+++ /dev/null
-/* Copyright (c) 1998, 1999 Thai Open Source Software Center Ltd\r
- See the file COPYING for copying permission.\r
-*/\r
-\r
-#include <stddef.h>\r
-\r
-#ifdef COMPILED_FROM_DSP\r
-#include "winconfig.h"\r
-#elif defined(MACOS_CLASSIC)\r
-#include "macconfig.h"\r
-#elif defined(__amigaos__)\r
-#include "amigaconfig.h"\r
-#elif defined(__WATCOMC__)\r
-#include "watcomconfig.h"\r
-#else\r
-#ifdef HAVE_EXPAT_CONFIG_H\r
-#include <expat_config.h>\r
-#endif\r
-#endif /* ndef COMPILED_FROM_DSP */\r
-\r
-#include "expat_external.h"\r
-#include "internal.h"\r
-#include "xmltok.h"\r
-#include "nametab.h"\r
-\r
-#ifdef XML_DTD\r
-#define IGNORE_SECTION_TOK_VTABLE , PREFIX(ignoreSectionTok)\r
-#else\r
-#define IGNORE_SECTION_TOK_VTABLE /* as nothing */\r
-#endif\r
-\r
-#define VTABLE1 \\r
- { PREFIX(prologTok), PREFIX(contentTok), \\r
- PREFIX(cdataSectionTok) IGNORE_SECTION_TOK_VTABLE }, \\r
- { PREFIX(attributeValueTok), PREFIX(entityValueTok) }, \\r
- PREFIX(sameName), \\r
- PREFIX(nameMatchesAscii), \\r
- PREFIX(nameLength), \\r
- PREFIX(skipS), \\r
- PREFIX(getAtts), \\r
- PREFIX(charRefNumber), \\r
- PREFIX(predefinedEntityName), \\r
- PREFIX(updatePosition), \\r
- PREFIX(isPublicId)\r
-\r
-#define VTABLE VTABLE1, PREFIX(toUtf8), PREFIX(toUtf16)\r
-\r
-#define UCS2_GET_NAMING(pages, hi, lo) \\r
- (namingBitmap[(pages[hi] << 3) + ((lo) >> 5)] & (1 << ((lo) & 0x1F)))\r
-\r
-/* A 2 byte UTF-8 representation splits the characters 11 bits between\r
- the bottom 5 and 6 bits of the bytes. We need 8 bits to index into\r
- pages, 3 bits to add to that index and 5 bits to generate the mask.\r
-*/\r
-#define UTF8_GET_NAMING2(pages, byte) \\r
- (namingBitmap[((pages)[(((byte)[0]) >> 2) & 7] << 3) \\r
- + ((((byte)[0]) & 3) << 1) \\r
- + ((((byte)[1]) >> 5) & 1)] \\r
- & (1 << (((byte)[1]) & 0x1F)))\r
-\r
-/* A 3 byte UTF-8 representation splits the characters 16 bits between\r
- the bottom 4, 6 and 6 bits of the bytes. We need 8 bits to index\r
- into pages, 3 bits to add to that index and 5 bits to generate the\r
- mask.\r
-*/\r
-#define UTF8_GET_NAMING3(pages, byte) \\r
- (namingBitmap[((pages)[((((byte)[0]) & 0xF) << 4) \\r
- + ((((byte)[1]) >> 2) & 0xF)] \\r
- << 3) \\r
- + ((((byte)[1]) & 3) << 1) \\r
- + ((((byte)[2]) >> 5) & 1)] \\r
- & (1 << (((byte)[2]) & 0x1F)))\r
-\r
-#define UTF8_GET_NAMING(pages, p, n) \\r
- ((n) == 2 \\r
- ? UTF8_GET_NAMING2(pages, (const unsigned char *)(p)) \\r
- : ((n) == 3 \\r
- ? UTF8_GET_NAMING3(pages, (const unsigned char *)(p)) \\r
- : 0))\r
-\r
-/* Detection of invalid UTF-8 sequences is based on Table 3.1B\r
- of Unicode 3.2: http://www.unicode.org/unicode/reports/tr28/\r
- with the additional restriction of not allowing the Unicode\r
- code points 0xFFFF and 0xFFFE (sequences EF,BF,BF and EF,BF,BE).\r
- Implementation details:\r
- (A & 0x80) == 0 means A < 0x80\r
- and\r
- (A & 0xC0) == 0xC0 means A > 0xBF\r
-*/\r
-\r
-#define UTF8_INVALID2(p) \\r
- ((*p) < 0xC2 || ((p)[1] & 0x80) == 0 || ((p)[1] & 0xC0) == 0xC0)\r
-\r
-#define UTF8_INVALID3(p) \\r
- (((p)[2] & 0x80) == 0 \\r
- || \\r
- ((*p) == 0xEF && (p)[1] == 0xBF \\r
- ? \\r
- (p)[2] > 0xBD \\r
- : \\r
- ((p)[2] & 0xC0) == 0xC0) \\r
- || \\r
- ((*p) == 0xE0 \\r
- ? \\r
- (p)[1] < 0xA0 || ((p)[1] & 0xC0) == 0xC0 \\r
- : \\r
- ((p)[1] & 0x80) == 0 \\r
- || \\r
- ((*p) == 0xED ? (p)[1] > 0x9F : ((p)[1] & 0xC0) == 0xC0)))\r
-\r
-#define UTF8_INVALID4(p) \\r
- (((p)[3] & 0x80) == 0 || ((p)[3] & 0xC0) == 0xC0 \\r
- || \\r
- ((p)[2] & 0x80) == 0 || ((p)[2] & 0xC0) == 0xC0 \\r
- || \\r
- ((*p) == 0xF0 \\r
- ? \\r
- (p)[1] < 0x90 || ((p)[1] & 0xC0) == 0xC0 \\r
- : \\r
- ((p)[1] & 0x80) == 0 \\r
- || \\r
- ((*p) == 0xF4 ? (p)[1] > 0x8F : ((p)[1] & 0xC0) == 0xC0)))\r
-\r
-static int PTRFASTCALL\r
-isNever(const ENCODING *enc, const char *p)\r
-{\r
- return 0;\r
-}\r
-\r
-static int PTRFASTCALL\r
-utf8_isName2(const ENCODING *enc, const char *p)\r
-{\r
- return UTF8_GET_NAMING2(namePages, (const unsigned char *)p);\r
-}\r
-\r
-static int PTRFASTCALL\r
-utf8_isName3(const ENCODING *enc, const char *p)\r
-{\r
- return UTF8_GET_NAMING3(namePages, (const unsigned char *)p);\r
-}\r
-\r
-#define utf8_isName4 isNever\r
-\r
-static int PTRFASTCALL\r
-utf8_isNmstrt2(const ENCODING *enc, const char *p)\r
-{\r
- return UTF8_GET_NAMING2(nmstrtPages, (const unsigned char *)p);\r
-}\r
-\r
-static int PTRFASTCALL\r
-utf8_isNmstrt3(const ENCODING *enc, const char *p)\r
-{\r
- return UTF8_GET_NAMING3(nmstrtPages, (const unsigned char *)p);\r
-}\r
-\r
-#define utf8_isNmstrt4 isNever\r
-\r
-static int PTRFASTCALL\r
-utf8_isInvalid2(const ENCODING *enc, const char *p)\r
-{\r
- return UTF8_INVALID2((const unsigned char *)p);\r
-}\r
-\r
-static int PTRFASTCALL\r
-utf8_isInvalid3(const ENCODING *enc, const char *p)\r
-{\r
- return UTF8_INVALID3((const unsigned char *)p);\r
-}\r
-\r
-static int PTRFASTCALL\r
-utf8_isInvalid4(const ENCODING *enc, const char *p)\r
-{\r
- return UTF8_INVALID4((const unsigned char *)p);\r
-}\r
-\r
-struct normal_encoding {\r
- ENCODING enc;\r
- unsigned char type[256];\r
-#ifdef XML_MIN_SIZE\r
- int (PTRFASTCALL *byteType)(const ENCODING *, const char *);\r
- int (PTRFASTCALL *isNameMin)(const ENCODING *, const char *);\r
- int (PTRFASTCALL *isNmstrtMin)(const ENCODING *, const char *);\r
- int (PTRFASTCALL *byteToAscii)(const ENCODING *, const char *);\r
- int (PTRCALL *charMatches)(const ENCODING *, const char *, int);\r
-#endif /* XML_MIN_SIZE */\r
- int (PTRFASTCALL *isName2)(const ENCODING *, const char *);\r
- int (PTRFASTCALL *isName3)(const ENCODING *, const char *);\r
- int (PTRFASTCALL *isName4)(const ENCODING *, const char *);\r
- int (PTRFASTCALL *isNmstrt2)(const ENCODING *, const char *);\r
- int (PTRFASTCALL *isNmstrt3)(const ENCODING *, const char *);\r
- int (PTRFASTCALL *isNmstrt4)(const ENCODING *, const char *);\r
- int (PTRFASTCALL *isInvalid2)(const ENCODING *, const char *);\r
- int (PTRFASTCALL *isInvalid3)(const ENCODING *, const char *);\r
- int (PTRFASTCALL *isInvalid4)(const ENCODING *, const char *);\r
-};\r
-\r
-#define AS_NORMAL_ENCODING(enc) ((const struct normal_encoding *) (enc))\r
-\r
-#ifdef XML_MIN_SIZE\r
-\r
-#define STANDARD_VTABLE(E) \\r
- E ## byteType, \\r
- E ## isNameMin, \\r
- E ## isNmstrtMin, \\r
- E ## byteToAscii, \\r
- E ## charMatches,\r
-\r
-#else\r
-\r
-#define STANDARD_VTABLE(E) /* as nothing */\r
-\r
-#endif\r
-\r
-#define NORMAL_VTABLE(E) \\r
- E ## isName2, \\r
- E ## isName3, \\r
- E ## isName4, \\r
- E ## isNmstrt2, \\r
- E ## isNmstrt3, \\r
- E ## isNmstrt4, \\r
- E ## isInvalid2, \\r
- E ## isInvalid3, \\r
- E ## isInvalid4\r
-\r
-static int FASTCALL checkCharRefNumber(int);\r
-\r
-#include "xmltok_impl.h"\r
-#include "ascii.h"\r
-\r
-#ifdef XML_MIN_SIZE\r
-#define sb_isNameMin isNever\r
-#define sb_isNmstrtMin isNever\r
-#endif\r
-\r
-#ifdef XML_MIN_SIZE\r
-#define MINBPC(enc) ((enc)->minBytesPerChar)\r
-#else\r
-/* minimum bytes per character */\r
-#define MINBPC(enc) 1\r
-#endif\r
-\r
-#define SB_BYTE_TYPE(enc, p) \\r
- (((struct normal_encoding *)(enc))->type[(unsigned char)*(p)])\r
-\r
-#ifdef XML_MIN_SIZE\r
-static int PTRFASTCALL\r
-sb_byteType(const ENCODING *enc, const char *p)\r
-{\r
- return SB_BYTE_TYPE(enc, p);\r
-}\r
-#define BYTE_TYPE(enc, p) \\r
- (AS_NORMAL_ENCODING(enc)->byteType(enc, p))\r
-#else\r
-#define BYTE_TYPE(enc, p) SB_BYTE_TYPE(enc, p)\r
-#endif\r
-\r
-#ifdef XML_MIN_SIZE\r
-#define BYTE_TO_ASCII(enc, p) \\r
- (AS_NORMAL_ENCODING(enc)->byteToAscii(enc, p))\r
-static int PTRFASTCALL\r
-sb_byteToAscii(const ENCODING *enc, const char *p)\r
-{\r
- return *p;\r
-}\r
-#else\r
-#define BYTE_TO_ASCII(enc, p) (*(p))\r
-#endif\r
-\r
-#define IS_NAME_CHAR(enc, p, n) \\r
- (AS_NORMAL_ENCODING(enc)->isName ## n(enc, p))\r
-#define IS_NMSTRT_CHAR(enc, p, n) \\r
- (AS_NORMAL_ENCODING(enc)->isNmstrt ## n(enc, p))\r
-#define IS_INVALID_CHAR(enc, p, n) \\r
- (AS_NORMAL_ENCODING(enc)->isInvalid ## n(enc, p))\r
-\r
-#ifdef XML_MIN_SIZE\r
-#define IS_NAME_CHAR_MINBPC(enc, p) \\r
- (AS_NORMAL_ENCODING(enc)->isNameMin(enc, p))\r
-#define IS_NMSTRT_CHAR_MINBPC(enc, p) \\r
- (AS_NORMAL_ENCODING(enc)->isNmstrtMin(enc, p))\r
-#else\r
-#define IS_NAME_CHAR_MINBPC(enc, p) (0)\r
-#define IS_NMSTRT_CHAR_MINBPC(enc, p) (0)\r
-#endif\r
-\r
-#ifdef XML_MIN_SIZE\r
-#define CHAR_MATCHES(enc, p, c) \\r
- (AS_NORMAL_ENCODING(enc)->charMatches(enc, p, c))\r
-static int PTRCALL\r
-sb_charMatches(const ENCODING *enc, const char *p, int c)\r
-{\r
- return *p == c;\r
-}\r
-#else\r
-/* c is an ASCII character */\r
-#define CHAR_MATCHES(enc, p, c) (*(p) == c)\r
-#endif\r
-\r
-#define PREFIX(ident) normal_ ## ident\r
-#define XML_TOK_IMPL_C\r
-#include "xmltok_impl.c"\r
-#undef XML_TOK_IMPL_C\r
-\r
-#undef MINBPC\r
-#undef BYTE_TYPE\r
-#undef BYTE_TO_ASCII\r
-#undef CHAR_MATCHES\r
-#undef IS_NAME_CHAR\r
-#undef IS_NAME_CHAR_MINBPC\r
-#undef IS_NMSTRT_CHAR\r
-#undef IS_NMSTRT_CHAR_MINBPC\r
-#undef IS_INVALID_CHAR\r
-\r
-enum { /* UTF8_cvalN is value of masked first byte of N byte sequence */\r
- UTF8_cval1 = 0x00,\r
- UTF8_cval2 = 0xc0,\r
- UTF8_cval3 = 0xe0,\r
- UTF8_cval4 = 0xf0\r
-};\r
-\r
-static void PTRCALL\r
-utf8_toUtf8(const ENCODING *enc,\r
- const char **fromP, const char *fromLim,\r
- char **toP, const char *toLim)\r
-{\r
- char *to;\r
- const char *from;\r
- if (fromLim - *fromP > toLim - *toP) {\r
- /* Avoid copying partial characters. */\r
- for (fromLim = *fromP + (toLim - *toP); fromLim > *fromP; fromLim--)\r
- if (((unsigned char)fromLim[-1] & 0xc0) != 0x80)\r
- break;\r
- }\r
- for (to = *toP, from = *fromP; from != fromLim; from++, to++)\r
- *to = *from;\r
- *fromP = from;\r
- *toP = to;\r
-}\r
-\r
-static void PTRCALL\r
-utf8_toUtf16(const ENCODING *enc,\r
- const char **fromP, const char *fromLim,\r
- unsigned short **toP, const unsigned short *toLim)\r
-{\r
- unsigned short *to = *toP;\r
- const char *from = *fromP;\r
- while (from != fromLim && to != toLim) {\r
- switch (((struct normal_encoding *)enc)->type[(unsigned char)*from]) {\r
- case BT_LEAD2:\r
- *to++ = (unsigned short)(((from[0] & 0x1f) << 6) | (from[1] & 0x3f));\r
- from += 2;\r
- break;\r
- case BT_LEAD3:\r
- *to++ = (unsigned short)(((from[0] & 0xf) << 12)\r
- | ((from[1] & 0x3f) << 6) | (from[2] & 0x3f));\r
- from += 3;\r
- break;\r
- case BT_LEAD4:\r
- {\r
- unsigned long n;\r
- if (to + 1 == toLim)\r
- goto after;\r
- n = ((from[0] & 0x7) << 18) | ((from[1] & 0x3f) << 12)\r
- | ((from[2] & 0x3f) << 6) | (from[3] & 0x3f);\r
- n -= 0x10000;\r
- to[0] = (unsigned short)((n >> 10) | 0xD800);\r
- to[1] = (unsigned short)((n & 0x3FF) | 0xDC00);\r
- to += 2;\r
- from += 4;\r
- }\r
- break;\r
- default:\r
- *to++ = *from++;\r
- break;\r
- }\r
- }\r
-after:\r
- *fromP = from;\r
- *toP = to;\r
-}\r
-\r
-#ifdef XML_NS\r
-static const struct normal_encoding utf8_encoding_ns = {\r
- { VTABLE1, utf8_toUtf8, utf8_toUtf16, 1, 1, 0 },\r
- {\r
-#include "asciitab.h"\r
-#include "utf8tab.h"\r
- },\r
- STANDARD_VTABLE(sb_) NORMAL_VTABLE(utf8_)\r
-};\r
-#endif\r
-\r
-static const struct normal_encoding utf8_encoding = {\r
- { VTABLE1, utf8_toUtf8, utf8_toUtf16, 1, 1, 0 },\r
- {\r
-#define BT_COLON BT_NMSTRT\r
-#include "asciitab.h"\r
-#undef BT_COLON\r
-#include "utf8tab.h"\r
- },\r
- STANDARD_VTABLE(sb_) NORMAL_VTABLE(utf8_)\r
-};\r
-\r
-#ifdef XML_NS\r
-\r
-static const struct normal_encoding internal_utf8_encoding_ns = {\r
- { VTABLE1, utf8_toUtf8, utf8_toUtf16, 1, 1, 0 },\r
- {\r
-#include "iasciitab.h"\r
-#include "utf8tab.h"\r
- },\r
- STANDARD_VTABLE(sb_) NORMAL_VTABLE(utf8_)\r
-};\r
-\r
-#endif\r
-\r
-static const struct normal_encoding internal_utf8_encoding = {\r
- { VTABLE1, utf8_toUtf8, utf8_toUtf16, 1, 1, 0 },\r
- {\r
-#define BT_COLON BT_NMSTRT\r
-#include "iasciitab.h"\r
-#undef BT_COLON\r
-#include "utf8tab.h"\r
- },\r
- STANDARD_VTABLE(sb_) NORMAL_VTABLE(utf8_)\r
-};\r
-\r
-static void PTRCALL\r
-latin1_toUtf8(const ENCODING *enc,\r
- const char **fromP, const char *fromLim,\r
- char **toP, const char *toLim)\r
-{\r
- for (;;) {\r
- unsigned char c;\r
- if (*fromP == fromLim)\r
- break;\r
- c = (unsigned char)**fromP;\r
- if (c & 0x80) {\r
- if (toLim - *toP < 2)\r
- break;\r
- *(*toP)++ = (char)((c >> 6) | UTF8_cval2);\r
- *(*toP)++ = (char)((c & 0x3f) | 0x80);\r
- (*fromP)++;\r
- }\r
- else {\r
- if (*toP == toLim)\r
- break;\r
- *(*toP)++ = *(*fromP)++;\r
- }\r
- }\r
-}\r
-\r
-static void PTRCALL\r
-latin1_toUtf16(const ENCODING *enc,\r
- const char **fromP, const char *fromLim,\r
- unsigned short **toP, const unsigned short *toLim)\r
-{\r
- while (*fromP != fromLim && *toP != toLim)\r
- *(*toP)++ = (unsigned char)*(*fromP)++;\r
-}\r
-\r
-#ifdef XML_NS\r
-\r
-static const struct normal_encoding latin1_encoding_ns = {\r
- { VTABLE1, latin1_toUtf8, latin1_toUtf16, 1, 0, 0 },\r
- {\r
-#include "asciitab.h"\r
-#include "latin1tab.h"\r
- },\r
- STANDARD_VTABLE(sb_)\r
-};\r
-\r
-#endif\r
-\r
-static const struct normal_encoding latin1_encoding = {\r
- { VTABLE1, latin1_toUtf8, latin1_toUtf16, 1, 0, 0 },\r
- {\r
-#define BT_COLON BT_NMSTRT\r
-#include "asciitab.h"\r
-#undef BT_COLON\r
-#include "latin1tab.h"\r
- },\r
- STANDARD_VTABLE(sb_)\r
-};\r
-\r
-static void PTRCALL\r
-ascii_toUtf8(const ENCODING *enc,\r
- const char **fromP, const char *fromLim,\r
- char **toP, const char *toLim)\r
-{\r
- while (*fromP != fromLim && *toP != toLim)\r
- *(*toP)++ = *(*fromP)++;\r
-}\r
-\r
-#ifdef XML_NS\r
-\r
-static const struct normal_encoding ascii_encoding_ns = {\r
- { VTABLE1, ascii_toUtf8, latin1_toUtf16, 1, 1, 0 },\r
- {\r
-#include "asciitab.h"\r
-/* BT_NONXML == 0 */\r
- },\r
- STANDARD_VTABLE(sb_)\r
-};\r
-\r
-#endif\r
-\r
-static const struct normal_encoding ascii_encoding = {\r
- { VTABLE1, ascii_toUtf8, latin1_toUtf16, 1, 1, 0 },\r
- {\r
-#define BT_COLON BT_NMSTRT\r
-#include "asciitab.h"\r
-#undef BT_COLON\r
-/* BT_NONXML == 0 */\r
- },\r
- STANDARD_VTABLE(sb_)\r
-};\r
-\r
-static int PTRFASTCALL\r
-unicode_byte_type(char hi, char lo)\r
-{\r
- switch ((unsigned char)hi) {\r
- case 0xD8: case 0xD9: case 0xDA: case 0xDB:\r
- return BT_LEAD4;\r
- case 0xDC: case 0xDD: case 0xDE: case 0xDF:\r
- return BT_TRAIL;\r
- case 0xFF:\r
- switch ((unsigned char)lo) {\r
- case 0xFF:\r
- case 0xFE:\r
- return BT_NONXML;\r
- }\r
- break;\r
- }\r
- return BT_NONASCII;\r
-}\r
-\r
-#define DEFINE_UTF16_TO_UTF8(E) \\r
-static void PTRCALL \\r
-E ## toUtf8(const ENCODING *enc, \\r
- const char **fromP, const char *fromLim, \\r
- char **toP, const char *toLim) \\r
-{ \\r
- const char *from; \\r
- for (from = *fromP; from != fromLim; from += 2) { \\r
- int plane; \\r
- unsigned char lo2; \\r
- unsigned char lo = GET_LO(from); \\r
- unsigned char hi = GET_HI(from); \\r
- switch (hi) { \\r
- case 0: \\r
- if (lo < 0x80) { \\r
- if (*toP == toLim) { \\r
- *fromP = from; \\r
- return; \\r
- } \\r
- *(*toP)++ = lo; \\r
- break; \\r
- } \\r
- /* fall through */ \\r
- case 0x1: case 0x2: case 0x3: \\r
- case 0x4: case 0x5: case 0x6: case 0x7: \\r
- if (toLim - *toP < 2) { \\r
- *fromP = from; \\r
- return; \\r
- } \\r
- *(*toP)++ = ((lo >> 6) | (hi << 2) | UTF8_cval2); \\r
- *(*toP)++ = ((lo & 0x3f) | 0x80); \\r
- break; \\r
- default: \\r
- if (toLim - *toP < 3) { \\r
- *fromP = from; \\r
- return; \\r
- } \\r
- /* 16 bits divided 4, 6, 6 amongst 3 bytes */ \\r
- *(*toP)++ = ((hi >> 4) | UTF8_cval3); \\r
- *(*toP)++ = (((hi & 0xf) << 2) | (lo >> 6) | 0x80); \\r
- *(*toP)++ = ((lo & 0x3f) | 0x80); \\r
- break; \\r
- case 0xD8: case 0xD9: case 0xDA: case 0xDB: \\r
- if (toLim - *toP < 4) { \\r
- *fromP = from; \\r
- return; \\r
- } \\r
- plane = (((hi & 0x3) << 2) | ((lo >> 6) & 0x3)) + 1; \\r
- *(*toP)++ = ((plane >> 2) | UTF8_cval4); \\r
- *(*toP)++ = (((lo >> 2) & 0xF) | ((plane & 0x3) << 4) | 0x80); \\r
- from += 2; \\r
- lo2 = GET_LO(from); \\r
- *(*toP)++ = (((lo & 0x3) << 4) \\r
- | ((GET_HI(from) & 0x3) << 2) \\r
- | (lo2 >> 6) \\r
- | 0x80); \\r
- *(*toP)++ = ((lo2 & 0x3f) | 0x80); \\r
- break; \\r
- } \\r
- } \\r
- *fromP = from; \\r
-}\r
-\r
-#define DEFINE_UTF16_TO_UTF16(E) \\r
-static void PTRCALL \\r
-E ## toUtf16(const ENCODING *enc, \\r
- const char **fromP, const char *fromLim, \\r
- unsigned short **toP, const unsigned short *toLim) \\r
-{ \\r
- /* Avoid copying first half only of surrogate */ \\r
- if (fromLim - *fromP > ((toLim - *toP) << 1) \\r
- && (GET_HI(fromLim - 2) & 0xF8) == 0xD8) \\r
- fromLim -= 2; \\r
- for (; *fromP != fromLim && *toP != toLim; *fromP += 2) \\r
- *(*toP)++ = (GET_HI(*fromP) << 8) | GET_LO(*fromP); \\r
-}\r
-\r
-#define SET2(ptr, ch) \\r
- (((ptr)[0] = ((ch) & 0xff)), ((ptr)[1] = ((ch) >> 8)))\r
-#define GET_LO(ptr) ((unsigned char)(ptr)[0])\r
-#define GET_HI(ptr) ((unsigned char)(ptr)[1])\r
-\r
-DEFINE_UTF16_TO_UTF8(little2_)\r
-DEFINE_UTF16_TO_UTF16(little2_)\r
-\r
-#undef SET2\r
-#undef GET_LO\r
-#undef GET_HI\r
-\r
-#define SET2(ptr, ch) \\r
- (((ptr)[0] = ((ch) >> 8)), ((ptr)[1] = ((ch) & 0xFF)))\r
-#define GET_LO(ptr) ((unsigned char)(ptr)[1])\r
-#define GET_HI(ptr) ((unsigned char)(ptr)[0])\r
-\r
-DEFINE_UTF16_TO_UTF8(big2_)\r
-DEFINE_UTF16_TO_UTF16(big2_)\r
-\r
-#undef SET2\r
-#undef GET_LO\r
-#undef GET_HI\r
-\r
-#define LITTLE2_BYTE_TYPE(enc, p) \\r
- ((p)[1] == 0 \\r
- ? ((struct normal_encoding *)(enc))->type[(unsigned char)*(p)] \\r
- : unicode_byte_type((p)[1], (p)[0]))\r
-#define LITTLE2_BYTE_TO_ASCII(enc, p) ((p)[1] == 0 ? (p)[0] : -1)\r
-#define LITTLE2_CHAR_MATCHES(enc, p, c) ((p)[1] == 0 && (p)[0] == c)\r
-#define LITTLE2_IS_NAME_CHAR_MINBPC(enc, p) \\r
- UCS2_GET_NAMING(namePages, (unsigned char)p[1], (unsigned char)p[0])\r
-#define LITTLE2_IS_NMSTRT_CHAR_MINBPC(enc, p) \\r
- UCS2_GET_NAMING(nmstrtPages, (unsigned char)p[1], (unsigned char)p[0])\r
-\r
-#ifdef XML_MIN_SIZE\r
-\r
-static int PTRFASTCALL\r
-little2_byteType(const ENCODING *enc, const char *p)\r
-{\r
- return LITTLE2_BYTE_TYPE(enc, p);\r
-}\r
-\r
-static int PTRFASTCALL\r
-little2_byteToAscii(const ENCODING *enc, const char *p)\r
-{\r
- return LITTLE2_BYTE_TO_ASCII(enc, p);\r
-}\r
-\r
-static int PTRCALL\r
-little2_charMatches(const ENCODING *enc, const char *p, int c)\r
-{\r
- return LITTLE2_CHAR_MATCHES(enc, p, c);\r
-}\r
-\r
-static int PTRFASTCALL\r
-little2_isNameMin(const ENCODING *enc, const char *p)\r
-{\r
- return LITTLE2_IS_NAME_CHAR_MINBPC(enc, p);\r
-}\r
-\r
-static int PTRFASTCALL\r
-little2_isNmstrtMin(const ENCODING *enc, const char *p)\r
-{\r
- return LITTLE2_IS_NMSTRT_CHAR_MINBPC(enc, p);\r
-}\r
-\r
-#undef VTABLE\r
-#define VTABLE VTABLE1, little2_toUtf8, little2_toUtf16\r
-\r
-#else /* not XML_MIN_SIZE */\r
-\r
-#undef PREFIX\r
-#define PREFIX(ident) little2_ ## ident\r
-#define MINBPC(enc) 2\r
-/* CHAR_MATCHES is guaranteed to have MINBPC bytes available. */\r
-#define BYTE_TYPE(enc, p) LITTLE2_BYTE_TYPE(enc, p)\r
-#define BYTE_TO_ASCII(enc, p) LITTLE2_BYTE_TO_ASCII(enc, p)\r
-#define CHAR_MATCHES(enc, p, c) LITTLE2_CHAR_MATCHES(enc, p, c)\r
-#define IS_NAME_CHAR(enc, p, n) 0\r
-#define IS_NAME_CHAR_MINBPC(enc, p) LITTLE2_IS_NAME_CHAR_MINBPC(enc, p)\r
-#define IS_NMSTRT_CHAR(enc, p, n) (0)\r
-#define IS_NMSTRT_CHAR_MINBPC(enc, p) LITTLE2_IS_NMSTRT_CHAR_MINBPC(enc, p)\r
-\r
-#define XML_TOK_IMPL_C\r
-#include "xmltok_impl.c"\r
-#undef XML_TOK_IMPL_C\r
-\r
-#undef MINBPC\r
-#undef BYTE_TYPE\r
-#undef BYTE_TO_ASCII\r
-#undef CHAR_MATCHES\r
-#undef IS_NAME_CHAR\r
-#undef IS_NAME_CHAR_MINBPC\r
-#undef IS_NMSTRT_CHAR\r
-#undef IS_NMSTRT_CHAR_MINBPC\r
-#undef IS_INVALID_CHAR\r
-\r
-#endif /* not XML_MIN_SIZE */\r
-\r
-#ifdef XML_NS\r
-\r
-static const struct normal_encoding little2_encoding_ns = {\r
- { VTABLE, 2, 0,\r
-#if BYTEORDER == 1234\r
- 1\r
-#else\r
- 0\r
-#endif\r
- },\r
- {\r
-#include "asciitab.h"\r
-#include "latin1tab.h"\r
- },\r
- STANDARD_VTABLE(little2_)\r
-};\r
-\r
-#endif\r
-\r
-static const struct normal_encoding little2_encoding = {\r
- { VTABLE, 2, 0,\r
-#if BYTEORDER == 1234\r
- 1\r
-#else\r
- 0\r
-#endif\r
- },\r
- {\r
-#define BT_COLON BT_NMSTRT\r
-#include "asciitab.h"\r
-#undef BT_COLON\r
-#include "latin1tab.h"\r
- },\r
- STANDARD_VTABLE(little2_)\r
-};\r
-\r
-#if BYTEORDER != 4321\r
-\r
-#ifdef XML_NS\r
-\r
-static const struct normal_encoding internal_little2_encoding_ns = {\r
- { VTABLE, 2, 0, 1 },\r
- {\r
-#include "iasciitab.h"\r
-#include "latin1tab.h"\r
- },\r
- STANDARD_VTABLE(little2_)\r
-};\r
-\r
-#endif\r
-\r
-static const struct normal_encoding internal_little2_encoding = {\r
- { VTABLE, 2, 0, 1 },\r
- {\r
-#define BT_COLON BT_NMSTRT\r
-#include "iasciitab.h"\r
-#undef BT_COLON\r
-#include "latin1tab.h"\r
- },\r
- STANDARD_VTABLE(little2_)\r
-};\r
-\r
-#endif\r
-\r
-\r
-#define BIG2_BYTE_TYPE(enc, p) \\r
- ((p)[0] == 0 \\r
- ? ((struct normal_encoding *)(enc))->type[(unsigned char)(p)[1]] \\r
- : unicode_byte_type((p)[0], (p)[1]))\r
-#define BIG2_BYTE_TO_ASCII(enc, p) ((p)[0] == 0 ? (p)[1] : -1)\r
-#define BIG2_CHAR_MATCHES(enc, p, c) ((p)[0] == 0 && (p)[1] == c)\r
-#define BIG2_IS_NAME_CHAR_MINBPC(enc, p) \\r
- UCS2_GET_NAMING(namePages, (unsigned char)p[0], (unsigned char)p[1])\r
-#define BIG2_IS_NMSTRT_CHAR_MINBPC(enc, p) \\r
- UCS2_GET_NAMING(nmstrtPages, (unsigned char)p[0], (unsigned char)p[1])\r
-\r
-#ifdef XML_MIN_SIZE\r
-\r
-static int PTRFASTCALL\r
-big2_byteType(const ENCODING *enc, const char *p)\r
-{\r
- return BIG2_BYTE_TYPE(enc, p);\r
-}\r
-\r
-static int PTRFASTCALL\r
-big2_byteToAscii(const ENCODING *enc, const char *p)\r
-{\r
- return BIG2_BYTE_TO_ASCII(enc, p);\r
-}\r
-\r
-static int PTRCALL\r
-big2_charMatches(const ENCODING *enc, const char *p, int c)\r
-{\r
- return BIG2_CHAR_MATCHES(enc, p, c);\r
-}\r
-\r
-static int PTRFASTCALL\r
-big2_isNameMin(const ENCODING *enc, const char *p)\r
-{\r
- return BIG2_IS_NAME_CHAR_MINBPC(enc, p);\r
-}\r
-\r
-static int PTRFASTCALL\r
-big2_isNmstrtMin(const ENCODING *enc, const char *p)\r
-{\r
- return BIG2_IS_NMSTRT_CHAR_MINBPC(enc, p);\r
-}\r
-\r
-#undef VTABLE\r
-#define VTABLE VTABLE1, big2_toUtf8, big2_toUtf16\r
-\r
-#else /* not XML_MIN_SIZE */\r
-\r
-#undef PREFIX\r
-#define PREFIX(ident) big2_ ## ident\r
-#define MINBPC(enc) 2\r
-/* CHAR_MATCHES is guaranteed to have MINBPC bytes available. */\r
-#define BYTE_TYPE(enc, p) BIG2_BYTE_TYPE(enc, p)\r
-#define BYTE_TO_ASCII(enc, p) BIG2_BYTE_TO_ASCII(enc, p)\r
-#define CHAR_MATCHES(enc, p, c) BIG2_CHAR_MATCHES(enc, p, c)\r
-#define IS_NAME_CHAR(enc, p, n) 0\r
-#define IS_NAME_CHAR_MINBPC(enc, p) BIG2_IS_NAME_CHAR_MINBPC(enc, p)\r
-#define IS_NMSTRT_CHAR(enc, p, n) (0)\r
-#define IS_NMSTRT_CHAR_MINBPC(enc, p) BIG2_IS_NMSTRT_CHAR_MINBPC(enc, p)\r
-\r
-#define XML_TOK_IMPL_C\r
-#include "xmltok_impl.c"\r
-#undef XML_TOK_IMPL_C\r
-\r
-#undef MINBPC\r
-#undef BYTE_TYPE\r
-#undef BYTE_TO_ASCII\r
-#undef CHAR_MATCHES\r
-#undef IS_NAME_CHAR\r
-#undef IS_NAME_CHAR_MINBPC\r
-#undef IS_NMSTRT_CHAR\r
-#undef IS_NMSTRT_CHAR_MINBPC\r
-#undef IS_INVALID_CHAR\r
-\r
-#endif /* not XML_MIN_SIZE */\r
-\r
-#ifdef XML_NS\r
-\r
-static const struct normal_encoding big2_encoding_ns = {\r
- { VTABLE, 2, 0,\r
-#if BYTEORDER == 4321\r
- 1\r
-#else\r
- 0\r
-#endif\r
- },\r
- {\r
-#include "asciitab.h"\r
-#include "latin1tab.h"\r
- },\r
- STANDARD_VTABLE(big2_)\r
-};\r
-\r
-#endif\r
-\r
-static const struct normal_encoding big2_encoding = {\r
- { VTABLE, 2, 0,\r
-#if BYTEORDER == 4321\r
- 1\r
-#else\r
- 0\r
-#endif\r
- },\r
- {\r
-#define BT_COLON BT_NMSTRT\r
-#include "asciitab.h"\r
-#undef BT_COLON\r
-#include "latin1tab.h"\r
- },\r
- STANDARD_VTABLE(big2_)\r
-};\r
-\r
-#if BYTEORDER != 1234\r
-\r
-#ifdef XML_NS\r
-\r
-static const struct normal_encoding internal_big2_encoding_ns = {\r
- { VTABLE, 2, 0, 1 },\r
- {\r
-#include "iasciitab.h"\r
-#include "latin1tab.h"\r
- },\r
- STANDARD_VTABLE(big2_)\r
-};\r
-\r
-#endif\r
-\r
-static const struct normal_encoding internal_big2_encoding = {\r
- { VTABLE, 2, 0, 1 },\r
- {\r
-#define BT_COLON BT_NMSTRT\r
-#include "iasciitab.h"\r
-#undef BT_COLON\r
-#include "latin1tab.h"\r
- },\r
- STANDARD_VTABLE(big2_)\r
-};\r
-\r
-#endif\r
-\r
-#undef PREFIX\r
-\r
-static int FASTCALL\r
-streqci(const char *s1, const char *s2)\r
-{\r
- for (;;) {\r
- char c1 = *s1++;\r
- char c2 = *s2++;\r
- if (ASCII_a <= c1 && c1 <= ASCII_z)\r
- c1 += ASCII_A - ASCII_a;\r
- if (ASCII_a <= c2 && c2 <= ASCII_z)\r
- c2 += ASCII_A - ASCII_a;\r
- if (c1 != c2)\r
- return 0;\r
- if (!c1)\r
- break;\r
- }\r
- return 1;\r
-}\r
-\r
-static void PTRCALL\r
-initUpdatePosition(const ENCODING *enc, const char *ptr,\r
- const char *end, POSITION *pos)\r
-{\r
- normal_updatePosition(&utf8_encoding.enc, ptr, end, pos);\r
-}\r
-\r
-static int\r
-toAscii(const ENCODING *enc, const char *ptr, const char *end)\r
-{\r
- char buf[1];\r
- char *p = buf;\r
- XmlUtf8Convert(enc, &ptr, end, &p, p + 1);\r
- if (p == buf)\r
- return -1;\r
- else\r
- return buf[0];\r
-}\r
-\r
-static int FASTCALL\r
-isSpace(int c)\r
-{\r
- switch (c) {\r
- case 0x20:\r
- case 0xD:\r
- case 0xA:\r
- case 0x9:\r
- return 1;\r
- }\r
- return 0;\r
-}\r
-\r
-/* Return 1 if there's just optional white space or there's an S\r
- followed by name=val.\r
-*/\r
-static int\r
-parsePseudoAttribute(const ENCODING *enc,\r
- const char *ptr,\r
- const char *end,\r
- const char **namePtr,\r
- const char **nameEndPtr,\r
- const char **valPtr,\r
- const char **nextTokPtr)\r
-{\r
- int c;\r
- char open;\r
- if (ptr == end) {\r
- *namePtr = NULL;\r
- return 1;\r
- }\r
- if (!isSpace(toAscii(enc, ptr, end))) {\r
- *nextTokPtr = ptr;\r
- return 0;\r
- }\r
- do {\r
- ptr += enc->minBytesPerChar;\r
- } while (isSpace(toAscii(enc, ptr, end)));\r
- if (ptr == end) {\r
- *namePtr = NULL;\r
- return 1;\r
- }\r
- *namePtr = ptr;\r
- for (;;) {\r
- c = toAscii(enc, ptr, end);\r
- if (c == -1) {\r
- *nextTokPtr = ptr;\r
- return 0;\r
- }\r
- if (c == ASCII_EQUALS) {\r
- *nameEndPtr = ptr;\r
- break;\r
- }\r
- if (isSpace(c)) {\r
- *nameEndPtr = ptr;\r
- do {\r
- ptr += enc->minBytesPerChar;\r
- } while (isSpace(c = toAscii(enc, ptr, end)));\r
- if (c != ASCII_EQUALS) {\r
- *nextTokPtr = ptr;\r
- return 0;\r
- }\r
- break;\r
- }\r
- ptr += enc->minBytesPerChar;\r
- }\r
- if (ptr == *namePtr) {\r
- *nextTokPtr = ptr;\r
- return 0;\r
- }\r
- ptr += enc->minBytesPerChar;\r
- c = toAscii(enc, ptr, end);\r
- while (isSpace(c)) {\r
- ptr += enc->minBytesPerChar;\r
- c = toAscii(enc, ptr, end);\r
- }\r
- if (c != ASCII_QUOT && c != ASCII_APOS) {\r
- *nextTokPtr = ptr;\r
- return 0;\r
- }\r
- open = (char)c;\r
- ptr += enc->minBytesPerChar;\r
- *valPtr = ptr;\r
- for (;; ptr += enc->minBytesPerChar) {\r
- c = toAscii(enc, ptr, end);\r
- if (c == open)\r
- break;\r
- if (!(ASCII_a <= c && c <= ASCII_z)\r
- && !(ASCII_A <= c && c <= ASCII_Z)\r
- && !(ASCII_0 <= c && c <= ASCII_9)\r
- && c != ASCII_PERIOD\r
- && c != ASCII_MINUS\r
- && c != ASCII_UNDERSCORE) {\r
- *nextTokPtr = ptr;\r
- return 0;\r
- }\r
- }\r
- *nextTokPtr = ptr + enc->minBytesPerChar;\r
- return 1;\r
-}\r
-\r
-static const char KW_version[] = {\r
- ASCII_v, ASCII_e, ASCII_r, ASCII_s, ASCII_i, ASCII_o, ASCII_n, '\0'\r
-};\r
-\r
-static const char KW_encoding[] = {\r
- ASCII_e, ASCII_n, ASCII_c, ASCII_o, ASCII_d, ASCII_i, ASCII_n, ASCII_g, '\0'\r
-};\r
-\r
-static const char KW_standalone[] = {\r
- ASCII_s, ASCII_t, ASCII_a, ASCII_n, ASCII_d, ASCII_a, ASCII_l, ASCII_o,\r
- ASCII_n, ASCII_e, '\0'\r
-};\r
-\r
-static const char KW_yes[] = {\r
- ASCII_y, ASCII_e, ASCII_s, '\0'\r
-};\r
-\r
-static const char KW_no[] = {\r
- ASCII_n, ASCII_o, '\0'\r
-};\r
-\r
-static int\r
-doParseXmlDecl(const ENCODING *(*encodingFinder)(const ENCODING *,\r
- const char *,\r
- const char *),\r
- int isGeneralTextEntity,\r
- const ENCODING *enc,\r
- const char *ptr,\r
- const char *end,\r
- const char **badPtr,\r
- const char **versionPtr,\r
- const char **versionEndPtr,\r
- const char **encodingName,\r
- const ENCODING **encoding,\r
- int *standalone)\r
-{\r
- const char *val = NULL;\r
- const char *name = NULL;\r
- const char *nameEnd = NULL;\r
- ptr += 5 * enc->minBytesPerChar;\r
- end -= 2 * enc->minBytesPerChar;\r
- if (!parsePseudoAttribute(enc, ptr, end, &name, &nameEnd, &val, &ptr)\r
- || !name) {\r
- *badPtr = ptr;\r
- return 0;\r
- }\r
- if (!XmlNameMatchesAscii(enc, name, nameEnd, KW_version)) {\r
- if (!isGeneralTextEntity) {\r
- *badPtr = name;\r
- return 0;\r
- }\r
- }\r
- else {\r
- if (versionPtr)\r
- *versionPtr = val;\r
- if (versionEndPtr)\r
- *versionEndPtr = ptr;\r
- if (!parsePseudoAttribute(enc, ptr, end, &name, &nameEnd, &val, &ptr)) {\r
- *badPtr = ptr;\r
- return 0;\r
- }\r
- if (!name) {\r
- if (isGeneralTextEntity) {\r
- /* a TextDecl must have an EncodingDecl */\r
- *badPtr = ptr;\r
- return 0;\r
- }\r
- return 1;\r
- }\r
- }\r
- if (XmlNameMatchesAscii(enc, name, nameEnd, KW_encoding)) {\r
- int c = toAscii(enc, val, end);\r
- if (!(ASCII_a <= c && c <= ASCII_z) && !(ASCII_A <= c && c <= ASCII_Z)) {\r
- *badPtr = val;\r
- return 0;\r
- }\r
- if (encodingName)\r
- *encodingName = val;\r
- if (encoding)\r
- *encoding = encodingFinder(enc, val, ptr - enc->minBytesPerChar);\r
- if (!parsePseudoAttribute(enc, ptr, end, &name, &nameEnd, &val, &ptr)) {\r
- *badPtr = ptr;\r
- return 0;\r
- }\r
- if (!name)\r
- return 1;\r
- }\r
- if (!XmlNameMatchesAscii(enc, name, nameEnd, KW_standalone)\r
- || isGeneralTextEntity) {\r
- *badPtr = name;\r
- return 0;\r
- }\r
- if (XmlNameMatchesAscii(enc, val, ptr - enc->minBytesPerChar, KW_yes)) {\r
- if (standalone)\r
- *standalone = 1;\r
- }\r
- else if (XmlNameMatchesAscii(enc, val, ptr - enc->minBytesPerChar, KW_no)) {\r
- if (standalone)\r
- *standalone = 0;\r
- }\r
- else {\r
- *badPtr = val;\r
- return 0;\r
- }\r
- while (isSpace(toAscii(enc, ptr, end)))\r
- ptr += enc->minBytesPerChar;\r
- if (ptr != end) {\r
- *badPtr = ptr;\r
- return 0;\r
- }\r
- return 1;\r
-}\r
-\r
-static int FASTCALL\r
-checkCharRefNumber(int result)\r
-{\r
- switch (result >> 8) {\r
- case 0xD8: case 0xD9: case 0xDA: case 0xDB:\r
- case 0xDC: case 0xDD: case 0xDE: case 0xDF:\r
- return -1;\r
- case 0:\r
- if (latin1_encoding.type[result] == BT_NONXML)\r
- return -1;\r
- break;\r
- case 0xFF:\r
- if (result == 0xFFFE || result == 0xFFFF)\r
- return -1;\r
- break;\r
- }\r
- return result;\r
-}\r
-\r
-int FASTCALL\r
-XmlUtf8Encode(int c, char *buf)\r
-{\r
- enum {\r
- /* minN is minimum legal resulting value for N byte sequence */\r
- min2 = 0x80,\r
- min3 = 0x800,\r
- min4 = 0x10000\r
- };\r
-\r
- if (c < 0)\r
- return 0;\r
- if (c < min2) {\r
- buf[0] = (char)(c | UTF8_cval1);\r
- return 1;\r
- }\r
- if (c < min3) {\r
- buf[0] = (char)((c >> 6) | UTF8_cval2);\r
- buf[1] = (char)((c & 0x3f) | 0x80);\r
- return 2;\r
- }\r
- if (c < min4) {\r
- buf[0] = (char)((c >> 12) | UTF8_cval3);\r
- buf[1] = (char)(((c >> 6) & 0x3f) | 0x80);\r
- buf[2] = (char)((c & 0x3f) | 0x80);\r
- return 3;\r
- }\r
- if (c < 0x110000) {\r
- buf[0] = (char)((c >> 18) | UTF8_cval4);\r
- buf[1] = (char)(((c >> 12) & 0x3f) | 0x80);\r
- buf[2] = (char)(((c >> 6) & 0x3f) | 0x80);\r
- buf[3] = (char)((c & 0x3f) | 0x80);\r
- return 4;\r
- }\r
- return 0;\r
-}\r
-\r
-int FASTCALL\r
-XmlUtf16Encode(int charNum, unsigned short *buf)\r
-{\r
- if (charNum < 0)\r
- return 0;\r
- if (charNum < 0x10000) {\r
- buf[0] = (unsigned short)charNum;\r
- return 1;\r
- }\r
- if (charNum < 0x110000) {\r
- charNum -= 0x10000;\r
- buf[0] = (unsigned short)((charNum >> 10) + 0xD800);\r
- buf[1] = (unsigned short)((charNum & 0x3FF) + 0xDC00);\r
- return 2;\r
- }\r
- return 0;\r
-}\r
-\r
-struct unknown_encoding {\r
- struct normal_encoding normal;\r
- CONVERTER convert;\r
- void *userData;\r
- unsigned short utf16[256];\r
- char utf8[256][4];\r
-};\r
-\r
-#define AS_UNKNOWN_ENCODING(enc) ((const struct unknown_encoding *) (enc))\r
-\r
-int\r
-XmlSizeOfUnknownEncoding(void)\r
-{\r
- return sizeof(struct unknown_encoding);\r
-}\r
-\r
-static int PTRFASTCALL\r
-unknown_isName(const ENCODING *enc, const char *p)\r
-{\r
- const struct unknown_encoding *uenc = AS_UNKNOWN_ENCODING(enc);\r
- int c = uenc->convert(uenc->userData, p);\r
- if (c & ~0xFFFF)\r
- return 0;\r
- return UCS2_GET_NAMING(namePages, c >> 8, c & 0xFF);\r
-}\r
-\r
-static int PTRFASTCALL\r
-unknown_isNmstrt(const ENCODING *enc, const char *p)\r
-{\r
- const struct unknown_encoding *uenc = AS_UNKNOWN_ENCODING(enc);\r
- int c = uenc->convert(uenc->userData, p);\r
- if (c & ~0xFFFF)\r
- return 0;\r
- return UCS2_GET_NAMING(nmstrtPages, c >> 8, c & 0xFF);\r
-}\r
-\r
-static int PTRFASTCALL\r
-unknown_isInvalid(const ENCODING *enc, const char *p)\r
-{\r
- const struct unknown_encoding *uenc = AS_UNKNOWN_ENCODING(enc);\r
- int c = uenc->convert(uenc->userData, p);\r
- return (c & ~0xFFFF) || checkCharRefNumber(c) < 0;\r
-}\r
-\r
-static void PTRCALL\r
-unknown_toUtf8(const ENCODING *enc,\r
- const char **fromP, const char *fromLim,\r
- char **toP, const char *toLim)\r
-{\r
- const struct unknown_encoding *uenc = AS_UNKNOWN_ENCODING(enc);\r
- char buf[XML_UTF8_ENCODE_MAX];\r
- for (;;) {\r
- const char *utf8;\r
- int n;\r
- if (*fromP == fromLim)\r
- break;\r
- utf8 = uenc->utf8[(unsigned char)**fromP];\r
- n = *utf8++;\r
- if (n == 0) {\r
- int c = uenc->convert(uenc->userData, *fromP);\r
- n = XmlUtf8Encode(c, buf);\r
- if (n > toLim - *toP)\r
- break;\r
- utf8 = buf;\r
- *fromP += (AS_NORMAL_ENCODING(enc)->type[(unsigned char)**fromP]\r
- - (BT_LEAD2 - 2));\r
- }\r
- else {\r
- if (n > toLim - *toP)\r
- break;\r
- (*fromP)++;\r
- }\r
- do {\r
- *(*toP)++ = *utf8++;\r
- } while (--n != 0);\r
- }\r
-}\r
-\r
-static void PTRCALL\r
-unknown_toUtf16(const ENCODING *enc,\r
- const char **fromP, const char *fromLim,\r
- unsigned short **toP, const unsigned short *toLim)\r
-{\r
- const struct unknown_encoding *uenc = AS_UNKNOWN_ENCODING(enc);\r
- while (*fromP != fromLim && *toP != toLim) {\r
- unsigned short c = uenc->utf16[(unsigned char)**fromP];\r
- if (c == 0) {\r
- c = (unsigned short)\r
- uenc->convert(uenc->userData, *fromP);\r
- *fromP += (AS_NORMAL_ENCODING(enc)->type[(unsigned char)**fromP]\r
- - (BT_LEAD2 - 2));\r
- }\r
- else\r
- (*fromP)++;\r
- *(*toP)++ = c;\r
- }\r
-}\r
-\r
-ENCODING *\r
-XmlInitUnknownEncoding(void *mem,\r
- int *table,\r
- CONVERTER convert,\r
- void *userData)\r
-{\r
- int i;\r
- struct unknown_encoding *e = (struct unknown_encoding *)mem;\r
- for (i = 0; i < (int)sizeof(struct normal_encoding); i++)\r
- ((char *)mem)[i] = ((char *)&latin1_encoding)[i];\r
- for (i = 0; i < 128; i++)\r
- if (latin1_encoding.type[i] != BT_OTHER\r
- && latin1_encoding.type[i] != BT_NONXML\r
- && table[i] != i)\r
- return 0;\r
- for (i = 0; i < 256; i++) {\r
- int c = table[i];\r
- if (c == -1) {\r
- e->normal.type[i] = BT_MALFORM;\r
- /* This shouldn't really get used. */\r
- e->utf16[i] = 0xFFFF;\r
- e->utf8[i][0] = 1;\r
- e->utf8[i][1] = 0;\r
- }\r
- else if (c < 0) {\r
- if (c < -4)\r
- return 0;\r
- e->normal.type[i] = (unsigned char)(BT_LEAD2 - (c + 2));\r
- e->utf8[i][0] = 0;\r
- e->utf16[i] = 0;\r
- }\r
- else if (c < 0x80) {\r
- if (latin1_encoding.type[c] != BT_OTHER\r
- && latin1_encoding.type[c] != BT_NONXML\r
- && c != i)\r
- return 0;\r
- e->normal.type[i] = latin1_encoding.type[c];\r
- e->utf8[i][0] = 1;\r
- e->utf8[i][1] = (char)c;\r
- e->utf16[i] = (unsigned short)(c == 0 ? 0xFFFF : c);\r
- }\r
- else if (checkCharRefNumber(c) < 0) {\r
- e->normal.type[i] = BT_NONXML;\r
- /* This shouldn't really get used. */\r
- e->utf16[i] = 0xFFFF;\r
- e->utf8[i][0] = 1;\r
- e->utf8[i][1] = 0;\r
- }\r
- else {\r
- if (c > 0xFFFF)\r
- return 0;\r
- if (UCS2_GET_NAMING(nmstrtPages, c >> 8, c & 0xff))\r
- e->normal.type[i] = BT_NMSTRT;\r
- else if (UCS2_GET_NAMING(namePages, c >> 8, c & 0xff))\r
- e->normal.type[i] = BT_NAME;\r
- else\r
- e->normal.type[i] = BT_OTHER;\r
- e->utf8[i][0] = (char)XmlUtf8Encode(c, e->utf8[i] + 1);\r
- e->utf16[i] = (unsigned short)c;\r
- }\r
- }\r
- e->userData = userData;\r
- e->convert = convert;\r
- if (convert) {\r
- e->normal.isName2 = unknown_isName;\r
- e->normal.isName3 = unknown_isName;\r
- e->normal.isName4 = unknown_isName;\r
- e->normal.isNmstrt2 = unknown_isNmstrt;\r
- e->normal.isNmstrt3 = unknown_isNmstrt;\r
- e->normal.isNmstrt4 = unknown_isNmstrt;\r
- e->normal.isInvalid2 = unknown_isInvalid;\r
- e->normal.isInvalid3 = unknown_isInvalid;\r
- e->normal.isInvalid4 = unknown_isInvalid;\r
- }\r
- e->normal.enc.utf8Convert = unknown_toUtf8;\r
- e->normal.enc.utf16Convert = unknown_toUtf16;\r
- return &(e->normal.enc);\r
-}\r
-\r
-/* If this enumeration is changed, getEncodingIndex and encodings\r
-must also be changed. */\r
-enum {\r
- UNKNOWN_ENC = -1,\r
- ISO_8859_1_ENC = 0,\r
- US_ASCII_ENC,\r
- UTF_8_ENC,\r
- UTF_16_ENC,\r
- UTF_16BE_ENC,\r
- UTF_16LE_ENC,\r
- /* must match encodingNames up to here */\r
- NO_ENC\r
-};\r
-\r
-static const char KW_ISO_8859_1[] = {\r
- ASCII_I, ASCII_S, ASCII_O, ASCII_MINUS, ASCII_8, ASCII_8, ASCII_5, ASCII_9,\r
- ASCII_MINUS, ASCII_1, '\0'\r
-};\r
-static const char KW_US_ASCII[] = {\r
- ASCII_U, ASCII_S, ASCII_MINUS, ASCII_A, ASCII_S, ASCII_C, ASCII_I, ASCII_I,\r
- '\0'\r
-};\r
-static const char KW_UTF_8[] = {\r
- ASCII_U, ASCII_T, ASCII_F, ASCII_MINUS, ASCII_8, '\0'\r
-};\r
-static const char KW_UTF_16[] = {\r
- ASCII_U, ASCII_T, ASCII_F, ASCII_MINUS, ASCII_1, ASCII_6, '\0'\r
-};\r
-static const char KW_UTF_16BE[] = {\r
- ASCII_U, ASCII_T, ASCII_F, ASCII_MINUS, ASCII_1, ASCII_6, ASCII_B, ASCII_E,\r
- '\0'\r
-};\r
-static const char KW_UTF_16LE[] = {\r
- ASCII_U, ASCII_T, ASCII_F, ASCII_MINUS, ASCII_1, ASCII_6, ASCII_L, ASCII_E,\r
- '\0'\r
-};\r
-\r
-static int FASTCALL\r
-getEncodingIndex(const char *name)\r
-{\r
- static const char * const encodingNames[] = {\r
- KW_ISO_8859_1,\r
- KW_US_ASCII,\r
- KW_UTF_8,\r
- KW_UTF_16,\r
- KW_UTF_16BE,\r
- KW_UTF_16LE,\r
- };\r
- int i;\r
- if (name == NULL)\r
- return NO_ENC;\r
- for (i = 0; i < (int)(sizeof(encodingNames)/sizeof(encodingNames[0])); i++)\r
- if (streqci(name, encodingNames[i]))\r
- return i;\r
- return UNKNOWN_ENC;\r
-}\r
-\r
-/* For binary compatibility, we store the index of the encoding\r
- specified at initialization in the isUtf16 member.\r
-*/\r
-\r
-#define INIT_ENC_INDEX(enc) ((int)(enc)->initEnc.isUtf16)\r
-#define SET_INIT_ENC_INDEX(enc, i) ((enc)->initEnc.isUtf16 = (char)i)\r
-\r
-/* This is what detects the encoding. encodingTable maps from\r
- encoding indices to encodings; INIT_ENC_INDEX(enc) is the index of\r
- the external (protocol) specified encoding; state is\r
- XML_CONTENT_STATE if we're parsing an external text entity, and\r
- XML_PROLOG_STATE otherwise.\r
-*/\r
-\r
-\r
-static int\r
-initScan(const ENCODING * const *encodingTable,\r
- const INIT_ENCODING *enc,\r
- int state,\r
- const char *ptr,\r
- const char *end,\r
- const char **nextTokPtr)\r
-{\r
- const ENCODING **encPtr;\r
-\r
- if (ptr == end)\r
- return XML_TOK_NONE;\r
- encPtr = enc->encPtr;\r
- if (ptr + 1 == end) {\r
- /* only a single byte available for auto-detection */\r
-#ifndef XML_DTD /* FIXME */\r
- /* a well-formed document entity must have more than one byte */\r
- if (state != XML_CONTENT_STATE)\r
- return XML_TOK_PARTIAL;\r
-#endif\r
- /* so we're parsing an external text entity... */\r
- /* if UTF-16 was externally specified, then we need at least 2 bytes */\r
- switch (INIT_ENC_INDEX(enc)) {\r
- case UTF_16_ENC:\r
- case UTF_16LE_ENC:\r
- case UTF_16BE_ENC:\r
- return XML_TOK_PARTIAL;\r
- }\r
- switch ((unsigned char)*ptr) {\r
- case 0xFE:\r
- case 0xFF:\r
- case 0xEF: /* possibly first byte of UTF-8 BOM */\r
- if (INIT_ENC_INDEX(enc) == ISO_8859_1_ENC\r
- && state == XML_CONTENT_STATE)\r
- break;\r
- /* fall through */\r
- case 0x00:\r
- case 0x3C:\r
- return XML_TOK_PARTIAL;\r
- }\r
- }\r
- else {\r
- switch (((unsigned char)ptr[0] << 8) | (unsigned char)ptr[1]) {\r
- case 0xFEFF:\r
- if (INIT_ENC_INDEX(enc) == ISO_8859_1_ENC\r
- && state == XML_CONTENT_STATE)\r
- break;\r
- *nextTokPtr = ptr + 2;\r
- *encPtr = encodingTable[UTF_16BE_ENC];\r
- return XML_TOK_BOM;\r
- /* 00 3C is handled in the default case */\r
- case 0x3C00:\r
- if ((INIT_ENC_INDEX(enc) == UTF_16BE_ENC\r
- || INIT_ENC_INDEX(enc) == UTF_16_ENC)\r
- && state == XML_CONTENT_STATE)\r
- break;\r
- *encPtr = encodingTable[UTF_16LE_ENC];\r
- return XmlTok(*encPtr, state, ptr, end, nextTokPtr);\r
- case 0xFFFE:\r
- if (INIT_ENC_INDEX(enc) == ISO_8859_1_ENC\r
- && state == XML_CONTENT_STATE)\r
- break;\r
- *nextTokPtr = ptr + 2;\r
- *encPtr = encodingTable[UTF_16LE_ENC];\r
- return XML_TOK_BOM;\r
- case 0xEFBB:\r
- /* Maybe a UTF-8 BOM (EF BB BF) */\r
- /* If there's an explicitly specified (external) encoding\r
- of ISO-8859-1 or some flavour of UTF-16\r
- and this is an external text entity,\r
- don't look for the BOM,\r
- because it might be a legal data.\r
- */\r
- if (state == XML_CONTENT_STATE) {\r
- int e = INIT_ENC_INDEX(enc);\r
- if (e == ISO_8859_1_ENC || e == UTF_16BE_ENC\r
- || e == UTF_16LE_ENC || e == UTF_16_ENC)\r
- break;\r
- }\r
- if (ptr + 2 == end)\r
- return XML_TOK_PARTIAL;\r
- if ((unsigned char)ptr[2] == 0xBF) {\r
- *nextTokPtr = ptr + 3;\r
- *encPtr = encodingTable[UTF_8_ENC];\r
- return XML_TOK_BOM;\r
- }\r
- break;\r
- default:\r
- if (ptr[0] == '\0') {\r
- /* 0 isn't a legal data character. Furthermore a document\r
- entity can only start with ASCII characters. So the only\r
- way this can fail to be big-endian UTF-16 is if it is an\r
- external parsed general entity that's labelled as\r
- UTF-16LE.\r
- */\r
- if (state == XML_CONTENT_STATE && INIT_ENC_INDEX(enc) == UTF_16LE_ENC)\r
- break;\r
- *encPtr = encodingTable[UTF_16BE_ENC];\r
- return XmlTok(*encPtr, state, ptr, end, nextTokPtr);\r
- }\r
- else if (ptr[1] == '\0') {\r
- /* We could recover here in the case:\r
- - parsing an external entity\r
- - second byte is 0\r
- - no externally specified encoding\r
- - no encoding declaration\r
- by assuming UTF-16LE. But we don't, because this would mean when\r
- presented just with a single byte, we couldn't reliably determine\r
- whether we needed further bytes.\r
- */\r
- if (state == XML_CONTENT_STATE)\r
- break;\r
- *encPtr = encodingTable[UTF_16LE_ENC];\r
- return XmlTok(*encPtr, state, ptr, end, nextTokPtr);\r
- }\r
- break;\r
- }\r
- }\r
- *encPtr = encodingTable[INIT_ENC_INDEX(enc)];\r
- return XmlTok(*encPtr, state, ptr, end, nextTokPtr);\r
-}\r
-\r
-\r
-#define NS(x) x\r
-#define ns(x) x\r
-#define XML_TOK_NS_C\r
-#include "xmltok_ns.c"\r
-#undef XML_TOK_NS_C\r
-#undef NS\r
-#undef ns\r
-\r
-#ifdef XML_NS\r
-\r
-#define NS(x) x ## NS\r
-#define ns(x) x ## _ns\r
-\r
-#define XML_TOK_NS_C\r
-#include "xmltok_ns.c"\r
-#undef XML_TOK_NS_C\r
-\r
-#undef NS\r
-#undef ns\r
-\r
-ENCODING *\r
-XmlInitUnknownEncodingNS(void *mem,\r
- int *table,\r
- CONVERTER convert,\r
- void *userData)\r
-{\r
- ENCODING *enc = XmlInitUnknownEncoding(mem, table, convert, userData);\r
- if (enc)\r
- ((struct normal_encoding *)enc)->type[ASCII_COLON] = BT_COLON;\r
- return enc;\r
-}\r
-\r
-#endif /* XML_NS */\r
projects / mirror_edk2.git / blobdiff