[3482] | 1 | /* |
---|
| 2 | www.sourceforge.net/projects/tinyxml |
---|
| 3 | Original code (2.0 and earlier )copyright (c) 2000-2002 Lee Thomason (www.grinninglizard.com) |
---|
| 4 | |
---|
| 5 | This software is provided 'as-is', without any express or implied |
---|
| 6 | warranty. In no event will the authors be held liable for any |
---|
| 7 | damages arising from the use of this software. |
---|
| 8 | |
---|
| 9 | Permission is granted to anyone to use this software for any |
---|
| 10 | purpose, including commercial applications, and to alter it and |
---|
| 11 | redistribute it freely, subject to the following restrictions: |
---|
| 12 | |
---|
| 13 | 1. The origin of this software must not be misrepresented; you must |
---|
| 14 | not claim that you wrote the original software. If you use this |
---|
| 15 | software in a product, an acknowledgment in the product documentation |
---|
| 16 | would be appreciated but is not required. |
---|
| 17 | |
---|
| 18 | 2. Altered source versions must be plainly marked as such, and |
---|
| 19 | must not be misrepresented as being the original software. |
---|
| 20 | |
---|
| 21 | 3. This notice may not be removed or altered from any source |
---|
| 22 | distribution. |
---|
| 23 | */ |
---|
| 24 | |
---|
| 25 | #include "tinyxml.h" |
---|
| 26 | #include <ctype.h> |
---|
| 27 | |
---|
| 28 | //#define DEBUG_PARSER |
---|
| 29 | |
---|
| 30 | // Note tha "PutString" hardcodes the same list. This |
---|
| 31 | // is less flexible than it appears. Changing the entries |
---|
| 32 | // or order will break putstring. |
---|
| 33 | TiXmlBase::Entity TiXmlBase::entity[ NUM_ENTITY ] = |
---|
| 34 | { |
---|
| 35 | { "&", 5, '&' }, |
---|
| 36 | { "<", 4, '<' }, |
---|
| 37 | { ">", 4, '>' }, |
---|
| 38 | { """, 6, '\"' }, |
---|
| 39 | { "'", 6, '\'' } |
---|
| 40 | }; |
---|
| 41 | |
---|
| 42 | // Bunch of unicode info at: |
---|
| 43 | // http://www.unicode.org/faq/utf_bom.html |
---|
| 44 | // Including the basic of this table, which determines the #bytes in the |
---|
| 45 | // sequence from the lead byte. 1 placed for invalid sequences -- |
---|
| 46 | // although the result will be junk, pass it through as much as possible. |
---|
| 47 | // Beware of the non-characters in UTF-8: |
---|
| 48 | // ef bb bf (Microsoft "lead bytes") |
---|
| 49 | // ef bf be |
---|
| 50 | // ef bf bf |
---|
| 51 | |
---|
| 52 | const char TIXML_UTF_LEAD_0 = (const char)0xef; |
---|
| 53 | const char TIXML_UTF_LEAD_1 = (const char)0xbb; |
---|
| 54 | const char TIXML_UTF_LEAD_2 = (const char)0xbf; |
---|
| 55 | |
---|
| 56 | const int TiXmlBase::utf8ByteTable[256] = |
---|
| 57 | { |
---|
| 58 | // 0 1 2 3 4 5 6 7 8 9 a b c d e f |
---|
| 59 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x00 |
---|
| 60 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x10 |
---|
| 61 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x20 |
---|
| 62 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x30 |
---|
| 63 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x40 |
---|
| 64 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x50 |
---|
| 65 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x60 |
---|
| 66 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x70 End of ASCII range |
---|
| 67 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x80 0x80 to 0xc1 invalid |
---|
| 68 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x90 |
---|
| 69 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0xa0 |
---|
| 70 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0xb0 |
---|
| 71 | 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, // 0xc0 0xc2 to 0xdf 2 byte |
---|
| 72 | 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, // 0xd0 |
---|
| 73 | 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, // 0xe0 0xe0 to 0xef 3 byte |
---|
| 74 | 4, 4, 4, 4, 4, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 // 0xf0 0xf0 to 0xf4 4 byte, 0xf5 and higher invalid |
---|
| 75 | }; |
---|
| 76 | |
---|
| 77 | |
---|
| 78 | void TiXmlBase::ConvertUTF32ToUTF8( unsigned long input, char* output, int* length ) |
---|
| 79 | { |
---|
| 80 | const unsigned long BYTE_MASK = 0xBF; |
---|
| 81 | const unsigned long BYTE_MARK = 0x80; |
---|
| 82 | const unsigned long FIRST_BYTE_MARK[7] = { 0x00, 0x00, 0xC0, 0xE0, 0xF0, 0xF8, 0xFC }; |
---|
| 83 | |
---|
| 84 | if (input < 0x80) |
---|
| 85 | *length = 1; |
---|
| 86 | else if ( input < 0x800 ) |
---|
| 87 | *length = 2; |
---|
| 88 | else if ( input < 0x10000 ) |
---|
| 89 | *length = 3; |
---|
| 90 | else if ( input < 0x200000 ) |
---|
| 91 | *length = 4; |
---|
| 92 | else |
---|
| 93 | { *length = 0; return; } // This code won't covert this correctly anyway. |
---|
| 94 | |
---|
| 95 | output += *length; |
---|
| 96 | |
---|
| 97 | // Scary scary fall throughs. |
---|
| 98 | switch (*length) |
---|
| 99 | { |
---|
| 100 | case 4: |
---|
| 101 | --output; |
---|
| 102 | *output = (char)((input | BYTE_MARK) & BYTE_MASK); |
---|
| 103 | input >>= 6; |
---|
| 104 | case 3: |
---|
| 105 | --output; |
---|
| 106 | *output = (char)((input | BYTE_MARK) & BYTE_MASK); |
---|
| 107 | input >>= 6; |
---|
| 108 | case 2: |
---|
| 109 | --output; |
---|
| 110 | *output = (char)((input | BYTE_MARK) & BYTE_MASK); |
---|
| 111 | input >>= 6; |
---|
| 112 | case 1: |
---|
| 113 | --output; |
---|
| 114 | *output = (char)(input | FIRST_BYTE_MARK[*length]); |
---|
| 115 | } |
---|
| 116 | } |
---|
| 117 | |
---|
| 118 | |
---|
| 119 | /*static*/ int TiXmlBase::IsAlpha( unsigned char anyByte, TiXmlEncoding ) |
---|
| 120 | { |
---|
| 121 | // This will only work for low-ascii, everything else is assumed to be a valid |
---|
| 122 | // letter. I'm not sure this is the best approach, but it is quite tricky trying |
---|
| 123 | // to figure out alhabetical vs. not across encoding. So take a very |
---|
| 124 | // conservative approach. |
---|
| 125 | |
---|
| 126 | // if ( encoding == TIXML_ENCODING_UTF8 ) |
---|
| 127 | // { |
---|
| 128 | if ( anyByte < 127 ) |
---|
| 129 | return isalpha( anyByte ); |
---|
| 130 | else |
---|
| 131 | return 1; // What else to do? The unicode set is huge...get the english ones right. |
---|
| 132 | // } |
---|
| 133 | // else |
---|
| 134 | // { |
---|
| 135 | // return isalpha( anyByte ); |
---|
| 136 | // } |
---|
| 137 | } |
---|
| 138 | |
---|
| 139 | |
---|
| 140 | /*static*/ int TiXmlBase::IsAlphaNum( unsigned char anyByte, TiXmlEncoding ) |
---|
| 141 | { |
---|
| 142 | // This will only work for low-ascii, everything else is assumed to be a valid |
---|
| 143 | // letter. I'm not sure this is the best approach, but it is quite tricky trying |
---|
| 144 | // to figure out alhabetical vs. not across encoding. So take a very |
---|
| 145 | // conservative approach. |
---|
| 146 | |
---|
| 147 | // if ( encoding == TIXML_ENCODING_UTF8 ) |
---|
| 148 | // { |
---|
| 149 | if ( anyByte < 127 ) |
---|
| 150 | return isalnum( anyByte ); |
---|
| 151 | else |
---|
| 152 | return 1; // What else to do? The unicode set is huge...get the english ones right. |
---|
| 153 | // } |
---|
| 154 | // else |
---|
| 155 | // { |
---|
| 156 | // return isalnum( anyByte ); |
---|
| 157 | // } |
---|
| 158 | } |
---|
| 159 | |
---|
| 160 | |
---|
| 161 | class TiXmlParsingData |
---|
| 162 | { |
---|
| 163 | friend class TiXmlDocument; |
---|
| 164 | public: |
---|
| 165 | void Stamp( const char* now, TiXmlEncoding encoding ); |
---|
| 166 | |
---|
| 167 | const TiXmlCursor& Cursor() { return cursor; } |
---|
| 168 | |
---|
| 169 | private: |
---|
| 170 | // Only used by the document! |
---|
| 171 | TiXmlParsingData( const char* start, int _tabsize, int row, int col ) |
---|
| 172 | { |
---|
| 173 | assert( start ); |
---|
| 174 | stamp = start; |
---|
| 175 | tabsize = _tabsize; |
---|
| 176 | cursor.row = row; |
---|
| 177 | cursor.col = col; |
---|
| 178 | } |
---|
| 179 | |
---|
| 180 | TiXmlCursor cursor; |
---|
| 181 | const char* stamp; |
---|
| 182 | int tabsize; |
---|
| 183 | }; |
---|
| 184 | |
---|
| 185 | |
---|
| 186 | void TiXmlParsingData::Stamp( const char* now, TiXmlEncoding encoding ) |
---|
| 187 | { |
---|
| 188 | assert( now ); |
---|
| 189 | |
---|
| 190 | // Do nothing if the tabsize is 0. |
---|
| 191 | if ( tabsize < 1 ) |
---|
| 192 | { |
---|
| 193 | return; |
---|
| 194 | } |
---|
| 195 | |
---|
| 196 | // Get the current row, column. |
---|
| 197 | int row = cursor.row; |
---|
| 198 | int col = cursor.col; |
---|
| 199 | const char* p = stamp; |
---|
| 200 | assert( p ); |
---|
| 201 | |
---|
| 202 | while ( p < now ) |
---|
| 203 | { |
---|
| 204 | // Code contributed by Fletcher Dunn: (modified by lee) |
---|
| 205 | switch (*p) { |
---|
| 206 | case 0: |
---|
| 207 | // We *should* never get here, but in case we do, don't |
---|
| 208 | // advance past the terminating null character, ever |
---|
| 209 | return; |
---|
| 210 | |
---|
| 211 | case '\r': |
---|
| 212 | // bump down to the next line |
---|
| 213 | ++row; |
---|
| 214 | col = 0; |
---|
| 215 | // Eat the character |
---|
| 216 | ++p; |
---|
| 217 | |
---|
| 218 | // Check for \r\n sequence, and treat this as a single character |
---|
| 219 | if (*p == '\n') { |
---|
| 220 | ++p; |
---|
| 221 | } |
---|
| 222 | break; |
---|
| 223 | |
---|
| 224 | case '\n': |
---|
| 225 | // bump down to the next line |
---|
| 226 | ++row; |
---|
| 227 | col = 0; |
---|
| 228 | |
---|
| 229 | // Eat the character |
---|
| 230 | ++p; |
---|
| 231 | |
---|
| 232 | // Check for \n\r sequence, and treat this as a single |
---|
| 233 | // character. (Yes, this bizarre thing does occur still |
---|
| 234 | // on some arcane platforms...) |
---|
| 235 | if (*p == '\r') { |
---|
| 236 | ++p; |
---|
| 237 | } |
---|
| 238 | break; |
---|
| 239 | |
---|
| 240 | case '\t': |
---|
| 241 | // Eat the character |
---|
| 242 | ++p; |
---|
| 243 | |
---|
| 244 | // Skip to next tab stop |
---|
| 245 | col = (col / tabsize + 1) * tabsize; |
---|
| 246 | break; |
---|
| 247 | |
---|
| 248 | case TIXML_UTF_LEAD_0: |
---|
| 249 | if ( encoding == TIXML_ENCODING_UTF8 ) |
---|
| 250 | { |
---|
| 251 | if ( *(p+1) && *(p+2) ) |
---|
| 252 | { |
---|
| 253 | // In these cases, don't advance the column. These are |
---|
| 254 | // 0-width spaces. |
---|
| 255 | if ( *(p+1)==TIXML_UTF_LEAD_1 && *(p+2)==TIXML_UTF_LEAD_2 ) |
---|
| 256 | p += 3; |
---|
| 257 | else if ( *(p+1)==(char)(0xbf) && *(p+2)==(char)(0xbe) ) |
---|
| 258 | p += 3; |
---|
| 259 | else if ( *(p+1)==(char)(0xbf) && *(p+2)==(char)(0xbf) ) |
---|
| 260 | p += 3; |
---|
| 261 | else |
---|
| 262 | { p +=3; ++col; } // A normal character. |
---|
| 263 | } |
---|
| 264 | } |
---|
| 265 | else |
---|
| 266 | { |
---|
| 267 | ++p; |
---|
| 268 | ++col; |
---|
| 269 | } |
---|
| 270 | break; |
---|
| 271 | |
---|
| 272 | default: |
---|
| 273 | if ( encoding == TIXML_ENCODING_UTF8 ) |
---|
| 274 | { |
---|
| 275 | // Eat the 1 to 4 byte utf8 character. |
---|
| 276 | int step = TiXmlBase::utf8ByteTable[*((unsigned char*)p)]; |
---|
| 277 | if ( step == 0 ) |
---|
| 278 | step = 1; // Error case from bad encoding, but handle gracefully. |
---|
| 279 | p += step; |
---|
| 280 | |
---|
| 281 | // Just advance one column, of course. |
---|
| 282 | ++col; |
---|
| 283 | } |
---|
| 284 | else |
---|
| 285 | { |
---|
| 286 | ++p; |
---|
| 287 | ++col; |
---|
| 288 | } |
---|
| 289 | break; |
---|
| 290 | } |
---|
| 291 | } |
---|
| 292 | cursor.row = row; |
---|
| 293 | cursor.col = col; |
---|
| 294 | assert( cursor.row >= -1 ); |
---|
| 295 | assert( cursor.col >= -1 ); |
---|
| 296 | stamp = p; |
---|
| 297 | assert( stamp ); |
---|
| 298 | } |
---|
| 299 | |
---|
| 300 | |
---|
| 301 | const char* TiXmlBase::SkipWhiteSpace( const char* p, TiXmlEncoding encoding ) |
---|
| 302 | { |
---|
| 303 | if ( !p || !*p ) |
---|
| 304 | { |
---|
| 305 | return 0; |
---|
| 306 | } |
---|
| 307 | if ( encoding == TIXML_ENCODING_UTF8 ) |
---|
| 308 | { |
---|
| 309 | while ( *p ) |
---|
| 310 | { |
---|
| 311 | // Skip the stupid Microsoft UTF-8 Byte order marks |
---|
| 312 | if ( *(p+0)==TIXML_UTF_LEAD_0 |
---|
| 313 | && *(p+1)==TIXML_UTF_LEAD_1 |
---|
| 314 | && *(p+2)==TIXML_UTF_LEAD_2 ) |
---|
| 315 | { |
---|
| 316 | p += 3; |
---|
| 317 | continue; |
---|
| 318 | } |
---|
| 319 | else if(*(p+0)==TIXML_UTF_LEAD_0 |
---|
| 320 | && *(p+1)==(const char) 0xbf |
---|
| 321 | && *(p+2)==(const char) 0xbe ) |
---|
| 322 | { |
---|
| 323 | p += 3; |
---|
| 324 | continue; |
---|
| 325 | } |
---|
| 326 | else if(*(p+0)==TIXML_UTF_LEAD_0 |
---|
| 327 | && *(p+1)==(const char) 0xbf |
---|
| 328 | && *(p+2)==(const char) 0xbf ) |
---|
| 329 | { |
---|
| 330 | p += 3; |
---|
| 331 | continue; |
---|
| 332 | } |
---|
| 333 | |
---|
| 334 | if ( IsWhiteSpace( *p ) || *p == '\n' || *p =='\r' ) // Still using old rules for white space. |
---|
| 335 | ++p; |
---|
| 336 | else |
---|
| 337 | break; |
---|
| 338 | } |
---|
| 339 | } |
---|
| 340 | else |
---|
| 341 | { |
---|
| 342 | while ( *p && IsWhiteSpace( *p ) || *p == '\n' || *p =='\r' ) |
---|
| 343 | ++p; |
---|
| 344 | } |
---|
| 345 | |
---|
| 346 | return p; |
---|
| 347 | } |
---|
| 348 | |
---|
| 349 | #ifdef TIXML_USE_STL |
---|
| 350 | /*static*/ bool TiXmlBase::StreamWhiteSpace( TIXML_ISTREAM * in, TIXML_STRING * tag ) |
---|
| 351 | { |
---|
| 352 | for( ;; ) |
---|
| 353 | { |
---|
| 354 | if ( !in->good() ) return false; |
---|
| 355 | |
---|
| 356 | int c = in->peek(); |
---|
| 357 | // At this scope, we can't get to a document. So fail silently. |
---|
| 358 | if ( !IsWhiteSpace( c ) || c <= 0 ) |
---|
| 359 | return true; |
---|
| 360 | |
---|
| 361 | *tag += (char) in->get(); |
---|
| 362 | } |
---|
| 363 | } |
---|
| 364 | |
---|
| 365 | /*static*/ bool TiXmlBase::StreamTo( TIXML_ISTREAM * in, int character, TIXML_STRING * tag ) |
---|
| 366 | { |
---|
| 367 | //assert( character > 0 && character < 128 ); // else it won't work in utf-8 |
---|
| 368 | while ( in->good() ) |
---|
| 369 | { |
---|
| 370 | int c = in->peek(); |
---|
| 371 | if ( c == character ) |
---|
| 372 | return true; |
---|
| 373 | if ( c <= 0 ) // Silent failure: can't get document at this scope |
---|
| 374 | return false; |
---|
| 375 | |
---|
| 376 | in->get(); |
---|
| 377 | *tag += (char) c; |
---|
| 378 | } |
---|
| 379 | return false; |
---|
| 380 | } |
---|
| 381 | #endif |
---|
| 382 | |
---|
| 383 | const char* TiXmlBase::ReadName( const char* p, TIXML_STRING * name, TiXmlEncoding encoding ) |
---|
| 384 | { |
---|
| 385 | *name = ""; |
---|
| 386 | assert( p ); |
---|
| 387 | |
---|
| 388 | // Names start with letters or underscores. |
---|
| 389 | // Of course, in unicode, tinyxml has no idea what a letter *is*. The |
---|
| 390 | // algorithm is generous. |
---|
| 391 | // |
---|
| 392 | // After that, they can be letters, underscores, numbers, |
---|
| 393 | // hyphens, or colons. (Colons are valid ony for namespaces, |
---|
| 394 | // but tinyxml can't tell namespaces from names.) |
---|
| 395 | if ( p && *p |
---|
| 396 | && ( IsAlpha( (unsigned char) *p, encoding ) || *p == '_' ) ) |
---|
| 397 | { |
---|
| 398 | while( p && *p |
---|
| 399 | && ( IsAlphaNum( (unsigned char ) *p, encoding ) |
---|
| 400 | || *p == '_' |
---|
| 401 | || *p == '-' |
---|
| 402 | || *p == '.' |
---|
| 403 | || *p == ':' ) ) |
---|
| 404 | { |
---|
| 405 | (*name) += *p; |
---|
| 406 | ++p; |
---|
| 407 | } |
---|
| 408 | return p; |
---|
| 409 | } |
---|
| 410 | return 0; |
---|
| 411 | } |
---|
| 412 | |
---|
| 413 | const char* TiXmlBase::GetEntity( const char* p, char* value, int* length, TiXmlEncoding encoding ) |
---|
| 414 | { |
---|
| 415 | // Presume an entity, and pull it out. |
---|
| 416 | TIXML_STRING ent; |
---|
| 417 | int i; |
---|
| 418 | *length = 0; |
---|
| 419 | |
---|
| 420 | if ( *(p+1) && *(p+1) == '#' && *(p+2) ) |
---|
| 421 | { |
---|
| 422 | unsigned long ucs = 0; |
---|
| 423 | //*ME: warning C4244: convert '__w64 int' to 'unsigned' |
---|
| 424 | //*ME: Use size_t instead of unsigned (pointer-arithmetic) |
---|
| 425 | size_t delta = 0; |
---|
| 426 | unsigned mult = 1; |
---|
| 427 | |
---|
| 428 | if ( *(p+2) == 'x' ) |
---|
| 429 | { |
---|
| 430 | // Hexadecimal. |
---|
| 431 | if ( !*(p+3) ) return 0; |
---|
| 432 | |
---|
| 433 | const char* q = p+3; |
---|
| 434 | q = strchr( q, ';' ); |
---|
| 435 | |
---|
| 436 | if ( !q || !*q ) return 0; |
---|
| 437 | |
---|
| 438 | delta = q-p; |
---|
| 439 | --q; |
---|
| 440 | |
---|
| 441 | while ( *q != 'x' ) |
---|
| 442 | { |
---|
| 443 | if ( *q >= '0' && *q <= '9' ) |
---|
| 444 | ucs += mult * (*q - '0'); |
---|
| 445 | else if ( *q >= 'a' && *q <= 'f' ) |
---|
| 446 | ucs += mult * (*q - 'a' + 10); |
---|
| 447 | else if ( *q >= 'A' && *q <= 'F' ) |
---|
| 448 | ucs += mult * (*q - 'A' + 10 ); |
---|
| 449 | else |
---|
| 450 | return 0; |
---|
| 451 | mult *= 16; |
---|
| 452 | --q; |
---|
| 453 | } |
---|
| 454 | } |
---|
| 455 | else |
---|
| 456 | { |
---|
| 457 | // Decimal. |
---|
| 458 | if ( !*(p+2) ) return 0; |
---|
| 459 | |
---|
| 460 | const char* q = p+2; |
---|
| 461 | q = strchr( q, ';' ); |
---|
| 462 | |
---|
| 463 | if ( !q || !*q ) return 0; |
---|
| 464 | |
---|
| 465 | delta = q-p; |
---|
| 466 | --q; |
---|
| 467 | |
---|
| 468 | while ( *q != '#' ) |
---|
| 469 | { |
---|
| 470 | if ( *q >= '0' && *q <= '9' ) |
---|
| 471 | ucs += mult * (*q - '0'); |
---|
| 472 | else |
---|
| 473 | return 0; |
---|
| 474 | mult *= 10; |
---|
| 475 | --q; |
---|
| 476 | } |
---|
| 477 | } |
---|
| 478 | if ( encoding == TIXML_ENCODING_UTF8 ) |
---|
| 479 | { |
---|
| 480 | // convert the UCS to UTF-8 |
---|
| 481 | ConvertUTF32ToUTF8( ucs, value, length ); |
---|
| 482 | } |
---|
| 483 | else |
---|
| 484 | { |
---|
| 485 | *value = (char)ucs; |
---|
| 486 | *length = 1; |
---|
| 487 | } |
---|
| 488 | return p + delta + 1; |
---|
| 489 | } |
---|
| 490 | |
---|
| 491 | // Now try to match it. |
---|
| 492 | for( i=0; i<NUM_ENTITY; ++i ) |
---|
| 493 | { |
---|
| 494 | if ( strncmp( entity[i].str, p, entity[i].strLength ) == 0 ) |
---|
| 495 | { |
---|
| 496 | assert( strlen( entity[i].str ) == entity[i].strLength ); |
---|
| 497 | *value = entity[i].chr; |
---|
| 498 | *length = 1; |
---|
| 499 | return ( p + entity[i].strLength ); |
---|
| 500 | } |
---|
| 501 | } |
---|
| 502 | |
---|
| 503 | // So it wasn't an entity, its unrecognized, or something like that. |
---|
| 504 | *value = *p; // Don't put back the last one, since we return it! |
---|
| 505 | return p+1; |
---|
| 506 | } |
---|
| 507 | |
---|
| 508 | |
---|
| 509 | bool TiXmlBase::StringEqual( const char* p, |
---|
| 510 | const char* tag, |
---|
| 511 | bool ignoreCase, |
---|
| 512 | TiXmlEncoding encoding ) |
---|
| 513 | { |
---|
| 514 | assert( p ); |
---|
| 515 | assert( tag ); |
---|
| 516 | if ( !p || !*p ) |
---|
| 517 | { |
---|
| 518 | assert( 0 ); |
---|
| 519 | return false; |
---|
| 520 | } |
---|
| 521 | |
---|
| 522 | const char* q = p; |
---|
| 523 | |
---|
| 524 | if ( ignoreCase ) |
---|
| 525 | { |
---|
| 526 | while ( *q && *tag && ToLower( *q, encoding ) == ToLower( *tag, encoding ) ) |
---|
| 527 | { |
---|
| 528 | ++q; |
---|
| 529 | ++tag; |
---|
| 530 | } |
---|
| 531 | |
---|
| 532 | if ( *tag == 0 ) |
---|
| 533 | return true; |
---|
| 534 | } |
---|
| 535 | else |
---|
| 536 | { |
---|
| 537 | while ( *q && *tag && *q == *tag ) |
---|
| 538 | { |
---|
| 539 | ++q; |
---|
| 540 | ++tag; |
---|
| 541 | } |
---|
| 542 | |
---|
| 543 | if ( *tag == 0 ) // Have we found the end of the tag, and everything equal? |
---|
| 544 | return true; |
---|
| 545 | } |
---|
| 546 | return false; |
---|
| 547 | } |
---|
| 548 | |
---|
| 549 | const char* TiXmlBase::ReadText( const char* p, |
---|
| 550 | TIXML_STRING * text, |
---|
| 551 | bool trimWhiteSpace, |
---|
| 552 | const char* endTag, |
---|
| 553 | bool caseInsensitive, |
---|
| 554 | TiXmlEncoding encoding ) |
---|
| 555 | { |
---|
| 556 | *text = ""; |
---|
| 557 | if ( !trimWhiteSpace // certain tags always keep whitespace |
---|
| 558 | || !condenseWhiteSpace ) // if true, whitespace is always kept |
---|
| 559 | { |
---|
| 560 | // Keep all the white space. |
---|
| 561 | while ( p && *p |
---|
| 562 | && !StringEqual( p, endTag, caseInsensitive, encoding ) |
---|
| 563 | ) |
---|
| 564 | { |
---|
| 565 | int len; |
---|
| 566 | char cArr[4] = { 0, 0, 0, 0 }; |
---|
| 567 | p = GetChar( p, cArr, &len, encoding ); |
---|
| 568 | text->append( cArr, len ); |
---|
| 569 | } |
---|
| 570 | } |
---|
| 571 | else |
---|
| 572 | { |
---|
| 573 | bool whitespace = false; |
---|
| 574 | |
---|
| 575 | // Remove leading white space: |
---|
| 576 | p = SkipWhiteSpace( p, encoding ); |
---|
| 577 | while ( p && *p |
---|
| 578 | && !StringEqual( p, endTag, caseInsensitive, encoding ) ) |
---|
| 579 | { |
---|
| 580 | if ( *p == '\r' || *p == '\n' ) |
---|
| 581 | { |
---|
| 582 | whitespace = true; |
---|
| 583 | ++p; |
---|
| 584 | } |
---|
| 585 | else if ( IsWhiteSpace( *p ) ) |
---|
| 586 | { |
---|
| 587 | whitespace = true; |
---|
| 588 | ++p; |
---|
| 589 | } |
---|
| 590 | else |
---|
| 591 | { |
---|
| 592 | // If we've found whitespace, add it before the |
---|
| 593 | // new character. Any whitespace just becomes a space. |
---|
| 594 | if ( whitespace ) |
---|
| 595 | { |
---|
| 596 | (*text) += ' '; |
---|
| 597 | whitespace = false; |
---|
| 598 | } |
---|
| 599 | int len; |
---|
| 600 | char cArr[4] = { 0, 0, 0, 0 }; |
---|
| 601 | p = GetChar( p, cArr, &len, encoding ); |
---|
| 602 | if ( len == 1 ) |
---|
| 603 | (*text) += cArr[0]; // more efficient |
---|
| 604 | else |
---|
| 605 | text->append( cArr, len ); |
---|
| 606 | } |
---|
| 607 | } |
---|
| 608 | } |
---|
| 609 | return p + strlen( endTag ); |
---|
| 610 | } |
---|
| 611 | |
---|
| 612 | #ifdef TIXML_USE_STL |
---|
| 613 | |
---|
| 614 | void TiXmlDocument::StreamIn( TIXML_ISTREAM * in, TIXML_STRING * tag ) |
---|
| 615 | { |
---|
| 616 | // The basic issue with a document is that we don't know what we're |
---|
| 617 | // streaming. Read something presumed to be a tag (and hope), then |
---|
| 618 | // identify it, and call the appropriate stream method on the tag. |
---|
| 619 | // |
---|
| 620 | // This "pre-streaming" will never read the closing ">" so the |
---|
| 621 | // sub-tag can orient itself. |
---|
| 622 | |
---|
| 623 | if ( !StreamTo( in, '<', tag ) ) |
---|
| 624 | { |
---|
| 625 | SetError( TIXML_ERROR_PARSING_EMPTY, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
| 626 | return; |
---|
| 627 | } |
---|
| 628 | |
---|
| 629 | while ( in->good() ) |
---|
| 630 | { |
---|
| 631 | int tagIndex = (int) tag->length(); |
---|
| 632 | while ( in->good() && in->peek() != '>' ) |
---|
| 633 | { |
---|
| 634 | int c = in->get(); |
---|
| 635 | if ( c <= 0 ) |
---|
| 636 | { |
---|
| 637 | SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
| 638 | break; |
---|
| 639 | } |
---|
| 640 | (*tag) += (char) c; |
---|
| 641 | } |
---|
| 642 | |
---|
| 643 | if ( in->good() ) |
---|
| 644 | { |
---|
| 645 | // We now have something we presume to be a node of |
---|
| 646 | // some sort. Identify it, and call the node to |
---|
| 647 | // continue streaming. |
---|
| 648 | TiXmlNode* node = Identify( tag->c_str() + tagIndex, TIXML_DEFAULT_ENCODING ); |
---|
| 649 | |
---|
| 650 | if ( node ) |
---|
| 651 | { |
---|
| 652 | node->StreamIn( in, tag ); |
---|
| 653 | bool isElement = node->ToElement() != 0; |
---|
| 654 | delete node; |
---|
| 655 | node = 0; |
---|
| 656 | |
---|
| 657 | // If this is the root element, we're done. Parsing will be |
---|
| 658 | // done by the >> operator. |
---|
| 659 | if ( isElement ) |
---|
| 660 | { |
---|
| 661 | return; |
---|
| 662 | } |
---|
| 663 | } |
---|
| 664 | else |
---|
| 665 | { |
---|
| 666 | SetError( TIXML_ERROR, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
| 667 | return; |
---|
| 668 | } |
---|
| 669 | } |
---|
| 670 | } |
---|
| 671 | // We should have returned sooner. |
---|
| 672 | SetError( TIXML_ERROR, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
| 673 | } |
---|
| 674 | |
---|
| 675 | #endif |
---|
| 676 | |
---|
| 677 | const char* TiXmlDocument::Parse( const char* p, TiXmlParsingData* prevData, TiXmlEncoding encoding ) |
---|
| 678 | { |
---|
| 679 | ClearError(); |
---|
| 680 | |
---|
| 681 | // Parse away, at the document level. Since a document |
---|
| 682 | // contains nothing but other tags, most of what happens |
---|
| 683 | // here is skipping white space. |
---|
| 684 | if ( !p || !*p ) |
---|
| 685 | { |
---|
| 686 | SetError( TIXML_ERROR_DOCUMENT_EMPTY, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
| 687 | return 0; |
---|
| 688 | } |
---|
| 689 | |
---|
| 690 | // Note that, for a document, this needs to come |
---|
| 691 | // before the while space skip, so that parsing |
---|
| 692 | // starts from the pointer we are given. |
---|
| 693 | location.Clear(); |
---|
| 694 | if ( prevData ) |
---|
| 695 | { |
---|
| 696 | location.row = prevData->cursor.row; |
---|
| 697 | location.col = prevData->cursor.col; |
---|
| 698 | } |
---|
| 699 | else |
---|
| 700 | { |
---|
| 701 | location.row = 0; |
---|
| 702 | location.col = 0; |
---|
| 703 | } |
---|
| 704 | TiXmlParsingData data( p, TabSize(), location.row, location.col ); |
---|
| 705 | location = data.Cursor(); |
---|
| 706 | |
---|
| 707 | if ( encoding == TIXML_ENCODING_UNKNOWN ) |
---|
| 708 | { |
---|
| 709 | // Check for the Microsoft UTF-8 lead bytes. |
---|
| 710 | if ( *(p+0) && *(p+0) == TIXML_UTF_LEAD_0 |
---|
| 711 | && *(p+1) && *(p+1) == TIXML_UTF_LEAD_1 |
---|
| 712 | && *(p+2) && *(p+2) == TIXML_UTF_LEAD_2 ) |
---|
| 713 | { |
---|
| 714 | encoding = TIXML_ENCODING_UTF8; |
---|
| 715 | } |
---|
| 716 | } |
---|
| 717 | |
---|
| 718 | p = SkipWhiteSpace( p, encoding ); |
---|
| 719 | if ( !p ) |
---|
| 720 | { |
---|
| 721 | SetError( TIXML_ERROR_DOCUMENT_EMPTY, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
| 722 | return 0; |
---|
| 723 | } |
---|
| 724 | |
---|
| 725 | while ( p && *p ) |
---|
| 726 | { |
---|
| 727 | TiXmlNode* node = Identify( p, encoding ); |
---|
| 728 | if ( node ) |
---|
| 729 | { |
---|
| 730 | p = node->Parse( p, &data, encoding ); |
---|
| 731 | LinkEndChild( node ); |
---|
| 732 | } |
---|
| 733 | else |
---|
| 734 | { |
---|
| 735 | break; |
---|
| 736 | } |
---|
| 737 | |
---|
| 738 | // Did we get encoding info? |
---|
| 739 | if ( encoding == TIXML_ENCODING_UNKNOWN |
---|
| 740 | && node->ToDeclaration() ) |
---|
| 741 | { |
---|
| 742 | TiXmlDeclaration* dec = node->ToDeclaration(); |
---|
| 743 | const char* enc = dec->Encoding(); |
---|
| 744 | assert( enc ); |
---|
| 745 | |
---|
| 746 | if ( *enc == 0 ) |
---|
| 747 | encoding = TIXML_ENCODING_UTF8; |
---|
| 748 | else if ( StringEqual( enc, "UTF-8", true, TIXML_ENCODING_UNKNOWN ) ) |
---|
| 749 | encoding = TIXML_ENCODING_UTF8; |
---|
| 750 | else if ( StringEqual( enc, "UTF8", true, TIXML_ENCODING_UNKNOWN ) ) |
---|
| 751 | encoding = TIXML_ENCODING_UTF8; // incorrect, but be nice |
---|
| 752 | else |
---|
| 753 | encoding = TIXML_ENCODING_LEGACY; |
---|
| 754 | } |
---|
| 755 | |
---|
| 756 | p = SkipWhiteSpace( p, encoding ); |
---|
| 757 | } |
---|
| 758 | |
---|
| 759 | // Was this empty? |
---|
| 760 | if ( !firstChild ) { |
---|
| 761 | SetError( TIXML_ERROR_DOCUMENT_EMPTY, 0, 0, encoding ); |
---|
| 762 | return 0; |
---|
| 763 | } |
---|
| 764 | |
---|
| 765 | // All is well. |
---|
| 766 | return p; |
---|
| 767 | } |
---|
| 768 | |
---|
| 769 | void TiXmlDocument::SetError( int err, const char* pError, TiXmlParsingData* data, TiXmlEncoding encoding ) |
---|
| 770 | { |
---|
| 771 | // The first error in a chain is more accurate - don't set again! |
---|
| 772 | if ( error ) |
---|
| 773 | return; |
---|
| 774 | |
---|
| 775 | assert( err > 0 && err < TIXML_ERROR_STRING_COUNT ); |
---|
| 776 | error = true; |
---|
| 777 | errorId = err; |
---|
| 778 | errorDesc = errorString[ errorId ]; |
---|
| 779 | |
---|
| 780 | errorLocation.Clear(); |
---|
| 781 | if ( pError && data ) |
---|
| 782 | { |
---|
| 783 | //TiXmlParsingData data( pError, prevData ); |
---|
| 784 | data->Stamp( pError, encoding ); |
---|
| 785 | errorLocation = data->Cursor(); |
---|
| 786 | } |
---|
| 787 | } |
---|
| 788 | |
---|
| 789 | |
---|
| 790 | TiXmlNode* TiXmlNode::Identify( const char* p, TiXmlEncoding encoding ) |
---|
| 791 | { |
---|
| 792 | TiXmlNode* returnNode = 0; |
---|
| 793 | |
---|
| 794 | p = SkipWhiteSpace( p, encoding ); |
---|
| 795 | if( !p || !*p || *p != '<' ) |
---|
| 796 | { |
---|
| 797 | return 0; |
---|
| 798 | } |
---|
| 799 | |
---|
| 800 | TiXmlDocument* doc = GetDocument(); |
---|
| 801 | p = SkipWhiteSpace( p, encoding ); |
---|
| 802 | |
---|
| 803 | if ( !p || !*p ) |
---|
| 804 | { |
---|
| 805 | return 0; |
---|
| 806 | } |
---|
| 807 | |
---|
| 808 | // What is this thing? |
---|
| 809 | // - Elements start with a letter or underscore, but xml is reserved. |
---|
| 810 | // - Comments: <!-- |
---|
| 811 | // - Decleration: <?xml |
---|
| 812 | // - Everthing else is unknown to tinyxml. |
---|
| 813 | // |
---|
| 814 | |
---|
| 815 | const char* xmlHeader = { "<?xml" }; |
---|
| 816 | const char* commentHeader = { "<!--" }; |
---|
| 817 | const char* dtdHeader = { "<!" }; |
---|
| 818 | |
---|
| 819 | if ( StringEqual( p, xmlHeader, true, encoding ) ) |
---|
| 820 | { |
---|
| 821 | #ifdef DEBUG_PARSER |
---|
| 822 | TIXML_LOG( "XML parsing Declaration\n" ); |
---|
| 823 | #endif |
---|
| 824 | returnNode = new TiXmlDeclaration(); |
---|
| 825 | } |
---|
| 826 | else if ( StringEqual( p, commentHeader, false, encoding ) ) |
---|
| 827 | { |
---|
| 828 | #ifdef DEBUG_PARSER |
---|
| 829 | TIXML_LOG( "XML parsing Comment\n" ); |
---|
| 830 | #endif |
---|
| 831 | returnNode = new TiXmlComment(); |
---|
| 832 | } |
---|
| 833 | else if ( StringEqual( p, dtdHeader, false, encoding ) ) |
---|
| 834 | { |
---|
| 835 | #ifdef DEBUG_PARSER |
---|
| 836 | TIXML_LOG( "XML parsing Unknown(1)\n" ); |
---|
| 837 | #endif |
---|
| 838 | returnNode = new TiXmlUnknown(); |
---|
| 839 | } |
---|
| 840 | else if ( IsAlpha( *(p+1), encoding ) |
---|
| 841 | || *(p+1) == '_' ) |
---|
| 842 | { |
---|
| 843 | #ifdef DEBUG_PARSER |
---|
| 844 | TIXML_LOG( "XML parsing Element\n" ); |
---|
| 845 | #endif |
---|
| 846 | returnNode = new TiXmlElement( "" ); |
---|
| 847 | } |
---|
| 848 | else |
---|
| 849 | { |
---|
| 850 | #ifdef DEBUG_PARSER |
---|
| 851 | TIXML_LOG( "XML parsing Unknown(2)\n" ); |
---|
| 852 | #endif |
---|
| 853 | returnNode = new TiXmlUnknown(); |
---|
| 854 | } |
---|
| 855 | |
---|
| 856 | if ( returnNode ) |
---|
| 857 | { |
---|
| 858 | // Set the parent, so it can report errors |
---|
| 859 | returnNode->parent = this; |
---|
| 860 | } |
---|
| 861 | else |
---|
| 862 | { |
---|
| 863 | if ( doc ) |
---|
| 864 | doc->SetError( TIXML_ERROR_OUT_OF_MEMORY, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
| 865 | } |
---|
| 866 | return returnNode; |
---|
| 867 | } |
---|
| 868 | |
---|
| 869 | #ifdef TIXML_USE_STL |
---|
| 870 | |
---|
| 871 | void TiXmlElement::StreamIn (TIXML_ISTREAM * in, TIXML_STRING * tag) |
---|
| 872 | { |
---|
| 873 | // We're called with some amount of pre-parsing. That is, some of "this" |
---|
| 874 | // element is in "tag". Go ahead and stream to the closing ">" |
---|
| 875 | while( in->good() ) |
---|
| 876 | { |
---|
| 877 | int c = in->get(); |
---|
| 878 | if ( c <= 0 ) |
---|
| 879 | { |
---|
| 880 | TiXmlDocument* document = GetDocument(); |
---|
| 881 | if ( document ) |
---|
| 882 | document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
| 883 | return; |
---|
| 884 | } |
---|
| 885 | (*tag) += (char) c ; |
---|
| 886 | |
---|
| 887 | if ( c == '>' ) |
---|
| 888 | break; |
---|
| 889 | } |
---|
| 890 | |
---|
| 891 | if ( tag->length() < 3 ) return; |
---|
| 892 | |
---|
| 893 | // Okay...if we are a "/>" tag, then we're done. We've read a complete tag. |
---|
| 894 | // If not, identify and stream. |
---|
| 895 | |
---|
| 896 | if ( tag->at( tag->length() - 1 ) == '>' |
---|
| 897 | && tag->at( tag->length() - 2 ) == '/' ) |
---|
| 898 | { |
---|
| 899 | // All good! |
---|
| 900 | return; |
---|
| 901 | } |
---|
| 902 | else if ( tag->at( tag->length() - 1 ) == '>' ) |
---|
| 903 | { |
---|
| 904 | // There is more. Could be: |
---|
| 905 | // text |
---|
| 906 | // closing tag |
---|
| 907 | // another node. |
---|
| 908 | for ( ;; ) |
---|
| 909 | { |
---|
| 910 | StreamWhiteSpace( in, tag ); |
---|
| 911 | |
---|
| 912 | // Do we have text? |
---|
| 913 | if ( in->good() && in->peek() != '<' ) |
---|
| 914 | { |
---|
| 915 | // Yep, text. |
---|
| 916 | TiXmlText text( "" ); |
---|
| 917 | text.StreamIn( in, tag ); |
---|
| 918 | |
---|
| 919 | // What follows text is a closing tag or another node. |
---|
| 920 | // Go around again and figure it out. |
---|
| 921 | continue; |
---|
| 922 | } |
---|
| 923 | |
---|
| 924 | // We now have either a closing tag...or another node. |
---|
| 925 | // We should be at a "<", regardless. |
---|
| 926 | if ( !in->good() ) return; |
---|
| 927 | assert( in->peek() == '<' ); |
---|
| 928 | int tagIndex = tag->length(); |
---|
| 929 | |
---|
| 930 | bool closingTag = false; |
---|
| 931 | bool firstCharFound = false; |
---|
| 932 | |
---|
| 933 | for( ;; ) |
---|
| 934 | { |
---|
| 935 | if ( !in->good() ) |
---|
| 936 | return; |
---|
| 937 | |
---|
| 938 | int c = in->peek(); |
---|
| 939 | if ( c <= 0 ) |
---|
| 940 | { |
---|
| 941 | TiXmlDocument* document = GetDocument(); |
---|
| 942 | if ( document ) |
---|
| 943 | document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
| 944 | return; |
---|
| 945 | } |
---|
| 946 | |
---|
| 947 | if ( c == '>' ) |
---|
| 948 | break; |
---|
| 949 | |
---|
| 950 | *tag += (char) c; |
---|
| 951 | in->get(); |
---|
| 952 | |
---|
| 953 | if ( !firstCharFound && c != '<' && !IsWhiteSpace( c ) ) |
---|
| 954 | { |
---|
| 955 | firstCharFound = true; |
---|
| 956 | if ( c == '/' ) |
---|
| 957 | closingTag = true; |
---|
| 958 | } |
---|
| 959 | } |
---|
| 960 | // If it was a closing tag, then read in the closing '>' to clean up the input stream. |
---|
| 961 | // If it was not, the streaming will be done by the tag. |
---|
| 962 | if ( closingTag ) |
---|
| 963 | { |
---|
| 964 | if ( !in->good() ) |
---|
| 965 | return; |
---|
| 966 | |
---|
| 967 | int c = in->get(); |
---|
| 968 | if ( c <= 0 ) |
---|
| 969 | { |
---|
| 970 | TiXmlDocument* document = GetDocument(); |
---|
| 971 | if ( document ) |
---|
| 972 | document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
| 973 | return; |
---|
| 974 | } |
---|
| 975 | assert( c == '>' ); |
---|
| 976 | *tag += (char) c; |
---|
| 977 | |
---|
| 978 | // We are done, once we've found our closing tag. |
---|
| 979 | return; |
---|
| 980 | } |
---|
| 981 | else |
---|
| 982 | { |
---|
| 983 | // If not a closing tag, id it, and stream. |
---|
| 984 | const char* tagloc = tag->c_str() + tagIndex; |
---|
| 985 | TiXmlNode* node = Identify( tagloc, TIXML_DEFAULT_ENCODING ); |
---|
| 986 | if ( !node ) |
---|
| 987 | return; |
---|
| 988 | node->StreamIn( in, tag ); |
---|
| 989 | delete node; |
---|
| 990 | node = 0; |
---|
| 991 | |
---|
| 992 | // No return: go around from the beginning: text, closing tag, or node. |
---|
| 993 | } |
---|
| 994 | } |
---|
| 995 | } |
---|
| 996 | } |
---|
| 997 | #endif |
---|
| 998 | |
---|
| 999 | const char* TiXmlElement::Parse( const char* p, TiXmlParsingData* data, TiXmlEncoding encoding ) |
---|
| 1000 | { |
---|
| 1001 | p = SkipWhiteSpace( p, encoding ); |
---|
| 1002 | TiXmlDocument* document = GetDocument(); |
---|
| 1003 | |
---|
| 1004 | if ( !p || !*p ) |
---|
| 1005 | { |
---|
| 1006 | if ( document ) document->SetError( TIXML_ERROR_PARSING_ELEMENT, 0, 0, encoding ); |
---|
| 1007 | return 0; |
---|
| 1008 | } |
---|
| 1009 | |
---|
| 1010 | // TiXmlParsingData data( p, prevData ); |
---|
| 1011 | if ( data ) |
---|
| 1012 | { |
---|
| 1013 | data->Stamp( p, encoding ); |
---|
| 1014 | location = data->Cursor(); |
---|
| 1015 | } |
---|
| 1016 | |
---|
| 1017 | if ( *p != '<' ) |
---|
| 1018 | { |
---|
| 1019 | if ( document ) document->SetError( TIXML_ERROR_PARSING_ELEMENT, p, data, encoding ); |
---|
| 1020 | return 0; |
---|
| 1021 | } |
---|
| 1022 | |
---|
| 1023 | p = SkipWhiteSpace( p+1, encoding ); |
---|
| 1024 | |
---|
| 1025 | // Read the name. |
---|
| 1026 | const char* pErr = p; |
---|
| 1027 | |
---|
| 1028 | p = ReadName( p, &value, encoding ); |
---|
| 1029 | if ( !p || !*p ) |
---|
| 1030 | { |
---|
| 1031 | if ( document ) document->SetError( TIXML_ERROR_FAILED_TO_READ_ELEMENT_NAME, pErr, data, encoding ); |
---|
| 1032 | return 0; |
---|
| 1033 | } |
---|
| 1034 | |
---|
| 1035 | TIXML_STRING endTag ("</"); |
---|
| 1036 | endTag += value; |
---|
| 1037 | endTag += ">"; |
---|
| 1038 | |
---|
| 1039 | // Check for and read attributes. Also look for an empty |
---|
| 1040 | // tag or an end tag. |
---|
| 1041 | while ( p && *p ) |
---|
| 1042 | { |
---|
| 1043 | pErr = p; |
---|
| 1044 | p = SkipWhiteSpace( p, encoding ); |
---|
| 1045 | if ( !p || !*p ) |
---|
| 1046 | { |
---|
| 1047 | if ( document ) document->SetError( TIXML_ERROR_READING_ATTRIBUTES, pErr, data, encoding ); |
---|
| 1048 | return 0; |
---|
| 1049 | } |
---|
| 1050 | if ( *p == '/' ) |
---|
| 1051 | { |
---|
| 1052 | ++p; |
---|
| 1053 | // Empty tag. |
---|
| 1054 | if ( *p != '>' ) |
---|
| 1055 | { |
---|
| 1056 | if ( document ) document->SetError( TIXML_ERROR_PARSING_EMPTY, p, data, encoding ); |
---|
| 1057 | return 0; |
---|
| 1058 | } |
---|
| 1059 | return (p+1); |
---|
| 1060 | } |
---|
| 1061 | else if ( *p == '>' ) |
---|
| 1062 | { |
---|
| 1063 | // Done with attributes (if there were any.) |
---|
| 1064 | // Read the value -- which can include other |
---|
| 1065 | // elements -- read the end tag, and return. |
---|
| 1066 | ++p; |
---|
| 1067 | p = ReadValue( p, data, encoding ); // Note this is an Element method, and will set the error if one happens. |
---|
| 1068 | if ( !p || !*p ) |
---|
| 1069 | return 0; |
---|
| 1070 | |
---|
| 1071 | // We should find the end tag now |
---|
| 1072 | if ( StringEqual( p, endTag.c_str(), false, encoding ) ) |
---|
| 1073 | { |
---|
| 1074 | p += endTag.length(); |
---|
| 1075 | return p; |
---|
| 1076 | } |
---|
| 1077 | else |
---|
| 1078 | { |
---|
| 1079 | if ( document ) document->SetError( TIXML_ERROR_READING_END_TAG, p, data, encoding ); |
---|
| 1080 | return 0; |
---|
| 1081 | } |
---|
| 1082 | } |
---|
| 1083 | else |
---|
| 1084 | { |
---|
| 1085 | // Try to read an attribute: |
---|
| 1086 | TiXmlAttribute* attrib = new TiXmlAttribute(); |
---|
| 1087 | if ( !attrib ) |
---|
| 1088 | { |
---|
| 1089 | if ( document ) document->SetError( TIXML_ERROR_OUT_OF_MEMORY, pErr, data, encoding ); |
---|
| 1090 | return 0; |
---|
| 1091 | } |
---|
| 1092 | |
---|
| 1093 | attrib->SetDocument( document ); |
---|
| 1094 | const char* pErr = p; |
---|
| 1095 | p = attrib->Parse( p, data, encoding ); |
---|
| 1096 | |
---|
| 1097 | if ( !p || !*p ) |
---|
| 1098 | { |
---|
| 1099 | if ( document ) document->SetError( TIXML_ERROR_PARSING_ELEMENT, pErr, data, encoding ); |
---|
| 1100 | delete attrib; |
---|
| 1101 | return 0; |
---|
| 1102 | } |
---|
| 1103 | |
---|
| 1104 | // Handle the strange case of double attributes: |
---|
| 1105 | TiXmlAttribute* node = attributeSet.Find( attrib->Name() ); |
---|
| 1106 | if ( node ) |
---|
| 1107 | { |
---|
| 1108 | node->SetValue( attrib->Value() ); |
---|
| 1109 | delete attrib; |
---|
| 1110 | return 0; |
---|
| 1111 | } |
---|
| 1112 | |
---|
| 1113 | attributeSet.Add( attrib ); |
---|
| 1114 | } |
---|
| 1115 | } |
---|
| 1116 | return p; |
---|
| 1117 | } |
---|
| 1118 | |
---|
| 1119 | |
---|
| 1120 | const char* TiXmlElement::ReadValue( const char* p, TiXmlParsingData* data, TiXmlEncoding encoding ) |
---|
| 1121 | { |
---|
| 1122 | TiXmlDocument* document = GetDocument(); |
---|
| 1123 | |
---|
| 1124 | const char* pWithWhiteSpace = p; |
---|
| 1125 | // Read in text and elements in any order. |
---|
| 1126 | p = SkipWhiteSpace( p, encoding ); |
---|
| 1127 | while ( p && *p ) |
---|
| 1128 | { |
---|
| 1129 | if ( *p != '<' ) |
---|
| 1130 | { |
---|
| 1131 | // Take what we have, make a text element. |
---|
| 1132 | TiXmlText* textNode = new TiXmlText( "" ); |
---|
| 1133 | |
---|
| 1134 | if ( !textNode ) |
---|
| 1135 | { |
---|
| 1136 | if ( document ) document->SetError( TIXML_ERROR_OUT_OF_MEMORY, 0, 0, encoding ); |
---|
| 1137 | return 0; |
---|
| 1138 | } |
---|
| 1139 | |
---|
| 1140 | if ( TiXmlBase::IsWhiteSpaceCondensed() ) |
---|
| 1141 | { |
---|
| 1142 | p = textNode->Parse( p, data, encoding ); |
---|
| 1143 | } |
---|
| 1144 | else |
---|
| 1145 | { |
---|
| 1146 | // Special case: we want to keep the white space |
---|
| 1147 | // so that leading spaces aren't removed. |
---|
| 1148 | p = textNode->Parse( pWithWhiteSpace, data, encoding ); |
---|
| 1149 | } |
---|
| 1150 | |
---|
| 1151 | if ( !textNode->Blank() ) |
---|
| 1152 | LinkEndChild( textNode ); |
---|
| 1153 | else |
---|
| 1154 | delete textNode; |
---|
| 1155 | } |
---|
| 1156 | else |
---|
| 1157 | { |
---|
| 1158 | // We hit a '<' |
---|
| 1159 | // Have we hit a new element or an end tag? |
---|
| 1160 | if ( StringEqual( p, "</", false, encoding ) ) |
---|
| 1161 | { |
---|
| 1162 | return p; |
---|
| 1163 | } |
---|
| 1164 | else |
---|
| 1165 | { |
---|
| 1166 | TiXmlNode* node = Identify( p, encoding ); |
---|
| 1167 | if ( node ) |
---|
| 1168 | { |
---|
| 1169 | p = node->Parse( p, data, encoding ); |
---|
| 1170 | LinkEndChild( node ); |
---|
| 1171 | } |
---|
| 1172 | else |
---|
| 1173 | { |
---|
| 1174 | return 0; |
---|
| 1175 | } |
---|
| 1176 | } |
---|
| 1177 | } |
---|
| 1178 | p = SkipWhiteSpace( p, encoding ); |
---|
| 1179 | } |
---|
| 1180 | |
---|
| 1181 | if ( !p ) |
---|
| 1182 | { |
---|
| 1183 | if ( document ) document->SetError( TIXML_ERROR_READING_ELEMENT_VALUE, 0, 0, encoding ); |
---|
| 1184 | } |
---|
| 1185 | return p; |
---|
| 1186 | } |
---|
| 1187 | |
---|
| 1188 | |
---|
| 1189 | #ifdef TIXML_USE_STL |
---|
| 1190 | void TiXmlUnknown::StreamIn( TIXML_ISTREAM * in, TIXML_STRING * tag ) |
---|
| 1191 | { |
---|
| 1192 | while ( in->good() ) |
---|
| 1193 | { |
---|
| 1194 | int c = in->get(); |
---|
| 1195 | if ( c <= 0 ) |
---|
| 1196 | { |
---|
| 1197 | TiXmlDocument* document = GetDocument(); |
---|
| 1198 | if ( document ) |
---|
| 1199 | document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
| 1200 | return; |
---|
| 1201 | } |
---|
| 1202 | (*tag) += (char) c; |
---|
| 1203 | |
---|
| 1204 | if ( c == '>' ) |
---|
| 1205 | { |
---|
| 1206 | // All is well. |
---|
| 1207 | return; |
---|
| 1208 | } |
---|
| 1209 | } |
---|
| 1210 | } |
---|
| 1211 | #endif |
---|
| 1212 | |
---|
| 1213 | |
---|
| 1214 | const char* TiXmlUnknown::Parse( const char* p, TiXmlParsingData* data, TiXmlEncoding encoding ) |
---|
| 1215 | { |
---|
| 1216 | TiXmlDocument* document = GetDocument(); |
---|
| 1217 | p = SkipWhiteSpace( p, encoding ); |
---|
| 1218 | |
---|
| 1219 | // TiXmlParsingData data( p, prevData ); |
---|
| 1220 | if ( data ) |
---|
| 1221 | { |
---|
| 1222 | data->Stamp( p, encoding ); |
---|
| 1223 | location = data->Cursor(); |
---|
| 1224 | } |
---|
| 1225 | if ( !p || !*p || *p != '<' ) |
---|
| 1226 | { |
---|
| 1227 | if ( document ) document->SetError( TIXML_ERROR_PARSING_UNKNOWN, p, data, encoding ); |
---|
| 1228 | return 0; |
---|
| 1229 | } |
---|
| 1230 | ++p; |
---|
| 1231 | value = ""; |
---|
| 1232 | |
---|
| 1233 | while ( p && *p && *p != '>' ) |
---|
| 1234 | { |
---|
| 1235 | value += *p; |
---|
| 1236 | ++p; |
---|
| 1237 | } |
---|
| 1238 | |
---|
| 1239 | if ( !p ) |
---|
| 1240 | { |
---|
| 1241 | if ( document ) document->SetError( TIXML_ERROR_PARSING_UNKNOWN, 0, 0, encoding ); |
---|
| 1242 | } |
---|
| 1243 | if ( *p == '>' ) |
---|
| 1244 | return p+1; |
---|
| 1245 | return p; |
---|
| 1246 | } |
---|
| 1247 | |
---|
| 1248 | #ifdef TIXML_USE_STL |
---|
| 1249 | void TiXmlComment::StreamIn( TIXML_ISTREAM * in, TIXML_STRING * tag ) |
---|
| 1250 | { |
---|
| 1251 | while ( in->good() ) |
---|
| 1252 | { |
---|
| 1253 | int c = in->get(); |
---|
| 1254 | if ( c <= 0 ) |
---|
| 1255 | { |
---|
| 1256 | TiXmlDocument* document = GetDocument(); |
---|
| 1257 | if ( document ) |
---|
| 1258 | document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
| 1259 | return; |
---|
| 1260 | } |
---|
| 1261 | |
---|
| 1262 | (*tag) += (char) c; |
---|
| 1263 | |
---|
| 1264 | if ( c == '>' |
---|
| 1265 | && tag->at( tag->length() - 2 ) == '-' |
---|
| 1266 | && tag->at( tag->length() - 3 ) == '-' ) |
---|
| 1267 | { |
---|
| 1268 | // All is well. |
---|
| 1269 | return; |
---|
| 1270 | } |
---|
| 1271 | } |
---|
| 1272 | } |
---|
| 1273 | #endif |
---|
| 1274 | |
---|
| 1275 | |
---|
| 1276 | const char* TiXmlComment::Parse( const char* p, TiXmlParsingData* data, TiXmlEncoding encoding ) |
---|
| 1277 | { |
---|
| 1278 | TiXmlDocument* document = GetDocument(); |
---|
| 1279 | value = ""; |
---|
| 1280 | |
---|
| 1281 | p = SkipWhiteSpace( p, encoding ); |
---|
| 1282 | |
---|
| 1283 | // TiXmlParsingData data( p, prevData ); |
---|
| 1284 | if ( data ) |
---|
| 1285 | { |
---|
| 1286 | data->Stamp( p, encoding ); |
---|
| 1287 | location = data->Cursor(); |
---|
| 1288 | } |
---|
| 1289 | const char* startTag = "<!--"; |
---|
| 1290 | const char* endTag = "-->"; |
---|
| 1291 | |
---|
| 1292 | if ( !StringEqual( p, startTag, false, encoding ) ) |
---|
| 1293 | { |
---|
| 1294 | document->SetError( TIXML_ERROR_PARSING_COMMENT, p, data, encoding ); |
---|
| 1295 | return 0; |
---|
| 1296 | } |
---|
| 1297 | p += strlen( startTag ); |
---|
| 1298 | p = ReadText( p, &value, false, endTag, false, encoding ); |
---|
| 1299 | return p; |
---|
| 1300 | } |
---|
| 1301 | |
---|
| 1302 | |
---|
| 1303 | const char* TiXmlAttribute::Parse( const char* p, TiXmlParsingData* data, TiXmlEncoding encoding ) |
---|
| 1304 | { |
---|
| 1305 | p = SkipWhiteSpace( p, encoding ); |
---|
| 1306 | if ( !p || !*p ) return 0; |
---|
| 1307 | |
---|
| 1308 | int tabsize = 4; |
---|
| 1309 | if ( document ) |
---|
| 1310 | tabsize = document->TabSize(); |
---|
| 1311 | |
---|
| 1312 | // TiXmlParsingData data( p, prevData ); |
---|
| 1313 | if ( data ) |
---|
| 1314 | { |
---|
| 1315 | data->Stamp( p, encoding ); |
---|
| 1316 | location = data->Cursor(); |
---|
| 1317 | } |
---|
| 1318 | // Read the name, the '=' and the value. |
---|
| 1319 | const char* pErr = p; |
---|
| 1320 | p = ReadName( p, &name, encoding ); |
---|
| 1321 | if ( !p || !*p ) |
---|
| 1322 | { |
---|
| 1323 | if ( document ) document->SetError( TIXML_ERROR_READING_ATTRIBUTES, pErr, data, encoding ); |
---|
| 1324 | return 0; |
---|
| 1325 | } |
---|
| 1326 | p = SkipWhiteSpace( p, encoding ); |
---|
| 1327 | if ( !p || !*p || *p != '=' ) |
---|
| 1328 | { |
---|
| 1329 | if ( document ) document->SetError( TIXML_ERROR_READING_ATTRIBUTES, p, data, encoding ); |
---|
| 1330 | return 0; |
---|
| 1331 | } |
---|
| 1332 | |
---|
| 1333 | ++p; // skip '=' |
---|
| 1334 | p = SkipWhiteSpace( p, encoding ); |
---|
| 1335 | if ( !p || !*p ) |
---|
| 1336 | { |
---|
| 1337 | if ( document ) document->SetError( TIXML_ERROR_READING_ATTRIBUTES, p, data, encoding ); |
---|
| 1338 | return 0; |
---|
| 1339 | } |
---|
| 1340 | |
---|
| 1341 | const char* end; |
---|
| 1342 | |
---|
| 1343 | if ( *p == '\'' ) |
---|
| 1344 | { |
---|
| 1345 | ++p; |
---|
| 1346 | end = "\'"; |
---|
| 1347 | p = ReadText( p, &value, false, end, false, encoding ); |
---|
| 1348 | } |
---|
| 1349 | else if ( *p == '"' ) |
---|
| 1350 | { |
---|
| 1351 | ++p; |
---|
| 1352 | end = "\""; |
---|
| 1353 | p = ReadText( p, &value, false, end, false, encoding ); |
---|
| 1354 | } |
---|
| 1355 | else |
---|
| 1356 | { |
---|
| 1357 | // All attribute values should be in single or double quotes. |
---|
| 1358 | // But this is such a common error that the parser will try |
---|
| 1359 | // its best, even without them. |
---|
| 1360 | value = ""; |
---|
| 1361 | while ( p && *p // existence |
---|
| 1362 | && !IsWhiteSpace( *p ) && *p != '\n' && *p != '\r' // whitespace |
---|
| 1363 | && *p != '/' && *p != '>' ) // tag end |
---|
| 1364 | { |
---|
| 1365 | value += *p; |
---|
| 1366 | ++p; |
---|
| 1367 | } |
---|
| 1368 | } |
---|
| 1369 | return p; |
---|
| 1370 | } |
---|
| 1371 | |
---|
| 1372 | #ifdef TIXML_USE_STL |
---|
| 1373 | void TiXmlText::StreamIn( TIXML_ISTREAM * in, TIXML_STRING * tag ) |
---|
| 1374 | { |
---|
| 1375 | while ( in->good() ) |
---|
| 1376 | { |
---|
| 1377 | int c = in->peek(); |
---|
| 1378 | if ( c == '<' ) |
---|
| 1379 | return; |
---|
| 1380 | if ( c <= 0 ) |
---|
| 1381 | { |
---|
| 1382 | TiXmlDocument* document = GetDocument(); |
---|
| 1383 | if ( document ) |
---|
| 1384 | document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
| 1385 | return; |
---|
| 1386 | } |
---|
| 1387 | |
---|
| 1388 | (*tag) += (char) c; |
---|
| 1389 | in->get(); |
---|
| 1390 | } |
---|
| 1391 | } |
---|
| 1392 | #endif |
---|
| 1393 | |
---|
| 1394 | const char* TiXmlText::Parse( const char* p, TiXmlParsingData* data, TiXmlEncoding encoding ) |
---|
| 1395 | { |
---|
| 1396 | value = ""; |
---|
| 1397 | // TiXmlParsingData data( p, prevData ); |
---|
| 1398 | if ( data ) |
---|
| 1399 | { |
---|
| 1400 | data->Stamp( p, encoding ); |
---|
| 1401 | location = data->Cursor(); |
---|
| 1402 | } |
---|
| 1403 | bool ignoreWhite = true; |
---|
| 1404 | |
---|
| 1405 | const char* end = "<"; |
---|
| 1406 | p = ReadText( p, &value, ignoreWhite, end, false, encoding ); |
---|
| 1407 | if ( p ) |
---|
| 1408 | return p-1; // don't truncate the '<' |
---|
| 1409 | return 0; |
---|
| 1410 | } |
---|
| 1411 | |
---|
| 1412 | #ifdef TIXML_USE_STL |
---|
| 1413 | void TiXmlDeclaration::StreamIn( TIXML_ISTREAM * in, TIXML_STRING * tag ) |
---|
| 1414 | { |
---|
| 1415 | while ( in->good() ) |
---|
| 1416 | { |
---|
| 1417 | int c = in->get(); |
---|
| 1418 | if ( c <= 0 ) |
---|
| 1419 | { |
---|
| 1420 | TiXmlDocument* document = GetDocument(); |
---|
| 1421 | if ( document ) |
---|
| 1422 | document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
| 1423 | return; |
---|
| 1424 | } |
---|
| 1425 | (*tag) += (char) c; |
---|
| 1426 | |
---|
| 1427 | if ( c == '>' ) |
---|
| 1428 | { |
---|
| 1429 | // All is well. |
---|
| 1430 | return; |
---|
| 1431 | } |
---|
| 1432 | } |
---|
| 1433 | } |
---|
| 1434 | #endif |
---|
| 1435 | |
---|
| 1436 | const char* TiXmlDeclaration::Parse( const char* p, TiXmlParsingData* data, TiXmlEncoding _encoding ) |
---|
| 1437 | { |
---|
| 1438 | p = SkipWhiteSpace( p, _encoding ); |
---|
| 1439 | // Find the beginning, find the end, and look for |
---|
| 1440 | // the stuff in-between. |
---|
| 1441 | TiXmlDocument* document = GetDocument(); |
---|
| 1442 | if ( !p || !*p || !StringEqual( p, "<?xml", true, _encoding ) ) |
---|
| 1443 | { |
---|
| 1444 | if ( document ) document->SetError( TIXML_ERROR_PARSING_DECLARATION, 0, 0, _encoding ); |
---|
| 1445 | return 0; |
---|
| 1446 | } |
---|
| 1447 | // TiXmlParsingData data( p, prevData ); |
---|
| 1448 | if ( data ) |
---|
| 1449 | { |
---|
| 1450 | data->Stamp( p, _encoding ); |
---|
| 1451 | location = data->Cursor(); |
---|
| 1452 | } |
---|
| 1453 | p += 5; |
---|
| 1454 | |
---|
| 1455 | version = ""; |
---|
| 1456 | encoding = ""; |
---|
| 1457 | standalone = ""; |
---|
| 1458 | |
---|
| 1459 | while ( p && *p ) |
---|
| 1460 | { |
---|
| 1461 | if ( *p == '>' ) |
---|
| 1462 | { |
---|
| 1463 | ++p; |
---|
| 1464 | return p; |
---|
| 1465 | } |
---|
| 1466 | |
---|
| 1467 | p = SkipWhiteSpace( p, _encoding ); |
---|
| 1468 | if ( StringEqual( p, "version", true, _encoding ) ) |
---|
| 1469 | { |
---|
| 1470 | TiXmlAttribute attrib; |
---|
| 1471 | p = attrib.Parse( p, data, _encoding ); |
---|
| 1472 | version = attrib.Value(); |
---|
| 1473 | } |
---|
| 1474 | else if ( StringEqual( p, "encoding", true, _encoding ) ) |
---|
| 1475 | { |
---|
| 1476 | TiXmlAttribute attrib; |
---|
| 1477 | p = attrib.Parse( p, data, _encoding ); |
---|
| 1478 | encoding = attrib.Value(); |
---|
| 1479 | } |
---|
| 1480 | else if ( StringEqual( p, "standalone", true, _encoding ) ) |
---|
| 1481 | { |
---|
| 1482 | TiXmlAttribute attrib; |
---|
| 1483 | p = attrib.Parse( p, data, _encoding ); |
---|
| 1484 | standalone = attrib.Value(); |
---|
| 1485 | } |
---|
| 1486 | else |
---|
| 1487 | { |
---|
| 1488 | // Read over whatever it is. |
---|
| 1489 | while( p && *p && *p != '>' && !IsWhiteSpace( *p ) ) |
---|
| 1490 | ++p; |
---|
| 1491 | } |
---|
| 1492 | } |
---|
| 1493 | return 0; |
---|
| 1494 | } |
---|
| 1495 | |
---|
| 1496 | bool TiXmlText::Blank() const |
---|
| 1497 | { |
---|
| 1498 | for ( unsigned i=0; i<value.length(); i++ ) |
---|
| 1499 | if ( !IsWhiteSpace( value[i] ) ) |
---|
| 1500 | return false; |
---|
| 1501 | return true; |
---|
| 1502 | } |
---|
| 1503 | |
---|