Skip to content

<regex>: Correct character translation in icase and collate mode #5553

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
StephanTLavavej merged 9 commits into from
Jun 14, 2025
Merged

<regex>: Correct character translation in icase and collate mode #5553

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
8c0b790
`<regex>`: Correct character translation in `icase` and `collate` mode
muellerj2 May 29, 2025
58e2560
fix call in `_Skip()`
muellerj2 May 29, 2025
b4f31fa
add tests to cover the "small range" optimization and the non-optimiz…
muellerj2 May 29, 2025
befd2b8
make added test wchar_t-only, reorder some tests, adjust regex variab…
muellerj2 May 31, 2025
8d2910b
Add _STD.
StephanTLavavej Jun 9, 2025
c6d119b
Include more headers.
StephanTLavavej Jun 9, 2025
e0a3e00
Explicitly mention private inheritance.
StephanTLavavej Jun 9, 2025
0d5f69d
`static_cast<unsigned int>(re.flags())`
StephanTLavavej Jun 9, 2025
a19f280
Combine check_match() default args.
StephanTLavavej Jun 10, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
86 changes: 58 additions & 28 deletions stl/inc/regex
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -228,34 +228,54 @@ struct _Lex_compare_memcmp_classify_pred<_Elem, _Elem, _Std_char_traits_lt<_Elem
: _Lex_compare_memcmp_classify_pred_for_std_char_traits_lt<_Elem> {};

template <class _RxTraits>
struct _Cmp_cs { // functor to compare two character values for equality
struct _Cmp_icase { // functor to compare for equality following case-insensitive translation of both characters
using _Elem = typename _RxTraits::char_type;
_STATIC_CALL_OPERATOR bool operator()(_Elem _Ex1, _Elem _Ex2) _CONST_CALL_OPERATOR {
return _Ex1 == _Ex2;

explicit _Cmp_icase(const _RxTraits& _Tr) noexcept : _Traits(_Tr) {}

bool operator()(_Elem _Ex1, _Elem _Ex2) const {
return _Traits.translate_nocase(_Ex1) == _Traits.translate_nocase(_Ex2);
}

const _RxTraits& _Traits;
};

template <class _RxTraits>
struct _Cmp_icase { // functor to compare for case-insensitive equality
struct _Cmp_collate { // functor to compare for equality following collating translation of both characters
using _Elem = typename _RxTraits::char_type;

explicit _Cmp_icase(const _RxTraits& _Tr) noexcept : _Traits(_Tr) {}
explicit _Cmp_collate(const _RxTraits& _Tr) noexcept : _Traits(_Tr) {}

bool operator()(_Elem _Ex1, _Elem _Ex2) const {
return _Traits.translate_nocase(_Ex1) == _Traits.translate_nocase(_Ex2);
return _Traits.translate(_Ex1) == _Traits.translate(_Ex2);
}

const _RxTraits& _Traits;
};

template <class _RxTraits>
struct _Cmp_collate { // functor to compare for locale-specific equality
struct _Cmp_icase_translateleft {
// functor to compare for equality following collating translation of the left character
using _Elem = typename _RxTraits::char_type;

explicit _Cmp_collate(const _RxTraits& _Tr) noexcept : _Traits(_Tr) {}
explicit _Cmp_icase_translateleft(const _RxTraits& _Tr) noexcept : _Traits(_Tr) {}

bool operator()(_Elem _Ex1, _Elem _Ex2) const {
return _Traits.translate(_Ex1) == _Traits.translate(_Ex2);
return _Traits.translate_nocase(_Ex1) == _Ex2;
}

const _RxTraits& _Traits;
};

template <class _RxTraits>
struct _Cmp_collate_translateleft {
// functor to compare for equality following collating translation of the left character
using _Elem = typename _RxTraits::char_type;

explicit _Cmp_collate_translateleft(const _RxTraits& _Tr) noexcept : _Traits(_Tr) {}

bool operator()(_Elem _Ex1, _Elem _Ex2) const {
return _Traits.translate(_Ex1) == _Ex2;
}

const _RxTraits& _Traits;
Expand Down Expand Up @@ -2933,10 +2953,6 @@ void _Builder2<_FwdIt, _Elem, _RxTraits>::_Add_class() { // add bracket expressi

template <class _FwdIt, class _Elem, class _RxTraits>
void _Builder2<_FwdIt, _Elem, _RxTraits>::_Add_char_to_bitmap(_Elem _Ch) { // add character to accelerator table
if (_Flags & regex_constants::icase) {
_Ch = _Traits.translate_nocase(_Ch);
}

_Node_class<_Elem, _RxTraits>* _Node = static_cast<_Node_class<_Elem, _RxTraits>*>(_Current);

if (!_Node->_Small) {
Expand All @@ -2948,10 +2964,6 @@ void _Builder2<_FwdIt, _Elem, _RxTraits>::_Add_char_to_bitmap(_Elem _Ch) { // ad

template <class _FwdIt, class _Elem, class _RxTraits>
void _Builder2<_FwdIt, _Elem, _RxTraits>::_Add_char_to_array(_Elem _Ch) { // append character to character array
if (_Flags & regex_constants::icase) {
_Ch = _Traits.translate_nocase(_Ch);
}

_Node_class<_Elem, _RxTraits>* _Node = static_cast<_Node_class<_Elem, _RxTraits>*>(_Current);
if (!_Node->_Large) {
_Node->_Large = new _Buf<_Elem>;
Expand All @@ -2962,6 +2974,12 @@ void _Builder2<_FwdIt, _Elem, _RxTraits>::_Add_char_to_array(_Elem _Ch) { // app

template <class _FwdIt, class _Elem, class _RxTraits>
void _Builder2<_FwdIt, _Elem, _RxTraits>::_Add_char_to_class(_Elem _Ch) { // add character to bracket expression
if (_Flags & regex_constants::icase) {
_Ch = _Traits.translate_nocase(_Ch);
} else if (_Flags & regex_constants::collate) {
_Ch = _Traits.translate(_Ch);
}

if (static_cast<typename _RxTraits::_Uelem>(_Ch) < _Bmp_max) {
_Add_char_to_bitmap(_Ch);
} else {
Expand Down Expand Up @@ -3607,23 +3625,33 @@ _BidIt1 _Cmp_chrange(_BidIt1 _Begin1, _BidIt1 _End1, _BidIt2 _Begin2, _BidIt2 _E
return _Res;
}
}

return _Begin2 == _End2 ? _Begin1 : _Res;
}

template <class _BidIt1, class _BidIt2, class _RxTraits>
_BidIt1 _Compare(_BidIt1 _Begin1, _BidIt1 _End1, _BidIt2 _Begin2, _BidIt2 _End2, const _RxTraits& _Traits,
regex_constants::syntax_option_type _Sflags) { // compare character ranges
_BidIt1 _Res = _End1;
_BidIt1 _Compare_translate_both(_BidIt1 _Begin1, _BidIt1 _End1, _BidIt2 _Begin2, _BidIt2 _End2,
const _RxTraits& _Traits, regex_constants::syntax_option_type _Sflags) {
// compare character ranges, translating characters in both ranges according to syntax options
if (_Sflags & regex_constants::icase) {
_Res = _Cmp_chrange(_Begin1, _End1, _Begin2, _End2, _Cmp_icase<_RxTraits>{_Traits});
return _STD _Cmp_chrange(_Begin1, _End1, _Begin2, _End2, _Cmp_icase<_RxTraits>{_Traits});
} else if (_Sflags & regex_constants::collate) {
_Res = _Cmp_chrange(_Begin1, _End1, _Begin2, _End2, _Cmp_collate<_RxTraits>{_Traits});
return _STD _Cmp_chrange(_Begin1, _End1, _Begin2, _End2, _Cmp_collate<_RxTraits>{_Traits});
} else {
_Res = _Cmp_chrange(_Begin1, _End1, _Begin2, _End2, _Cmp_cs<_RxTraits>{});
return _STD _Cmp_chrange(_Begin1, _End1, _Begin2, _End2, equal_to<typename _RxTraits::char_type>{});
}
}

return _Res;
template <class _BidIt1, class _BidIt2, class _RxTraits>
_BidIt1 _Compare_translate_left(_BidIt1 _Begin1, _BidIt1 _End1, _BidIt2 _Begin2, _BidIt2 _End2,
const _RxTraits& _Traits, regex_constants::syntax_option_type _Sflags) {
// compare character ranges, translating characters in the left range according to syntax options
if (_Sflags & regex_constants::icase) {
return _STD _Cmp_chrange(_Begin1, _End1, _Begin2, _End2, _Cmp_icase_translateleft<_RxTraits>{_Traits});
} else if (_Sflags & regex_constants::collate) {
return _STD _Cmp_chrange(_Begin1, _End1, _Begin2, _End2, _Cmp_collate_translateleft<_RxTraits>{_Traits});
} else {
return _STD _Cmp_chrange(_Begin1, _End1, _Begin2, _End2, equal_to<typename _RxTraits::char_type>{});
}
}

template <class _Elem>
Expand Down Expand Up @@ -3896,7 +3924,7 @@ bool _Matcher2<_BidIt, _Elem, _RxTraits, _It, _Alloc>::_Match_pat(_Node_base* _N
{ // check for string match
_Node_str<_Elem>* _Node = static_cast<_Node_str<_Elem>*>(_Nx);
_It _Res0;
if ((_Res0 = _Compare(_Tgt_state._Cur, _End, _Node->_Data._Str(),
if ((_Res0 = _STD _Compare_translate_left(_Tgt_state._Cur, _End, _Node->_Data._Str(),
_Node->_Data._Str() + _Node->_Data._Size(), _Traits, _Sflags))
!= _Tgt_state._Cur) {
_Tgt_state._Cur = _Res0;
Expand Down Expand Up @@ -3971,7 +3999,8 @@ bool _Matcher2<_BidIt, _Elem, _RxTraits, _It, _Alloc>::_Match_pat(_Node_base* _N
_It _Bx = _Tgt_state._Grps[_Node->_Idx]._Begin;
_It _Ex = _Tgt_state._Grps[_Node->_Idx]._End;
if (_Bx != _Ex // _Bx == _Ex for zero-length match
&& (_Res0 = _Compare(_Tgt_state._Cur, _End, _Bx, _Ex, _Traits, _Sflags)) == _Tgt_state._Cur) {
&& (_Res0 = _STD _Compare_translate_both(_Tgt_state._Cur, _End, _Bx, _Ex, _Traits, _Sflags))
== _Tgt_state._Cur) {
_Failed = true;
} else {
_Tgt_state._Cur = _Res0;
Expand Down Expand Up @@ -4084,7 +4113,8 @@ _BidIt _Matcher2<_BidIt, _Elem, _RxTraits, _It, _Alloc>::_Skip(_BidIt _First_arg
_Node_str<_Elem>* _Node = static_cast<_Node_str<_Elem>*>(_Nx);
for (; _First_arg != _Last; ++_First_arg) { // look for starting match
_BidIt _Next = _First_arg;
if (_Compare(_First_arg, ++_Next, _Node->_Data._Str(), _Node->_Data._Str() + 1, _Traits, _Sflags)
if (_STD _Compare_translate_left(
_First_arg, ++_Next, _Node->_Data._Str(), _Node->_Data._Str() + 1, _Traits, _Sflags)
!= _First_arg) {
break;
}
Expand Down
1 change: 1 addition & 0 deletions tests/std/test.lst
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -265,6 +265,7 @@ tests\GH_005315_destructor_tombstones
tests\GH_005402_string_with_volatile_range
tests\GH_005421_vector_algorithms_integer_class_type_iterator
tests\GH_005472_do_not_overlap
tests\GH_005553_regex_character_translation
tests\LWG2381_num_get_floating_point
tests\LWG2597_complex_branch_cut
tests\LWG3018_shared_ptr_function
Expand Down
4 changes: 4 additions & 0 deletions tests/std/tests/GH_005553_regex_character_translation/env.lst
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,4 @@
# Copyright (c) Microsoft Corporation.
# SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception

RUNALL_INCLUDE ..\usual_matrix.lst
Loading