diff options
Diffstat (limited to 'include/regex')
| -rw-r--r-- | include/regex | 135 |
1 files changed, 89 insertions, 46 deletions
diff --git a/include/regex b/include/regex index bd83d7c10ca73..26efac1c620dd 100644 --- a/include/regex +++ b/include/regex @@ -1,10 +1,9 @@ // -*- C++ -*- //===--------------------------- regex ------------------------------------===// // -// The LLVM Compiler Infrastructure -// -// This file is dual licensed under the MIT and the University of Illinois Open -// Source Licenses. See LICENSE.TXT for details. +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception // //===----------------------------------------------------------------------===// @@ -678,7 +677,7 @@ public: regex_constants::match_flag_type m = regex_constants::match_default); regex_iterator(_BidirectionalIterator __a, _BidirectionalIterator __b, const regex_type&& __re, - regex_constants::match_flag_type __m + regex_constants::match_flag_type __m = regex_constants::match_default) = delete; // C++14 regex_iterator(const regex_iterator&); regex_iterator& operator=(const regex_iterator&); @@ -794,7 +793,11 @@ enum syntax_option_type nosubs = 1 << 1, optimize = 1 << 2, collate = 1 << 3, +#ifdef _LIBCPP_ABI_REGEX_CONSTANTS_NONZERO + ECMAScript = 1 << 9, +#else ECMAScript = 0, +#endif basic = 1 << 4, extended = 1 << 5, awk = 1 << 6, @@ -802,6 +805,16 @@ enum syntax_option_type egrep = 1 << 8 }; +inline _LIBCPP_CONSTEXPR +syntax_option_type __get_grammar(syntax_option_type __g) +{ +#ifdef _LIBCPP_ABI_REGEX_CONSTANTS_NONZERO + return static_cast<syntax_option_type>(__g & 0x3F0); +#else + return static_cast<syntax_option_type>(__g & 0x1F0); +#endif +} + inline _LIBCPP_INLINE_VISIBILITY _LIBCPP_CONSTEXPR syntax_option_type @@ -988,16 +1001,7 @@ public: typedef locale locale_type; typedef ctype_base::mask char_class_type; -#if defined(__mips__) && defined(__GLIBC__) - static const char_class_type __regex_word = static_cast<char_class_type>(_ISbit(15)); -#elif defined(__NetBSD__) - // NetBSD defines classes up to 0x2000 - // see sys/ctype_bits.h, _CTYPE_Q - static const char_class_type __regex_word = 0x8000; -#else - static const char_class_type __regex_word = 0x80; -#endif - + static const char_class_type __regex_word = ctype_base::__regex_word; private: locale __loc_; const ctype<char_type>* __ct_; @@ -2293,7 +2297,7 @@ public: else { if (__b.size() != 1 || __e.size() != 1) - __throw_regex_error<regex_constants::error_collate>(); + __throw_regex_error<regex_constants::error_range>(); if (__icase_) { __b[0] = __traits_.translate_nocase(__b[0]); @@ -2527,19 +2531,27 @@ public: // construct/copy/destroy: _LIBCPP_INLINE_VISIBILITY basic_regex() - : __flags_(), __marked_count_(0), __loop_count_(0), __open_count_(0), + : __flags_(regex_constants::ECMAScript), __marked_count_(0), __loop_count_(0), __open_count_(0), __end_(0) {} _LIBCPP_INLINE_VISIBILITY explicit basic_regex(const value_type* __p, flag_type __f = regex_constants::ECMAScript) : __flags_(__f), __marked_count_(0), __loop_count_(0), __open_count_(0), __end_(0) - {__parse(__p, __p + __traits_.length(__p));} + { + if (__get_grammar(__flags_) == 0) __flags_ |= regex_constants::ECMAScript; + __parse(__p, __p + __traits_.length(__p)); + } + _LIBCPP_INLINE_VISIBILITY basic_regex(const value_type* __p, size_t __len, flag_type __f = regex_constants::ECMAScript) : __flags_(__f), __marked_count_(0), __loop_count_(0), __open_count_(0), __end_(0) - {__parse(__p, __p + __len);} + { + if (__get_grammar(__flags_) == 0) __flags_ |= regex_constants::ECMAScript; + __parse(__p, __p + __len); + } + // basic_regex(const basic_regex&) = default; // basic_regex(basic_regex&&) = default; template <class _ST, class _SA> @@ -2548,21 +2560,31 @@ public: flag_type __f = regex_constants::ECMAScript) : __flags_(__f), __marked_count_(0), __loop_count_(0), __open_count_(0), __end_(0) - {__parse(__p.begin(), __p.end());} + { + if (__get_grammar(__flags_) == 0) __flags_ |= regex_constants::ECMAScript; + __parse(__p.begin(), __p.end()); + } + template <class _ForwardIterator> _LIBCPP_INLINE_VISIBILITY basic_regex(_ForwardIterator __first, _ForwardIterator __last, flag_type __f = regex_constants::ECMAScript) : __flags_(__f), __marked_count_(0), __loop_count_(0), __open_count_(0), __end_(0) - {__parse(__first, __last);} + { + if (__get_grammar(__flags_) == 0) __flags_ |= regex_constants::ECMAScript; + __parse(__first, __last); + } #ifndef _LIBCPP_CXX03_LANG _LIBCPP_INLINE_VISIBILITY basic_regex(initializer_list<value_type> __il, flag_type __f = regex_constants::ECMAScript) : __flags_(__f), __marked_count_(0), __loop_count_(0), __open_count_(0), __end_(0) - {__parse(__il.begin(), __il.end());} + { + if (__get_grammar(__flags_) == 0) __flags_ |= regex_constants::ECMAScript; + __parse(__il.begin(), __il.end()); + } #endif // _LIBCPP_CXX03_LANG // ~basic_regex() = default; @@ -3044,7 +3066,7 @@ basic_regex<_CharT, _Traits>::__parse(_ForwardIterator __first, __h.release(); __end_ = __start_.get(); } - switch (__flags_ & 0x1F0) + switch (__get_grammar(__flags_)) { case ECMAScript: __first = __parse_ecma_exp(__first, __last); @@ -3379,7 +3401,7 @@ basic_regex<_CharT, _Traits>::__parse_BACKREF(_ForwardIterator __first, if (__temp != __last) { if (*__first == '\\') - { + { int __val = __traits_.value(*__temp, 10); if (__val >= 1 && __val <= 9) { @@ -3513,7 +3535,7 @@ basic_regex<_CharT, _Traits>::__parse_QUOTED_CHAR_ERE(_ForwardIterator __first, __first = ++__temp; break; default: - if ((__flags_ & 0x1F0) == awk) + if (__get_grammar(__flags_) == awk) __first = __parse_awk_escape(++__first, __last); break; } @@ -3597,7 +3619,7 @@ basic_regex<_CharT, _Traits>::__parse_ERE_dupl_symbol(_ForwardIterator __first, { if (__first != __last) { - unsigned __grammar = __flags_ & 0x1F0; + unsigned __grammar = __get_grammar(__flags_); switch (*__first) { case '*': @@ -3718,7 +3740,7 @@ basic_regex<_CharT, _Traits>::__parse_bracket_expression(_ForwardIterator __firs // __ml owned by *this if (__first == __last) __throw_regex_error<regex_constants::error_brack>(); - if ((__flags_ & 0x1F0) != ECMAScript && *__first == ']') + if (__get_grammar(__flags_) != ECMAScript && *__first == ']') { __ml->__add_char(']'); ++__first; @@ -3779,7 +3801,7 @@ basic_regex<_CharT, _Traits>::__parse_expression_term(_ForwardIterator __first, else if (*__temp == '.') __first = __parse_collating_symbol(++__temp, __last, __start_range); } - unsigned __grammar = __flags_ & 0x1F0; + unsigned __grammar = __get_grammar(__flags_); if (__start_range.empty()) { if ((__grammar == ECMAScript || __grammar == awk) && *__first == '\\') @@ -4076,7 +4098,7 @@ basic_regex<_CharT, _Traits>::__parse_DUP_COUNT(_ForwardIterator __first, if ( __val != -1 ) { __c = __val; - for (++__first; + for (++__first; __first != __last && ( __val = __traits_.value(*__first, 10)) != -1; ++__first) { @@ -4456,7 +4478,7 @@ basic_regex<_CharT, _Traits>::__parse_character_escape(_ForwardIterator __first, case 'c': if ((__t = _VSTD::next(__first)) != __last) { - if (('A' <= *__t && *__t <= 'Z') || + if (('A' <= *__t && *__t <= 'Z') || ('a' <= *__t && *__t <= 'z')) { if (__str) @@ -4465,7 +4487,7 @@ basic_regex<_CharT, _Traits>::__parse_character_escape(_ForwardIterator __first, __push_char(_CharT(*__t % 32)); __first = ++__t; } - else + else __throw_regex_error<regex_constants::error_escape>(); } else @@ -5265,21 +5287,41 @@ public: // element access: _LIBCPP_INLINE_VISIBILITY difference_type length(size_type __sub = 0) const - {return (*this)[__sub].length();} + { + _LIBCPP_ASSERT(ready(), "match_results::length() called when not ready"); + return (*this)[__sub].length(); + } _LIBCPP_INLINE_VISIBILITY difference_type position(size_type __sub = 0) const - {return _VSTD::distance(__position_start_, (*this)[__sub].first);} + { + _LIBCPP_ASSERT(ready(), "match_results::position() called when not ready"); + return _VSTD::distance(__position_start_, (*this)[__sub].first); + } _LIBCPP_INLINE_VISIBILITY string_type str(size_type __sub = 0) const - {return (*this)[__sub].str();} + { + _LIBCPP_ASSERT(ready(), "match_results::str() called when not ready"); + return (*this)[__sub].str(); + } _LIBCPP_INLINE_VISIBILITY const_reference operator[](size_type __n) const - {return __n < __matches_.size() ? __matches_[__n] : __unmatched_;} + { + _LIBCPP_ASSERT(ready(), "match_results::operator[]() called when not ready"); + return __n < __matches_.size() ? __matches_[__n] : __unmatched_; + } _LIBCPP_INLINE_VISIBILITY - const_reference prefix() const {return __prefix_;} + const_reference prefix() const + { + _LIBCPP_ASSERT(ready(), "match_results::prefix() called when not ready"); + return __prefix_; + } _LIBCPP_INLINE_VISIBILITY - const_reference suffix() const {return __suffix_;} + const_reference suffix() const + { + _LIBCPP_ASSERT(ready(), "match_results::suffix() called when not ready"); + return __suffix_; + } _LIBCPP_INLINE_VISIBILITY const_iterator begin() const {return empty() ? __matches_.end() : __matches_.begin();} @@ -5417,6 +5459,7 @@ match_results<_BidirectionalIterator, _Allocator>::format(_OutputIter __output_i const char_type* __fmt_first, const char_type* __fmt_last, regex_constants::match_flag_type __flags) const { + _LIBCPP_ASSERT(ready(), "match_results::format() called when not ready"); if (__flags & regex_constants::format_sed) { for (; __fmt_first != __fmt_last; ++__fmt_first) @@ -5849,7 +5892,7 @@ basic_regex<_CharT, _Traits>::__match_at_start( match_results<const _CharT*, _Allocator>& __m, regex_constants::match_flag_type __flags, bool __at_first) const { - if ((__flags_ & 0x1F0) == ECMAScript) + if (__get_grammar(__flags_) == ECMAScript) return __match_at_start_ecma(__first, __last, __m, __flags, __at_first); if (mark_count() == 0) return __match_at_start_posix_nosubs(__first, __last, __m, __flags, __at_first); @@ -5866,7 +5909,7 @@ basic_regex<_CharT, _Traits>::__search( { __m.__init(1 + mark_count(), __first, __last, __flags & regex_constants::__no_update_pos); - if (__match_at_start(__first, __last, __m, __flags, + if (__match_at_start(__first, __last, __m, __flags, !(__flags & regex_constants::__no_update_pos))) { __m.__prefix_.second = __m[0].first; @@ -6012,7 +6055,7 @@ bool regex_search(const basic_string<_Cp, _ST, _SA>&& __s, match_results<typename basic_string<_Cp, _ST, _SA>::const_iterator, _Ap>&, const basic_regex<_Cp, _Tp>& __e, - regex_constants::match_flag_type __flags = regex_constants::match_default) = delete; + regex_constants::match_flag_type __flags = regex_constants::match_default) = delete; #endif // regex_match @@ -6076,7 +6119,7 @@ bool regex_match(const basic_string<_CharT, _ST, _SA>&& __s, match_results<typename basic_string<_CharT, _ST, _SA>::const_iterator, _Allocator>& __m, const basic_regex<_CharT, _Traits>& __e, - regex_constants::match_flag_type __flags = regex_constants::match_default) = delete; + regex_constants::match_flag_type __flags = regex_constants::match_default) = delete; #endif template <class _CharT, class _Traits> @@ -6129,7 +6172,7 @@ public: #if _LIBCPP_STD_VER > 11 regex_iterator(_BidirectionalIterator __a, _BidirectionalIterator __b, const regex_type&& __re, - regex_constants::match_flag_type __m + regex_constants::match_flag_type __m = regex_constants::match_default) = delete; #endif @@ -6140,7 +6183,7 @@ public: _LIBCPP_INLINE_VISIBILITY reference operator*() const {return __match_;} _LIBCPP_INLINE_VISIBILITY - pointer operator->() const {return &__match_;} + pointer operator->() const {return _VSTD::addressof(__match_);} regex_iterator& operator++(); _LIBCPP_INLINE_VISIBILITY @@ -6164,7 +6207,7 @@ regex_iterator<_BidirectionalIterator, _CharT, _Traits>:: const regex_type& __re, regex_constants::match_flag_type __m) : __begin_(__a), __end_(__b), - __pregex_(&__re), + __pregex_(_VSTD::addressof(__re)), __flags_(__m) { _VSTD::regex_search(__begin_, __end_, __match_, *__pregex_, __flags_); @@ -6325,7 +6368,7 @@ private: __result_ = &__position_->prefix(); else __result_ = &(*__position_)[__subs_[__n_]]; - } + } }; template <class _BidirectionalIterator, class _CharT, class _Traits> @@ -6405,7 +6448,7 @@ regex_token_iterator<_BidirectionalIterator, _CharT, _Traits>:: regex_constants::match_flag_type __m) : __position_(__a, __b, __re, __m), __n_(0), - __subs_(__submatches, __submatches + _Np) + __subs_(begin(__submatches), end(__submatches)) { __init(__a, __b); } |