source: trunk/src/helpers/xmltok.c@ 132

Last change on this file since 132 was 98, checked in by umoeller, 24 years ago

Misc updates.
  • Property svn:eol-style set to CRLF
  • Property svn:keywords set to Author Date Id Revision
File size: 46.3 KB
Line 
1/*
2 * Copyright (c) 1998, 1999 Thai Open Source Software Center Ltd
3 * See the file COPYING for copying permission.
4 */
5
6/* #ifdef COMPILED_FROM_DSP
7 * # include "winconfig.h"
8 * #else
9 * # include <config.h>
10 * #endif
11 */
12
13#include <memory.h>
14
15#include "expat\expat_setup.h" // V0.9.9 (2001-02-10) [umoeller]
16
17#pragma info(norea, nogen)
18 // disable "statement unreachable" and "missing break statement"
19 // this code generates those options HEAVILY
20
21#include "expat\xmltok.h"
22#include "expat\nametab.h"
23
24#ifdef XML_DTD
25#define IGNORE_SECTION_TOK_VTABLE , PREFIX(ignoreSectionTok)
26#else
27#define IGNORE_SECTION_TOK_VTABLE /* as nothing */
28#endif
29
30#define VTABLE1 \
31 { PREFIX(prologTok), PREFIX(contentTok), \
32 PREFIX(cdataSectionTok) IGNORE_SECTION_TOK_VTABLE }, \
33 { PREFIX(attributeValueTok), PREFIX(entityValueTok) }, \
34 PREFIX(sameName), \
35 PREFIX(nameMatchesAscii), \
36 PREFIX(nameLength), \
37 PREFIX(skipS), \
38 PREFIX(getAtts), \
39 PREFIX(charRefNumber), \
40 PREFIX(predefinedEntityName), \
41 PREFIX(updatePosition), \
42 PREFIX(isPublicId)
43
44#define VTABLE VTABLE1, PREFIX(toUtf8), PREFIX(toUtf16)
45
46#define UCS2_GET_NAMING(pages, hi, lo) \
47 (namingBitmap[(pages[hi] << 3) + ((lo) >> 5)] & (1 << ((lo) & 0x1F)))
48
49/* A 2 byte UTF-8 representation splits the characters 11 bits
50 * between the bottom 5 and 6 bits of the bytes.
51 * We need 8 bits to index into pages, 3 bits to add to that index and
52 * 5 bits to generate the mask. */
53#define UTF8_GET_NAMING2(pages, byte) \
54 (namingBitmap[((pages)[(((byte)[0]) >> 2) & 7] << 3) \
55 + ((((byte)[0]) & 3) << 1) \
56 + ((((byte)[1]) >> 5) & 1)] \
57 & (1 << (((byte)[1]) & 0x1F)))
58
59/* A 3 byte UTF-8 representation splits the characters 16 bits
60 * between the bottom 4, 6 and 6 bits of the bytes.
61 * We need 8 bits to index into pages, 3 bits to add to that index and
62 * 5 bits to generate the mask. */
63#define UTF8_GET_NAMING3(pages, byte) \
64 (namingBitmap[((pages)[((((byte)[0]) & 0xF) << 4) \
65 + ((((byte)[1]) >> 2) & 0xF)] \
66 << 3) \
67 + ((((byte)[1]) & 3) << 1) \
68 + ((((byte)[2]) >> 5) & 1)] \
69 & (1 << (((byte)[2]) & 0x1F)))
70
71#define UTF8_GET_NAMING(pages, p, n) \
72 ((n) == 2 \
73 ? UTF8_GET_NAMING2(pages, (const unsigned char *)(p)) \
74 : ((n) == 3 \
75 ? UTF8_GET_NAMING3(pages, (const unsigned char *)(p)) \
76 : 0))
77
78#define UTF8_INVALID3(p) \
79 ((*p) == 0xED \
80 ? (((p)[1] & 0x20) != 0) \
81 : ((*p) == 0xEF \
82 ? ((p)[1] == 0xBF && ((p)[2] == 0xBF || (p)[2] == 0xBE)) \
83 : 0))
84
85#define UTF8_INVALID4(p) ((*p) == 0xF4 && ((p)[1] & 0x30) != 0)
86
87static
88int EXPATENTRY isNever(const ENCODING * enc, const char *p)
89{
90 return 0;
91}
92
93static
94int EXPATENTRY utf8_isName2(const ENCODING * enc, const char *p)
95{
96 return UTF8_GET_NAMING2(namePages, (const unsigned char *)p);
97}
98
99static
100int EXPATENTRY utf8_isName3(const ENCODING * enc, const char *p)
101{
102 return UTF8_GET_NAMING3(namePages, (const unsigned char *)p);
103}
104
105#define utf8_isName4 isNever
106
107static
108int EXPATENTRY utf8_isNmstrt2(const ENCODING * enc, const char *p)
109{
110 return UTF8_GET_NAMING2(nmstrtPages, (const unsigned char *)p);
111}
112
113static
114int EXPATENTRY utf8_isNmstrt3(const ENCODING * enc, const char *p)
115{
116 return UTF8_GET_NAMING3(nmstrtPages, (const unsigned char *)p);
117}
118
119#define utf8_isNmstrt4 isNever
120
121#define utf8_isInvalid2 isNever
122
123static
124int EXPATENTRY utf8_isInvalid3(const ENCODING * enc, const char *p)
125{
126 return UTF8_INVALID3((const unsigned char *)p);
127}
128
129static
130int EXPATENTRY utf8_isInvalid4(const ENCODING * enc, const char *p)
131{
132 return UTF8_INVALID4((const unsigned char *)p);
133}
134
135struct normal_encoding
136{
137 ENCODING enc;
138 unsigned char type[256];
139#ifdef XML_MIN_SIZE
140 int (* EXPATENTRY byteType) (const ENCODING *, const char *);
141 int (* EXPATENTRY isNameMin) (const ENCODING *, const char *);
142 int (* EXPATENTRY isNmstrtMin) (const ENCODING *, const char *);
143 int (* EXPATENTRY byteToAscii) (const ENCODING *, const char *);
144 int (* EXPATENTRY charMatches) (const ENCODING *, const char *, int);
145#endif /* XML_MIN_SIZE */
146 int (* EXPATENTRY isName2) (const ENCODING *, const char *);
147 int (* EXPATENTRY isName3) (const ENCODING *, const char *);
148 int (* EXPATENTRY isName4) (const ENCODING *, const char *);
149 int (* EXPATENTRY isNmstrt2) (const ENCODING *, const char *);
150 int (* EXPATENTRY isNmstrt3) (const ENCODING *, const char *);
151 int (* EXPATENTRY isNmstrt4) (const ENCODING *, const char *);
152 int (* EXPATENTRY isInvalid2) (const ENCODING *, const char *);
153 int (* EXPATENTRY isInvalid3) (const ENCODING *, const char *);
154 int (* EXPATENTRY isInvalid4) (const ENCODING *, const char *);
155};
156
157#ifdef XML_MIN_SIZE
158
159#define STANDARD_VTABLE(E) \
160 E ## byteType, \
161 E ## isNameMin, \
162 E ## isNmstrtMin, \
163 E ## byteToAscii, \
164 E ## charMatches,
165
166#else
167
168#define STANDARD_VTABLE(E) /* as nothing */
169
170#endif
171
172#define NORMAL_VTABLE(E) \
173 E ## isName2, \
174 E ## isName3, \
175 E ## isName4, \
176 E ## isNmstrt2, \
177 E ## isNmstrt3, \
178 E ## isNmstrt4, \
179 E ## isInvalid2, \
180 E ## isInvalid3, \
181 E ## isInvalid4
182
183static int checkCharRefNumber(int);
184
185#include "expat\xmltok_impl.h"
186#include "expat\ascii.h"
187
188#ifdef XML_MIN_SIZE
189#define sb_isNameMin isNever
190#define sb_isNmstrtMin isNever
191#endif
192
193#ifdef XML_MIN_SIZE
194#define MINBPC(enc) ((enc)->minBytesPerChar)
195#else
196/* minimum bytes per character */
197#define MINBPC(enc) 1
198#endif
199
200#define SB_BYTE_TYPE(enc, p) \
201 (((struct normal_encoding *)(enc))->type[(unsigned char)*(p)])
202
203#ifdef XML_MIN_SIZE
204static int EXPATENTRY sb_byteType(const ENCODING * enc, const char *p)
205{
206 return SB_BYTE_TYPE(enc, p);
207}
208#define BYTE_TYPE(enc, p) \
209 (((const struct normal_encoding *)(enc))->byteType(enc, p))
210#else
211#define BYTE_TYPE(enc, p) SB_BYTE_TYPE(enc, p)
212#endif
213
214#ifdef XML_MIN_SIZE
215#define BYTE_TO_ASCII(enc, p) \
216 (((const struct normal_encoding *)(enc))->byteToAscii(enc, p))
217static
218int EXPATENTRY sb_byteToAscii(const ENCODING * enc, const char *p)
219{
220 return *p;
221}
222#else
223#define BYTE_TO_ASCII(enc, p) (*(p))
224#endif
225
226#define IS_NAME_CHAR(enc, p, n) \
227 (((const struct normal_encoding *)(enc))->isName ## n(enc, p))
228#define IS_NMSTRT_CHAR(enc, p, n) \
229 (((const struct normal_encoding *)(enc))->isNmstrt ## n(enc, p))
230#define IS_INVALID_CHAR(enc, p, n) \
231 (((const struct normal_encoding *)(enc))->isInvalid ## n(enc, p))
232
233#ifdef XML_MIN_SIZE
234#define IS_NAME_CHAR_MINBPC(enc, p) \
235 (((const struct normal_encoding *)(enc))->isNameMin(enc, p))
236#define IS_NMSTRT_CHAR_MINBPC(enc, p) \
237 (((const struct normal_encoding *)(enc))->isNmstrtMin(enc, p))
238#else
239#define IS_NAME_CHAR_MINBPC(enc, p) (0)
240#define IS_NMSTRT_CHAR_MINBPC(enc, p) (0)
241#endif
242
243#ifdef XML_MIN_SIZE
244#define CHAR_MATCHES(enc, p, c) \
245 (((const struct normal_encoding *)(enc))->charMatches(enc, p, c))
246static
247int EXPATENTRY sb_charMatches(const ENCODING * enc, const char *p, int c)
248{
249 return *p == c;
250}
251#else
252/* c is an ASCII character */
253#define CHAR_MATCHES(enc, p, c) (*(p) == c)
254#endif
255
256#define PREFIX(ident) normal_ ## ident
257#include "xmltok_impl.c"
258
259#undef MINBPC
260#undef BYTE_TYPE
261#undef BYTE_TO_ASCII
262#undef CHAR_MATCHES
263#undef IS_NAME_CHAR
264#undef IS_NAME_CHAR_MINBPC
265#undef IS_NMSTRT_CHAR
266#undef IS_NMSTRT_CHAR_MINBPC
267#undef IS_INVALID_CHAR
268
269enum
270{ /* UTF8_cvalN is value of masked first byte of N byte sequence */
271 UTF8_cval1 = 0x00,
272 UTF8_cval2 = 0xc0,
273 UTF8_cval3 = 0xe0,
274 UTF8_cval4 = 0xf0
275};
276
277static void EXPATENTRY utf8_toUtf8(const ENCODING * enc,
278 const char **fromP,
279 const char *fromLim,
280 char **toP,
281 const char *toLim)
282{
283 char *to;
284 const char *from;
285
286 if (fromLim - *fromP > toLim - *toP)
287 {
288 /* Avoid copying partial characters. */
289 for (fromLim = *fromP + (toLim - *toP);
290 fromLim > *fromP;
291 fromLim--)
292 if (((unsigned char)fromLim[-1] & 0xc0) != 0x80)
293 break;
294 }
295 for (to = *toP, from = *fromP;
296 from != fromLim;
297 from++, to++)
298 *to = *from;
299 *fromP = from;
300 *toP = to;
301}
302
303static void EXPATENTRY utf8_toUtf16(const ENCODING * enc,
304 const char **fromP,
305 const char *fromLim,
306 unsigned short **toP,
307 const unsigned short *toLim)
308{
309 unsigned short *to = *toP;
310 const char *from = *fromP;
311
312 while (from != fromLim && to != toLim)
313 {
314 switch (((struct normal_encoding *)enc)->type[(unsigned char)*from])
315 {
316 case BT_LEAD2:
317 *to++ = ((from[0] & 0x1f) << 6) | (from[1] & 0x3f);
318 from += 2;
319 break;
320 case BT_LEAD3:
321 *to++ = ((from[0] & 0xf) << 12) | ((from[1] & 0x3f) << 6) | (from[2] & 0x3f);
322 from += 3;
323 break;
324 case BT_LEAD4:
325 {
326 unsigned long n;
327
328 if (to + 1 == toLim)
329 break;
330 n = ((from[0] & 0x7) << 18) | ((from[1] & 0x3f) << 12) | ((from[2] & 0x3f) << 6) | (from[3] & 0x3f);
331 n -= 0x10000;
332 to[0] = (unsigned short)((n >> 10) | 0xD800);
333 to[1] = (unsigned short)((n & 0x3FF) | 0xDC00);
334 to += 2;
335 from += 4;
336 }
337 break;
338 default:
339 *to++ = *from++;
340 break;
341 }
342 }
343 *fromP = from;
344 *toP = to;
345}
346
347#ifdef XML_NS
348static const struct normal_encoding utf8_encoding_ns =
349{
350 {VTABLE1, utf8_toUtf8, utf8_toUtf16, 1, 1, 0},
351 {
352#include "expat\asciitab.h"
353#include "expat\utf8tab.h"
354 },
355 STANDARD_VTABLE(sb_) NORMAL_VTABLE(utf8_)
356};
357
358#endif
359
360static const struct normal_encoding utf8_encoding =
361{
362 {VTABLE1, utf8_toUtf8, utf8_toUtf16, 1, 1, 0},
363 {
364#define BT_COLON BT_NMSTRT
365#include "expat\asciitab.h"
366#undef BT_COLON
367#include "expat\utf8tab.h"
368 },
369 STANDARD_VTABLE(sb_) NORMAL_VTABLE(utf8_)
370};
371
372#ifdef XML_NS
373
374static const struct normal_encoding internal_utf8_encoding_ns =
375{
376 {VTABLE1, utf8_toUtf8, utf8_toUtf16, 1, 1, 0},
377 {
378#include "expat\iasciitab.h"
379#include "expat\utf8tab.h"
380 },
381 STANDARD_VTABLE(sb_) NORMAL_VTABLE(utf8_)
382};
383
384#endif
385
386static const struct normal_encoding internal_utf8_encoding =
387{
388 {VTABLE1, utf8_toUtf8, utf8_toUtf16, 1, 1, 0},
389 {
390#define BT_COLON BT_NMSTRT
391#include "expat\iasciitab.h"
392#undef BT_COLON
393#include "expat\utf8tab.h"
394 },
395 STANDARD_VTABLE(sb_) NORMAL_VTABLE(utf8_)
396};
397
398static void EXPATENTRY latin1_toUtf8(const ENCODING * enc,
399 const char **fromP,
400 const char *fromLim,
401 char **toP,
402 const char *toLim)
403{
404 for (;;)
405 {
406 unsigned char c;
407
408 if (*fromP == fromLim)
409 break;
410 c = (unsigned char)**fromP;
411 if (c & 0x80)
412 {
413 if (toLim - *toP < 2)
414 break;
415 *(*toP)++ = ((c >> 6) | UTF8_cval2);
416 *(*toP)++ = ((c & 0x3f) | 0x80);
417 (*fromP)++;
418 }
419 else
420 {
421 if (*toP == toLim)
422 break;
423 *(*toP)++ = *(*fromP)++;
424 }
425 }
426}
427
428static void EXPATENTRY latin1_toUtf16(const ENCODING * enc,
429 const char **fromP,
430 const char *fromLim,
431 unsigned short **toP,
432 const unsigned short *toLim)
433{
434 while (*fromP != fromLim && *toP != toLim)
435 *(*toP)++ = (unsigned char)*(*fromP)++;
436}
437
438#ifdef XML_NS
439
440static const struct normal_encoding latin1_encoding_ns =
441{
442 {VTABLE1, latin1_toUtf8, latin1_toUtf16, 1, 0, 0},
443 {
444#include "expat\asciitab.h"
445#include "expat\latin1tab.h"
446 },
447 STANDARD_VTABLE(sb_)
448};
449
450#endif
451
452static const struct normal_encoding latin1_encoding =
453{
454 {VTABLE1, latin1_toUtf8, latin1_toUtf16, 1, 0, 0},
455 {
456#define BT_COLON BT_NMSTRT
457#include "expat\asciitab.h"
458#undef BT_COLON
459#include "expat\latin1tab.h"
460 },
461 STANDARD_VTABLE(sb_)
462};
463
464static void EXPATENTRY ascii_toUtf8(const ENCODING * enc,
465 const char **fromP,
466 const char *fromLim,
467 char **toP,
468 const char *toLim)
469{
470 while (*fromP != fromLim && *toP != toLim)
471 *(*toP)++ = *(*fromP)++;
472}
473
474#ifdef XML_NS
475
476static const struct normal_encoding ascii_encoding_ns =
477{
478 {VTABLE1, ascii_toUtf8, latin1_toUtf16, 1, 1, 0},
479 {
480#include "expat\asciitab.h"
481/* BT_NONXML == 0 */
482 },
483 STANDARD_VTABLE(sb_)
484};
485
486#endif
487
488static const struct normal_encoding ascii_encoding =
489{
490 {VTABLE1, ascii_toUtf8, latin1_toUtf16, 1, 1, 0},
491 {
492#define BT_COLON BT_NMSTRT
493#include "expat\asciitab.h"
494#undef BT_COLON
495/* BT_NONXML == 0 */
496 },
497 STANDARD_VTABLE(sb_)
498};
499
500static int unicode_byte_type(char hi, char lo)
501{
502 switch ((unsigned char)hi)
503 {
504 case 0xD8:
505 case 0xD9:
506 case 0xDA:
507 case 0xDB:
508 return BT_LEAD4;
509 case 0xDC:
510 case 0xDD:
511 case 0xDE:
512 case 0xDF:
513 return BT_TRAIL;
514 case 0xFF:
515 switch ((unsigned char)lo)
516 {
517 case 0xFF:
518 case 0xFE:
519 return BT_NONXML;
520 }
521 break;
522 }
523 return BT_NONASCII;
524}
525
526#define DEFINE_UTF16_TO_UTF8(E) \
527static void EXPATENTRY E ## toUtf8(const ENCODING *enc, \
528 const char **fromP, const char *fromLim, \
529 char **toP, const char *toLim) \
530{ \
531 const char *from; \
532 for (from = *fromP; from != fromLim; from += 2) { \
533 int plane; \
534 unsigned char lo2; \
535 unsigned char lo = GET_LO(from); \
536 unsigned char hi = GET_HI(from); \
537 switch (hi) { \
538 case 0: \
539 if (lo < 0x80) { \
540 if (*toP == toLim) { \
541 *fromP = from; \
542 return; \
543 } \
544 *(*toP)++ = lo; \
545 break; \
546 } \
547 /* fall through */ \
548 case 0x1: case 0x2: case 0x3: \
549 case 0x4: case 0x5: case 0x6: case 0x7: \
550 if (toLim - *toP < 2) { \
551 *fromP = from; \
552 return; \
553 } \
554 *(*toP)++ = ((lo >> 6) | (hi << 2) | UTF8_cval2); \
555 *(*toP)++ = ((lo & 0x3f) | 0x80); \
556 break; \
557 default: \
558 if (toLim - *toP < 3) { \
559 *fromP = from; \
560 return; \
561 } \
562 /* 16 bits divided 4, 6, 6 amongst 3 bytes */ \
563 *(*toP)++ = ((hi >> 4) | UTF8_cval3); \
564 *(*toP)++ = (((hi & 0xf) << 2) | (lo >> 6) | 0x80); \
565 *(*toP)++ = ((lo & 0x3f) | 0x80); \
566 break; \
567 case 0xD8: case 0xD9: case 0xDA: case 0xDB: \
568 if (toLim - *toP < 4) { \
569 *fromP = from; \
570 return; \
571 } \
572 plane = (((hi & 0x3) << 2) | ((lo >> 6) & 0x3)) + 1; \
573 *(*toP)++ = ((plane >> 2) | UTF8_cval4); \
574 *(*toP)++ = (((lo >> 2) & 0xF) | ((plane & 0x3) << 4) | 0x80); \
575 from += 2; \
576 lo2 = GET_LO(from); \
577 *(*toP)++ = (((lo & 0x3) << 4) \
578 | ((GET_HI(from) & 0x3) << 2) \
579 | (lo2 >> 6) \
580 | 0x80); \
581 *(*toP)++ = ((lo2 & 0x3f) | 0x80); \
582 break; \
583 } \
584 } \
585 *fromP = from; \
586}
587
588#define DEFINE_UTF16_TO_UTF16(E) \
589static void EXPATENTRY E ## toUtf16(const ENCODING *enc, \
590 const char **fromP, const char *fromLim, \
591 unsigned short **toP, const unsigned short *toLim) \
592{ \
593 /* Avoid copying first half only of surrogate */ \
594 if (fromLim - *fromP > ((toLim - *toP) << 1) \
595 && (GET_HI(fromLim - 2) & 0xF8) == 0xD8) \
596 fromLim -= 2; \
597 for (; *fromP != fromLim && *toP != toLim; *fromP += 2) \
598 *(*toP)++ = (GET_HI(*fromP) << 8) | GET_LO(*fromP); \
599}
600
601#define SET2(ptr, ch) \
602 (((ptr)[0] = ((ch) & 0xff)), ((ptr)[1] = ((ch) >> 8)))
603#define GET_LO(ptr) ((unsigned char)(ptr)[0])
604#define GET_HI(ptr) ((unsigned char)(ptr)[1])
605
606DEFINE_UTF16_TO_UTF8(little2_)
607DEFINE_UTF16_TO_UTF16(little2_)
608
609#undef SET2
610#undef GET_LO
611#undef GET_HI
612
613#define SET2(ptr, ch) \
614 (((ptr)[0] = ((ch) >> 8)), ((ptr)[1] = ((ch) & 0xFF)))
615#define GET_LO(ptr) ((unsigned char)(ptr)[1])
616#define GET_HI(ptr) ((unsigned char)(ptr)[0])
617
618DEFINE_UTF16_TO_UTF8(big2_)
619DEFINE_UTF16_TO_UTF16(big2_)
620
621#undef SET2
622#undef GET_LO
623#undef GET_HI
624
625#define LITTLE2_BYTE_TYPE(enc, p) \
626 ((p)[1] == 0 \
627 ? ((struct normal_encoding *)(enc))->type[(unsigned char)*(p)] \
628 : unicode_byte_type((p)[1], (p)[0]))
629#define LITTLE2_BYTE_TO_ASCII(enc, p) ((p)[1] == 0 ? (p)[0] : -1)
630#define LITTLE2_CHAR_MATCHES(enc, p, c) ((p)[1] == 0 && (p)[0] == c)
631#define LITTLE2_IS_NAME_CHAR_MINBPC(enc, p) \
632 UCS2_GET_NAMING(namePages, (unsigned char)p[1], (unsigned char)p[0])
633#define LITTLE2_IS_NMSTRT_CHAR_MINBPC(enc, p) \
634 UCS2_GET_NAMING(nmstrtPages, (unsigned char)p[1], (unsigned char)p[0])
635
636#ifdef XML_MIN_SIZE
637static int EXPATENTRY little2_byteType(const ENCODING * enc, const char *p)
638{
639 return LITTLE2_BYTE_TYPE(enc, p);
640}
641
642static int EXPATENTRY little2_byteToAscii(const ENCODING * enc, const char *p)
643{
644 return LITTLE2_BYTE_TO_ASCII(enc, p);
645}
646
647static int EXPATENTRY little2_charMatches(const ENCODING * enc, const char *p, int c)
648{
649 return LITTLE2_CHAR_MATCHES(enc, p, c);
650}
651
652static int EXPATENTRY little2_isNameMin(const ENCODING * enc, const char *p)
653{
654 return LITTLE2_IS_NAME_CHAR_MINBPC(enc, p);
655}
656
657static int EXPATENTRY little2_isNmstrtMin(const ENCODING * enc, const char *p)
658{
659 return LITTLE2_IS_NMSTRT_CHAR_MINBPC(enc, p);
660}
661
662#undef VTABLE
663#define VTABLE VTABLE1, little2_toUtf8, little2_toUtf16
664
665#else /* not XML_MIN_SIZE */
666
667#undef PREFIX
668#define PREFIX(ident) little2_ ## ident
669#define MINBPC(enc) 2
670/* CHAR_MATCHES is guaranteed to have MINBPC bytes available. */
671#define BYTE_TYPE(enc, p) LITTLE2_BYTE_TYPE(enc, p)
672#define BYTE_TO_ASCII(enc, p) LITTLE2_BYTE_TO_ASCII(enc, p)
673#define CHAR_MATCHES(enc, p, c) LITTLE2_CHAR_MATCHES(enc, p, c)
674#define IS_NAME_CHAR(enc, p, n) 0
675#define IS_NAME_CHAR_MINBPC(enc, p) LITTLE2_IS_NAME_CHAR_MINBPC(enc, p)
676#define IS_NMSTRT_CHAR(enc, p, n) (0)
677#define IS_NMSTRT_CHAR_MINBPC(enc, p) LITTLE2_IS_NMSTRT_CHAR_MINBPC(enc, p)
678
679#include "xmltok_impl.c"
680
681#undef MINBPC
682#undef BYTE_TYPE
683#undef BYTE_TO_ASCII
684#undef CHAR_MATCHES
685#undef IS_NAME_CHAR
686#undef IS_NAME_CHAR_MINBPC
687#undef IS_NMSTRT_CHAR
688#undef IS_NMSTRT_CHAR_MINBPC
689#undef IS_INVALID_CHAR
690
691#endif /* not XML_MIN_SIZE */
692
693#ifdef XML_NS
694
695 static const struct normal_encoding little2_encoding_ns =
696 {
697 {VTABLE, 2, 0,
698#if XML_BYTE_ORDER == 12
699 1
700#else
701 0
702#endif
703 },
704 {
705#include "expat\asciitab.h"
706#include "expat\latin1tab.h"
707 },
708 STANDARD_VTABLE(little2_)
709};
710
711#endif
712
713 static const struct normal_encoding little2_encoding =
714 {
715 {VTABLE, 2, 0,
716#if XML_BYTE_ORDER == 12
717 1
718#else
719 0
720#endif
721 },
722 {
723#define BT_COLON BT_NMSTRT
724#include "expat\asciitab.h"
725#undef BT_COLON
726#include "expat\latin1tab.h"
727 },
728 STANDARD_VTABLE(little2_)
729};
730
731#if XML_BYTE_ORDER != 21
732
733#ifdef XML_NS
734
735 static const struct normal_encoding internal_little2_encoding_ns =
736 {
737 {VTABLE, 2, 0, 1},
738 {
739#include "expat\iasciitab.h"
740#include "expat\latin1tab.h"
741 },
742 STANDARD_VTABLE(little2_)
743};
744
745#endif
746
747 static const struct normal_encoding internal_little2_encoding =
748 {
749 {VTABLE, 2, 0, 1},
750 {
751#define BT_COLON BT_NMSTRT
752#include "expat\iasciitab.h"
753#undef BT_COLON
754#include "expat\latin1tab.h"
755 },
756 STANDARD_VTABLE(little2_)
757};
758
759#endif
760
761
762#define BIG2_BYTE_TYPE(enc, p) \
763 ((p)[0] == 0 \
764 ? ((struct normal_encoding *)(enc))->type[(unsigned char)(p)[1]] \
765 : unicode_byte_type((p)[0], (p)[1]))
766#define BIG2_BYTE_TO_ASCII(enc, p) ((p)[0] == 0 ? (p)[1] : -1)
767#define BIG2_CHAR_MATCHES(enc, p, c) ((p)[0] == 0 && (p)[1] == c)
768#define BIG2_IS_NAME_CHAR_MINBPC(enc, p) \
769 UCS2_GET_NAMING(namePages, (unsigned char)p[0], (unsigned char)p[1])
770#define BIG2_IS_NMSTRT_CHAR_MINBPC(enc, p) \
771 UCS2_GET_NAMING(nmstrtPages, (unsigned char)p[0], (unsigned char)p[1])
772
773#ifdef XML_MIN_SIZE
774
775static int EXPATENTRY big2_byteType(const ENCODING * enc, const char *p)
776{
777 return BIG2_BYTE_TYPE(enc, p);
778}
779
780static int EXPATENTRY big2_byteToAscii(const ENCODING * enc, const char *p)
781{
782 return BIG2_BYTE_TO_ASCII(enc, p);
783}
784
785static int EXPATENTRY big2_charMatches(const ENCODING * enc, const char *p, int c)
786{
787 return BIG2_CHAR_MATCHES(enc, p, c);
788}
789
790static int EXPATENTRY big2_isNameMin(const ENCODING * enc, const char *p)
791{
792 return BIG2_IS_NAME_CHAR_MINBPC(enc, p);
793}
794
795static int EXPATENTRY big2_isNmstrtMin(const ENCODING * enc, const char *p)
796{
797 return BIG2_IS_NMSTRT_CHAR_MINBPC(enc, p);
798}
799
800#undef VTABLE
801#define VTABLE VTABLE1, big2_toUtf8, big2_toUtf16
802
803#else /* not XML_MIN_SIZE */
804
805#undef PREFIX
806#define PREFIX(ident) big2_ ## ident
807#define MINBPC(enc) 2
808/* CHAR_MATCHES is guaranteed to have MINBPC bytes available. */
809#define BYTE_TYPE(enc, p) BIG2_BYTE_TYPE(enc, p)
810#define BYTE_TO_ASCII(enc, p) BIG2_BYTE_TO_ASCII(enc, p)
811#define CHAR_MATCHES(enc, p, c) BIG2_CHAR_MATCHES(enc, p, c)
812#define IS_NAME_CHAR(enc, p, n) 0
813#define IS_NAME_CHAR_MINBPC(enc, p) BIG2_IS_NAME_CHAR_MINBPC(enc, p)
814#define IS_NMSTRT_CHAR(enc, p, n) (0)
815#define IS_NMSTRT_CHAR_MINBPC(enc, p) BIG2_IS_NMSTRT_CHAR_MINBPC(enc, p)
816
817#include "xmltok_impl.c"
818
819#undef MINBPC
820#undef BYTE_TYPE
821#undef BYTE_TO_ASCII
822#undef CHAR_MATCHES
823#undef IS_NAME_CHAR
824#undef IS_NAME_CHAR_MINBPC
825#undef IS_NMSTRT_CHAR
826#undef IS_NMSTRT_CHAR_MINBPC
827#undef IS_INVALID_CHAR
828
829#endif /* not XML_MIN_SIZE */
830
831#ifdef XML_NS
832
833 static const struct normal_encoding big2_encoding_ns =
834 {
835 {VTABLE, 2, 0,
836#if XML_BYTE_ORDER == 21
837 1
838#else
839 0
840#endif
841 },
842 {
843#include "expat\asciitab.h"
844#include "expat\latin1tab.h"
845 },
846 STANDARD_VTABLE(big2_)
847};
848
849#endif
850
851 static const struct normal_encoding big2_encoding =
852 {
853 {VTABLE, 2, 0,
854#if XML_BYTE_ORDER == 21
855 1
856#else
857 0
858#endif
859 },
860 {
861#define BT_COLON BT_NMSTRT
862#include "expat\asciitab.h"
863#undef BT_COLON
864#include "expat\latin1tab.h"
865 },
866 STANDARD_VTABLE(big2_)
867};
868
869#if XML_BYTE_ORDER != 12
870
871#ifdef XML_NS
872
873 static const struct normal_encoding internal_big2_encoding_ns =
874 {
875 {VTABLE, 2, 0, 1},
876 {
877#include "expat\iasciitab.h"
878#include "expat\latin1tab.h"
879 },
880 STANDARD_VTABLE(big2_)
881};
882
883#endif
884
885 static const struct normal_encoding internal_big2_encoding =
886 {
887 {VTABLE, 2, 0, 1},
888 {
889#define BT_COLON BT_NMSTRT
890#include "expat\iasciitab.h"
891#undef BT_COLON
892#include "expat\latin1tab.h"
893 },
894 STANDARD_VTABLE(big2_)
895};
896
897#endif
898
899#undef PREFIX
900
901static int streqci(const char *s1, const char *s2)
902{
903 for (;;)
904 {
905 char c1 = *s1++;
906 char c2 = *s2++;
907
908 if (ASCII_a <= c1 && c1 <= ASCII_z)
909 c1 += ASCII_A - ASCII_a;
910 if (ASCII_a <= c2 && c2 <= ASCII_z)
911 c2 += ASCII_A - ASCII_a;
912 if (c1 != c2)
913 return 0;
914 if (!c1)
915 break;
916 }
917 return 1;
918}
919
920static void EXPATENTRY initUpdatePosition(const ENCODING * enc, const char *ptr,
921 const char *end, POSITION * pos)
922{
923 normal_updatePosition(&utf8_encoding.enc, ptr, end, pos);
924}
925
926static int EXPATENTRY toAscii(const ENCODING * enc, const char *ptr, const char *end)
927{
928 char buf[1];
929 char *p = buf;
930
931 XmlUtf8Convert(enc, &ptr, end, &p, p + 1);
932 if (p == buf)
933 return -1;
934 else
935 return buf[0];
936}
937
938static int EXPATENTRY isSpace(int c)
939{
940 switch (c)
941 {
942 case 0x20:
943 case 0xD:
944 case 0xA:
945 case 0x9:
946 return 1;
947 }
948 return 0;
949}
950
951/* Return 1 if there's just optional white space
952 * or there's an S followed by name=val. */
953static int EXPATENTRY parsePseudoAttribute(const ENCODING * enc,
954 const char *ptr,
955 const char *end,
956 const char **namePtr,
957 const char **nameEndPtr,
958 const char **valPtr,
959 const char **nextTokPtr)
960{
961 int c;
962 char open;
963
964 if (ptr == end)
965 {
966 *namePtr = 0;
967 return 1;
968 }
969 if (!isSpace(toAscii(enc, ptr, end)))
970 {
971 *nextTokPtr = ptr;
972 return 0;
973 }
974 do
975 {
976 ptr += enc->minBytesPerChar;
977 }
978 while (isSpace(toAscii(enc, ptr, end)));
979 if (ptr == end)
980 {
981 *namePtr = 0;
982 return 1;
983 }
984 *namePtr = ptr;
985 for (;;)
986 {
987 c = toAscii(enc, ptr, end);
988 if (c == -1)
989 {
990 *nextTokPtr = ptr;
991 return 0;
992 }
993 if (c == ASCII_EQUALS)
994 {
995 *nameEndPtr = ptr;
996 break;
997 }
998 if (isSpace(c))
999 {
1000 *nameEndPtr = ptr;
1001 do
1002 {
1003 ptr += enc->minBytesPerChar;
1004 }
1005 while (isSpace(c = toAscii(enc, ptr, end)));
1006 if (c != ASCII_EQUALS)
1007 {
1008 *nextTokPtr = ptr;
1009 return 0;
1010 }
1011 break;
1012 }
1013 ptr += enc->minBytesPerChar;
1014 }
1015 if (ptr == *namePtr)
1016 {
1017 *nextTokPtr = ptr;
1018 return 0;
1019 }
1020 ptr += enc->minBytesPerChar;
1021 c = toAscii(enc, ptr, end);
1022 while (isSpace(c))
1023 {
1024 ptr += enc->minBytesPerChar;
1025 c = toAscii(enc, ptr, end);
1026 }
1027 if (c != ASCII_QUOT && c != ASCII_APOS)
1028 {
1029 *nextTokPtr = ptr;
1030 return 0;
1031 }
1032 open = c;
1033 ptr += enc->minBytesPerChar;
1034 *valPtr = ptr;
1035 for (;; ptr += enc->minBytesPerChar)
1036 {
1037 c = toAscii(enc, ptr, end);
1038 if (c == open)
1039 break;
1040 if (!(ASCII_a <= c && c <= ASCII_z)
1041 && !(ASCII_A <= c && c <= ASCII_Z)
1042 && !(ASCII_0 <= c && c <= ASCII_9)
1043 && c != ASCII_PERIOD
1044 && c != ASCII_MINUS
1045 && c != ASCII_UNDERSCORE)
1046 {
1047 *nextTokPtr = ptr;
1048 return 0;
1049 }
1050 }
1051 *nextTokPtr = ptr + enc->minBytesPerChar;
1052 return 1;
1053}
1054
1055static const char KW_version[] =
1056{
1057 ASCII_v, ASCII_e, ASCII_r, ASCII_s, ASCII_i, ASCII_o, ASCII_n, '\0'
1058};
1059
1060static const char KW_encoding[] =
1061{
1062 ASCII_e, ASCII_n, ASCII_c, ASCII_o, ASCII_d, ASCII_i, ASCII_n, ASCII_g, '\0'
1063};
1064
1065static const char KW_standalone[] =
1066{
1067 ASCII_s, ASCII_t, ASCII_a, ASCII_n, ASCII_d, ASCII_a, ASCII_l, ASCII_o, ASCII_n, ASCII_e, '\0'
1068};
1069
1070static const char KW_yes[] =
1071{
1072 ASCII_y, ASCII_e, ASCII_s, '\0'
1073};
1074
1075static const char KW_no[] =
1076{
1077 ASCII_n, ASCII_o, '\0'
1078};
1079
1080static int doParseXmlDecl(const ENCODING * (*encodingFinder) (const ENCODING *,
1081 const char *,
1082 const char *),
1083 int isGeneralTextEntity,
1084 const ENCODING * enc,
1085 const char *ptr,
1086 const char *end,
1087 const char **badPtr,
1088 const char **versionPtr,
1089 const char **versionEndPtr,
1090 const char **encodingName,
1091 const ENCODING ** encoding,
1092 int *standalone)
1093{
1094 const char *val = 0;
1095 const char *name = 0;
1096 const char *nameEnd = 0;
1097
1098 ptr += 5 * enc->minBytesPerChar;
1099 end -= 2 * enc->minBytesPerChar;
1100 if (!parsePseudoAttribute(enc, ptr, end, &name, &nameEnd, &val, &ptr) || !name)
1101 {
1102 *badPtr = ptr;
1103 return 0;
1104 }
1105 if (!XmlNameMatchesAscii(enc, name, nameEnd, KW_version))
1106 {
1107 if (!isGeneralTextEntity)
1108 {
1109 *badPtr = name;
1110 return 0;
1111 }
1112 }
1113 else
1114 {
1115 if (versionPtr)
1116 *versionPtr = val;
1117 if (versionEndPtr)
1118 *versionEndPtr = ptr;
1119 if (!parsePseudoAttribute(enc, ptr, end, &name, &nameEnd, &val, &ptr))
1120 {
1121 *badPtr = ptr;
1122 return 0;
1123 }
1124 if (!name)
1125 {
1126 if (isGeneralTextEntity)
1127 {
1128 /* a TextDecl must have an EncodingDecl */
1129 *badPtr = ptr;
1130 return 0;
1131 }
1132 return 1;
1133 }
1134 }
1135 if (XmlNameMatchesAscii(enc, name, nameEnd, KW_encoding))
1136 {
1137 int c = toAscii(enc, val, end);
1138
1139 if (!(ASCII_a <= c && c <= ASCII_z) && !(ASCII_A <= c && c <= ASCII_Z))
1140 {
1141 *badPtr = val;
1142 return 0;
1143 }
1144 if (encodingName)
1145 *encodingName = val;
1146 if (encoding)
1147 *encoding = encodingFinder(enc, val, ptr - enc->minBytesPerChar);
1148 if (!parsePseudoAttribute(enc, ptr, end, &name, &nameEnd, &val, &ptr))
1149 {
1150 *badPtr = ptr;
1151 return 0;
1152 }
1153 if (!name)
1154 return 1;
1155 }
1156 if (!XmlNameMatchesAscii(enc, name, nameEnd, KW_standalone) || isGeneralTextEntity)
1157 {
1158 *badPtr = name;
1159 return 0;
1160 }
1161 if (XmlNameMatchesAscii(enc, val, ptr - enc->minBytesPerChar, KW_yes))
1162 {
1163 if (standalone)
1164 *standalone = 1;
1165 }
1166 else if (XmlNameMatchesAscii(enc, val, ptr - enc->minBytesPerChar, KW_no))
1167 {
1168 if (standalone)
1169 *standalone = 0;
1170 }
1171 else
1172 {
1173 *badPtr = val;
1174 return 0;
1175 }
1176 while (isSpace(toAscii(enc, ptr, end)))
1177 ptr += enc->minBytesPerChar;
1178 if (ptr != end)
1179 {
1180 *badPtr = ptr;
1181 return 0;
1182 }
1183 return 1;
1184}
1185
1186static int checkCharRefNumber(int result)
1187{
1188 switch (result >> 8)
1189 {
1190 case 0xD8:
1191 case 0xD9:
1192 case 0xDA:
1193 case 0xDB:
1194 case 0xDC:
1195 case 0xDD:
1196 case 0xDE:
1197 case 0xDF:
1198 return -1;
1199 case 0:
1200 if (latin1_encoding.type[result] == BT_NONXML)
1201 return -1;
1202 break;
1203 case 0xFF:
1204 if (result == 0xFFFE || result == 0xFFFF)
1205 return -1;
1206 break;
1207 }
1208 return result;
1209}
1210
1211int XmlUtf8Encode(int c, char *buf)
1212{
1213 enum
1214 {
1215 /* minN is minimum legal resulting value for N byte sequence */
1216 min2 = 0x80,
1217 min3 = 0x800,
1218 min4 = 0x10000
1219 };
1220
1221 if (c < 0)
1222 return 0;
1223 if (c < min2)
1224 {
1225 buf[0] = (c | UTF8_cval1);
1226 return 1;
1227 }
1228 if (c < min3)
1229 {
1230 buf[0] = ((c >> 6) | UTF8_cval2);
1231 buf[1] = ((c & 0x3f) | 0x80);
1232 return 2;
1233 }
1234 if (c < min4)
1235 {
1236 buf[0] = ((c >> 12) | UTF8_cval3);
1237 buf[1] = (((c >> 6) & 0x3f) | 0x80);
1238 buf[2] = ((c & 0x3f) | 0x80);
1239 return 3;
1240 }
1241 if (c < 0x110000)
1242 {
1243 buf[0] = ((c >> 18) | UTF8_cval4);
1244 buf[1] = (((c >> 12) & 0x3f) | 0x80);
1245 buf[2] = (((c >> 6) & 0x3f) | 0x80);
1246 buf[3] = ((c & 0x3f) | 0x80);
1247 return 4;
1248 }
1249 return 0;
1250}
1251
1252int XmlUtf16Encode(int charNum, unsigned short *buf)
1253{
1254 if (charNum < 0)
1255 return 0;
1256 if (charNum < 0x10000)
1257 {
1258 buf[0] = charNum;
1259 return 1;
1260 }
1261 if (charNum < 0x110000)
1262 {
1263 charNum -= 0x10000;
1264 buf[0] = (charNum >> 10) + 0xD800;
1265 buf[1] = (charNum & 0x3FF) + 0xDC00;
1266 return 2;
1267 }
1268 return 0;
1269}
1270
1271struct unknown_encoding
1272{
1273 struct normal_encoding normal;
1274 int (*convert) (void *userData, const char *p);
1275 void *userData;
1276 unsigned short utf16[256];
1277 char utf8[256][4];
1278};
1279
1280int XmlSizeOfUnknownEncoding(void)
1281{
1282 return sizeof(struct unknown_encoding);
1283}
1284
1285static int EXPATENTRY unknown_isName(const ENCODING * enc, const char *p)
1286{
1287 int c = ((const struct unknown_encoding *)enc)
1288 ->convert(((const struct unknown_encoding *)enc)->userData, p);
1289
1290 if (c & ~0xFFFF)
1291 return 0;
1292 return UCS2_GET_NAMING(namePages, c >> 8, c & 0xFF);
1293}
1294
1295static int EXPATENTRY unknown_isNmstrt(const ENCODING * enc, const char *p)
1296{
1297 int c = ((const struct unknown_encoding *)enc)
1298 ->convert(((const struct unknown_encoding *)enc)->userData, p);
1299
1300 if (c & ~0xFFFF)
1301 return 0;
1302 return UCS2_GET_NAMING(nmstrtPages, c >> 8, c & 0xFF);
1303}
1304
1305static int EXPATENTRY unknown_isInvalid(const ENCODING * enc, const char *p)
1306{
1307 int c = ((const struct unknown_encoding *)enc)
1308 ->convert(((const struct unknown_encoding *)enc)->userData, p);
1309
1310 return (c & ~0xFFFF) || checkCharRefNumber(c) < 0;
1311}
1312
1313static void EXPATENTRY unknown_toUtf8(const ENCODING * enc,
1314 const char **fromP,
1315 const char *fromLim,
1316 char **toP,
1317 const char *toLim)
1318{
1319 char buf[XML_UTF8_ENCODE_MAX];
1320
1321 for (;;)
1322 {
1323 const char *utf8;
1324 int n;
1325
1326 if (*fromP == fromLim)
1327 break;
1328 utf8 = ((const struct unknown_encoding *)enc)->utf8[(unsigned char)**fromP];
1329 n = *utf8++;
1330 if (n == 0)
1331 {
1332 int c = ((const struct unknown_encoding *)enc)
1333 ->convert(((const struct unknown_encoding *)enc)->userData, *fromP);
1334
1335 n = XmlUtf8Encode(c, buf);
1336 if (n > toLim - *toP)
1337 break;
1338 utf8 = buf;
1339 *fromP += ((const struct normal_encoding *)enc)->type[(unsigned char)**fromP]
1340 - (BT_LEAD2 - 2);
1341 }
1342 else
1343 {
1344 if (n > toLim - *toP)
1345 break;
1346 (*fromP)++;
1347 }
1348 do
1349 {
1350 *(*toP)++ = *utf8++;
1351 }
1352 while (--n != 0);
1353 }
1354}
1355
1356static void EXPATENTRY unknown_toUtf16(const ENCODING * enc,
1357 const char **fromP,
1358 const char *fromLim,
1359 unsigned short **toP,
1360 const unsigned short *toLim)
1361{
1362 while (*fromP != fromLim && *toP != toLim)
1363 {
1364 unsigned short c
1365 = ((const struct unknown_encoding *)enc)->utf16[(unsigned char)**fromP];
1366
1367 if (c == 0)
1368 {
1369 c = (unsigned short)((const struct unknown_encoding *)enc)
1370 ->convert(((const struct unknown_encoding *)enc)->userData, *fromP);
1371 *fromP += ((const struct normal_encoding *)enc)->type[(unsigned char)**fromP]
1372 - (BT_LEAD2 - 2);
1373 }
1374 else
1375 (*fromP)++;
1376 *(*toP)++ = c;
1377 }
1378}
1379
1380/*
1381 *@@ XmlInitUnknownEncoding:
1382 *
1383 *@@changed V0.9.14 (2001-08-09) [umoeller]: couple of performance hacks
1384 */
1385
1386ENCODING* XmlInitUnknownEncoding(void *mem,
1387 int *table,
1388 int (*convert) (void *userData, const char *p),
1389 void *userData)
1390{
1391 int i;
1392 struct unknown_encoding *e = (struct unknown_encoding*)mem;
1393
1394 // gee, isn't this a regular memcpy?!?
1395 /* for (i = 0;
1396 i < (int)sizeof(struct normal_encoding);
1397 i++)
1398 ((char *)mem)[i] = ((char *)&latin1_encoding)[i]; */
1399
1400 // replaced the above with this V0.9.14 (2001-08-09) [umoeller]
1401 memcpy(mem, &latin1_encoding, sizeof(struct normal_encoding));
1402
1403 for (i = 0; i < 128; i++)
1404 if ( latin1_encoding.type[i] != BT_OTHER
1405 && latin1_encoding.type[i] != BT_NONXML
1406 && table[i] != i
1407 )
1408 return 0;
1409
1410 for (i = 0; i < 256; i++)
1411 {
1412 int c = table[i];
1413
1414 if (c == -1)
1415 {
1416 e->normal.type[i] = BT_MALFORM;
1417 /* This shouldn't really get used. */
1418 e->utf16[i] = 0xFFFF;
1419 e->utf8[i][0] = 1;
1420 e->utf8[i][1] = 0;
1421 }
1422 else if (c < 0)
1423 {
1424 if (c < -4)
1425 return 0;
1426 e->normal.type[i] = BT_LEAD2 - (c + 2);
1427 e->utf8[i][0] = 0;
1428 e->utf16[i] = 0;
1429 }
1430 else if (c < 0x80)
1431 {
1432 if ( latin1_encoding.type[c] != BT_OTHER
1433 && latin1_encoding.type[c] != BT_NONXML
1434 && c != i
1435 )
1436 return 0;
1437 e->normal.type[i] = latin1_encoding.type[c];
1438 e->utf8[i][0] = 1;
1439 e->utf8[i][1] = (char)c;
1440 e->utf16[i] = c == 0 ? 0xFFFF : c;
1441 }
1442 else if (checkCharRefNumber(c) < 0)
1443 {
1444 e->normal.type[i] = BT_NONXML;
1445 /* This shouldn't really get used. */
1446 e->utf16[i] = 0xFFFF;
1447 e->utf8[i][0] = 1;
1448 e->utf8[i][1] = 0;
1449 }
1450 else
1451 {
1452 if (c > 0xFFFF)
1453 return 0;
1454 if (UCS2_GET_NAMING(nmstrtPages, c >> 8, c & 0xff))
1455 e->normal.type[i] = BT_NMSTRT;
1456 else if (UCS2_GET_NAMING(namePages, c >> 8, c & 0xff))
1457 e->normal.type[i] = BT_NAME;
1458 else
1459 e->normal.type[i] = BT_OTHER;
1460 e->utf8[i][0] = (char)XmlUtf8Encode(c, e->utf8[i] + 1);
1461 e->utf16[i] = c;
1462 }
1463 }
1464 e->userData = userData;
1465 e->convert = convert;
1466 if (convert)
1467 {
1468 e->normal.isName2 = unknown_isName;
1469 e->normal.isName3 = unknown_isName;
1470 e->normal.isName4 = unknown_isName;
1471 e->normal.isNmstrt2 = unknown_isNmstrt;
1472 e->normal.isNmstrt3 = unknown_isNmstrt;
1473 e->normal.isNmstrt4 = unknown_isNmstrt;
1474 e->normal.isInvalid2 = unknown_isInvalid;
1475 e->normal.isInvalid3 = unknown_isInvalid;
1476 e->normal.isInvalid4 = unknown_isInvalid;
1477 }
1478 e->normal.enc.utf8Convert = unknown_toUtf8;
1479 e->normal.enc.utf16Convert = unknown_toUtf16;
1480 return &(e->normal.enc);
1481}
1482
1483/* If this enumeration is changed, getEncodingIndex and encodings
1484 * must also be changed. */
1485enum
1486{
1487 UNKNOWN_ENC = -1,
1488 ISO_8859_1_ENC = 0,
1489 US_ASCII_ENC,
1490 UTF_8_ENC,
1491 UTF_16_ENC,
1492 UTF_16BE_ENC,
1493 UTF_16LE_ENC,
1494 /* must match encodingNames up to here */
1495 NO_ENC
1496};
1497
1498static const char KW_ISO_8859_1[] =
1499{
1500 ASCII_I, ASCII_S, ASCII_O, ASCII_MINUS, ASCII_8, ASCII_8, ASCII_5, ASCII_9, ASCII_MINUS, ASCII_1, '\0'
1501};
1502static const char KW_US_ASCII[] =
1503{
1504 ASCII_U, ASCII_S, ASCII_MINUS, ASCII_A, ASCII_S, ASCII_C, ASCII_I, ASCII_I, '\0'
1505};
1506static const char KW_UTF_8[] =
1507{
1508 ASCII_U, ASCII_T, ASCII_F, ASCII_MINUS, ASCII_8, '\0'
1509};
1510static const char KW_UTF_16[] =
1511{
1512 ASCII_U, ASCII_T, ASCII_F, ASCII_MINUS, ASCII_1, ASCII_6, '\0'
1513};
1514static const char KW_UTF_16BE[] =
1515{
1516 ASCII_U, ASCII_T, ASCII_F, ASCII_MINUS, ASCII_1, ASCII_6, ASCII_B, ASCII_E, '\0'
1517};
1518static const char KW_UTF_16LE[] =
1519{
1520 ASCII_U, ASCII_T, ASCII_F, ASCII_MINUS, ASCII_1, ASCII_6, ASCII_L, ASCII_E, '\0'
1521};
1522
1523static int getEncodingIndex(const char *name)
1524{
1525 static const char *encodingNames[] =
1526 {
1527 KW_ISO_8859_1,
1528 KW_US_ASCII,
1529 KW_UTF_8,
1530 KW_UTF_16,
1531 KW_UTF_16BE,
1532 KW_UTF_16LE,
1533 };
1534 int i;
1535
1536 if (name == 0)
1537 return NO_ENC;
1538 for (i = 0;
1539 i < (int)(sizeof(encodingNames) / sizeof(encodingNames[0]));
1540 i++)
1541 if (streqci(name, encodingNames[i]))
1542 return i;
1543 return UNKNOWN_ENC;
1544}
1545
1546/* For binary compatibility, we store the index of the encoding specified
1547 * at initialization in the isUtf16 member. */
1548
1549#define INIT_ENC_INDEX(enc) ((int)(enc)->initEnc.isUtf16)
1550#define SET_INIT_ENC_INDEX(enc, i) ((enc)->initEnc.isUtf16 = (char)i)
1551
1552/* This is what detects the encoding.
1553 * encodingTable maps from encoding indices to encodings;
1554 * INIT_ENC_INDEX(enc) is the index of the external (protocol) specified encoding;
1555 * state is XML_CONTENT_STATE if we're parsing an external text entity,
1556 * and XML_PROLOG_STATE otherwise.
1557 */
1558
1559
1560static int EXPATENTRY initScan(const ENCODING ** encodingTable,
1561 const INIT_ENCODING * enc,
1562 int state,
1563 const char *ptr,
1564 const char *end,
1565 const char **nextTokPtr)
1566{
1567 const ENCODING **encPtr;
1568
1569 if (ptr == end)
1570 return XML_TOK_NONE;
1571 encPtr = enc->encPtr;
1572 if (ptr + 1 == end)
1573 {
1574 /* only a single byte available for auto-detection */
1575#ifndef XML_DTD /* FIXME */
1576 /* a well-formed document entity must have more than one byte */
1577 if (state != XML_CONTENT_STATE)
1578 return XML_TOK_PARTIAL;
1579#endif
1580 /* so we're parsing an external text entity... */
1581 /* if UTF-16 was externally specified, then we need at least 2 bytes */
1582 switch (INIT_ENC_INDEX(enc))
1583 {
1584 case UTF_16_ENC:
1585 case UTF_16LE_ENC:
1586 case UTF_16BE_ENC:
1587 return XML_TOK_PARTIAL;
1588 }
1589 switch ((unsigned char)*ptr)
1590 {
1591 case 0xFE:
1592 case 0xFF:
1593 case 0xEF: /* possibly first byte of UTF-8 BOM */
1594 if (INIT_ENC_INDEX(enc) == ISO_8859_1_ENC
1595 && state == XML_CONTENT_STATE)
1596 break;
1597 /* fall through */
1598 case 0x00:
1599 case 0x3C:
1600 return XML_TOK_PARTIAL;
1601 }
1602 }
1603 else
1604 {
1605 switch (((unsigned char)ptr[0] << 8) | (unsigned char)ptr[1])
1606 {
1607 case 0xFEFF:
1608 if (INIT_ENC_INDEX(enc) == ISO_8859_1_ENC
1609 && state == XML_CONTENT_STATE)
1610 break;
1611 *nextTokPtr = ptr + 2;
1612 *encPtr = encodingTable[UTF_16BE_ENC];
1613 return XML_TOK_BOM;
1614 /* 00 3C is handled in the default case */
1615 case 0x3C00:
1616 if ((INIT_ENC_INDEX(enc) == UTF_16BE_ENC
1617 || INIT_ENC_INDEX(enc) == UTF_16_ENC)
1618 && state == XML_CONTENT_STATE)
1619 break;
1620 *encPtr = encodingTable[UTF_16LE_ENC];
1621 return XmlTok(*encPtr, state, ptr, end, nextTokPtr);
1622 case 0xFFFE:
1623 if (INIT_ENC_INDEX(enc) == ISO_8859_1_ENC
1624 && state == XML_CONTENT_STATE)
1625 break;
1626 *nextTokPtr = ptr + 2;
1627 *encPtr = encodingTable[UTF_16LE_ENC];
1628 return XML_TOK_BOM;
1629 case 0xEFBB:
1630 /* Maybe a UTF-8 BOM (EF BB BF) */
1631 /* If there's an explicitly specified (external) encoding
1632 * of ISO-8859-1 or some flavour of UTF-16
1633 * and this is an external text entity,
1634 * don't look for the BOM,
1635 * because it might be a legal data. */
1636 if (state == XML_CONTENT_STATE)
1637 {
1638 int e = INIT_ENC_INDEX(enc);
1639
1640 if (e == ISO_8859_1_ENC || e == UTF_16BE_ENC || e == UTF_16LE_ENC || e == UTF_16_ENC)
1641 break;
1642 }
1643 if (ptr + 2 == end)
1644 return XML_TOK_PARTIAL;
1645 if ((unsigned char)ptr[2] == 0xBF)
1646 {
1647 *nextTokPtr = ptr + 3;
1648 *encPtr = encodingTable[UTF_8_ENC];
1649 return XML_TOK_BOM;
1650 }
1651 break;
1652 default:
1653 if (ptr[0] == '\0')
1654 {
1655 /* 0 isn't a legal data character. Furthermore a document entity can only
1656 * start with ASCII characters. So the only way this can fail to be big-endian
1657 * UTF-16 if it it's an external parsed general entity that's labelled as
1658 * UTF-16LE. */
1659 if (state == XML_CONTENT_STATE && INIT_ENC_INDEX(enc) == UTF_16LE_ENC)
1660 break;
1661 *encPtr = encodingTable[UTF_16BE_ENC];
1662 return XmlTok(*encPtr, state, ptr, end, nextTokPtr);
1663 }
1664 else if (ptr[1] == '\0')
1665 {
1666 /* We could recover here in the case:
1667 * - parsing an external entity
1668 * - second byte is 0
1669 * - no externally specified encoding
1670 * - no encoding declaration
1671 * by assuming UTF-16LE. But we don't, because this would mean when
1672 * presented just with a single byte, we couldn't reliably determine
1673 * whether we needed further bytes. */
1674 if (state == XML_CONTENT_STATE)
1675 break;
1676 *encPtr = encodingTable[UTF_16LE_ENC];
1677 return XmlTok(*encPtr, state, ptr, end, nextTokPtr);
1678 }
1679 break;
1680 }
1681 }
1682 *encPtr = encodingTable[INIT_ENC_INDEX(enc)];
1683 return XmlTok(*encPtr, state, ptr, end, nextTokPtr);
1684}
1685
1686
1687#define NS(x) x
1688#define ns(x) x
1689#include "xmltok_ns.c"
1690#undef NS
1691#undef ns
1692
1693#ifdef XML_NS
1694
1695#define NS(x) x ## NS
1696#define ns(x) x ## _ns
1697
1698#include "xmltok_ns.c"
1699
1700#undef NS
1701#undef ns
1702
1703ENCODING * XmlInitUnknownEncodingNS(void *mem,
1704 int *table,
1705 int (* EXPATENTRY convert) (void *userData, const char *p),
1706 void *userData)
1707{
1708 ENCODING *enc = XmlInitUnknownEncoding(mem, table, convert, userData);
1709
1710 if (enc)
1711 ((struct normal_encoding *)enc)->type[ASCII_COLON] = BT_COLON;
1712 return enc;
1713}
1714
1715#endif /* XML_NS */
Note: See TracBrowser for help on using the repository browser.