1 /* Copyright (c) 1998, 1999 Thai Open Source Software Center Ltd
2 See the file COPYING for copying permission.
9 #elif defined(MACOS_CLASSIC)
10 #include "macconfig.h"
11 #elif defined(__amigaos__)
12 #include "amigaconfig.h"
13 #elif defined(__WATCOMC__)
14 #include "watcomconfig.h"
16 #ifdef HAVE_EXPAT_CONFIG_H
17 #include <expat_config.h>
19 #endif /* ndef WIN32 */
21 #include "expat_external.h"
27 #define IGNORE_SECTION_TOK_VTABLE , PREFIX(ignoreSectionTok)
29 #define IGNORE_SECTION_TOK_VTABLE /* as nothing */
33 { PREFIX(prologTok), PREFIX(contentTok), \
34 PREFIX(cdataSectionTok) IGNORE_SECTION_TOK_VTABLE }, \
35 { PREFIX(attributeValueTok), PREFIX(entityValueTok) }, \
37 PREFIX(nameMatchesAscii), \
41 PREFIX(charRefNumber), \
42 PREFIX(predefinedEntityName), \
43 PREFIX(updatePosition), \
46 #define VTABLE VTABLE1, PREFIX(toUtf8), PREFIX(toUtf16)
48 #define UCS2_GET_NAMING(pages, hi, lo) \
49 (namingBitmap[(pages[hi] << 3) + ((lo) >> 5)] & (1u << ((lo) & 0x1F)))
51 /* A 2 byte UTF-8 representation splits the characters 11 bits between
52 the bottom 5 and 6 bits of the bytes. We need 8 bits to index into
53 pages, 3 bits to add to that index and 5 bits to generate the mask.
55 #define UTF8_GET_NAMING2(pages, byte) \
56 (namingBitmap[((pages)[(((byte)[0]) >> 2) & 7] << 3) \
57 + ((((byte)[0]) & 3) << 1) \
58 + ((((byte)[1]) >> 5) & 1)] \
59 & (1u << (((byte)[1]) & 0x1F)))
61 /* A 3 byte UTF-8 representation splits the characters 16 bits between
62 the bottom 4, 6 and 6 bits of the bytes. We need 8 bits to index
63 into pages, 3 bits to add to that index and 5 bits to generate the
66 #define UTF8_GET_NAMING3(pages, byte) \
67 (namingBitmap[((pages)[((((byte)[0]) & 0xF) << 4) \
68 + ((((byte)[1]) >> 2) & 0xF)] \
70 + ((((byte)[1]) & 3) << 1) \
71 + ((((byte)[2]) >> 5) & 1)] \
72 & (1u << (((byte)[2]) & 0x1F)))
74 #define UTF8_GET_NAMING(pages, p, n) \
76 ? UTF8_GET_NAMING2(pages, (const unsigned char *)(p)) \
78 ? UTF8_GET_NAMING3(pages, (const unsigned char *)(p)) \
81 /* Detection of invalid UTF-8 sequences is based on Table 3.1B
82 of Unicode 3.2: http://www.unicode.org/unicode/reports/tr28/
83 with the additional restriction of not allowing the Unicode
84 code points 0xFFFF and 0xFFFE (sequences EF,BF,BF and EF,BF,BE).
85 Implementation details:
86 (A & 0x80) == 0 means A < 0x80
88 (A & 0xC0) == 0xC0 means A > 0xBF
91 #define UTF8_INVALID2(p) \
92 ((*p) < 0xC2 || ((p)[1] & 0x80) == 0 || ((p)[1] & 0xC0) == 0xC0)
94 #define UTF8_INVALID3(p) \
95 (((p)[2] & 0x80) == 0 \
97 ((*p) == 0xEF && (p)[1] == 0xBF \
101 ((p)[2] & 0xC0) == 0xC0) \
105 (p)[1] < 0xA0 || ((p)[1] & 0xC0) == 0xC0 \
107 ((p)[1] & 0x80) == 0 \
109 ((*p) == 0xED ? (p)[1] > 0x9F : ((p)[1] & 0xC0) == 0xC0)))
111 #define UTF8_INVALID4(p) \
112 (((p)[3] & 0x80) == 0 || ((p)[3] & 0xC0) == 0xC0 \
114 ((p)[2] & 0x80) == 0 || ((p)[2] & 0xC0) == 0xC0 \
118 (p)[1] < 0x90 || ((p)[1] & 0xC0) == 0xC0 \
120 ((p)[1] & 0x80) == 0 \
122 ((*p) == 0xF4 ? (p)[1] > 0x8F : ((p)[1] & 0xC0) == 0xC0)))
124 static int PTRFASTCALL
125 isNever(const ENCODING *UNUSED_P(enc), const char *UNUSED_P(p))
130 static int PTRFASTCALL
131 utf8_isName2(const ENCODING *UNUSED_P(enc), const char *p)
133 return UTF8_GET_NAMING2(namePages, (const unsigned char *)p);
136 static int PTRFASTCALL
137 utf8_isName3(const ENCODING *UNUSED_P(enc), const char *p)
139 return UTF8_GET_NAMING3(namePages, (const unsigned char *)p);
142 #define utf8_isName4 isNever
144 static int PTRFASTCALL
145 utf8_isNmstrt2(const ENCODING *UNUSED_P(enc), const char *p)
147 return UTF8_GET_NAMING2(nmstrtPages, (const unsigned char *)p);
150 static int PTRFASTCALL
151 utf8_isNmstrt3(const ENCODING *UNUSED_P(enc), const char *p)
153 return UTF8_GET_NAMING3(nmstrtPages, (const unsigned char *)p);
156 #define utf8_isNmstrt4 isNever
158 static int PTRFASTCALL
159 utf8_isInvalid2(const ENCODING *UNUSED_P(enc), const char *p)
161 return UTF8_INVALID2((const unsigned char *)p);
164 static int PTRFASTCALL
165 utf8_isInvalid3(const ENCODING *UNUSED_P(enc), const char *p)
167 return UTF8_INVALID3((const unsigned char *)p);
170 static int PTRFASTCALL
171 utf8_isInvalid4(const ENCODING *UNUSED_P(enc), const char *p)
173 return UTF8_INVALID4((const unsigned char *)p);
176 struct normal_encoding {
178 unsigned char type[256];
180 int (PTRFASTCALL *byteType)(const ENCODING *, const char *);
181 int (PTRFASTCALL *isNameMin)(const ENCODING *, const char *);
182 int (PTRFASTCALL *isNmstrtMin)(const ENCODING *, const char *);
183 int (PTRFASTCALL *byteToAscii)(const ENCODING *, const char *);
184 int (PTRCALL *charMatches)(const ENCODING *, const char *, int);
185 #endif /* XML_MIN_SIZE */
186 int (PTRFASTCALL *isName2)(const ENCODING *, const char *);
187 int (PTRFASTCALL *isName3)(const ENCODING *, const char *);
188 int (PTRFASTCALL *isName4)(const ENCODING *, const char *);
189 int (PTRFASTCALL *isNmstrt2)(const ENCODING *, const char *);
190 int (PTRFASTCALL *isNmstrt3)(const ENCODING *, const char *);
191 int (PTRFASTCALL *isNmstrt4)(const ENCODING *, const char *);
192 int (PTRFASTCALL *isInvalid2)(const ENCODING *, const char *);
193 int (PTRFASTCALL *isInvalid3)(const ENCODING *, const char *);
194 int (PTRFASTCALL *isInvalid4)(const ENCODING *, const char *);
197 #define AS_NORMAL_ENCODING(enc) ((const struct normal_encoding *) (enc))
201 #define STANDARD_VTABLE(E) \
210 #define STANDARD_VTABLE(E) /* as nothing */
214 #define NORMAL_VTABLE(E) \
225 #define NULL_VTABLE \
226 /* isName2 */ NULL, \
227 /* isName3 */ NULL, \
228 /* isName4 */ NULL, \
229 /* isNmstrt2 */ NULL, \
230 /* isNmstrt3 */ NULL, \
231 /* isNmstrt4 */ NULL, \
232 /* isInvalid2 */ NULL, \
233 /* isInvalid3 */ NULL, \
234 /* isInvalid4 */ NULL
236 static int FASTCALL checkCharRefNumber(int);
238 #include "xmltok_impl.h"
242 #define sb_isNameMin isNever
243 #define sb_isNmstrtMin isNever
247 #define MINBPC(enc) ((enc)->minBytesPerChar)
249 /* minimum bytes per character */
250 #define MINBPC(enc) 1
253 #define SB_BYTE_TYPE(enc, p) \
254 (((struct normal_encoding *)(enc))->type[(unsigned char)*(p)])
257 static int PTRFASTCALL
258 sb_byteType(const ENCODING *enc, const char *p)
260 return SB_BYTE_TYPE(enc, p);
262 #define BYTE_TYPE(enc, p) \
263 (AS_NORMAL_ENCODING(enc)->byteType(enc, p))
265 #define BYTE_TYPE(enc, p) SB_BYTE_TYPE(enc, p)
269 #define BYTE_TO_ASCII(enc, p) \
270 (AS_NORMAL_ENCODING(enc)->byteToAscii(enc, p))
271 static int PTRFASTCALL
272 sb_byteToAscii(const ENCODING *enc, const char *p)
277 #define BYTE_TO_ASCII(enc, p) (*(p))
280 #define IS_NAME_CHAR(enc, p, n) \
281 (AS_NORMAL_ENCODING(enc)->isName ## n(enc, p))
282 #define IS_NMSTRT_CHAR(enc, p, n) \
283 (AS_NORMAL_ENCODING(enc)->isNmstrt ## n(enc, p))
284 #define IS_INVALID_CHAR(enc, p, n) \
285 (AS_NORMAL_ENCODING(enc)->isInvalid ## n(enc, p))
288 #define IS_NAME_CHAR_MINBPC(enc, p) \
289 (AS_NORMAL_ENCODING(enc)->isNameMin(enc, p))
290 #define IS_NMSTRT_CHAR_MINBPC(enc, p) \
291 (AS_NORMAL_ENCODING(enc)->isNmstrtMin(enc, p))
293 #define IS_NAME_CHAR_MINBPC(enc, p) (0)
294 #define IS_NMSTRT_CHAR_MINBPC(enc, p) (0)
298 #define CHAR_MATCHES(enc, p, c) \
299 (AS_NORMAL_ENCODING(enc)->charMatches(enc, p, c))
301 sb_charMatches(const ENCODING *enc, const char *p, int c)
306 /* c is an ASCII character */
307 #define CHAR_MATCHES(enc, p, c) (*(p) == c)
310 #define PREFIX(ident) normal_ ## ident
311 #define XML_TOK_IMPL_C
312 #include "xmltok_impl.c"
313 #undef XML_TOK_IMPL_C
320 #undef IS_NAME_CHAR_MINBPC
321 #undef IS_NMSTRT_CHAR
322 #undef IS_NMSTRT_CHAR_MINBPC
323 #undef IS_INVALID_CHAR
325 enum { /* UTF8_cvalN is value of masked first byte of N byte sequence */
333 align_limit_to_full_utf8_characters(const char * from, const char ** fromLimRef)
335 const char * fromLim = *fromLimRef;
337 for (; fromLim > from; fromLim--, walked++) {
338 const unsigned char prev = (unsigned char)fromLim[-1];
339 if ((prev & 0xf8u) == 0xf0u) { /* 4-byte character, lead by 0b11110xxx byte */
340 if (walked + 1 >= 4) {
346 } else if ((prev & 0xf0u) == 0xe0u) { /* 3-byte character, lead by 0b1110xxxx byte */
347 if (walked + 1 >= 3) {
353 } else if ((prev & 0xe0u) == 0xc0u) { /* 2-byte character, lead by 0b110xxxxx byte */
354 if (walked + 1 >= 2) {
360 } else if ((prev & 0x80u) == 0x00u) { /* 1-byte character, matching 0b0xxxxxxx */
364 *fromLimRef = fromLim;
367 static enum XML_Convert_Result PTRCALL
368 utf8_toUtf8(const ENCODING *UNUSED_P(enc),
369 const char **fromP, const char *fromLim,
370 char **toP, const char *toLim)
372 enum XML_Convert_Result res = XML_CONVERT_COMPLETED;
375 if (fromLim - *fromP > toLim - *toP) {
376 /* Avoid copying partial characters. */
377 res = XML_CONVERT_OUTPUT_EXHAUSTED;
378 fromLim = *fromP + (toLim - *toP);
379 align_limit_to_full_utf8_characters(*fromP, &fromLim);
381 for (to = *toP, from = *fromP; (from < fromLim) && (to < toLim); from++, to++)
386 if ((to == toLim) && (from < fromLim))
387 return XML_CONVERT_OUTPUT_EXHAUSTED;
392 static enum XML_Convert_Result PTRCALL
393 utf8_toUtf16(const ENCODING *enc,
394 const char **fromP, const char *fromLim,
395 unsigned short **toP, const unsigned short *toLim)
397 enum XML_Convert_Result res = XML_CONVERT_COMPLETED;
398 unsigned short *to = *toP;
399 const char *from = *fromP;
400 while (from < fromLim && to < toLim) {
401 switch (((struct normal_encoding *)enc)->type[(unsigned char)*from]) {
403 if (fromLim - from < 2) {
404 res = XML_CONVERT_INPUT_INCOMPLETE;
407 *to++ = (unsigned short)(((from[0] & 0x1f) << 6) | (from[1] & 0x3f));
411 if (fromLim - from < 3) {
412 res = XML_CONVERT_INPUT_INCOMPLETE;
415 *to++ = (unsigned short)(((from[0] & 0xf) << 12)
416 | ((from[1] & 0x3f) << 6) | (from[2] & 0x3f));
422 if (toLim - to < 2) {
423 res = XML_CONVERT_OUTPUT_EXHAUSTED;
426 if (fromLim - from < 4) {
427 res = XML_CONVERT_INPUT_INCOMPLETE;
430 n = ((from[0] & 0x7) << 18) | ((from[1] & 0x3f) << 12)
431 | ((from[2] & 0x3f) << 6) | (from[3] & 0x3f);
433 to[0] = (unsigned short)((n >> 10) | 0xD800);
434 to[1] = (unsigned short)((n & 0x3FF) | 0xDC00);
451 static const struct normal_encoding utf8_encoding_ns = {
452 { VTABLE1, utf8_toUtf8, utf8_toUtf16, 1, 1, 0 },
454 #include "asciitab.h"
457 STANDARD_VTABLE(sb_) NORMAL_VTABLE(utf8_)
461 static const struct normal_encoding utf8_encoding = {
462 { VTABLE1, utf8_toUtf8, utf8_toUtf16, 1, 1, 0 },
464 #define BT_COLON BT_NMSTRT
465 #include "asciitab.h"
469 STANDARD_VTABLE(sb_) NORMAL_VTABLE(utf8_)
474 static const struct normal_encoding internal_utf8_encoding_ns = {
475 { VTABLE1, utf8_toUtf8, utf8_toUtf16, 1, 1, 0 },
477 #include "iasciitab.h"
480 STANDARD_VTABLE(sb_) NORMAL_VTABLE(utf8_)
485 static const struct normal_encoding internal_utf8_encoding = {
486 { VTABLE1, utf8_toUtf8, utf8_toUtf16, 1, 1, 0 },
488 #define BT_COLON BT_NMSTRT
489 #include "iasciitab.h"
493 STANDARD_VTABLE(sb_) NORMAL_VTABLE(utf8_)
496 static enum XML_Convert_Result PTRCALL
497 latin1_toUtf8(const ENCODING *UNUSED_P(enc),
498 const char **fromP, const char *fromLim,
499 char **toP, const char *toLim)
503 if (*fromP == fromLim)
504 return XML_CONVERT_COMPLETED;
505 c = (unsigned char)**fromP;
507 if (toLim - *toP < 2)
508 return XML_CONVERT_OUTPUT_EXHAUSTED;
509 *(*toP)++ = (char)((c >> 6) | UTF8_cval2);
510 *(*toP)++ = (char)((c & 0x3f) | 0x80);
515 return XML_CONVERT_OUTPUT_EXHAUSTED;
516 *(*toP)++ = *(*fromP)++;
521 static enum XML_Convert_Result PTRCALL
522 latin1_toUtf16(const ENCODING *UNUSED_P(enc),
523 const char **fromP, const char *fromLim,
524 unsigned short **toP, const unsigned short *toLim)
526 while (*fromP < fromLim && *toP < toLim)
527 *(*toP)++ = (unsigned char)*(*fromP)++;
529 if ((*toP == toLim) && (*fromP < fromLim))
530 return XML_CONVERT_OUTPUT_EXHAUSTED;
532 return XML_CONVERT_COMPLETED;
537 static const struct normal_encoding latin1_encoding_ns = {
538 { VTABLE1, latin1_toUtf8, latin1_toUtf16, 1, 0, 0 },
540 #include "asciitab.h"
541 #include "latin1tab.h"
543 STANDARD_VTABLE(sb_) NULL_VTABLE
548 static const struct normal_encoding latin1_encoding = {
549 { VTABLE1, latin1_toUtf8, latin1_toUtf16, 1, 0, 0 },
551 #define BT_COLON BT_NMSTRT
552 #include "asciitab.h"
554 #include "latin1tab.h"
556 STANDARD_VTABLE(sb_) NULL_VTABLE
559 static enum XML_Convert_Result PTRCALL
560 ascii_toUtf8(const ENCODING *UNUSED_P(enc),
561 const char **fromP, const char *fromLim,
562 char **toP, const char *toLim)
564 while (*fromP < fromLim && *toP < toLim)
565 *(*toP)++ = *(*fromP)++;
567 if ((*toP == toLim) && (*fromP < fromLim))
568 return XML_CONVERT_OUTPUT_EXHAUSTED;
570 return XML_CONVERT_COMPLETED;
575 static const struct normal_encoding ascii_encoding_ns = {
576 { VTABLE1, ascii_toUtf8, latin1_toUtf16, 1, 1, 0 },
578 #include "asciitab.h"
581 STANDARD_VTABLE(sb_) NULL_VTABLE
586 static const struct normal_encoding ascii_encoding = {
587 { VTABLE1, ascii_toUtf8, latin1_toUtf16, 1, 1, 0 },
589 #define BT_COLON BT_NMSTRT
590 #include "asciitab.h"
594 STANDARD_VTABLE(sb_) NULL_VTABLE
597 static int PTRFASTCALL
598 unicode_byte_type(char hi, char lo)
600 switch ((unsigned char)hi) {
601 case 0xD8: case 0xD9: case 0xDA: case 0xDB:
603 case 0xDC: case 0xDD: case 0xDE: case 0xDF:
606 switch ((unsigned char)lo) {
616 #define DEFINE_UTF16_TO_UTF8(E) \
617 static enum XML_Convert_Result PTRCALL \
618 E ## toUtf8(const ENCODING *UNUSED_P(enc), \
619 const char **fromP, const char *fromLim, \
620 char **toP, const char *toLim) \
622 const char *from = *fromP; \
623 fromLim = from + (((fromLim - from) >> 1) << 1); /* shrink to even */ \
624 for (; from < fromLim; from += 2) { \
627 unsigned char lo = GET_LO(from); \
628 unsigned char hi = GET_HI(from); \
632 if (*toP == toLim) { \
634 return XML_CONVERT_OUTPUT_EXHAUSTED; \
640 case 0x1: case 0x2: case 0x3: \
641 case 0x4: case 0x5: case 0x6: case 0x7: \
642 if (toLim - *toP < 2) { \
644 return XML_CONVERT_OUTPUT_EXHAUSTED; \
646 *(*toP)++ = ((lo >> 6) | (hi << 2) | UTF8_cval2); \
647 *(*toP)++ = ((lo & 0x3f) | 0x80); \
650 if (toLim - *toP < 3) { \
652 return XML_CONVERT_OUTPUT_EXHAUSTED; \
654 /* 16 bits divided 4, 6, 6 amongst 3 bytes */ \
655 *(*toP)++ = ((hi >> 4) | UTF8_cval3); \
656 *(*toP)++ = (((hi & 0xf) << 2) | (lo >> 6) | 0x80); \
657 *(*toP)++ = ((lo & 0x3f) | 0x80); \
659 case 0xD8: case 0xD9: case 0xDA: case 0xDB: \
660 if (toLim - *toP < 4) { \
662 return XML_CONVERT_OUTPUT_EXHAUSTED; \
664 if (fromLim - from < 4) { \
666 return XML_CONVERT_INPUT_INCOMPLETE; \
668 plane = (((hi & 0x3) << 2) | ((lo >> 6) & 0x3)) + 1; \
669 *(*toP)++ = ((plane >> 2) | UTF8_cval4); \
670 *(*toP)++ = (((lo >> 2) & 0xF) | ((plane & 0x3) << 4) | 0x80); \
672 lo2 = GET_LO(from); \
673 *(*toP)++ = (((lo & 0x3) << 4) \
674 | ((GET_HI(from) & 0x3) << 2) \
677 *(*toP)++ = ((lo2 & 0x3f) | 0x80); \
682 if (from < fromLim) \
683 return XML_CONVERT_INPUT_INCOMPLETE; \
685 return XML_CONVERT_COMPLETED; \
688 #define DEFINE_UTF16_TO_UTF16(E) \
689 static enum XML_Convert_Result PTRCALL \
690 E ## toUtf16(const ENCODING *UNUSED_P(enc), \
691 const char **fromP, const char *fromLim, \
692 unsigned short **toP, const unsigned short *toLim) \
694 enum XML_Convert_Result res = XML_CONVERT_COMPLETED; \
695 fromLim = *fromP + (((fromLim - *fromP) >> 1) << 1); /* shrink to even */ \
696 /* Avoid copying first half only of surrogate */ \
697 if (fromLim - *fromP > ((toLim - *toP) << 1) \
698 && (GET_HI(fromLim - 2) & 0xF8) == 0xD8) { \
700 res = XML_CONVERT_INPUT_INCOMPLETE; \
702 for (; *fromP < fromLim && *toP < toLim; *fromP += 2) \
703 *(*toP)++ = (GET_HI(*fromP) << 8) | GET_LO(*fromP); \
704 if ((*toP == toLim) && (*fromP < fromLim)) \
705 return XML_CONVERT_OUTPUT_EXHAUSTED; \
710 #define SET2(ptr, ch) \
711 (((ptr)[0] = ((ch) & 0xff)), ((ptr)[1] = ((ch) >> 8)))
712 #define GET_LO(ptr) ((unsigned char)(ptr)[0])
713 #define GET_HI(ptr) ((unsigned char)(ptr)[1])
715 DEFINE_UTF16_TO_UTF8(little2_)
716 DEFINE_UTF16_TO_UTF16(little2_)
722 #define SET2(ptr, ch) \
723 (((ptr)[0] = ((ch) >> 8)), ((ptr)[1] = ((ch) & 0xFF)))
724 #define GET_LO(ptr) ((unsigned char)(ptr)[1])
725 #define GET_HI(ptr) ((unsigned char)(ptr)[0])
727 DEFINE_UTF16_TO_UTF8(big2_)
728 DEFINE_UTF16_TO_UTF16(big2_)
734 #define LITTLE2_BYTE_TYPE(enc, p) \
736 ? ((struct normal_encoding *)(enc))->type[(unsigned char)*(p)] \
737 : unicode_byte_type((p)[1], (p)[0]))
738 #define LITTLE2_BYTE_TO_ASCII(enc, p) ((p)[1] == 0 ? (p)[0] : -1)
739 #define LITTLE2_CHAR_MATCHES(enc, p, c) ((p)[1] == 0 && (p)[0] == c)
740 #define LITTLE2_IS_NAME_CHAR_MINBPC(enc, p) \
741 UCS2_GET_NAMING(namePages, (unsigned char)p[1], (unsigned char)p[0])
742 #define LITTLE2_IS_NMSTRT_CHAR_MINBPC(enc, p) \
743 UCS2_GET_NAMING(nmstrtPages, (unsigned char)p[1], (unsigned char)p[0])
747 static int PTRFASTCALL
748 little2_byteType(const ENCODING *enc, const char *p)
750 return LITTLE2_BYTE_TYPE(enc, p);
753 static int PTRFASTCALL
754 little2_byteToAscii(const ENCODING *enc, const char *p)
756 return LITTLE2_BYTE_TO_ASCII(enc, p);
760 little2_charMatches(const ENCODING *enc, const char *p, int c)
762 return LITTLE2_CHAR_MATCHES(enc, p, c);
765 static int PTRFASTCALL
766 little2_isNameMin(const ENCODING *enc, const char *p)
768 return LITTLE2_IS_NAME_CHAR_MINBPC(enc, p);
771 static int PTRFASTCALL
772 little2_isNmstrtMin(const ENCODING *enc, const char *p)
774 return LITTLE2_IS_NMSTRT_CHAR_MINBPC(enc, p);
778 #define VTABLE VTABLE1, little2_toUtf8, little2_toUtf16
780 #else /* not XML_MIN_SIZE */
783 #define PREFIX(ident) little2_ ## ident
784 #define MINBPC(enc) 2
785 /* CHAR_MATCHES is guaranteed to have MINBPC bytes available. */
786 #define BYTE_TYPE(enc, p) LITTLE2_BYTE_TYPE(enc, p)
787 #define BYTE_TO_ASCII(enc, p) LITTLE2_BYTE_TO_ASCII(enc, p)
788 #define CHAR_MATCHES(enc, p, c) LITTLE2_CHAR_MATCHES(enc, p, c)
789 #define IS_NAME_CHAR(enc, p, n) 0
790 #define IS_NAME_CHAR_MINBPC(enc, p) LITTLE2_IS_NAME_CHAR_MINBPC(enc, p)
791 #define IS_NMSTRT_CHAR(enc, p, n) (0)
792 #define IS_NMSTRT_CHAR_MINBPC(enc, p) LITTLE2_IS_NMSTRT_CHAR_MINBPC(enc, p)
794 #define XML_TOK_IMPL_C
795 #include "xmltok_impl.c"
796 #undef XML_TOK_IMPL_C
803 #undef IS_NAME_CHAR_MINBPC
804 #undef IS_NMSTRT_CHAR
805 #undef IS_NMSTRT_CHAR_MINBPC
806 #undef IS_INVALID_CHAR
808 #endif /* not XML_MIN_SIZE */
812 static const struct normal_encoding little2_encoding_ns = {
814 #if BYTEORDER == 1234
821 #include "asciitab.h"
822 #include "latin1tab.h"
824 STANDARD_VTABLE(little2_) NULL_VTABLE
829 static const struct normal_encoding little2_encoding = {
831 #if BYTEORDER == 1234
838 #define BT_COLON BT_NMSTRT
839 #include "asciitab.h"
841 #include "latin1tab.h"
843 STANDARD_VTABLE(little2_) NULL_VTABLE
846 #if BYTEORDER != 4321
850 static const struct normal_encoding internal_little2_encoding_ns = {
853 #include "iasciitab.h"
854 #include "latin1tab.h"
856 STANDARD_VTABLE(little2_) NULL_VTABLE
861 static const struct normal_encoding internal_little2_encoding = {
864 #define BT_COLON BT_NMSTRT
865 #include "iasciitab.h"
867 #include "latin1tab.h"
869 STANDARD_VTABLE(little2_) NULL_VTABLE
875 #define BIG2_BYTE_TYPE(enc, p) \
877 ? ((struct normal_encoding *)(enc))->type[(unsigned char)(p)[1]] \
878 : unicode_byte_type((p)[0], (p)[1]))
879 #define BIG2_BYTE_TO_ASCII(enc, p) ((p)[0] == 0 ? (p)[1] : -1)
880 #define BIG2_CHAR_MATCHES(enc, p, c) ((p)[0] == 0 && (p)[1] == c)
881 #define BIG2_IS_NAME_CHAR_MINBPC(enc, p) \
882 UCS2_GET_NAMING(namePages, (unsigned char)p[0], (unsigned char)p[1])
883 #define BIG2_IS_NMSTRT_CHAR_MINBPC(enc, p) \
884 UCS2_GET_NAMING(nmstrtPages, (unsigned char)p[0], (unsigned char)p[1])
888 static int PTRFASTCALL
889 big2_byteType(const ENCODING *enc, const char *p)
891 return BIG2_BYTE_TYPE(enc, p);
894 static int PTRFASTCALL
895 big2_byteToAscii(const ENCODING *enc, const char *p)
897 return BIG2_BYTE_TO_ASCII(enc, p);
901 big2_charMatches(const ENCODING *enc, const char *p, int c)
903 return BIG2_CHAR_MATCHES(enc, p, c);
906 static int PTRFASTCALL
907 big2_isNameMin(const ENCODING *enc, const char *p)
909 return BIG2_IS_NAME_CHAR_MINBPC(enc, p);
912 static int PTRFASTCALL
913 big2_isNmstrtMin(const ENCODING *enc, const char *p)
915 return BIG2_IS_NMSTRT_CHAR_MINBPC(enc, p);
919 #define VTABLE VTABLE1, big2_toUtf8, big2_toUtf16
921 #else /* not XML_MIN_SIZE */
924 #define PREFIX(ident) big2_ ## ident
925 #define MINBPC(enc) 2
926 /* CHAR_MATCHES is guaranteed to have MINBPC bytes available. */
927 #define BYTE_TYPE(enc, p) BIG2_BYTE_TYPE(enc, p)
928 #define BYTE_TO_ASCII(enc, p) BIG2_BYTE_TO_ASCII(enc, p)
929 #define CHAR_MATCHES(enc, p, c) BIG2_CHAR_MATCHES(enc, p, c)
930 #define IS_NAME_CHAR(enc, p, n) 0
931 #define IS_NAME_CHAR_MINBPC(enc, p) BIG2_IS_NAME_CHAR_MINBPC(enc, p)
932 #define IS_NMSTRT_CHAR(enc, p, n) (0)
933 #define IS_NMSTRT_CHAR_MINBPC(enc, p) BIG2_IS_NMSTRT_CHAR_MINBPC(enc, p)
935 #define XML_TOK_IMPL_C
936 #include "xmltok_impl.c"
937 #undef XML_TOK_IMPL_C
944 #undef IS_NAME_CHAR_MINBPC
945 #undef IS_NMSTRT_CHAR
946 #undef IS_NMSTRT_CHAR_MINBPC
947 #undef IS_INVALID_CHAR
949 #endif /* not XML_MIN_SIZE */
953 static const struct normal_encoding big2_encoding_ns = {
955 #if BYTEORDER == 4321
962 #include "asciitab.h"
963 #include "latin1tab.h"
965 STANDARD_VTABLE(big2_) NULL_VTABLE
970 static const struct normal_encoding big2_encoding = {
972 #if BYTEORDER == 4321
979 #define BT_COLON BT_NMSTRT
980 #include "asciitab.h"
982 #include "latin1tab.h"
984 STANDARD_VTABLE(big2_) NULL_VTABLE
987 #if BYTEORDER != 1234
991 static const struct normal_encoding internal_big2_encoding_ns = {
994 #include "iasciitab.h"
995 #include "latin1tab.h"
997 STANDARD_VTABLE(big2_) NULL_VTABLE
1002 static const struct normal_encoding internal_big2_encoding = {
1003 { VTABLE, 2, 0, 1 },
1005 #define BT_COLON BT_NMSTRT
1006 #include "iasciitab.h"
1008 #include "latin1tab.h"
1010 STANDARD_VTABLE(big2_) NULL_VTABLE
1018 streqci(const char *s1, const char *s2)
1023 if (ASCII_a <= c1 && c1 <= ASCII_z)
1024 c1 += ASCII_A - ASCII_a;
1025 if (ASCII_a <= c2 && c2 <= ASCII_z)
1026 c2 += ASCII_A - ASCII_a;
1036 initUpdatePosition(const ENCODING *UNUSED_P(enc), const char *ptr,
1037 const char *end, POSITION *pos)
1039 normal_updatePosition(&utf8_encoding.enc, ptr, end, pos);
1043 toAscii(const ENCODING *enc, const char *ptr, const char *end)
1047 XmlUtf8Convert(enc, &ptr, end, &p, p + 1);
1067 /* Return 1 if there's just optional white space or there's an S
1068 followed by name=val.
1071 parsePseudoAttribute(const ENCODING *enc,
1074 const char **namePtr,
1075 const char **nameEndPtr,
1076 const char **valPtr,
1077 const char **nextTokPtr)
1085 if (!isSpace(toAscii(enc, ptr, end))) {
1090 ptr += enc->minBytesPerChar;
1091 } while (isSpace(toAscii(enc, ptr, end)));
1098 c = toAscii(enc, ptr, end);
1103 if (c == ASCII_EQUALS) {
1110 ptr += enc->minBytesPerChar;
1111 } while (isSpace(c = toAscii(enc, ptr, end)));
1112 if (c != ASCII_EQUALS) {
1118 ptr += enc->minBytesPerChar;
1120 if (ptr == *namePtr) {
1124 ptr += enc->minBytesPerChar;
1125 c = toAscii(enc, ptr, end);
1126 while (isSpace(c)) {
1127 ptr += enc->minBytesPerChar;
1128 c = toAscii(enc, ptr, end);
1130 if (c != ASCII_QUOT && c != ASCII_APOS) {
1135 ptr += enc->minBytesPerChar;
1137 for (;; ptr += enc->minBytesPerChar) {
1138 c = toAscii(enc, ptr, end);
1141 if (!(ASCII_a <= c && c <= ASCII_z)
1142 && !(ASCII_A <= c && c <= ASCII_Z)
1143 && !(ASCII_0 <= c && c <= ASCII_9)
1144 && c != ASCII_PERIOD
1146 && c != ASCII_UNDERSCORE) {
1151 *nextTokPtr = ptr + enc->minBytesPerChar;
1155 static const char KW_version[] = {
1156 ASCII_v, ASCII_e, ASCII_r, ASCII_s, ASCII_i, ASCII_o, ASCII_n, '\0'
1159 static const char KW_encoding[] = {
1160 ASCII_e, ASCII_n, ASCII_c, ASCII_o, ASCII_d, ASCII_i, ASCII_n, ASCII_g, '\0'
1163 static const char KW_standalone[] = {
1164 ASCII_s, ASCII_t, ASCII_a, ASCII_n, ASCII_d, ASCII_a, ASCII_l, ASCII_o,
1165 ASCII_n, ASCII_e, '\0'
1168 static const char KW_yes[] = {
1169 ASCII_y, ASCII_e, ASCII_s, '\0'
1172 static const char KW_no[] = {
1173 ASCII_n, ASCII_o, '\0'
1177 doParseXmlDecl(const ENCODING *(*encodingFinder)(const ENCODING *,
1180 int isGeneralTextEntity,
1181 const ENCODING *enc,
1184 const char **badPtr,
1185 const char **versionPtr,
1186 const char **versionEndPtr,
1187 const char **encodingName,
1188 const ENCODING **encoding,
1191 const char *val = NULL;
1192 const char *name = NULL;
1193 const char *nameEnd = NULL;
1194 ptr += 5 * enc->minBytesPerChar;
1195 end -= 2 * enc->minBytesPerChar;
1196 if (!parsePseudoAttribute(enc, ptr, end, &name, &nameEnd, &val, &ptr)
1201 if (!XmlNameMatchesAscii(enc, name, nameEnd, KW_version)) {
1202 if (!isGeneralTextEntity) {
1211 *versionEndPtr = ptr;
1212 if (!parsePseudoAttribute(enc, ptr, end, &name, &nameEnd, &val, &ptr)) {
1217 if (isGeneralTextEntity) {
1218 /* a TextDecl must have an EncodingDecl */
1225 if (XmlNameMatchesAscii(enc, name, nameEnd, KW_encoding)) {
1226 int c = toAscii(enc, val, end);
1227 if (!(ASCII_a <= c && c <= ASCII_z) && !(ASCII_A <= c && c <= ASCII_Z)) {
1232 *encodingName = val;
1234 *encoding = encodingFinder(enc, val, ptr - enc->minBytesPerChar);
1235 if (!parsePseudoAttribute(enc, ptr, end, &name, &nameEnd, &val, &ptr)) {
1242 if (!XmlNameMatchesAscii(enc, name, nameEnd, KW_standalone)
1243 || isGeneralTextEntity) {
1247 if (XmlNameMatchesAscii(enc, val, ptr - enc->minBytesPerChar, KW_yes)) {
1251 else if (XmlNameMatchesAscii(enc, val, ptr - enc->minBytesPerChar, KW_no)) {
1259 while (isSpace(toAscii(enc, ptr, end)))
1260 ptr += enc->minBytesPerChar;
1269 checkCharRefNumber(int result)
1271 switch (result >> 8) {
1272 case 0xD8: case 0xD9: case 0xDA: case 0xDB:
1273 case 0xDC: case 0xDD: case 0xDE: case 0xDF:
1276 if (latin1_encoding.type[result] == BT_NONXML)
1280 if (result == 0xFFFE || result == 0xFFFF)
1288 XmlUtf8Encode(int c, char *buf)
1291 /* minN is minimum legal resulting value for N byte sequence */
1300 buf[0] = (char)(c | UTF8_cval1);
1304 buf[0] = (char)((c >> 6) | UTF8_cval2);
1305 buf[1] = (char)((c & 0x3f) | 0x80);
1309 buf[0] = (char)((c >> 12) | UTF8_cval3);
1310 buf[1] = (char)(((c >> 6) & 0x3f) | 0x80);
1311 buf[2] = (char)((c & 0x3f) | 0x80);
1315 buf[0] = (char)((c >> 18) | UTF8_cval4);
1316 buf[1] = (char)(((c >> 12) & 0x3f) | 0x80);
1317 buf[2] = (char)(((c >> 6) & 0x3f) | 0x80);
1318 buf[3] = (char)((c & 0x3f) | 0x80);
1325 XmlUtf16Encode(int charNum, unsigned short *buf)
1329 if (charNum < 0x10000) {
1330 buf[0] = (unsigned short)charNum;
1333 if (charNum < 0x110000) {
1335 buf[0] = (unsigned short)((charNum >> 10) + 0xD800);
1336 buf[1] = (unsigned short)((charNum & 0x3FF) + 0xDC00);
1342 struct unknown_encoding {
1343 struct normal_encoding normal;
1346 unsigned short utf16[256];
1350 #define AS_UNKNOWN_ENCODING(enc) ((const struct unknown_encoding *) (enc))
1353 XmlSizeOfUnknownEncoding(void)
1355 return sizeof(struct unknown_encoding);
1358 static int PTRFASTCALL
1359 unknown_isName(const ENCODING *enc, const char *p)
1361 const struct unknown_encoding *uenc = AS_UNKNOWN_ENCODING(enc);
1362 int c = uenc->convert(uenc->userData, p);
1365 return UCS2_GET_NAMING(namePages, c >> 8, c & 0xFF);
1368 static int PTRFASTCALL
1369 unknown_isNmstrt(const ENCODING *enc, const char *p)
1371 const struct unknown_encoding *uenc = AS_UNKNOWN_ENCODING(enc);
1372 int c = uenc->convert(uenc->userData, p);
1375 return UCS2_GET_NAMING(nmstrtPages, c >> 8, c & 0xFF);
1378 static int PTRFASTCALL
1379 unknown_isInvalid(const ENCODING *enc, const char *p)
1381 const struct unknown_encoding *uenc = AS_UNKNOWN_ENCODING(enc);
1382 int c = uenc->convert(uenc->userData, p);
1383 return (c & ~0xFFFF) || checkCharRefNumber(c) < 0;
1386 static enum XML_Convert_Result PTRCALL
1387 unknown_toUtf8(const ENCODING *enc,
1388 const char **fromP, const char *fromLim,
1389 char **toP, const char *toLim)
1391 const struct unknown_encoding *uenc = AS_UNKNOWN_ENCODING(enc);
1392 char buf[XML_UTF8_ENCODE_MAX];
1396 if (*fromP == fromLim)
1397 return XML_CONVERT_COMPLETED;
1398 utf8 = uenc->utf8[(unsigned char)**fromP];
1401 int c = uenc->convert(uenc->userData, *fromP);
1402 n = XmlUtf8Encode(c, buf);
1403 if (n > toLim - *toP)
1404 return XML_CONVERT_OUTPUT_EXHAUSTED;
1406 *fromP += (AS_NORMAL_ENCODING(enc)->type[(unsigned char)**fromP]
1410 if (n > toLim - *toP)
1411 return XML_CONVERT_OUTPUT_EXHAUSTED;
1415 *(*toP)++ = *utf8++;
1420 static enum XML_Convert_Result PTRCALL
1421 unknown_toUtf16(const ENCODING *enc,
1422 const char **fromP, const char *fromLim,
1423 unsigned short **toP, const unsigned short *toLim)
1425 const struct unknown_encoding *uenc = AS_UNKNOWN_ENCODING(enc);
1426 while (*fromP < fromLim && *toP < toLim) {
1427 unsigned short c = uenc->utf16[(unsigned char)**fromP];
1429 c = (unsigned short)
1430 uenc->convert(uenc->userData, *fromP);
1431 *fromP += (AS_NORMAL_ENCODING(enc)->type[(unsigned char)**fromP]
1439 if ((*toP == toLim) && (*fromP < fromLim))
1440 return XML_CONVERT_OUTPUT_EXHAUSTED;
1442 return XML_CONVERT_COMPLETED;
1446 XmlInitUnknownEncoding(void *mem,
1452 struct unknown_encoding *e = (struct unknown_encoding *)mem;
1453 for (i = 0; i < (int)sizeof(struct normal_encoding); i++)
1454 ((char *)mem)[i] = ((char *)&latin1_encoding)[i];
1455 for (i = 0; i < 128; i++)
1456 if (latin1_encoding.type[i] != BT_OTHER
1457 && latin1_encoding.type[i] != BT_NONXML
1460 for (i = 0; i < 256; i++) {
1463 e->normal.type[i] = BT_MALFORM;
1464 /* This shouldn't really get used. */
1465 e->utf16[i] = 0xFFFF;
1472 e->normal.type[i] = (unsigned char)(BT_LEAD2 - (c + 2));
1476 else if (c < 0x80) {
1477 if (latin1_encoding.type[c] != BT_OTHER
1478 && latin1_encoding.type[c] != BT_NONXML
1481 e->normal.type[i] = latin1_encoding.type[c];
1483 e->utf8[i][1] = (char)c;
1484 e->utf16[i] = (unsigned short)(c == 0 ? 0xFFFF : c);
1486 else if (checkCharRefNumber(c) < 0) {
1487 e->normal.type[i] = BT_NONXML;
1488 /* This shouldn't really get used. */
1489 e->utf16[i] = 0xFFFF;
1496 if (UCS2_GET_NAMING(nmstrtPages, c >> 8, c & 0xff))
1497 e->normal.type[i] = BT_NMSTRT;
1498 else if (UCS2_GET_NAMING(namePages, c >> 8, c & 0xff))
1499 e->normal.type[i] = BT_NAME;
1501 e->normal.type[i] = BT_OTHER;
1502 e->utf8[i][0] = (char)XmlUtf8Encode(c, e->utf8[i] + 1);
1503 e->utf16[i] = (unsigned short)c;
1506 e->userData = userData;
1507 e->convert = convert;
1509 e->normal.isName2 = unknown_isName;
1510 e->normal.isName3 = unknown_isName;
1511 e->normal.isName4 = unknown_isName;
1512 e->normal.isNmstrt2 = unknown_isNmstrt;
1513 e->normal.isNmstrt3 = unknown_isNmstrt;
1514 e->normal.isNmstrt4 = unknown_isNmstrt;
1515 e->normal.isInvalid2 = unknown_isInvalid;
1516 e->normal.isInvalid3 = unknown_isInvalid;
1517 e->normal.isInvalid4 = unknown_isInvalid;
1519 e->normal.enc.utf8Convert = unknown_toUtf8;
1520 e->normal.enc.utf16Convert = unknown_toUtf16;
1521 return &(e->normal.enc);
1524 /* If this enumeration is changed, getEncodingIndex and encodings
1525 must also be changed. */
1534 /* must match encodingNames up to here */
1538 static const char KW_ISO_8859_1[] = {
1539 ASCII_I, ASCII_S, ASCII_O, ASCII_MINUS, ASCII_8, ASCII_8, ASCII_5, ASCII_9,
1540 ASCII_MINUS, ASCII_1, '\0'
1542 static const char KW_US_ASCII[] = {
1543 ASCII_U, ASCII_S, ASCII_MINUS, ASCII_A, ASCII_S, ASCII_C, ASCII_I, ASCII_I,
1546 static const char KW_UTF_8[] = {
1547 ASCII_U, ASCII_T, ASCII_F, ASCII_MINUS, ASCII_8, '\0'
1549 static const char KW_UTF_16[] = {
1550 ASCII_U, ASCII_T, ASCII_F, ASCII_MINUS, ASCII_1, ASCII_6, '\0'
1552 static const char KW_UTF_16BE[] = {
1553 ASCII_U, ASCII_T, ASCII_F, ASCII_MINUS, ASCII_1, ASCII_6, ASCII_B, ASCII_E,
1556 static const char KW_UTF_16LE[] = {
1557 ASCII_U, ASCII_T, ASCII_F, ASCII_MINUS, ASCII_1, ASCII_6, ASCII_L, ASCII_E,
1562 getEncodingIndex(const char *name)
1564 static const char * const encodingNames[] = {
1575 for (i = 0; i < (int)(sizeof(encodingNames)/sizeof(encodingNames[0])); i++)
1576 if (streqci(name, encodingNames[i]))
1581 /* For binary compatibility, we store the index of the encoding
1582 specified at initialization in the isUtf16 member.
1585 #define INIT_ENC_INDEX(enc) ((int)(enc)->initEnc.isUtf16)
1586 #define SET_INIT_ENC_INDEX(enc, i) ((enc)->initEnc.isUtf16 = (char)i)
1588 /* This is what detects the encoding. encodingTable maps from
1589 encoding indices to encodings; INIT_ENC_INDEX(enc) is the index of
1590 the external (protocol) specified encoding; state is
1591 XML_CONTENT_STATE if we're parsing an external text entity, and
1592 XML_PROLOG_STATE otherwise.
1597 initScan(const ENCODING * const *encodingTable,
1598 const INIT_ENCODING *enc,
1602 const char **nextTokPtr)
1604 const ENCODING **encPtr;
1607 return XML_TOK_NONE;
1608 encPtr = enc->encPtr;
1609 if (ptr + 1 == end) {
1610 /* only a single byte available for auto-detection */
1611 #ifndef XML_DTD /* FIXME */
1612 /* a well-formed document entity must have more than one byte */
1613 if (state != XML_CONTENT_STATE)
1614 return XML_TOK_PARTIAL;
1616 /* so we're parsing an external text entity... */
1617 /* if UTF-16 was externally specified, then we need at least 2 bytes */
1618 switch (INIT_ENC_INDEX(enc)) {
1622 return XML_TOK_PARTIAL;
1624 switch ((unsigned char)*ptr) {
1627 case 0xEF: /* possibly first byte of UTF-8 BOM */
1628 if (INIT_ENC_INDEX(enc) == ISO_8859_1_ENC
1629 && state == XML_CONTENT_STATE)
1634 return XML_TOK_PARTIAL;
1638 switch (((unsigned char)ptr[0] << 8) | (unsigned char)ptr[1]) {
1640 if (INIT_ENC_INDEX(enc) == ISO_8859_1_ENC
1641 && state == XML_CONTENT_STATE)
1643 *nextTokPtr = ptr + 2;
1644 *encPtr = encodingTable[UTF_16BE_ENC];
1646 /* 00 3C is handled in the default case */
1648 if ((INIT_ENC_INDEX(enc) == UTF_16BE_ENC
1649 || INIT_ENC_INDEX(enc) == UTF_16_ENC)
1650 && state == XML_CONTENT_STATE)
1652 *encPtr = encodingTable[UTF_16LE_ENC];
1653 return XmlTok(*encPtr, state, ptr, end, nextTokPtr);
1655 if (INIT_ENC_INDEX(enc) == ISO_8859_1_ENC
1656 && state == XML_CONTENT_STATE)
1658 *nextTokPtr = ptr + 2;
1659 *encPtr = encodingTable[UTF_16LE_ENC];
1662 /* Maybe a UTF-8 BOM (EF BB BF) */
1663 /* If there's an explicitly specified (external) encoding
1664 of ISO-8859-1 or some flavour of UTF-16
1665 and this is an external text entity,
1666 don't look for the BOM,
1667 because it might be a legal data.
1669 if (state == XML_CONTENT_STATE) {
1670 int e = INIT_ENC_INDEX(enc);
1671 if (e == ISO_8859_1_ENC || e == UTF_16BE_ENC
1672 || e == UTF_16LE_ENC || e == UTF_16_ENC)
1676 return XML_TOK_PARTIAL;
1677 if ((unsigned char)ptr[2] == 0xBF) {
1678 *nextTokPtr = ptr + 3;
1679 *encPtr = encodingTable[UTF_8_ENC];
1684 if (ptr[0] == '\0') {
1685 /* 0 isn't a legal data character. Furthermore a document
1686 entity can only start with ASCII characters. So the only
1687 way this can fail to be big-endian UTF-16 if it it's an
1688 external parsed general entity that's labelled as
1691 if (state == XML_CONTENT_STATE && INIT_ENC_INDEX(enc) == UTF_16LE_ENC)
1693 *encPtr = encodingTable[UTF_16BE_ENC];
1694 return XmlTok(*encPtr, state, ptr, end, nextTokPtr);
1696 else if (ptr[1] == '\0') {
1697 /* We could recover here in the case:
1698 - parsing an external entity
1700 - no externally specified encoding
1701 - no encoding declaration
1702 by assuming UTF-16LE. But we don't, because this would mean when
1703 presented just with a single byte, we couldn't reliably determine
1704 whether we needed further bytes.
1706 if (state == XML_CONTENT_STATE)
1708 *encPtr = encodingTable[UTF_16LE_ENC];
1709 return XmlTok(*encPtr, state, ptr, end, nextTokPtr);
1714 *encPtr = encodingTable[INIT_ENC_INDEX(enc)];
1715 return XmlTok(*encPtr, state, ptr, end, nextTokPtr);
1721 #define XML_TOK_NS_C
1722 #include "xmltok_ns.c"
1729 #define NS(x) x ## NS
1730 #define ns(x) x ## _ns
1732 #define XML_TOK_NS_C
1733 #include "xmltok_ns.c"
1740 XmlInitUnknownEncodingNS(void *mem,
1745 ENCODING *enc = XmlInitUnknownEncoding(mem, table, convert, userData);
1747 ((struct normal_encoding *)enc)->type[ASCII_COLON] = BT_COLON;