_unicode[unicode_sgml[m]].unicode[0];
if (e1 < e2)
i = m + 1;
else if (e1 > e2)
j = m;
else {
auto r = strncmp(sgml_unicode[unicode_sgml[m]].unicode + 1, _countof(sgml_unicode[0].unicode) - 1, entity + offset, count - offset);
if (r < 0)
i = m + 1;
else if (r > 0)
j = m;
else {
for (; i < m && sgml_unicode[unicode_sgml[m - 1]].unicode[0] == e2 && strncmp(sgml_unicode[unicode_sgml[m - 1]].unicode + 1, _countof(sgml_unicode[0].unicode) - 1, entity + offset, count - offset) == 0; m--);
return sgml_unicode[unicode_sgml[m]].sgml;
}
}
}
return nullptr;
}
inline const char* chr2sgml(_In_reads_or_z_(count) const utf32_t* entity, _In_ size_t count)
{
_Assume_(entity && count);
utf32_t e2 = entity[0];
for (size_t i = 0, j = _countof(unicode_sgml); i < j; ) {
size_t m = (i + j) / 2;
auto e1 = sgml_unicode[unicode_sgml[m]].unicode[0];
if (e1 < e2) if (e1 < e2)
i = m + 1; i = m + 1;
else if (e1 > e2) else if (e1 > e2)
@ -521,20 +518,37 @@ namespace stdex
} }
return nullptr; return nullptr;
} }
inline utf32_t wstr_to_utf32(_In_reads_(end) const utf16_t* src, _Inout_ size_t& i, _In_ size_t end)
{
_Assume_(i < end);
if (i + 1 >= end || !is_surrogate_pair(src + i))
return src[i++];
utf32_t unicode = surrogate_pair_to_ucs4(src + i);
i += 2;
return unicode;
}
inline utf32_t wstr_to_utf32(_In_reads_(end) const utf32_t* src, _Inout_ size_t& i, _In_ size_t end)
{
_Assume_(i < end);
return src[i++];
}
/// \endcond /// \endcond
/// ///
/// Convert Unicode string (UTF-16 on Windows) to SGML and append to string /// Convert Unicode string to SGML and append to string
/// ///
/// \param[in,out] dst String to append SGML to /// \param[in,out] dst String to append SGML to
/// \param[in] src Unicode string /// \param[in] src Unicode string
/// \param[in] count_src Unicode string character count limit /// \param[in] count_src Unicode string character count limit
/// \param[in] what Bitwise flag of stdex::sgml_* constants that force extra characters otherwise not converted to SGML /// \param[in] what Bitwise flag of stdex::sgml_* constants that force extra characters otherwise not converted to SGML
/// ///
template <class TR_to = std::char_traits<char>, class AX_to = std::allocator<char>> template <class T_from = wchar_t, class TR_to = std::char_traits<char>, class AX_to = std::allocator<char>>
inline void str2sgmlcat( void str2sgmlcat(
_Inout_ std::basic_string<char, TR_to, AX_to>& dst, _Inout_ std::basic_string<char, TR_to, AX_to>& dst,
_In_reads_or_z_opt_(count_src) const wchar_t* src, _In_ size_t count_src, _In_reads_or_z_opt_(count_src) const T_from* src, _In_ size_t count_src,
_In_ int what = 0) _In_ int what = 0)
{ {
_Assume_(src || !count_src); _Assume_(src || !count_src);
@ -559,18 +573,18 @@ namespace stdex
size_t n = glyphlen(src + i, count_src - i); size_t n = glyphlen(src + i, count_src - i);
if (n == 1 && if (n == 1 &&
do_ascii && is7bit(src[i]) && do_ascii && is7bit(src[i]) &&
src[i] != L'&' && src[i] != '&' &&
(do_quot || (src[i] != L'"')) && (do_quot || (src[i] != '"')) &&
(do_apos || (src[i] != L'\'')) && (do_apos || (src[i] != '\'')) &&
(do_lt_gt || (src[i] != L'<' && src[i] != L'>')) && (do_lt_gt || (src[i] != '<' && src[i] != '>')) &&
(do_bsol || (src[i] != L'\\')) && (do_bsol || (src[i] != '\\')) &&
(do_dollar || (src[i] != L'$')) && (do_dollar || (src[i] != '$')) &&
(do_percnt || (src[i] != L'%')) && (do_percnt || (src[i] != '%')) &&
(do_commat || (src[i] != L'@')) && (do_commat || (src[i] != '@')) &&
(do_num || (src[i] != L'#')) && (do_num || (src[i] != '#')) &&
(do_lpar_rpar || (src[i] != L'(' && src[i] != L')')) && (do_lpar_rpar || (src[i] != '(' && src[i] != ')')) &&
(do_lcub_rcub || (src[i] != L'{' && src[i] != L'}')) && (do_lcub_rcub || (src[i] != '{' && src[i] != '}')) &&
(do_lsqb_rsqb || (src[i] != L'[' && src[i] != L']'))) (do_lsqb_rsqb || (src[i] != '[' && src[i] != ']')))
{ {
// 7-bit ASCII and no desire to encode it as an SGML entity. // 7-bit ASCII and no desire to encode it as an SGML entity.
dst.append(1, static_cast<char>(src[i++])); dst.append(1, static_cast<char>(src[i++]));
@ -589,7 +603,7 @@ namespace stdex
dst.append(1, static_cast<char>(src[i++])); dst.append(1, static_cast<char>(src[i++]));
else { else {
char tmp[3 + 8 + 1 + 1]; char tmp[3 + 8 + 1 + 1];
snprintf(tmp, _countof(tmp), "&#x%x;", src[i++]); snprintf(tmp, _countof(tmp), "&#x%x;", static_cast<unsigned int>(src[i++]));
dst.append(tmp); dst.append(tmp);
} }
} }
@ -606,19 +620,8 @@ namespace stdex
else if (is7bit(src[i])) else if (is7bit(src[i]))
dst.append(1, static_cast<char>(src[i++])); dst.append(1, static_cast<char>(src[i++]));
else { else {
utf32_t unicode;
#ifdef _WIN32
if (i + 1 < end && is_surrogate_pair(src + i)) {
unicode = surrogate_pair_to_ucs4(src + i);
i += 2;
}
else
#endif
{
unicode = src[i++];
}
char tmp[3 + 8 + 1 + 1]; char tmp[3 + 8 + 1 + 1];
snprintf(tmp, _countof(tmp), "&#x%x;", static_cast<unsigned int>(unicode)); snprintf(tmp, _countof(tmp), "&#x%x;", static_cast<unsigned int>(wstr_to_utf32(src, i, end)));
dst.append(tmp); dst.append(tmp);
} }
} }
@ -628,23 +631,23 @@ namespace stdex
} }
/// ///
/// Convert Unicode string (UTF-16 on Windows) to SGML and append to string /// Convert Unicode string to SGML and append to string
/// ///
/// \param[in,out] dst String to append SGML to /// \param[in,out] dst String to append SGML to
/// \param[in] src Unicode string /// \param[in] src Unicode string
/// \param[in] what Bitwise flag of stdex::sgml_* constants that force extra characters otherwise not converted to SGML /// \param[in] what Bitwise flag of stdex::sgml_* constants that force extra characters otherwise not converted to SGML
/// ///
template <class TR_to = std::char_traits<char>, class AX_to = std::allocator<char>> template <class T_from = wchar_t, class TR_to = std::char_traits<char>, class AX_to = std::allocator<char>>
void str2sgmlcat( void str2sgmlcat(
_Inout_ std::basic_string<char, TR_to, AX_to>& dst, _Inout_ std::basic_string<char, TR_to, AX_to>& dst,
_In_ const std::basic_string_view<wchar_t, std::char_traits<wchar_t>> src, _In_ const std::basic_string_view<T_from, std::char_traits<T_from>> src,
_In_ int what = 0) _In_ int what = 0)
{ {
str2sgmlcat(dst, src.data(), src.size(), what); str2sgmlcat(dst, src.data(), src.size(), what);
} }
/// ///
/// Convert Unicode string (UTF-16 on Windows) to SGML and append to string /// Convert Unicode string to SGML and append to string
/// ///
/// \param[in,out] dst String to append SGML to /// \param[in,out] dst String to append SGML to
/// \param[in] count_dst SGML string character count limit. Function throws std::invalid_argument if there is not enough space in SGML string (including space for zero-terminator). /// \param[in] count_dst SGML string character count limit. Function throws std::invalid_argument if there is not enough space in SGML string (including space for zero-terminator).
@ -654,9 +657,10 @@ namespace stdex
/// ///
/// \return Final length of SGML string in code points excluding zero-terminator /// \return Final length of SGML string in code points excluding zero-terminator
/// ///
inline size_t str2sgmlcat( template <class T_from = wchar_t>
size_t str2sgmlcat(
_Inout_cap_(count_dst) char* dst, _In_ size_t count_dst, _Inout_cap_(count_dst) char* dst, _In_ size_t count_dst,
_In_reads_or_z_opt_(count_src) const wchar_t* src, _In_ size_t count_src, _In_reads_or_z_opt_(count_src) const T_from* src, _In_ size_t count_src,
_In_ int what = 0) _In_ int what = 0)
{ {
_Assume_(dst || !count_dst); _Assume_(dst || !count_dst);
@ -683,18 +687,18 @@ namespace stdex
size_t n = glyphlen(src + i, count_src - i); size_t n = glyphlen(src + i, count_src - i);
if (n == 1 && if (n == 1 &&
do_ascii && is7bit(src[i]) && do_ascii && is7bit(src[i]) &&
src[i] != L'&' && src[i] != '&' &&
(do_quot || (src[i] != L'"')) && (do_quot || (src[i] != '"')) &&
(do_apos || (src[i] != L'\'')) && (do_apos || (src[i] != '\'')) &&
(do_lt_gt || (src[i] != L'<' && src[i] != L'>')) && (do_lt_gt || (src[i] != '<' && src[i] != '>')) &&
(do_bsol || (src[i] != L'\\')) && (do_bsol || (src[i] != '\\')) &&
(do_dollar || (src[i] != L'$')) && (do_dollar || (src[i] != '$')) &&
(do_percnt || (src[i] != L'%')) && (do_percnt || (src[i] != '%')) &&
(do_commat || (src[i] != L'@')) && (do_commat || (src[i] != '@')) &&
(do_num || (src[i] != L'#')) && (do_num || (src[i] != '#')) &&
(do_lpar_rpar || (src[i] != L'(' && src[i] != L')')) && (do_lpar_rpar || (src[i] != '(' && src[i] != ')')) &&
(do_lcub_rcub || (src[i] != L'{' && src[i] != L'}')) && (do_lcub_rcub || (src[i] != '{' && src[i] != '}')) &&
(do_lsqb_rsqb || (src[i] != L'[' && src[i] != L']'))) (do_lsqb_rsqb || (src[i] != '[' && src[i] != ']')))
{ {
// 7-bit ASCII and no desire to encode it as an SGML entity. // 7-bit ASCII and no desire to encode it as an SGML entity.
if (j + 1 >= count_dst) if (j + 1 >= count_dst)
@ -721,7 +725,7 @@ namespace stdex
} }
else { else {
char tmp[3 + 8 + 1 + 1]; char tmp[3 + 8 + 1 + 1];
int m = snprintf(tmp, _countof(tmp), "&#x%x;", src[i++]); int m = snprintf(tmp, _countof(tmp), "&#x%x;", static_cast<unsigned int>(src[i++]));
_Assume_(m >= 0); _Assume_(m >= 0);
if (static_cast<size_t>(m) >= count_dst) if (static_cast<size_t>(m) >= count_dst)
throw buffer_overrun; throw buffer_overrun;
@ -748,19 +752,8 @@ namespace stdex
dst[j++] = static_cast<char>(src[i++]); dst[j++] = static_cast<char>(src[i++]);
} }
else { else {
utf32_t unicode;
#ifdef _WIN32
if (i + 1 < end && is_surrogate_pair(src + i)) {
unicode = surrogate_pair_to_ucs4(src + i);
i += 2;
}
else
#endif
{
unicode = src[i++];
}
char tmp[3 + 8 + 1 + 1]; char tmp[3 + 8 + 1 + 1];
int m = snprintf(tmp, _countof(tmp), "&#x%x;", static_cast<unsigned int>(unicode)); int m = snprintf(tmp, _countof(tmp), "&#x%x;", static_cast<unsigned int>(wstr_to_utf32(src, i, end)));
_Assume_(m >= 0); _Assume_(m >= 0);
if (static_cast<size_t>(m) >= count_dst) if (static_cast<size_t>(m) >= count_dst)
throw buffer_overrun; throw buffer_overrun;
@ -778,17 +771,17 @@ namespace stdex
} }
/// ///
/// Convert Unicode string (UTF-16 on Windows) to SGML /// Convert Unicode string to SGML
/// ///
/// \param[in,out] dst String to write SGML to /// \param[in,out] dst String to write SGML to
/// \param[in] src Unicode string /// \param[in] src Unicode string
/// \param[in] count_src Unicode string character count limit /// \param[in] count_src Unicode string character count limit
/// \param[in] what Bitwise flag of stdex::sgml_* constants that force extra characters otherwise not converted to SGML /// \param[in] what Bitwise flag of stdex::sgml_* constants that force extra characters otherwise not converted to SGML
/// ///
template <class TR_to = std::char_traits<char>, class AX_to = std::allocator<char>> template <class T_from = wchar_t, class TR_to = std::char_traits<char>, class AX_to = std::allocator<char>>
inline void str2sgmlcpy( void str2sgmlcpy(
_Inout_ std::basic_string<char, TR_to, AX_to>& dst, _Inout_ std::basic_string<char, TR_to, AX_to>& dst,
_In_reads_or_z_opt_(count_src) const wchar_t* src, _In_ size_t count_src, _In_reads_or_z_opt_(count_src) const T_from* src, _In_ size_t count_src,
_In_ int what = 0) _In_ int what = 0)
{ {
dst.clear(); dst.clear();
@ -796,23 +789,23 @@ namespace stdex
} }
/// ///
/// Convert Unicode string (UTF-16 on Windows) to SGML /// Convert Unicode string to SGML
/// ///
/// \param[in,out] dst String to write SGML to /// \param[in,out] dst String to write SGML to
/// \param[in] src Unicode string /// \param[in] src Unicode string
/// \param[in] what Bitwise flag of stdex::sgml_* constants that force extra characters otherwise not converted to SGML /// \param[in] what Bitwise flag of stdex::sgml_* constants that force extra characters otherwise not converted to SGML
/// ///
template <class TR_to = std::char_traits<char>, class AX_to = std::allocator<char>> template <class T_from = wchar_t, class TR_to = std::char_traits<char>, class AX_to = std::allocator<char>>
void str2sgmlcpy( void str2sgmlcpy(
_Inout_ std::basic_string<char, TR_to, AX_to>& dst, _Inout_ std::basic_string<char, TR_to, AX_to>& dst,
_In_ const std::basic_string_view<wchar_t, std::char_traits<wchar_t>> src, _In_ const std::basic_string_view<T_from, std::char_traits<T_from>> src,
_In_ int what = 0) _In_ int what = 0)
{ {
str2sgmlcpy(dst, src.data(), src.size(), what); str2sgmlcpy(dst, src.data(), src.size(), what);
} }
/// ///
/// Convert Unicode string (UTF-16 on Windows) to SGML /// Convert Unicode string to SGML
/// ///
/// \param[in,out] dst String to write SGML to /// \param[in,out] dst String to write SGML to
/// \param[in] count_dst SGML string character count limit. Function throws std::invalid_argument if there is not enough space in SGML string (including space for zero-terminator). /// \param[in] count_dst SGML string character count limit. Function throws std::invalid_argument if there is not enough space in SGML string (including space for zero-terminator).
@ -822,9 +815,10 @@ namespace stdex
/// ///
/// \return Final length of SGML string in code points excluding zero-terminator /// \return Final length of SGML string in code points excluding zero-terminator
/// ///
inline size_t str2sgmlcpy( template <class T_from = wchar_t>
size_t str2sgmlcpy(
_Inout_cap_(count_dst) char* dst, _In_ size_t count_dst, _Inout_cap_(count_dst) char* dst, _In_ size_t count_dst,
_In_reads_or_z_opt_(count_src) const wchar_t* src, _In_ size_t count_src, _In_reads_or_z_opt_(count_src) const T_from* src, _In_ size_t count_src,
_In_ int what = 0) _In_ int what = 0)
{ {
_Assume_(dst || !count_dst); _Assume_(dst || !count_dst);
@ -834,7 +828,7 @@ namespace stdex
} }
/// ///
/// Convert Unicode string (UTF-16 on Windows) to SGML string /// Convert Unicode string to SGML string
/// ///
/// \param[in] src Unicode string /// \param[in] src Unicode string
/// \param[in] count_src Unicode string character count limit /// \param[in] count_src Unicode string character count limit
@ -842,8 +836,9 @@ namespace stdex
/// ///
/// \return SGML string /// \return SGML string
/// ///
inline std::string str2sgml( template <class T_from = wchar_t>
_In_reads_or_z_opt_(count_src) const wchar_t* src, _In_ size_t count_src, std::string str2sgml(
_In_reads_or_z_opt_(count_src) const T_from* src, _In_ size_t count_src,
_In_ int what = 0) _In_ int what = 0)
{ {
std::string dst; std::string dst;
@ -852,15 +847,16 @@ namespace stdex
} }
/// ///
/// Convert Unicode string (UTF-16 on Windows) to SGML string /// Convert Unicode string to SGML string
/// ///
/// \param[in] src Unicode string /// \param[in] src Unicode string
/// \param[in] what Bitwise flag of stdex::sgml_* constants that force extra characters otherwise not converted to SGML /// \param[in] what Bitwise flag of stdex::sgml_* constants that force extra characters otherwise not converted to SGML
/// ///
/// \return SGML string /// \return SGML string
/// ///
inline std::string str2sgml( template <class T_from = wchar_t>
_In_ const std::basic_string_view<wchar_t, std::char_traits<wchar_t>> src, std::string str2sgml(
_In_ const std::basic_string_view<T_from, std::char_traits<T_from>> src,
_In_ int what = 0) _In_ int what = 0)
{ {
return str2sgml(src.data(), src.size(), what); return str2sgml(src.data(), src.size(), what);

File diff suppressed because it is too large Load Diff

View File

@ -207,15 +207,28 @@ namespace stdex
/// \param[in] glyph Start of a glyph /// \param[in] glyph Start of a glyph
/// \param[in] count Code unit limit /// \param[in] count Code unit limit
/// ///
inline size_t glyphlen(_In_reads_or_z_opt_(count) const wchar_t* glyph, _In_ size_t count) inline size_t glyphlen(_In_reads_or_z_opt_(count) const utf16_t* glyph, _In_ size_t count)
{ {
_Assume_(glyph || !count); _Assume_(glyph || !count);
if (count) { if (count) {
#ifdef _WIN32
size_t i = count < 2 || !is_surrogate_pair(glyph) ? 1 : 2; size_t i = count < 2 || !is_surrogate_pair(glyph) ? 1 : 2;
#else for (; i < count && iscombining(glyph[i]); ++i);
return i;
}
return 0;
}
///
/// Return number of code units the glyph represents
///
/// \param[in] glyph Start of a glyph
/// \param[in] count Code unit limit
///
inline size_t glyphlen(_In_reads_or_z_opt_(count) const utf32_t* glyph, _In_ size_t count)
{
_Assume_(glyph || !count);
if (count) {
size_t i = 1; size_t i = 1;
#endif
for (; i < count && iscombining(glyph[i]); ++i); for (; i < count && iscombining(glyph[i]); ++i);
return i; return i;
} }
@ -228,17 +241,28 @@ namespace stdex
/// \param[in] str Start of a string /// \param[in] str Start of a string
/// \param[in] count Length of a string in code units /// \param[in] count Length of a string in code units
/// ///
inline size_t glyphrlen(_In_reads_or_z_opt_(count) const wchar_t* str, _In_ size_t count) inline size_t glyphrlen(_In_reads_or_z_opt_(count) const utf16_t* str, _In_ size_t count)
{ {
_Assume_(count && str && str[count - 1]); _Assume_(count && str && str[count - 1]);
for (size_t i = count; i--;) { for (size_t i = count; i--;) {
if (!iscombining(str[i])) { if (!iscombining(str[i]))
#ifdef _WIN32
return count - (!is_low_surrogate(str[i]) || i == 0 || !is_high_surrogate(str[i - 1]) ? i : i - 1); return count - (!is_low_surrogate(str[i]) || i == 0 || !is_high_surrogate(str[i - 1]) ? i : i - 1);
#else
return count - i;
#endif
} }
return count;
}
///
/// Return number of code units the last glyph in the string represents
///
/// \param[in] str Start of a string
/// \param[in] count Length of a string in code units
///
inline size_t glyphrlen(_In_reads_or_z_opt_(count) const utf32_t* str, _In_ size_t count)
{
_Assume_(count && str && str[count - 1]);
for (size_t i = count; i--;) {
if (!iscombining(str[i]))
return count - (i == 0 ? i : i - 1);
} }
return count; return count;
} }
@ -928,6 +952,61 @@ namespace stdex
return strncmp(str1, N1, str2, N2); return strncmp(str1, N1, str2, N2);
} }
///
/// Binary compare two strings
///
/// \param[in] str1 String 1
/// \param[in] count1 String 1 code unit count limit
/// \param[in] str2 String 2
/// \param[in] count2 String 2 code unit count limit
///
/// \return Negative if str1<str2; positive if str1>str2; zero if str1==str2
///
inline int strncmp(
_In_reads_or_z_opt_(count1) const utf32_t* str1, _In_ size_t count1,
_In_reads_or_z_opt_(count2) const utf16_t* str2, _In_ size_t count2)
{
_Assume_(str1 || !count1);
_Assume_(str2 || !count2);
size_t i, j, j_next; utf32_t a, b;
for (i = 0, j = 0; i < count1 && j < count2; ++i, j = j_next) {
a = str1[i];
if (!a)
break;
if (j + 1 >= count2 || !is_surrogate_pair(&str2[j])) {
b = static_cast<utf32_t>(str2[j]);
j_next = j + 1;
}
else {
b = surrogate_pair_to_ucs4(&str2[j]);
j_next = j + 2;
}
if (!b)
break;
if (a > b) return +1;
if (a < b) return -1;
}
if (i < count1 && str1[i]) return +1;
if (j < count2 && str2[j]) return -1;
return 0;
}
///
/// Binary compare two strings
///
/// \param[in] str1 String 1
/// \param[in] str2 String 2
///
/// \return Negative if str1<str2; positive if str1>str2; zero if str1==str2
///
template <size_t N1, size_t N2>
int strncmp(
_In_ const utf32_t (&str1)[N1],
_In_ const utf16_t (&str2)[N2])
{
return strncmp(str1, N1, str2, N2);
}
/// ///
/// Binary compare two strings in reverse direction /// Binary compare two strings in reverse direction
/// ///