| 1 | /* |
|---|
| 2 | * |
|---|
| 3 | * Copyright (c) 2004 |
|---|
| 4 | * John Maddock |
|---|
| 5 | * |
|---|
| 6 | * Use, modification and distribution are subject to the |
|---|
| 7 | * Boost Software License, Version 1.0. (See accompanying file |
|---|
| 8 | * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) |
|---|
| 9 | * |
|---|
| 10 | */ |
|---|
| 11 | |
|---|
| 12 | /* |
|---|
| 13 | * LOCATION: see http://www.boost.org for most recent version. |
|---|
| 14 | * FILE regex_traits_defaults.hpp |
|---|
| 15 | * VERSION see <boost/version.hpp> |
|---|
| 16 | * DESCRIPTION: Declares API's for access to regex_traits default properties. |
|---|
| 17 | */ |
|---|
| 18 | |
|---|
| 19 | #ifndef BOOST_REGEX_TRAITS_DEFAULTS_HPP_INCLUDED |
|---|
| 20 | #define BOOST_REGEX_TRAITS_DEFAULTS_HPP_INCLUDED |
|---|
| 21 | |
|---|
| 22 | #ifdef BOOST_HAS_ABI_HEADERS |
|---|
| 23 | # include BOOST_ABI_PREFIX |
|---|
| 24 | #endif |
|---|
| 25 | |
|---|
| 26 | #ifndef BOOST_REGEX_SYNTAX_TYPE_HPP |
|---|
| 27 | #include <boost/regex/v4/syntax_type.hpp> |
|---|
| 28 | #endif |
|---|
| 29 | #ifndef BOOST_REGEX_ERROR_TYPE_HPP |
|---|
| 30 | #include <boost/regex/v4/error_type.hpp> |
|---|
| 31 | #endif |
|---|
| 32 | |
|---|
| 33 | #ifdef BOOST_NO_STDC_NAMESPACE |
|---|
| 34 | namespace std{ |
|---|
| 35 | using ::strlen; |
|---|
| 36 | } |
|---|
| 37 | #endif |
|---|
| 38 | |
|---|
| 39 | namespace boost{ namespace re_detail{ |
|---|
| 40 | |
|---|
| 41 | |
|---|
| 42 | // |
|---|
| 43 | // helpers to suppress warnings: |
|---|
| 44 | // |
|---|
| 45 | template <class charT> |
|---|
| 46 | inline bool is_extended(charT c) |
|---|
| 47 | { return c > 256; } |
|---|
| 48 | inline bool is_extended(char) |
|---|
| 49 | { return false; } |
|---|
| 50 | |
|---|
| 51 | |
|---|
| 52 | BOOST_REGEX_DECL const char* BOOST_REGEX_CALL get_default_syntax(regex_constants::syntax_type n); |
|---|
| 53 | BOOST_REGEX_DECL const char* BOOST_REGEX_CALL get_default_error_string(regex_constants::error_type n); |
|---|
| 54 | BOOST_REGEX_DECL regex_constants::syntax_type BOOST_REGEX_CALL get_default_syntax_type(char c); |
|---|
| 55 | BOOST_REGEX_DECL regex_constants::escape_syntax_type BOOST_REGEX_CALL get_default_escape_syntax_type(char c); |
|---|
| 56 | |
|---|
| 57 | // is charT c a combining character? |
|---|
| 58 | BOOST_REGEX_DECL bool BOOST_REGEX_CALL is_combining_implementation(uint_least16_t s); |
|---|
| 59 | |
|---|
| 60 | template <class charT> |
|---|
| 61 | inline bool is_combining(charT c) |
|---|
| 62 | { |
|---|
| 63 | return (c <= static_cast<charT>(0)) ? false : ((c >= static_cast<charT>((std::numeric_limits<uint_least16_t>::max)())) ? false : is_combining_implementation(static_cast<unsigned short>(c))); |
|---|
| 64 | } |
|---|
| 65 | template <> |
|---|
| 66 | inline bool is_combining<char>(char) |
|---|
| 67 | { |
|---|
| 68 | return false; |
|---|
| 69 | } |
|---|
| 70 | template <> |
|---|
| 71 | inline bool is_combining<signed char>(signed char) |
|---|
| 72 | { |
|---|
| 73 | return false; |
|---|
| 74 | } |
|---|
| 75 | template <> |
|---|
| 76 | inline bool is_combining<unsigned char>(unsigned char) |
|---|
| 77 | { |
|---|
| 78 | return false; |
|---|
| 79 | } |
|---|
| 80 | #ifndef __hpux // can't use WCHAR_MIN/MAX in pp-directives. |
|---|
| 81 | #ifdef _MSC_VER |
|---|
| 82 | template<> |
|---|
| 83 | inline bool is_combining<wchar_t>(wchar_t c) |
|---|
| 84 | { |
|---|
| 85 | return is_combining_implementation(static_cast<unsigned short>(c)); |
|---|
| 86 | } |
|---|
| 87 | #elif !defined(__DECCXX) && !defined(__osf__) && !defined(__OSF__) && defined(WCHAR_MIN) && (WCHAR_MIN == 0) && !defined(BOOST_NO_INTRINSIC_WCHAR_T) |
|---|
| 88 | #if defined(WCHAR_MAX) && (WCHAR_MAX <= USHRT_MAX) |
|---|
| 89 | template<> |
|---|
| 90 | inline bool is_combining<wchar_t>(wchar_t c) |
|---|
| 91 | { |
|---|
| 92 | return is_combining_implementation(static_cast<unsigned short>(c)); |
|---|
| 93 | } |
|---|
| 94 | #else |
|---|
| 95 | template<> |
|---|
| 96 | inline bool is_combining<wchar_t>(wchar_t c) |
|---|
| 97 | { |
|---|
| 98 | return (c >= (std::numeric_limits<uint_least16_t>::max)()) ? false : is_combining_implementation(static_cast<unsigned short>(c)); |
|---|
| 99 | } |
|---|
| 100 | #endif |
|---|
| 101 | #endif |
|---|
| 102 | #endif |
|---|
| 103 | |
|---|
| 104 | // |
|---|
| 105 | // is a charT c a line separator? |
|---|
| 106 | // |
|---|
| 107 | template <class charT> |
|---|
| 108 | inline bool is_separator(charT c) |
|---|
| 109 | { |
|---|
| 110 | return BOOST_REGEX_MAKE_BOOL( |
|---|
| 111 | (c == static_cast<charT>('\n')) |
|---|
| 112 | || (c == static_cast<charT>('\r')) |
|---|
| 113 | || (c == static_cast<charT>('\f')) |
|---|
| 114 | || (static_cast<boost::uint16_t>(c) == 0x2028u) |
|---|
| 115 | || (static_cast<boost::uint16_t>(c) == 0x2029u) |
|---|
| 116 | || (static_cast<boost::uint16_t>(c) == 0x85u)); |
|---|
| 117 | } |
|---|
| 118 | template <> |
|---|
| 119 | inline bool is_separator<char>(char c) |
|---|
| 120 | { |
|---|
| 121 | return BOOST_REGEX_MAKE_BOOL((c == '\n') || (c == '\r') || (c == '\f')); |
|---|
| 122 | } |
|---|
| 123 | |
|---|
| 124 | // |
|---|
| 125 | // get a default collating element: |
|---|
| 126 | // |
|---|
| 127 | BOOST_REGEX_DECL std::string BOOST_REGEX_CALL lookup_default_collate_name(const std::string& name); |
|---|
| 128 | |
|---|
| 129 | // |
|---|
| 130 | // get the id of a character clasification, the individual |
|---|
| 131 | // traits classes then transform that id into a bitmask: |
|---|
| 132 | // |
|---|
| 133 | template <class charT> |
|---|
| 134 | struct character_pointer_range |
|---|
| 135 | { |
|---|
| 136 | const charT* p1; |
|---|
| 137 | const charT* p2; |
|---|
| 138 | |
|---|
| 139 | bool operator < (const character_pointer_range& r)const |
|---|
| 140 | { |
|---|
| 141 | return std::lexicographical_compare(p1, p2, r.p1, r.p2); |
|---|
| 142 | } |
|---|
| 143 | bool operator == (const character_pointer_range& r)const |
|---|
| 144 | { |
|---|
| 145 | // Not only do we check that the ranges are of equal size before |
|---|
| 146 | // calling std::equal, but there is no other algorithm available: |
|---|
| 147 | // not even a non-standard MS one. So forward to unchecked_equal |
|---|
| 148 | // in the MS case. |
|---|
| 149 | return ((p2 - p1) == (r.p2 - r.p1)) && re_detail::equal(p1, p2, r.p1); |
|---|
| 150 | } |
|---|
| 151 | }; |
|---|
| 152 | template <class charT> |
|---|
| 153 | int get_default_class_id(const charT* p1, const charT* p2) |
|---|
| 154 | { |
|---|
| 155 | static const charT data[72] = { |
|---|
| 156 | 'a', 'l', 'n', 'u', 'm', |
|---|
| 157 | 'a', 'l', 'p', 'h', 'a', |
|---|
| 158 | 'b', 'l', 'a', 'n', 'k', |
|---|
| 159 | 'c', 'n', 't', 'r', 'l', |
|---|
| 160 | 'd', 'i', 'g', 'i', 't', |
|---|
| 161 | 'g', 'r', 'a', 'p', 'h', |
|---|
| 162 | 'l', 'o', 'w', 'e', 'r', |
|---|
| 163 | 'p', 'r', 'i', 'n', 't', |
|---|
| 164 | 'p', 'u', 'n', 'c', 't', |
|---|
| 165 | 's', 'p', 'a', 'c', 'e', |
|---|
| 166 | 'u', 'n', 'i', 'c', 'o', 'd', 'e', |
|---|
| 167 | 'u', 'p', 'p', 'e', 'r', |
|---|
| 168 | 'w', 'o', 'r', 'd', |
|---|
| 169 | 'x', 'd', 'i', 'g', 'i', 't', |
|---|
| 170 | }; |
|---|
| 171 | |
|---|
| 172 | static const character_pointer_range<charT> ranges[19] = |
|---|
| 173 | { |
|---|
| 174 | {data+0, data+5,}, // alnum |
|---|
| 175 | {data+5, data+10,}, // alpha |
|---|
| 176 | {data+10, data+15,}, // blank |
|---|
| 177 | {data+15, data+20,}, // cntrl |
|---|
| 178 | {data+20, data+21,}, // d |
|---|
| 179 | {data+20, data+25,}, // digit |
|---|
| 180 | {data+25, data+30,}, // graph |
|---|
| 181 | {data+30, data+31,}, // l |
|---|
| 182 | {data+30, data+35,}, // lower |
|---|
| 183 | {data+35, data+40,}, // print |
|---|
| 184 | {data+40, data+45,}, // punct |
|---|
| 185 | {data+45, data+46,}, // s |
|---|
| 186 | {data+45, data+50,}, // space |
|---|
| 187 | {data+57, data+58,}, // u |
|---|
| 188 | {data+50, data+57,}, // unicode |
|---|
| 189 | {data+57, data+62,}, // upper |
|---|
| 190 | {data+62, data+63,}, // w |
|---|
| 191 | {data+62, data+66,}, // word |
|---|
| 192 | {data+66, data+72,}, // xdigit |
|---|
| 193 | }; |
|---|
| 194 | static const character_pointer_range<charT>* ranges_begin = ranges; |
|---|
| 195 | static const character_pointer_range<charT>* ranges_end = ranges + (sizeof(ranges)/sizeof(ranges[0])); |
|---|
| 196 | |
|---|
| 197 | character_pointer_range<charT> t = { p1, p2, }; |
|---|
| 198 | const character_pointer_range<charT>* p = std::lower_bound(ranges_begin, ranges_end, t); |
|---|
| 199 | if((p != ranges_end) && (t == *p)) |
|---|
| 200 | return static_cast<int>(p - ranges); |
|---|
| 201 | return -1; |
|---|
| 202 | } |
|---|
| 203 | |
|---|
| 204 | // |
|---|
| 205 | // helper functions: |
|---|
| 206 | // |
|---|
| 207 | template <class charT> |
|---|
| 208 | std::ptrdiff_t global_length(const charT* p) |
|---|
| 209 | { |
|---|
| 210 | std::ptrdiff_t n = 0; |
|---|
| 211 | while(*p) |
|---|
| 212 | { |
|---|
| 213 | ++p; |
|---|
| 214 | ++n; |
|---|
| 215 | } |
|---|
| 216 | return n; |
|---|
| 217 | } |
|---|
| 218 | template<> |
|---|
| 219 | inline std::ptrdiff_t global_length<char>(const char* p) |
|---|
| 220 | { |
|---|
| 221 | return (std::strlen)(p); |
|---|
| 222 | } |
|---|
| 223 | #ifndef BOOST_NO_WREGEX |
|---|
| 224 | template<> |
|---|
| 225 | inline std::ptrdiff_t global_length<wchar_t>(const wchar_t* p) |
|---|
| 226 | { |
|---|
| 227 | return (std::wcslen)(p); |
|---|
| 228 | } |
|---|
| 229 | #endif |
|---|
| 230 | template <class charT> |
|---|
| 231 | inline charT BOOST_REGEX_CALL global_lower(charT c) |
|---|
| 232 | { |
|---|
| 233 | return c; |
|---|
| 234 | } |
|---|
| 235 | template <class charT> |
|---|
| 236 | inline charT BOOST_REGEX_CALL global_upper(charT c) |
|---|
| 237 | { |
|---|
| 238 | return c; |
|---|
| 239 | } |
|---|
| 240 | |
|---|
| 241 | BOOST_REGEX_DECL char BOOST_REGEX_CALL do_global_lower(char c); |
|---|
| 242 | BOOST_REGEX_DECL char BOOST_REGEX_CALL do_global_upper(char c); |
|---|
| 243 | #ifndef BOOST_NO_WREGEX |
|---|
| 244 | BOOST_REGEX_DECL wchar_t BOOST_REGEX_CALL do_global_lower(wchar_t c); |
|---|
| 245 | BOOST_REGEX_DECL wchar_t BOOST_REGEX_CALL do_global_upper(wchar_t c); |
|---|
| 246 | #endif |
|---|
| 247 | #ifdef BOOST_REGEX_HAS_OTHER_WCHAR_T |
|---|
| 248 | BOOST_REGEX_DECL unsigned short BOOST_REGEX_CALL do_global_lower(unsigned short c); |
|---|
| 249 | BOOST_REGEX_DECL unsigned short BOOST_REGEX_CALL do_global_upper(unsigned short c); |
|---|
| 250 | #endif |
|---|
| 251 | // |
|---|
| 252 | // This sucks: declare template specialisations of global_lower/global_upper |
|---|
| 253 | // that just forward to the non-template implementation functions. We do |
|---|
| 254 | // this because there is one compiler (Compaq Tru64 C++) that doesn't seem |
|---|
| 255 | // to differentiate between templates and non-template overloads.... |
|---|
| 256 | // what's more, the primary template, plus all overloads have to be |
|---|
| 257 | // defined in the same translation unit (if one is inline they all must be) |
|---|
| 258 | // otherwise the "local template instantiation" compiler option can pick |
|---|
| 259 | // the wrong instantiation when linking: |
|---|
| 260 | // |
|---|
| 261 | template<> inline char BOOST_REGEX_CALL global_lower<char>(char c){ return do_global_lower(c); } |
|---|
| 262 | template<> inline char BOOST_REGEX_CALL global_upper<char>(char c){ return do_global_upper(c); } |
|---|
| 263 | #ifndef BOOST_NO_WREGEX |
|---|
| 264 | template<> inline wchar_t BOOST_REGEX_CALL global_lower<wchar_t>(wchar_t c){ return do_global_lower(c); } |
|---|
| 265 | template<> inline wchar_t BOOST_REGEX_CALL global_upper<wchar_t>(wchar_t c){ return do_global_upper(c); } |
|---|
| 266 | #endif |
|---|
| 267 | #ifdef BOOST_REGEX_HAS_OTHER_WCHAR_T |
|---|
| 268 | template<> inline unsigned short BOOST_REGEX_CALL global_lower<unsigned short>(unsigned short c){ return do_global_lower(c); } |
|---|
| 269 | template<> inline unsigned short BOOST_REGEX_CALL global_upper<unsigned short>(unsigned short c){ return do_global_upper(c); } |
|---|
| 270 | #endif |
|---|
| 271 | |
|---|
| 272 | template <class charT> |
|---|
| 273 | int global_value(charT c) |
|---|
| 274 | { |
|---|
| 275 | static const charT zero = '0'; |
|---|
| 276 | static const charT nine = '9'; |
|---|
| 277 | static const charT a = 'a'; |
|---|
| 278 | static const charT f = 'f'; |
|---|
| 279 | static const charT A = 'A'; |
|---|
| 280 | static const charT F = 'F'; |
|---|
| 281 | |
|---|
| 282 | if(c > f) return -1; |
|---|
| 283 | if(c >= a) return 10 + (c - a); |
|---|
| 284 | if(c > F) return -1; |
|---|
| 285 | if(c >= A) return 10 + (c - A); |
|---|
| 286 | if(c > nine) return -1; |
|---|
| 287 | if(c >= zero) return c - zero; |
|---|
| 288 | return -1; |
|---|
| 289 | } |
|---|
| 290 | template <class charT, class traits> |
|---|
| 291 | int global_toi(const charT*& p1, const charT* p2, int radix, const traits& t) |
|---|
| 292 | { |
|---|
| 293 | (void)t; // warning suppression |
|---|
| 294 | int next_value = t.value(*p1, radix); |
|---|
| 295 | if((p1 == p2) || (next_value < 0) || (next_value >= radix)) |
|---|
| 296 | return -1; |
|---|
| 297 | int result = 0; |
|---|
| 298 | while(p1 != p2) |
|---|
| 299 | { |
|---|
| 300 | next_value = t.value(*p1, radix); |
|---|
| 301 | if((next_value < 0) || (next_value >= radix)) |
|---|
| 302 | break; |
|---|
| 303 | result *= radix; |
|---|
| 304 | result += next_value; |
|---|
| 305 | ++p1; |
|---|
| 306 | } |
|---|
| 307 | return result; |
|---|
| 308 | } |
|---|
| 309 | |
|---|
| 310 | } // re_detail |
|---|
| 311 | } // boost |
|---|
| 312 | |
|---|
| 313 | #ifdef BOOST_HAS_ABI_HEADERS |
|---|
| 314 | # include BOOST_ABI_SUFFIX |
|---|
| 315 | #endif |
|---|
| 316 | |
|---|
| 317 | #endif |
|---|