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 <ctype.h> |
---|
26 | #include <stddef.h> |
---|
27 | |
---|
28 | #include "tinyxml.h" |
---|
29 | |
---|
30 | //#define DEBUG_PARSER |
---|
31 | #if defined( DEBUG_PARSER ) |
---|
32 | # if defined( DEBUG ) && defined( _MSC_VER ) |
---|
33 | # include <windows.h> |
---|
34 | # define TIXML_LOG OutputDebugString |
---|
35 | # else |
---|
36 | # define TIXML_LOG printf |
---|
37 | # endif |
---|
38 | #endif |
---|
39 | |
---|
40 | // Note tha "PutString" hardcodes the same list. This |
---|
41 | // is less flexible than it appears. Changing the entries |
---|
42 | // or order will break putstring. |
---|
43 | TiXmlBase::Entity TiXmlBase::entity[ NUM_ENTITY ] = |
---|
44 | { |
---|
45 | { "&", 5, '&' }, |
---|
46 | { "<", 4, '<' }, |
---|
47 | { ">", 4, '>' }, |
---|
48 | { """, 6, '\"' }, |
---|
49 | { "'", 6, '\'' } |
---|
50 | }; |
---|
51 | |
---|
52 | // Bunch of unicode info at: |
---|
53 | // http://www.unicode.org/faq/utf_bom.html |
---|
54 | // Including the basic of this table, which determines the #bytes in the |
---|
55 | // sequence from the lead byte. 1 placed for invalid sequences -- |
---|
56 | // although the result will be junk, pass it through as much as possible. |
---|
57 | // Beware of the non-characters in UTF-8: |
---|
58 | // ef bb bf (Microsoft "lead bytes") |
---|
59 | // ef bf be |
---|
60 | // ef bf bf |
---|
61 | |
---|
62 | const unsigned char TIXML_UTF_LEAD_0 = 0xefU; |
---|
63 | const unsigned char TIXML_UTF_LEAD_1 = 0xbbU; |
---|
64 | const unsigned char TIXML_UTF_LEAD_2 = 0xbfU; |
---|
65 | |
---|
66 | const int TiXmlBase::utf8ByteTable[256] = |
---|
67 | { |
---|
68 | // 0 1 2 3 4 5 6 7 8 9 a b c d e f |
---|
69 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x00 |
---|
70 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x10 |
---|
71 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x20 |
---|
72 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x30 |
---|
73 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x40 |
---|
74 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x50 |
---|
75 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x60 |
---|
76 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x70 End of ASCII range |
---|
77 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x80 0x80 to 0xc1 invalid |
---|
78 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0x90 |
---|
79 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0xa0 |
---|
80 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 0xb0 |
---|
81 | 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, // 0xc0 0xc2 to 0xdf 2 byte |
---|
82 | 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, // 0xd0 |
---|
83 | 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, // 0xe0 0xe0 to 0xef 3 byte |
---|
84 | 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 |
---|
85 | }; |
---|
86 | |
---|
87 | |
---|
88 | void TiXmlBase::ConvertUTF32ToUTF8( unsigned long input, char* output, int* length ) |
---|
89 | { |
---|
90 | const unsigned long BYTE_MASK = 0xBF; |
---|
91 | const unsigned long BYTE_MARK = 0x80; |
---|
92 | const unsigned long FIRST_BYTE_MARK[7] = { 0x00, 0x00, 0xC0, 0xE0, 0xF0, 0xF8, 0xFC }; |
---|
93 | |
---|
94 | if (input < 0x80) |
---|
95 | *length = 1; |
---|
96 | else if ( input < 0x800 ) |
---|
97 | *length = 2; |
---|
98 | else if ( input < 0x10000 ) |
---|
99 | *length = 3; |
---|
100 | else if ( input < 0x200000 ) |
---|
101 | *length = 4; |
---|
102 | else |
---|
103 | { *length = 0; return; } // This code won't covert this correctly anyway. |
---|
104 | |
---|
105 | output += *length; |
---|
106 | |
---|
107 | // Scary scary fall throughs. |
---|
108 | switch (*length) |
---|
109 | { |
---|
110 | case 4: |
---|
111 | --output; |
---|
112 | *output = (char)((input | BYTE_MARK) & BYTE_MASK); |
---|
113 | input >>= 6; |
---|
114 | case 3: |
---|
115 | --output; |
---|
116 | *output = (char)((input | BYTE_MARK) & BYTE_MASK); |
---|
117 | input >>= 6; |
---|
118 | case 2: |
---|
119 | --output; |
---|
120 | *output = (char)((input | BYTE_MARK) & BYTE_MASK); |
---|
121 | input >>= 6; |
---|
122 | case 1: |
---|
123 | --output; |
---|
124 | *output = (char)(input | FIRST_BYTE_MARK[*length]); |
---|
125 | } |
---|
126 | } |
---|
127 | |
---|
128 | |
---|
129 | /*static*/ int TiXmlBase::IsAlpha( unsigned char anyByte, TiXmlEncoding /*encoding*/ ) |
---|
130 | { |
---|
131 | // This will only work for low-ascii, everything else is assumed to be a valid |
---|
132 | // letter. I'm not sure this is the best approach, but it is quite tricky trying |
---|
133 | // to figure out alhabetical vs. not across encoding. So take a very |
---|
134 | // conservative approach. |
---|
135 | |
---|
136 | // if ( encoding == TIXML_ENCODING_UTF8 ) |
---|
137 | // { |
---|
138 | if ( anyByte < 127 ) |
---|
139 | return isalpha( anyByte ); |
---|
140 | else |
---|
141 | return 1; // What else to do? The unicode set is huge...get the english ones right. |
---|
142 | // } |
---|
143 | // else |
---|
144 | // { |
---|
145 | // return isalpha( anyByte ); |
---|
146 | // } |
---|
147 | } |
---|
148 | |
---|
149 | |
---|
150 | /*static*/ int TiXmlBase::IsAlphaNum( unsigned char anyByte, TiXmlEncoding /*encoding*/ ) |
---|
151 | { |
---|
152 | // This will only work for low-ascii, everything else is assumed to be a valid |
---|
153 | // letter. I'm not sure this is the best approach, but it is quite tricky trying |
---|
154 | // to figure out alhabetical vs. not across encoding. So take a very |
---|
155 | // conservative approach. |
---|
156 | |
---|
157 | // if ( encoding == TIXML_ENCODING_UTF8 ) |
---|
158 | // { |
---|
159 | if ( anyByte < 127 ) |
---|
160 | return isalnum( anyByte ); |
---|
161 | else |
---|
162 | return 1; // What else to do? The unicode set is huge...get the english ones right. |
---|
163 | // } |
---|
164 | // else |
---|
165 | // { |
---|
166 | // return isalnum( anyByte ); |
---|
167 | // } |
---|
168 | } |
---|
169 | |
---|
170 | |
---|
171 | class TiXmlParsingData |
---|
172 | { |
---|
173 | friend class TiXmlDocument; |
---|
174 | public: |
---|
175 | void Stamp( const char* now, TiXmlEncoding encoding ); |
---|
176 | |
---|
177 | const TiXmlCursor& Cursor() { return cursor; } |
---|
178 | |
---|
179 | private: |
---|
180 | // Only used by the document! |
---|
181 | TiXmlParsingData( const char* start, int _tabsize, int row, int col ) |
---|
182 | { |
---|
183 | assert( start ); |
---|
184 | stamp = start; |
---|
185 | tabsize = _tabsize; |
---|
186 | cursor.row = row; |
---|
187 | cursor.col = col; |
---|
188 | } |
---|
189 | |
---|
190 | TiXmlCursor cursor; |
---|
191 | const char* stamp; |
---|
192 | int tabsize; |
---|
193 | }; |
---|
194 | |
---|
195 | |
---|
196 | void TiXmlParsingData::Stamp( const char* now, TiXmlEncoding encoding ) |
---|
197 | { |
---|
198 | assert( now ); |
---|
199 | |
---|
200 | // Do nothing if the tabsize is 0. |
---|
201 | if ( tabsize < 1 ) |
---|
202 | { |
---|
203 | return; |
---|
204 | } |
---|
205 | |
---|
206 | // Get the current row, column. |
---|
207 | int row = cursor.row; |
---|
208 | int col = cursor.col; |
---|
209 | const char* p = stamp; |
---|
210 | assert( p ); |
---|
211 | |
---|
212 | while ( p < now ) |
---|
213 | { |
---|
214 | // Treat p as unsigned, so we have a happy compiler. |
---|
215 | const unsigned char* pU = (const unsigned char*)p; |
---|
216 | |
---|
217 | // Code contributed by Fletcher Dunn: (modified by lee) |
---|
218 | switch (*pU) { |
---|
219 | case 0: |
---|
220 | // We *should* never get here, but in case we do, don't |
---|
221 | // advance past the terminating null character, ever |
---|
222 | return; |
---|
223 | |
---|
224 | case '\r': |
---|
225 | // bump down to the next line |
---|
226 | ++row; |
---|
227 | col = 0; |
---|
228 | // Eat the character |
---|
229 | ++p; |
---|
230 | |
---|
231 | // Check for \r\n sequence, and treat this as a single character |
---|
232 | if (*p == '\n') { |
---|
233 | ++p; |
---|
234 | } |
---|
235 | break; |
---|
236 | |
---|
237 | case '\n': |
---|
238 | // bump down to the next line |
---|
239 | ++row; |
---|
240 | col = 0; |
---|
241 | |
---|
242 | // Eat the character |
---|
243 | ++p; |
---|
244 | |
---|
245 | // Check for \n\r sequence, and treat this as a single |
---|
246 | // character. (Yes, this bizarre thing does occur still |
---|
247 | // on some arcane platforms...) |
---|
248 | if (*p == '\r') { |
---|
249 | ++p; |
---|
250 | } |
---|
251 | break; |
---|
252 | |
---|
253 | case '\t': |
---|
254 | // Eat the character |
---|
255 | ++p; |
---|
256 | |
---|
257 | // Skip to next tab stop |
---|
258 | col = (col / tabsize + 1) * tabsize; |
---|
259 | break; |
---|
260 | |
---|
261 | case TIXML_UTF_LEAD_0: |
---|
262 | if ( encoding == TIXML_ENCODING_UTF8 ) |
---|
263 | { |
---|
264 | if ( *(p+1) && *(p+2) ) |
---|
265 | { |
---|
266 | // In these cases, don't advance the column. These are |
---|
267 | // 0-width spaces. |
---|
268 | if ( *(pU+1)==TIXML_UTF_LEAD_1 && *(pU+2)==TIXML_UTF_LEAD_2 ) |
---|
269 | p += 3; |
---|
270 | else if ( *(pU+1)==0xbfU && *(pU+2)==0xbeU ) |
---|
271 | p += 3; |
---|
272 | else if ( *(pU+1)==0xbfU && *(pU+2)==0xbfU ) |
---|
273 | p += 3; |
---|
274 | else |
---|
275 | { p +=3; ++col; } // A normal character. |
---|
276 | } |
---|
277 | } |
---|
278 | else |
---|
279 | { |
---|
280 | ++p; |
---|
281 | ++col; |
---|
282 | } |
---|
283 | break; |
---|
284 | |
---|
285 | default: |
---|
286 | if ( encoding == TIXML_ENCODING_UTF8 ) |
---|
287 | { |
---|
288 | // Eat the 1 to 4 byte utf8 character. |
---|
289 | int step = TiXmlBase::utf8ByteTable[*((const unsigned char*)p)]; |
---|
290 | if ( step == 0 ) |
---|
291 | step = 1; // Error case from bad encoding, but handle gracefully. |
---|
292 | p += step; |
---|
293 | |
---|
294 | // Just advance one column, of course. |
---|
295 | ++col; |
---|
296 | } |
---|
297 | else |
---|
298 | { |
---|
299 | ++p; |
---|
300 | ++col; |
---|
301 | } |
---|
302 | break; |
---|
303 | } |
---|
304 | } |
---|
305 | cursor.row = row; |
---|
306 | cursor.col = col; |
---|
307 | assert( cursor.row >= -1 ); |
---|
308 | assert( cursor.col >= -1 ); |
---|
309 | stamp = p; |
---|
310 | assert( stamp ); |
---|
311 | } |
---|
312 | |
---|
313 | |
---|
314 | const char* TiXmlBase::SkipWhiteSpace( const char* p, TiXmlEncoding encoding ) |
---|
315 | { |
---|
316 | if ( !p || !*p ) |
---|
317 | { |
---|
318 | return 0; |
---|
319 | } |
---|
320 | if ( encoding == TIXML_ENCODING_UTF8 ) |
---|
321 | { |
---|
322 | while ( *p ) |
---|
323 | { |
---|
324 | const unsigned char* pU = (const unsigned char*)p; |
---|
325 | |
---|
326 | // Skip the stupid Microsoft UTF-8 Byte order marks |
---|
327 | if ( *(pU+0)==TIXML_UTF_LEAD_0 |
---|
328 | && *(pU+1)==TIXML_UTF_LEAD_1 |
---|
329 | && *(pU+2)==TIXML_UTF_LEAD_2 ) |
---|
330 | { |
---|
331 | p += 3; |
---|
332 | continue; |
---|
333 | } |
---|
334 | else if(*(pU+0)==TIXML_UTF_LEAD_0 |
---|
335 | && *(pU+1)==0xbfU |
---|
336 | && *(pU+2)==0xbeU ) |
---|
337 | { |
---|
338 | p += 3; |
---|
339 | continue; |
---|
340 | } |
---|
341 | else if(*(pU+0)==TIXML_UTF_LEAD_0 |
---|
342 | && *(pU+1)==0xbfU |
---|
343 | && *(pU+2)==0xbfU ) |
---|
344 | { |
---|
345 | p += 3; |
---|
346 | continue; |
---|
347 | } |
---|
348 | |
---|
349 | if ( IsWhiteSpace( *p ) || *p == '\n' || *p =='\r' ) // Still using old rules for white space. |
---|
350 | ++p; |
---|
351 | else |
---|
352 | break; |
---|
353 | } |
---|
354 | } |
---|
355 | else |
---|
356 | { |
---|
357 | while ( *p && IsWhiteSpace( *p ) || *p == '\n' || *p =='\r' ) |
---|
358 | ++p; |
---|
359 | } |
---|
360 | |
---|
361 | return p; |
---|
362 | } |
---|
363 | |
---|
364 | #ifdef TIXML_USE_STL |
---|
365 | /*static*/ bool TiXmlBase::StreamWhiteSpace( std::istream * in, TIXML_STRING * tag ) |
---|
366 | { |
---|
367 | for( ;; ) |
---|
368 | { |
---|
369 | if ( !in->good() ) return false; |
---|
370 | |
---|
371 | int c = in->peek(); |
---|
372 | // At this scope, we can't get to a document. So fail silently. |
---|
373 | if ( !IsWhiteSpace( c ) || c <= 0 ) |
---|
374 | return true; |
---|
375 | |
---|
376 | *tag += (char) in->get(); |
---|
377 | } |
---|
378 | } |
---|
379 | |
---|
380 | /*static*/ bool TiXmlBase::StreamTo( std::istream * in, int character, TIXML_STRING * tag ) |
---|
381 | { |
---|
382 | //assert( character > 0 && character < 128 ); // else it won't work in utf-8 |
---|
383 | while ( in->good() ) |
---|
384 | { |
---|
385 | int c = in->peek(); |
---|
386 | if ( c == character ) |
---|
387 | return true; |
---|
388 | if ( c <= 0 ) // Silent failure: can't get document at this scope |
---|
389 | return false; |
---|
390 | |
---|
391 | in->get(); |
---|
392 | *tag += (char) c; |
---|
393 | } |
---|
394 | return false; |
---|
395 | } |
---|
396 | #endif |
---|
397 | |
---|
398 | // One of TinyXML's more performance demanding functions. Try to keep the memory overhead down. The |
---|
399 | // "assign" optimization removes over 10% of the execution time. |
---|
400 | // |
---|
401 | const char* TiXmlBase::ReadName( const char* p, TIXML_STRING * name, TiXmlEncoding encoding ) |
---|
402 | { |
---|
403 | // Oddly, not supported on some comilers, |
---|
404 | //name->clear(); |
---|
405 | // So use this: |
---|
406 | *name = ""; |
---|
407 | assert( p ); |
---|
408 | |
---|
409 | // Names start with letters or underscores. |
---|
410 | // Of course, in unicode, tinyxml has no idea what a letter *is*. The |
---|
411 | // algorithm is generous. |
---|
412 | // |
---|
413 | // After that, they can be letters, underscores, numbers, |
---|
414 | // hyphens, or colons. (Colons are valid ony for namespaces, |
---|
415 | // but tinyxml can't tell namespaces from names.) |
---|
416 | if ( p && *p |
---|
417 | && ( IsAlpha( (unsigned char) *p, encoding ) || *p == '_' ) ) |
---|
418 | { |
---|
419 | const char* start = p; |
---|
420 | while( p && *p |
---|
421 | && ( IsAlphaNum( (unsigned char ) *p, encoding ) |
---|
422 | || *p == '_' |
---|
423 | || *p == '-' |
---|
424 | || *p == '.' |
---|
425 | || *p == ':' ) ) |
---|
426 | { |
---|
427 | //(*name) += *p; // expensive |
---|
428 | ++p; |
---|
429 | } |
---|
430 | if ( p-start > 0 ) { |
---|
431 | name->assign( start, p-start ); |
---|
432 | } |
---|
433 | return p; |
---|
434 | } |
---|
435 | return 0; |
---|
436 | } |
---|
437 | |
---|
438 | const char* TiXmlBase::GetEntity( const char* p, char* value, int* length, TiXmlEncoding encoding ) |
---|
439 | { |
---|
440 | // Presume an entity, and pull it out. |
---|
441 | TIXML_STRING ent; |
---|
442 | int i; |
---|
443 | *length = 0; |
---|
444 | |
---|
445 | if ( *(p+1) && *(p+1) == '#' && *(p+2) ) |
---|
446 | { |
---|
447 | unsigned long ucs = 0; |
---|
448 | ptrdiff_t delta = 0; |
---|
449 | unsigned mult = 1; |
---|
450 | |
---|
451 | if ( *(p+2) == 'x' ) |
---|
452 | { |
---|
453 | // Hexadecimal. |
---|
454 | if ( !*(p+3) ) return 0; |
---|
455 | |
---|
456 | const char* q = p+3; |
---|
457 | q = strchr( q, ';' ); |
---|
458 | |
---|
459 | if ( !q || !*q ) return 0; |
---|
460 | |
---|
461 | delta = q-p; |
---|
462 | --q; |
---|
463 | |
---|
464 | while ( *q != 'x' ) |
---|
465 | { |
---|
466 | if ( *q >= '0' && *q <= '9' ) |
---|
467 | ucs += mult * (*q - '0'); |
---|
468 | else if ( *q >= 'a' && *q <= 'f' ) |
---|
469 | ucs += mult * (*q - 'a' + 10); |
---|
470 | else if ( *q >= 'A' && *q <= 'F' ) |
---|
471 | ucs += mult * (*q - 'A' + 10 ); |
---|
472 | else |
---|
473 | return 0; |
---|
474 | mult *= 16; |
---|
475 | --q; |
---|
476 | } |
---|
477 | } |
---|
478 | else |
---|
479 | { |
---|
480 | // Decimal. |
---|
481 | if ( !*(p+2) ) return 0; |
---|
482 | |
---|
483 | const char* q = p+2; |
---|
484 | q = strchr( q, ';' ); |
---|
485 | |
---|
486 | if ( !q || !*q ) return 0; |
---|
487 | |
---|
488 | delta = q-p; |
---|
489 | --q; |
---|
490 | |
---|
491 | while ( *q != '#' ) |
---|
492 | { |
---|
493 | if ( *q >= '0' && *q <= '9' ) |
---|
494 | ucs += mult * (*q - '0'); |
---|
495 | else |
---|
496 | return 0; |
---|
497 | mult *= 10; |
---|
498 | --q; |
---|
499 | } |
---|
500 | } |
---|
501 | if ( encoding == TIXML_ENCODING_UTF8 ) |
---|
502 | { |
---|
503 | // convert the UCS to UTF-8 |
---|
504 | ConvertUTF32ToUTF8( ucs, value, length ); |
---|
505 | } |
---|
506 | else |
---|
507 | { |
---|
508 | *value = (char)ucs; |
---|
509 | *length = 1; |
---|
510 | } |
---|
511 | return p + delta + 1; |
---|
512 | } |
---|
513 | |
---|
514 | // Now try to match it. |
---|
515 | for( i=0; i<NUM_ENTITY; ++i ) |
---|
516 | { |
---|
517 | if ( strncmp( entity[i].str, p, entity[i].strLength ) == 0 ) |
---|
518 | { |
---|
519 | assert( strlen( entity[i].str ) == entity[i].strLength ); |
---|
520 | *value = entity[i].chr; |
---|
521 | *length = 1; |
---|
522 | return ( p + entity[i].strLength ); |
---|
523 | } |
---|
524 | } |
---|
525 | |
---|
526 | // So it wasn't an entity, its unrecognized, or something like that. |
---|
527 | *value = *p; // Don't put back the last one, since we return it! |
---|
528 | //*length = 1; // Leave unrecognized entities - this doesn't really work. |
---|
529 | // Just writes strange XML. |
---|
530 | return p+1; |
---|
531 | } |
---|
532 | |
---|
533 | |
---|
534 | bool TiXmlBase::StringEqual( const char* p, |
---|
535 | const char* tag, |
---|
536 | bool ignoreCase, |
---|
537 | TiXmlEncoding encoding ) |
---|
538 | { |
---|
539 | assert( p ); |
---|
540 | assert( tag ); |
---|
541 | if ( !p || !*p ) |
---|
542 | { |
---|
543 | assert( 0 ); |
---|
544 | return false; |
---|
545 | } |
---|
546 | |
---|
547 | const char* q = p; |
---|
548 | |
---|
549 | if ( ignoreCase ) |
---|
550 | { |
---|
551 | while ( *q && *tag && ToLower( *q, encoding ) == ToLower( *tag, encoding ) ) |
---|
552 | { |
---|
553 | ++q; |
---|
554 | ++tag; |
---|
555 | } |
---|
556 | |
---|
557 | if ( *tag == 0 ) |
---|
558 | return true; |
---|
559 | } |
---|
560 | else |
---|
561 | { |
---|
562 | while ( *q && *tag && *q == *tag ) |
---|
563 | { |
---|
564 | ++q; |
---|
565 | ++tag; |
---|
566 | } |
---|
567 | |
---|
568 | if ( *tag == 0 ) // Have we found the end of the tag, and everything equal? |
---|
569 | return true; |
---|
570 | } |
---|
571 | return false; |
---|
572 | } |
---|
573 | |
---|
574 | const char* TiXmlBase::ReadText( const char* p, |
---|
575 | TIXML_STRING * text, |
---|
576 | bool trimWhiteSpace, |
---|
577 | const char* endTag, |
---|
578 | bool caseInsensitive, |
---|
579 | TiXmlEncoding encoding ) |
---|
580 | { |
---|
581 | *text = ""; |
---|
582 | if ( !trimWhiteSpace // certain tags always keep whitespace |
---|
583 | || !condenseWhiteSpace ) // if true, whitespace is always kept |
---|
584 | { |
---|
585 | // Keep all the white space. |
---|
586 | while ( p && *p |
---|
587 | && !StringEqual( p, endTag, caseInsensitive, encoding ) |
---|
588 | ) |
---|
589 | { |
---|
590 | int len; |
---|
591 | char cArr[4] = { 0, 0, 0, 0 }; |
---|
592 | p = GetChar( p, cArr, &len, encoding ); |
---|
593 | text->append( cArr, len ); |
---|
594 | } |
---|
595 | } |
---|
596 | else |
---|
597 | { |
---|
598 | bool whitespace = false; |
---|
599 | |
---|
600 | // Remove leading white space: |
---|
601 | p = SkipWhiteSpace( p, encoding ); |
---|
602 | while ( p && *p |
---|
603 | && !StringEqual( p, endTag, caseInsensitive, encoding ) ) |
---|
604 | { |
---|
605 | if ( *p == '\r' || *p == '\n' ) |
---|
606 | { |
---|
607 | whitespace = true; |
---|
608 | ++p; |
---|
609 | } |
---|
610 | else if ( IsWhiteSpace( *p ) ) |
---|
611 | { |
---|
612 | whitespace = true; |
---|
613 | ++p; |
---|
614 | } |
---|
615 | else |
---|
616 | { |
---|
617 | // If we've found whitespace, add it before the |
---|
618 | // new character. Any whitespace just becomes a space. |
---|
619 | if ( whitespace ) |
---|
620 | { |
---|
621 | (*text) += ' '; |
---|
622 | whitespace = false; |
---|
623 | } |
---|
624 | int len; |
---|
625 | char cArr[4] = { 0, 0, 0, 0 }; |
---|
626 | p = GetChar( p, cArr, &len, encoding ); |
---|
627 | if ( len == 1 ) |
---|
628 | (*text) += cArr[0]; // more efficient |
---|
629 | else |
---|
630 | text->append( cArr, len ); |
---|
631 | } |
---|
632 | } |
---|
633 | } |
---|
634 | if ( p ) |
---|
635 | p += strlen( endTag ); |
---|
636 | return p; |
---|
637 | } |
---|
638 | |
---|
639 | #ifdef TIXML_USE_STL |
---|
640 | |
---|
641 | void TiXmlDocument::StreamIn( std::istream * in, TIXML_STRING * tag ) |
---|
642 | { |
---|
643 | // The basic issue with a document is that we don't know what we're |
---|
644 | // streaming. Read something presumed to be a tag (and hope), then |
---|
645 | // identify it, and call the appropriate stream method on the tag. |
---|
646 | // |
---|
647 | // This "pre-streaming" will never read the closing ">" so the |
---|
648 | // sub-tag can orient itself. |
---|
649 | |
---|
650 | if ( !StreamTo( in, '<', tag ) ) |
---|
651 | { |
---|
652 | SetError( TIXML_ERROR_PARSING_EMPTY, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
653 | return; |
---|
654 | } |
---|
655 | |
---|
656 | while ( in->good() ) |
---|
657 | { |
---|
658 | int tagIndex = (int) tag->length(); |
---|
659 | while ( in->good() && in->peek() != '>' ) |
---|
660 | { |
---|
661 | int c = in->get(); |
---|
662 | if ( c <= 0 ) |
---|
663 | { |
---|
664 | SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
665 | break; |
---|
666 | } |
---|
667 | (*tag) += (char) c; |
---|
668 | } |
---|
669 | |
---|
670 | if ( in->good() ) |
---|
671 | { |
---|
672 | // We now have something we presume to be a node of |
---|
673 | // some sort. Identify it, and call the node to |
---|
674 | // continue streaming. |
---|
675 | TiXmlNode* node = Identify( tag->c_str() + tagIndex, TIXML_DEFAULT_ENCODING ); |
---|
676 | |
---|
677 | if ( node ) |
---|
678 | { |
---|
679 | node->StreamIn( in, tag ); |
---|
680 | bool isElement = node->ToElement() != 0; |
---|
681 | delete node; |
---|
682 | node = 0; |
---|
683 | |
---|
684 | // If this is the root element, we're done. Parsing will be |
---|
685 | // done by the >> operator. |
---|
686 | if ( isElement ) |
---|
687 | { |
---|
688 | return; |
---|
689 | } |
---|
690 | } |
---|
691 | else |
---|
692 | { |
---|
693 | SetError( TIXML_ERROR, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
694 | return; |
---|
695 | } |
---|
696 | } |
---|
697 | } |
---|
698 | // We should have returned sooner. |
---|
699 | SetError( TIXML_ERROR, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
700 | } |
---|
701 | |
---|
702 | #endif |
---|
703 | |
---|
704 | const char* TiXmlDocument::Parse( const char* p, TiXmlParsingData* prevData, TiXmlEncoding encoding ) |
---|
705 | { |
---|
706 | ClearError(); |
---|
707 | |
---|
708 | // Parse away, at the document level. Since a document |
---|
709 | // contains nothing but other tags, most of what happens |
---|
710 | // here is skipping white space. |
---|
711 | if ( !p || !*p ) |
---|
712 | { |
---|
713 | SetError( TIXML_ERROR_DOCUMENT_EMPTY, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
714 | return 0; |
---|
715 | } |
---|
716 | |
---|
717 | // Note that, for a document, this needs to come |
---|
718 | // before the while space skip, so that parsing |
---|
719 | // starts from the pointer we are given. |
---|
720 | location.Clear(); |
---|
721 | if ( prevData ) |
---|
722 | { |
---|
723 | location.row = prevData->cursor.row; |
---|
724 | location.col = prevData->cursor.col; |
---|
725 | } |
---|
726 | else |
---|
727 | { |
---|
728 | location.row = 0; |
---|
729 | location.col = 0; |
---|
730 | } |
---|
731 | TiXmlParsingData data( p, TabSize(), location.row, location.col ); |
---|
732 | location = data.Cursor(); |
---|
733 | |
---|
734 | if ( encoding == TIXML_ENCODING_UNKNOWN ) |
---|
735 | { |
---|
736 | // Check for the Microsoft UTF-8 lead bytes. |
---|
737 | const unsigned char* pU = (const unsigned char*)p; |
---|
738 | if ( *(pU+0) && *(pU+0) == TIXML_UTF_LEAD_0 |
---|
739 | && *(pU+1) && *(pU+1) == TIXML_UTF_LEAD_1 |
---|
740 | && *(pU+2) && *(pU+2) == TIXML_UTF_LEAD_2 ) |
---|
741 | { |
---|
742 | encoding = TIXML_ENCODING_UTF8; |
---|
743 | useMicrosoftBOM = true; |
---|
744 | } |
---|
745 | } |
---|
746 | |
---|
747 | p = SkipWhiteSpace( p, encoding ); |
---|
748 | if ( !p ) |
---|
749 | { |
---|
750 | SetError( TIXML_ERROR_DOCUMENT_EMPTY, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
751 | return 0; |
---|
752 | } |
---|
753 | |
---|
754 | while ( p && *p ) |
---|
755 | { |
---|
756 | TiXmlNode* node = Identify( p, encoding ); |
---|
757 | if ( node ) |
---|
758 | { |
---|
759 | p = node->Parse( p, &data, encoding ); |
---|
760 | LinkEndChild( node ); |
---|
761 | } |
---|
762 | else |
---|
763 | { |
---|
764 | break; |
---|
765 | } |
---|
766 | |
---|
767 | // Did we get encoding info? |
---|
768 | if ( encoding == TIXML_ENCODING_UNKNOWN |
---|
769 | && node->ToDeclaration() ) |
---|
770 | { |
---|
771 | TiXmlDeclaration* dec = node->ToDeclaration(); |
---|
772 | const char* enc = dec->Encoding(); |
---|
773 | assert( enc ); |
---|
774 | |
---|
775 | if ( *enc == 0 ) |
---|
776 | encoding = TIXML_ENCODING_UTF8; |
---|
777 | else if ( StringEqual( enc, "UTF-8", true, TIXML_ENCODING_UNKNOWN ) ) |
---|
778 | encoding = TIXML_ENCODING_UTF8; |
---|
779 | else if ( StringEqual( enc, "UTF8", true, TIXML_ENCODING_UNKNOWN ) ) |
---|
780 | encoding = TIXML_ENCODING_UTF8; // incorrect, but be nice |
---|
781 | else |
---|
782 | encoding = TIXML_ENCODING_LEGACY; |
---|
783 | } |
---|
784 | |
---|
785 | p = SkipWhiteSpace( p, encoding ); |
---|
786 | } |
---|
787 | |
---|
788 | // Was this empty? |
---|
789 | if ( !firstChild ) { |
---|
790 | SetError( TIXML_ERROR_DOCUMENT_EMPTY, 0, 0, encoding ); |
---|
791 | return 0; |
---|
792 | } |
---|
793 | |
---|
794 | // All is well. |
---|
795 | return p; |
---|
796 | } |
---|
797 | |
---|
798 | void TiXmlDocument::SetError( int err, const char* pError, TiXmlParsingData* data, TiXmlEncoding encoding ) |
---|
799 | { |
---|
800 | // The first error in a chain is more accurate - don't set again! |
---|
801 | if ( error ) |
---|
802 | return; |
---|
803 | |
---|
804 | assert( err > 0 && err < TIXML_ERROR_STRING_COUNT ); |
---|
805 | error = true; |
---|
806 | errorId = err; |
---|
807 | errorDesc = errorString[ errorId ]; |
---|
808 | |
---|
809 | errorLocation.Clear(); |
---|
810 | if ( pError && data ) |
---|
811 | { |
---|
812 | data->Stamp( pError, encoding ); |
---|
813 | errorLocation = data->Cursor(); |
---|
814 | } |
---|
815 | } |
---|
816 | |
---|
817 | |
---|
818 | TiXmlNode* TiXmlNode::Identify( const char* p, TiXmlEncoding encoding ) |
---|
819 | { |
---|
820 | TiXmlNode* returnNode = 0; |
---|
821 | |
---|
822 | p = SkipWhiteSpace( p, encoding ); |
---|
823 | if( !p || !*p || *p != '<' ) |
---|
824 | { |
---|
825 | return 0; |
---|
826 | } |
---|
827 | |
---|
828 | TiXmlDocument* doc = GetDocument(); |
---|
829 | p = SkipWhiteSpace( p, encoding ); |
---|
830 | |
---|
831 | if ( !p || !*p ) |
---|
832 | { |
---|
833 | return 0; |
---|
834 | } |
---|
835 | |
---|
836 | // What is this thing? |
---|
837 | // - Elements start with a letter or underscore, but xml is reserved. |
---|
838 | // - Comments: <!-- |
---|
839 | // - Decleration: <?xml |
---|
840 | // - Everthing else is unknown to tinyxml. |
---|
841 | // |
---|
842 | |
---|
843 | const char* xmlHeader = { "<?xml" }; |
---|
844 | const char* commentHeader = { "<!--" }; |
---|
845 | const char* dtdHeader = { "<!" }; |
---|
846 | const char* cdataHeader = { "<![CDATA[" }; |
---|
847 | |
---|
848 | if ( StringEqual( p, xmlHeader, true, encoding ) ) |
---|
849 | { |
---|
850 | #ifdef DEBUG_PARSER |
---|
851 | TIXML_LOG( "XML parsing Declaration\n" ); |
---|
852 | #endif |
---|
853 | returnNode = new TiXmlDeclaration(); |
---|
854 | } |
---|
855 | else if ( StringEqual( p, commentHeader, false, encoding ) ) |
---|
856 | { |
---|
857 | #ifdef DEBUG_PARSER |
---|
858 | TIXML_LOG( "XML parsing Comment\n" ); |
---|
859 | #endif |
---|
860 | returnNode = new TiXmlComment(); |
---|
861 | } |
---|
862 | else if ( StringEqual( p, cdataHeader, false, encoding ) ) |
---|
863 | { |
---|
864 | #ifdef DEBUG_PARSER |
---|
865 | TIXML_LOG( "XML parsing CDATA\n" ); |
---|
866 | #endif |
---|
867 | TiXmlText* text = new TiXmlText( "" ); |
---|
868 | text->SetCDATA( true ); |
---|
869 | returnNode = text; |
---|
870 | } |
---|
871 | else if ( StringEqual( p, dtdHeader, false, encoding ) ) |
---|
872 | { |
---|
873 | #ifdef DEBUG_PARSER |
---|
874 | TIXML_LOG( "XML parsing Unknown(1)\n" ); |
---|
875 | #endif |
---|
876 | returnNode = new TiXmlUnknown(); |
---|
877 | } |
---|
878 | else if ( IsAlpha( *(p+1), encoding ) |
---|
879 | || *(p+1) == '_' ) |
---|
880 | { |
---|
881 | #ifdef DEBUG_PARSER |
---|
882 | TIXML_LOG( "XML parsing Element\n" ); |
---|
883 | #endif |
---|
884 | returnNode = new TiXmlElement( "" ); |
---|
885 | } |
---|
886 | else |
---|
887 | { |
---|
888 | #ifdef DEBUG_PARSER |
---|
889 | TIXML_LOG( "XML parsing Unknown(2)\n" ); |
---|
890 | #endif |
---|
891 | returnNode = new TiXmlUnknown(); |
---|
892 | } |
---|
893 | |
---|
894 | if ( returnNode ) |
---|
895 | { |
---|
896 | // Set the parent, so it can report errors |
---|
897 | returnNode->parent = this; |
---|
898 | } |
---|
899 | else |
---|
900 | { |
---|
901 | if ( doc ) |
---|
902 | doc->SetError( TIXML_ERROR_OUT_OF_MEMORY, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
903 | } |
---|
904 | return returnNode; |
---|
905 | } |
---|
906 | |
---|
907 | #ifdef TIXML_USE_STL |
---|
908 | |
---|
909 | void TiXmlElement::StreamIn (std::istream * in, TIXML_STRING * tag) |
---|
910 | { |
---|
911 | // We're called with some amount of pre-parsing. That is, some of "this" |
---|
912 | // element is in "tag". Go ahead and stream to the closing ">" |
---|
913 | while( in->good() ) |
---|
914 | { |
---|
915 | int c = in->get(); |
---|
916 | if ( c <= 0 ) |
---|
917 | { |
---|
918 | TiXmlDocument* document = GetDocument(); |
---|
919 | if ( document ) |
---|
920 | document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
921 | return; |
---|
922 | } |
---|
923 | (*tag) += (char) c ; |
---|
924 | |
---|
925 | if ( c == '>' ) |
---|
926 | break; |
---|
927 | } |
---|
928 | |
---|
929 | if ( tag->length() < 3 ) return; |
---|
930 | |
---|
931 | // Okay...if we are a "/>" tag, then we're done. We've read a complete tag. |
---|
932 | // If not, identify and stream. |
---|
933 | |
---|
934 | if ( tag->at( tag->length() - 1 ) == '>' |
---|
935 | && tag->at( tag->length() - 2 ) == '/' ) |
---|
936 | { |
---|
937 | // All good! |
---|
938 | return; |
---|
939 | } |
---|
940 | else if ( tag->at( tag->length() - 1 ) == '>' ) |
---|
941 | { |
---|
942 | // There is more. Could be: |
---|
943 | // text |
---|
944 | // cdata text (which looks like another node) |
---|
945 | // closing tag |
---|
946 | // another node. |
---|
947 | for ( ;; ) |
---|
948 | { |
---|
949 | StreamWhiteSpace( in, tag ); |
---|
950 | |
---|
951 | // Do we have text? |
---|
952 | if ( in->good() && in->peek() != '<' ) |
---|
953 | { |
---|
954 | // Yep, text. |
---|
955 | TiXmlText text( "" ); |
---|
956 | text.StreamIn( in, tag ); |
---|
957 | |
---|
958 | // What follows text is a closing tag or another node. |
---|
959 | // Go around again and figure it out. |
---|
960 | continue; |
---|
961 | } |
---|
962 | |
---|
963 | // We now have either a closing tag...or another node. |
---|
964 | // We should be at a "<", regardless. |
---|
965 | if ( !in->good() ) return; |
---|
966 | assert( in->peek() == '<' ); |
---|
967 | int tagIndex = (int) tag->length(); |
---|
968 | |
---|
969 | bool closingTag = false; |
---|
970 | bool firstCharFound = false; |
---|
971 | |
---|
972 | for( ;; ) |
---|
973 | { |
---|
974 | if ( !in->good() ) |
---|
975 | return; |
---|
976 | |
---|
977 | int c = in->peek(); |
---|
978 | if ( c <= 0 ) |
---|
979 | { |
---|
980 | TiXmlDocument* document = GetDocument(); |
---|
981 | if ( document ) |
---|
982 | document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
983 | return; |
---|
984 | } |
---|
985 | |
---|
986 | if ( c == '>' ) |
---|
987 | break; |
---|
988 | |
---|
989 | *tag += (char) c; |
---|
990 | in->get(); |
---|
991 | |
---|
992 | // Early out if we find the CDATA id. |
---|
993 | if ( c == '[' && tag->size() >= 9 ) |
---|
994 | { |
---|
995 | size_t len = tag->size(); |
---|
996 | const char* start = tag->c_str() + len - 9; |
---|
997 | if ( strcmp( start, "<![CDATA[" ) == 0 ) { |
---|
998 | assert( !closingTag ); |
---|
999 | break; |
---|
1000 | } |
---|
1001 | } |
---|
1002 | |
---|
1003 | if ( !firstCharFound && c != '<' && !IsWhiteSpace( c ) ) |
---|
1004 | { |
---|
1005 | firstCharFound = true; |
---|
1006 | if ( c == '/' ) |
---|
1007 | closingTag = true; |
---|
1008 | } |
---|
1009 | } |
---|
1010 | // If it was a closing tag, then read in the closing '>' to clean up the input stream. |
---|
1011 | // If it was not, the streaming will be done by the tag. |
---|
1012 | if ( closingTag ) |
---|
1013 | { |
---|
1014 | if ( !in->good() ) |
---|
1015 | return; |
---|
1016 | |
---|
1017 | int c = in->get(); |
---|
1018 | if ( c <= 0 ) |
---|
1019 | { |
---|
1020 | TiXmlDocument* document = GetDocument(); |
---|
1021 | if ( document ) |
---|
1022 | document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
1023 | return; |
---|
1024 | } |
---|
1025 | assert( c == '>' ); |
---|
1026 | *tag += (char) c; |
---|
1027 | |
---|
1028 | // We are done, once we've found our closing tag. |
---|
1029 | return; |
---|
1030 | } |
---|
1031 | else |
---|
1032 | { |
---|
1033 | // If not a closing tag, id it, and stream. |
---|
1034 | const char* tagloc = tag->c_str() + tagIndex; |
---|
1035 | TiXmlNode* node = Identify( tagloc, TIXML_DEFAULT_ENCODING ); |
---|
1036 | if ( !node ) |
---|
1037 | return; |
---|
1038 | node->StreamIn( in, tag ); |
---|
1039 | delete node; |
---|
1040 | node = 0; |
---|
1041 | |
---|
1042 | // No return: go around from the beginning: text, closing tag, or node. |
---|
1043 | } |
---|
1044 | } |
---|
1045 | } |
---|
1046 | } |
---|
1047 | #endif |
---|
1048 | |
---|
1049 | const char* TiXmlElement::Parse( const char* p, TiXmlParsingData* data, TiXmlEncoding encoding ) |
---|
1050 | { |
---|
1051 | p = SkipWhiteSpace( p, encoding ); |
---|
1052 | TiXmlDocument* document = GetDocument(); |
---|
1053 | |
---|
1054 | if ( !p || !*p ) |
---|
1055 | { |
---|
1056 | if ( document ) document->SetError( TIXML_ERROR_PARSING_ELEMENT, 0, 0, encoding ); |
---|
1057 | return 0; |
---|
1058 | } |
---|
1059 | |
---|
1060 | if ( data ) |
---|
1061 | { |
---|
1062 | data->Stamp( p, encoding ); |
---|
1063 | location = data->Cursor(); |
---|
1064 | } |
---|
1065 | |
---|
1066 | if ( *p != '<' ) |
---|
1067 | { |
---|
1068 | if ( document ) document->SetError( TIXML_ERROR_PARSING_ELEMENT, p, data, encoding ); |
---|
1069 | return 0; |
---|
1070 | } |
---|
1071 | |
---|
1072 | p = SkipWhiteSpace( p+1, encoding ); |
---|
1073 | |
---|
1074 | // Read the name. |
---|
1075 | const char* pErr = p; |
---|
1076 | |
---|
1077 | p = ReadName( p, &value, encoding ); |
---|
1078 | if ( !p || !*p ) |
---|
1079 | { |
---|
1080 | if ( document ) document->SetError( TIXML_ERROR_FAILED_TO_READ_ELEMENT_NAME, pErr, data, encoding ); |
---|
1081 | return 0; |
---|
1082 | } |
---|
1083 | |
---|
1084 | TIXML_STRING endTag ("</"); |
---|
1085 | endTag += value; |
---|
1086 | endTag += ">"; |
---|
1087 | |
---|
1088 | // Check for and read attributes. Also look for an empty |
---|
1089 | // tag or an end tag. |
---|
1090 | while ( p && *p ) |
---|
1091 | { |
---|
1092 | pErr = p; |
---|
1093 | p = SkipWhiteSpace( p, encoding ); |
---|
1094 | if ( !p || !*p ) |
---|
1095 | { |
---|
1096 | if ( document ) document->SetError( TIXML_ERROR_READING_ATTRIBUTES, pErr, data, encoding ); |
---|
1097 | return 0; |
---|
1098 | } |
---|
1099 | if ( *p == '/' ) |
---|
1100 | { |
---|
1101 | ++p; |
---|
1102 | // Empty tag. |
---|
1103 | if ( *p != '>' ) |
---|
1104 | { |
---|
1105 | if ( document ) document->SetError( TIXML_ERROR_PARSING_EMPTY, p, data, encoding ); |
---|
1106 | return 0; |
---|
1107 | } |
---|
1108 | return (p+1); |
---|
1109 | } |
---|
1110 | else if ( *p == '>' ) |
---|
1111 | { |
---|
1112 | // Done with attributes (if there were any.) |
---|
1113 | // Read the value -- which can include other |
---|
1114 | // elements -- read the end tag, and return. |
---|
1115 | ++p; |
---|
1116 | p = ReadValue( p, data, encoding ); // Note this is an Element method, and will set the error if one happens. |
---|
1117 | if ( !p || !*p ) |
---|
1118 | return 0; |
---|
1119 | |
---|
1120 | // We should find the end tag now |
---|
1121 | if ( StringEqual( p, endTag.c_str(), false, encoding ) ) |
---|
1122 | { |
---|
1123 | p += endTag.length(); |
---|
1124 | return p; |
---|
1125 | } |
---|
1126 | else |
---|
1127 | { |
---|
1128 | if ( document ) document->SetError( TIXML_ERROR_READING_END_TAG, p, data, encoding ); |
---|
1129 | return 0; |
---|
1130 | } |
---|
1131 | } |
---|
1132 | else |
---|
1133 | { |
---|
1134 | // Try to read an attribute: |
---|
1135 | TiXmlAttribute* attrib = new TiXmlAttribute(); |
---|
1136 | if ( !attrib ) |
---|
1137 | { |
---|
1138 | if ( document ) document->SetError( TIXML_ERROR_OUT_OF_MEMORY, pErr, data, encoding ); |
---|
1139 | return 0; |
---|
1140 | } |
---|
1141 | |
---|
1142 | attrib->SetDocument( document ); |
---|
1143 | pErr = p; |
---|
1144 | p = attrib->Parse( p, data, encoding ); |
---|
1145 | |
---|
1146 | if ( !p || !*p ) |
---|
1147 | { |
---|
1148 | if ( document ) document->SetError( TIXML_ERROR_PARSING_ELEMENT, pErr, data, encoding ); |
---|
1149 | delete attrib; |
---|
1150 | return 0; |
---|
1151 | } |
---|
1152 | |
---|
1153 | // Handle the strange case of double attributes: |
---|
1154 | #ifdef TIXML_USE_STL |
---|
1155 | TiXmlAttribute* node = attributeSet.Find( attrib->NameTStr() ); |
---|
1156 | #else |
---|
1157 | TiXmlAttribute* node = attributeSet.Find( attrib->Name() ); |
---|
1158 | #endif |
---|
1159 | if ( node ) |
---|
1160 | { |
---|
1161 | node->SetValue( attrib->Value() ); |
---|
1162 | delete attrib; |
---|
1163 | return 0; |
---|
1164 | } |
---|
1165 | |
---|
1166 | attributeSet.Add( attrib ); |
---|
1167 | } |
---|
1168 | } |
---|
1169 | return p; |
---|
1170 | } |
---|
1171 | |
---|
1172 | |
---|
1173 | const char* TiXmlElement::ReadValue( const char* p, TiXmlParsingData* data, TiXmlEncoding encoding ) |
---|
1174 | { |
---|
1175 | TiXmlDocument* document = GetDocument(); |
---|
1176 | |
---|
1177 | // Read in text and elements in any order. |
---|
1178 | const char* pWithWhiteSpace = p; |
---|
1179 | p = SkipWhiteSpace( p, encoding ); |
---|
1180 | |
---|
1181 | while ( p && *p ) |
---|
1182 | { |
---|
1183 | if ( *p != '<' ) |
---|
1184 | { |
---|
1185 | // Take what we have, make a text element. |
---|
1186 | TiXmlText* textNode = new TiXmlText( "" ); |
---|
1187 | |
---|
1188 | if ( !textNode ) |
---|
1189 | { |
---|
1190 | if ( document ) document->SetError( TIXML_ERROR_OUT_OF_MEMORY, 0, 0, encoding ); |
---|
1191 | return 0; |
---|
1192 | } |
---|
1193 | |
---|
1194 | if ( TiXmlBase::IsWhiteSpaceCondensed() ) |
---|
1195 | { |
---|
1196 | p = textNode->Parse( p, data, encoding ); |
---|
1197 | } |
---|
1198 | else |
---|
1199 | { |
---|
1200 | // Special case: we want to keep the white space |
---|
1201 | // so that leading spaces aren't removed. |
---|
1202 | p = textNode->Parse( pWithWhiteSpace, data, encoding ); |
---|
1203 | } |
---|
1204 | |
---|
1205 | if ( !textNode->Blank() ) |
---|
1206 | LinkEndChild( textNode ); |
---|
1207 | else |
---|
1208 | delete textNode; |
---|
1209 | } |
---|
1210 | else |
---|
1211 | { |
---|
1212 | // We hit a '<' |
---|
1213 | // Have we hit a new element or an end tag? This could also be |
---|
1214 | // a TiXmlText in the "CDATA" style. |
---|
1215 | if ( StringEqual( p, "</", false, encoding ) ) |
---|
1216 | { |
---|
1217 | return p; |
---|
1218 | } |
---|
1219 | else |
---|
1220 | { |
---|
1221 | TiXmlNode* node = Identify( p, encoding ); |
---|
1222 | if ( node ) |
---|
1223 | { |
---|
1224 | p = node->Parse( p, data, encoding ); |
---|
1225 | LinkEndChild( node ); |
---|
1226 | } |
---|
1227 | else |
---|
1228 | { |
---|
1229 | return 0; |
---|
1230 | } |
---|
1231 | } |
---|
1232 | } |
---|
1233 | pWithWhiteSpace = p; |
---|
1234 | p = SkipWhiteSpace( p, encoding ); |
---|
1235 | } |
---|
1236 | |
---|
1237 | if ( !p ) |
---|
1238 | { |
---|
1239 | if ( document ) document->SetError( TIXML_ERROR_READING_ELEMENT_VALUE, 0, 0, encoding ); |
---|
1240 | } |
---|
1241 | return p; |
---|
1242 | } |
---|
1243 | |
---|
1244 | |
---|
1245 | #ifdef TIXML_USE_STL |
---|
1246 | void TiXmlUnknown::StreamIn( std::istream * in, TIXML_STRING * tag ) |
---|
1247 | { |
---|
1248 | while ( in->good() ) |
---|
1249 | { |
---|
1250 | int c = in->get(); |
---|
1251 | if ( c <= 0 ) |
---|
1252 | { |
---|
1253 | TiXmlDocument* document = GetDocument(); |
---|
1254 | if ( document ) |
---|
1255 | document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
1256 | return; |
---|
1257 | } |
---|
1258 | (*tag) += (char) c; |
---|
1259 | |
---|
1260 | if ( c == '>' ) |
---|
1261 | { |
---|
1262 | // All is well. |
---|
1263 | return; |
---|
1264 | } |
---|
1265 | } |
---|
1266 | } |
---|
1267 | #endif |
---|
1268 | |
---|
1269 | |
---|
1270 | const char* TiXmlUnknown::Parse( const char* p, TiXmlParsingData* data, TiXmlEncoding encoding ) |
---|
1271 | { |
---|
1272 | TiXmlDocument* document = GetDocument(); |
---|
1273 | p = SkipWhiteSpace( p, encoding ); |
---|
1274 | |
---|
1275 | if ( data ) |
---|
1276 | { |
---|
1277 | data->Stamp( p, encoding ); |
---|
1278 | location = data->Cursor(); |
---|
1279 | } |
---|
1280 | if ( !p || !*p || *p != '<' ) |
---|
1281 | { |
---|
1282 | if ( document ) document->SetError( TIXML_ERROR_PARSING_UNKNOWN, p, data, encoding ); |
---|
1283 | return 0; |
---|
1284 | } |
---|
1285 | ++p; |
---|
1286 | value = ""; |
---|
1287 | |
---|
1288 | while ( p && *p && *p != '>' ) |
---|
1289 | { |
---|
1290 | value += *p; |
---|
1291 | ++p; |
---|
1292 | } |
---|
1293 | |
---|
1294 | if ( !p ) |
---|
1295 | { |
---|
1296 | if ( document ) document->SetError( TIXML_ERROR_PARSING_UNKNOWN, 0, 0, encoding ); |
---|
1297 | } |
---|
1298 | if ( *p == '>' ) |
---|
1299 | return p+1; |
---|
1300 | return p; |
---|
1301 | } |
---|
1302 | |
---|
1303 | #ifdef TIXML_USE_STL |
---|
1304 | void TiXmlComment::StreamIn( std::istream * in, TIXML_STRING * tag ) |
---|
1305 | { |
---|
1306 | while ( in->good() ) |
---|
1307 | { |
---|
1308 | int c = in->get(); |
---|
1309 | if ( c <= 0 ) |
---|
1310 | { |
---|
1311 | TiXmlDocument* document = GetDocument(); |
---|
1312 | if ( document ) |
---|
1313 | document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
1314 | return; |
---|
1315 | } |
---|
1316 | |
---|
1317 | (*tag) += (char) c; |
---|
1318 | |
---|
1319 | if ( c == '>' |
---|
1320 | && tag->at( tag->length() - 2 ) == '-' |
---|
1321 | && tag->at( tag->length() - 3 ) == '-' ) |
---|
1322 | { |
---|
1323 | // All is well. |
---|
1324 | return; |
---|
1325 | } |
---|
1326 | } |
---|
1327 | } |
---|
1328 | #endif |
---|
1329 | |
---|
1330 | |
---|
1331 | const char* TiXmlComment::Parse( const char* p, TiXmlParsingData* data, TiXmlEncoding encoding ) |
---|
1332 | { |
---|
1333 | TiXmlDocument* document = GetDocument(); |
---|
1334 | value = ""; |
---|
1335 | |
---|
1336 | p = SkipWhiteSpace( p, encoding ); |
---|
1337 | |
---|
1338 | if ( data ) |
---|
1339 | { |
---|
1340 | data->Stamp( p, encoding ); |
---|
1341 | location = data->Cursor(); |
---|
1342 | } |
---|
1343 | const char* startTag = "<!--"; |
---|
1344 | const char* endTag = "-->"; |
---|
1345 | |
---|
1346 | if ( !StringEqual( p, startTag, false, encoding ) ) |
---|
1347 | { |
---|
1348 | document->SetError( TIXML_ERROR_PARSING_COMMENT, p, data, encoding ); |
---|
1349 | return 0; |
---|
1350 | } |
---|
1351 | p += strlen( startTag ); |
---|
1352 | p = ReadText( p, &value, false, endTag, false, encoding ); |
---|
1353 | return p; |
---|
1354 | } |
---|
1355 | |
---|
1356 | |
---|
1357 | const char* TiXmlAttribute::Parse( const char* p, TiXmlParsingData* data, TiXmlEncoding encoding ) |
---|
1358 | { |
---|
1359 | p = SkipWhiteSpace( p, encoding ); |
---|
1360 | if ( !p || !*p ) return 0; |
---|
1361 | |
---|
1362 | // int tabsize = 4; |
---|
1363 | // if ( document ) |
---|
1364 | // tabsize = document->TabSize(); |
---|
1365 | |
---|
1366 | if ( data ) |
---|
1367 | { |
---|
1368 | data->Stamp( p, encoding ); |
---|
1369 | location = data->Cursor(); |
---|
1370 | } |
---|
1371 | // Read the name, the '=' and the value. |
---|
1372 | const char* pErr = p; |
---|
1373 | p = ReadName( p, &name, encoding ); |
---|
1374 | if ( !p || !*p ) |
---|
1375 | { |
---|
1376 | if ( document ) document->SetError( TIXML_ERROR_READING_ATTRIBUTES, pErr, data, encoding ); |
---|
1377 | return 0; |
---|
1378 | } |
---|
1379 | p = SkipWhiteSpace( p, encoding ); |
---|
1380 | if ( !p || !*p || *p != '=' ) |
---|
1381 | { |
---|
1382 | if ( document ) document->SetError( TIXML_ERROR_READING_ATTRIBUTES, p, data, encoding ); |
---|
1383 | return 0; |
---|
1384 | } |
---|
1385 | |
---|
1386 | ++p; // skip '=' |
---|
1387 | p = SkipWhiteSpace( p, encoding ); |
---|
1388 | if ( !p || !*p ) |
---|
1389 | { |
---|
1390 | if ( document ) document->SetError( TIXML_ERROR_READING_ATTRIBUTES, p, data, encoding ); |
---|
1391 | return 0; |
---|
1392 | } |
---|
1393 | |
---|
1394 | const char* end; |
---|
1395 | const char SINGLE_QUOTE = '\''; |
---|
1396 | const char DOUBLE_QUOTE = '\"'; |
---|
1397 | |
---|
1398 | if ( *p == SINGLE_QUOTE ) |
---|
1399 | { |
---|
1400 | ++p; |
---|
1401 | end = "\'"; // single quote in string |
---|
1402 | p = ReadText( p, &value, false, end, false, encoding ); |
---|
1403 | } |
---|
1404 | else if ( *p == DOUBLE_QUOTE ) |
---|
1405 | { |
---|
1406 | ++p; |
---|
1407 | end = "\""; // double quote in string |
---|
1408 | p = ReadText( p, &value, false, end, false, encoding ); |
---|
1409 | } |
---|
1410 | else |
---|
1411 | { |
---|
1412 | // All attribute values should be in single or double quotes. |
---|
1413 | // But this is such a common error that the parser will try |
---|
1414 | // its best, even without them. |
---|
1415 | value = ""; |
---|
1416 | while ( p && *p // existence |
---|
1417 | && !IsWhiteSpace( *p ) && *p != '\n' && *p != '\r' // whitespace |
---|
1418 | && *p != '/' && *p != '>' ) // tag end |
---|
1419 | { |
---|
1420 | if ( *p == SINGLE_QUOTE || *p == DOUBLE_QUOTE ) { |
---|
1421 | // [ 1451649 ] Attribute values with trailing quotes not handled correctly |
---|
1422 | // We did not have an opening quote but seem to have a |
---|
1423 | // closing one. Give up and throw an error. |
---|
1424 | if ( document ) document->SetError( TIXML_ERROR_READING_ATTRIBUTES, p, data, encoding ); |
---|
1425 | return 0; |
---|
1426 | } |
---|
1427 | value += *p; |
---|
1428 | ++p; |
---|
1429 | } |
---|
1430 | } |
---|
1431 | return p; |
---|
1432 | } |
---|
1433 | |
---|
1434 | #ifdef TIXML_USE_STL |
---|
1435 | void TiXmlText::StreamIn( std::istream * in, TIXML_STRING * tag ) |
---|
1436 | { |
---|
1437 | while ( in->good() ) |
---|
1438 | { |
---|
1439 | int c = in->peek(); |
---|
1440 | if ( !cdata && (c == '<' ) ) |
---|
1441 | { |
---|
1442 | return; |
---|
1443 | } |
---|
1444 | if ( c <= 0 ) |
---|
1445 | { |
---|
1446 | TiXmlDocument* document = GetDocument(); |
---|
1447 | if ( document ) |
---|
1448 | document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
1449 | return; |
---|
1450 | } |
---|
1451 | |
---|
1452 | (*tag) += (char) c; |
---|
1453 | in->get(); // "commits" the peek made above |
---|
1454 | |
---|
1455 | if ( cdata && c == '>' && tag->size() >= 3 ) { |
---|
1456 | size_t len = tag->size(); |
---|
1457 | if ( (*tag)[len-2] == ']' && (*tag)[len-3] == ']' ) { |
---|
1458 | // terminator of cdata. |
---|
1459 | return; |
---|
1460 | } |
---|
1461 | } |
---|
1462 | } |
---|
1463 | } |
---|
1464 | #endif |
---|
1465 | |
---|
1466 | const char* TiXmlText::Parse( const char* p, TiXmlParsingData* data, TiXmlEncoding encoding ) |
---|
1467 | { |
---|
1468 | value = ""; |
---|
1469 | TiXmlDocument* document = GetDocument(); |
---|
1470 | |
---|
1471 | if ( data ) |
---|
1472 | { |
---|
1473 | data->Stamp( p, encoding ); |
---|
1474 | location = data->Cursor(); |
---|
1475 | } |
---|
1476 | |
---|
1477 | const char* const startTag = "<![CDATA["; |
---|
1478 | const char* const endTag = "]]>"; |
---|
1479 | |
---|
1480 | if ( cdata || StringEqual( p, startTag, false, encoding ) ) |
---|
1481 | { |
---|
1482 | cdata = true; |
---|
1483 | |
---|
1484 | if ( !StringEqual( p, startTag, false, encoding ) ) |
---|
1485 | { |
---|
1486 | document->SetError( TIXML_ERROR_PARSING_CDATA, p, data, encoding ); |
---|
1487 | return 0; |
---|
1488 | } |
---|
1489 | p += strlen( startTag ); |
---|
1490 | |
---|
1491 | // Keep all the white space, ignore the encoding, etc. |
---|
1492 | while ( p && *p |
---|
1493 | && !StringEqual( p, endTag, false, encoding ) |
---|
1494 | ) |
---|
1495 | { |
---|
1496 | value += *p; |
---|
1497 | ++p; |
---|
1498 | } |
---|
1499 | |
---|
1500 | TIXML_STRING dummy; |
---|
1501 | p = ReadText( p, &dummy, false, endTag, false, encoding ); |
---|
1502 | return p; |
---|
1503 | } |
---|
1504 | else |
---|
1505 | { |
---|
1506 | bool ignoreWhite = true; |
---|
1507 | |
---|
1508 | const char* end = "<"; |
---|
1509 | p = ReadText( p, &value, ignoreWhite, end, false, encoding ); |
---|
1510 | if ( p ) |
---|
1511 | return p-1; // don't truncate the '<' |
---|
1512 | return 0; |
---|
1513 | } |
---|
1514 | } |
---|
1515 | |
---|
1516 | #ifdef TIXML_USE_STL |
---|
1517 | void TiXmlDeclaration::StreamIn( std::istream * in, TIXML_STRING * tag ) |
---|
1518 | { |
---|
1519 | while ( in->good() ) |
---|
1520 | { |
---|
1521 | int c = in->get(); |
---|
1522 | if ( c <= 0 ) |
---|
1523 | { |
---|
1524 | TiXmlDocument* document = GetDocument(); |
---|
1525 | if ( document ) |
---|
1526 | document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN ); |
---|
1527 | return; |
---|
1528 | } |
---|
1529 | (*tag) += (char) c; |
---|
1530 | |
---|
1531 | if ( c == '>' ) |
---|
1532 | { |
---|
1533 | // All is well. |
---|
1534 | return; |
---|
1535 | } |
---|
1536 | } |
---|
1537 | } |
---|
1538 | #endif |
---|
1539 | |
---|
1540 | const char* TiXmlDeclaration::Parse( const char* p, TiXmlParsingData* data, TiXmlEncoding _encoding ) |
---|
1541 | { |
---|
1542 | p = SkipWhiteSpace( p, _encoding ); |
---|
1543 | // Find the beginning, find the end, and look for |
---|
1544 | // the stuff in-between. |
---|
1545 | TiXmlDocument* document = GetDocument(); |
---|
1546 | if ( !p || !*p || !StringEqual( p, "<?xml", true, _encoding ) ) |
---|
1547 | { |
---|
1548 | if ( document ) document->SetError( TIXML_ERROR_PARSING_DECLARATION, 0, 0, _encoding ); |
---|
1549 | return 0; |
---|
1550 | } |
---|
1551 | if ( data ) |
---|
1552 | { |
---|
1553 | data->Stamp( p, _encoding ); |
---|
1554 | location = data->Cursor(); |
---|
1555 | } |
---|
1556 | p += 5; |
---|
1557 | |
---|
1558 | version = ""; |
---|
1559 | encoding = ""; |
---|
1560 | standalone = ""; |
---|
1561 | |
---|
1562 | while ( p && *p ) |
---|
1563 | { |
---|
1564 | if ( *p == '>' ) |
---|
1565 | { |
---|
1566 | ++p; |
---|
1567 | return p; |
---|
1568 | } |
---|
1569 | |
---|
1570 | p = SkipWhiteSpace( p, _encoding ); |
---|
1571 | if ( StringEqual( p, "version", true, _encoding ) ) |
---|
1572 | { |
---|
1573 | TiXmlAttribute attrib; |
---|
1574 | p = attrib.Parse( p, data, _encoding ); |
---|
1575 | version = attrib.Value(); |
---|
1576 | } |
---|
1577 | else if ( StringEqual( p, "encoding", true, _encoding ) ) |
---|
1578 | { |
---|
1579 | TiXmlAttribute attrib; |
---|
1580 | p = attrib.Parse( p, data, _encoding ); |
---|
1581 | encoding = attrib.Value(); |
---|
1582 | } |
---|
1583 | else if ( StringEqual( p, "standalone", true, _encoding ) ) |
---|
1584 | { |
---|
1585 | TiXmlAttribute attrib; |
---|
1586 | p = attrib.Parse( p, data, _encoding ); |
---|
1587 | standalone = attrib.Value(); |
---|
1588 | } |
---|
1589 | else |
---|
1590 | { |
---|
1591 | // Read over whatever it is. |
---|
1592 | while( p && *p && *p != '>' && !IsWhiteSpace( *p ) ) |
---|
1593 | ++p; |
---|
1594 | } |
---|
1595 | } |
---|
1596 | return 0; |
---|
1597 | } |
---|
1598 | |
---|
1599 | bool TiXmlText::Blank() const |
---|
1600 | { |
---|
1601 | for ( unsigned i=0; i<value.length(); i++ ) |
---|
1602 | if ( !IsWhiteSpace( value[i] ) ) |
---|
1603 | return false; |
---|
1604 | return true; |
---|
1605 | } |
---|
1606 | |
---|