xmltok.c

俄罗斯人开发的大名鼎鼎的Pocket Pc 阅读器haaliread的源代码,visual c

      /* fall through */ \
    case 0x1: case 0x2: case 0x3: \
    case 0x4: case 0x5: case 0x6: case 0x7: \
      if (toLim -  *toP < 2) { \
        *fromP = from; \
        return; \
      } \
      *(*toP)++ = ((lo >> 6) | (hi << 2) |  UTF8_cval2); \
      *(*toP)++ = ((lo & 0x3f) | 0x80); \
      break; \
    default: \
      if (toLim -  *toP < 3)  { \
        *fromP = from; \
        return; \
      } \
      /* 16 bits divided 4, 6, 6 amongst 3 bytes */ \
      *(*toP)++ = ((hi >> 4) | UTF8_cval3); \
      *(*toP)++ = (((hi & 0xf) << 2) | (lo >> 6) | 0x80); \
      *(*toP)++ = ((lo & 0x3f) | 0x80); \
      break; \
    case 0xD8: case 0xD9: case 0xDA: case 0xDB: \
      if (toLim -  *toP < 4) { \
        *fromP = from; \
        return; \
      } \
      plane = (((hi & 0x3) << 2) | ((lo >> 6) & 0x3)) + 1; \
      *(*toP)++ = ((plane >> 2) | UTF8_cval4); \
      *(*toP)++ = (((lo >> 2) & 0xF) | ((plane & 0x3) << 4) | 0x80); \
      from += 2; \
      lo2 = GET_LO(from); \
      *(*toP)++ = (((lo & 0x3) << 4) \
                   | ((GET_HI(from) & 0x3) << 2) \
                   | (lo2 >> 6) \
                   | 0x80); \
      *(*toP)++ = ((lo2 & 0x3f) | 0x80); \
      break; \
    } \
  } \
  *fromP = from; \
}

#define DEFINE_UTF16_TO_UTF16(E) \
static void  PTRCALL \
E ## toUtf16(const ENCODING *enc, \
             const char **fromP, const char *fromLim, \
             unsigned short **toP, const unsigned short *toLim) \
{ \
  /* Avoid copying first half only of surrogate */ \
  if (fromLim - *fromP > ((toLim - *toP) << 1) \
      && (GET_HI(fromLim - 2) & 0xF8) == 0xD8) \
    fromLim -= 2; \
  for (; *fromP != fromLim && *toP != toLim; *fromP += 2) \
    *(*toP)++ = (GET_HI(*fromP) << 8) | GET_LO(*fromP); \
}

#define SET2(ptr, ch) \
  (((ptr)[0] = ((ch) & 0xff)), ((ptr)[1] = ((ch) >> 8)))
#define GET_LO(ptr) ((unsigned char)(ptr)[0])
#define GET_HI(ptr) ((unsigned char)(ptr)[1])

DEFINE_UTF16_TO_UTF8(little2_)
DEFINE_UTF16_TO_UTF16(little2_)

#undef SET2
#undef GET_LO
#undef GET_HI

#define SET2(ptr, ch) \
  (((ptr)[0] = ((ch) >> 8)), ((ptr)[1] = ((ch) & 0xFF)))
#define GET_LO(ptr) ((unsigned char)(ptr)[1])
#define GET_HI(ptr) ((unsigned char)(ptr)[0])

DEFINE_UTF16_TO_UTF8(big2_)
DEFINE_UTF16_TO_UTF16(big2_)

#undef SET2
#undef GET_LO
#undef GET_HI

#define LITTLE2_BYTE_TYPE(enc, p) \
 ((p)[1] == 0 \
  ? ((struct normal_encoding *)(enc))->type[(unsigned char)*(p)] \
  : unicode_byte_type((p)[1], (p)[0]))
#define LITTLE2_BYTE_TO_ASCII(enc, p) ((p)[1] == 0 ? (p)[0] : -1)
#define LITTLE2_CHAR_MATCHES(enc, p, c) ((p)[1] == 0 && (p)[0] == c)
#define LITTLE2_IS_NAME_CHAR_MINBPC(enc, p) \
  UCS2_GET_NAMING(namePages, (unsigned char)p[1], (unsigned char)p[0])
#define LITTLE2_IS_NMSTRT_CHAR_MINBPC(enc, p) \
  UCS2_GET_NAMING(nmstrtPages, (unsigned char)p[1], (unsigned char)p[0])

#ifdef XML_MIN_SIZE

static int PTRFASTCALL
little2_byteType(const ENCODING *enc, const char *p)
{
  return LITTLE2_BYTE_TYPE(enc, p);
}

static int PTRFASTCALL
little2_byteToAscii(const ENCODING *enc, const char *p)
{
  return LITTLE2_BYTE_TO_ASCII(enc, p);
}

static int PTRCALL
little2_charMatches(const ENCODING *enc, const char *p, int c)
{
  return LITTLE2_CHAR_MATCHES(enc, p, c);
}

static int PTRFASTCALL
little2_isNameMin(const ENCODING *enc, const char *p)
{
  return LITTLE2_IS_NAME_CHAR_MINBPC(enc, p);
}

static int PTRFASTCALL
little2_isNmstrtMin(const ENCODING *enc, const char *p)
{
  return LITTLE2_IS_NMSTRT_CHAR_MINBPC(enc, p);
}

#undef VTABLE
#define VTABLE VTABLE1, little2_toUtf8, little2_toUtf16

#else /* not XML_MIN_SIZE */

#undef PREFIX
#define PREFIX(ident) little2_ ## ident
#define MINBPC(enc) 2
/* CHAR_MATCHES is guaranteed to have MINBPC bytes available. */
#define BYTE_TYPE(enc, p) LITTLE2_BYTE_TYPE(enc, p)
#define BYTE_TO_ASCII(enc, p) LITTLE2_BYTE_TO_ASCII(enc, p)
#define CHAR_MATCHES(enc, p, c) LITTLE2_CHAR_MATCHES(enc, p, c)
#define IS_NAME_CHAR(enc, p, n) 0
#define IS_NAME_CHAR_MINBPC(enc, p) LITTLE2_IS_NAME_CHAR_MINBPC(enc, p)
#define IS_NMSTRT_CHAR(enc, p, n) (0)
#define IS_NMSTRT_CHAR_MINBPC(enc, p) LITTLE2_IS_NMSTRT_CHAR_MINBPC(enc, p)

#include "xmltok_impl.c"

#undef MINBPC
#undef BYTE_TYPE
#undef BYTE_TO_ASCII
#undef CHAR_MATCHES
#undef IS_NAME_CHAR
#undef IS_NAME_CHAR_MINBPC
#undef IS_NMSTRT_CHAR
#undef IS_NMSTRT_CHAR_MINBPC
#undef IS_INVALID_CHAR

#endif /* not XML_MIN_SIZE */

#ifdef XML_NS

static const struct normal_encoding little2_encoding_ns = {
  { VTABLE, 2, 0,
#if BYTEORDER == 1234
    1
#else
    0
#endif
  },
  {
#include "asciitab.h"
#include "latin1tab.h"
  },
  STANDARD_VTABLE(little2_)
};

#endif

static const struct normal_encoding little2_encoding = {
  { VTABLE, 2, 0,
#if BYTEORDER == 1234
    1
#else
    0
#endif
  },
  {
#define BT_COLON BT_NMSTRT
#include "asciitab.h"
#undef BT_COLON
#include "latin1tab.h"
  },
  STANDARD_VTABLE(little2_)
};

#if BYTEORDER != 4321

#ifdef XML_NS

static const struct normal_encoding internal_little2_encoding_ns = {
  { VTABLE, 2, 0, 1 },
  {
#include "iasciitab.h"
#include "latin1tab.h"
  },
  STANDARD_VTABLE(little2_)
};

#endif

static const struct normal_encoding internal_little2_encoding = {
  { VTABLE, 2, 0, 1 },
  {
#define BT_COLON BT_NMSTRT
#include "iasciitab.h"
#undef BT_COLON
#include "latin1tab.h"
  },
  STANDARD_VTABLE(little2_)
};

#endif


#define BIG2_BYTE_TYPE(enc, p) \
 ((p)[0] == 0 \
  ? ((struct normal_encoding *)(enc))->type[(unsigned char)(p)[1]] \
  : unicode_byte_type((p)[0], (p)[1]))
#define BIG2_BYTE_TO_ASCII(enc, p) ((p)[0] == 0 ? (p)[1] : -1)
#define BIG2_CHAR_MATCHES(enc, p, c) ((p)[0] == 0 && (p)[1] == c)
#define BIG2_IS_NAME_CHAR_MINBPC(enc, p) \
  UCS2_GET_NAMING(namePages, (unsigned char)p[0], (unsigned char)p[1])
#define BIG2_IS_NMSTRT_CHAR_MINBPC(enc, p) \
  UCS2_GET_NAMING(nmstrtPages, (unsigned char)p[0], (unsigned char)p[1])

#ifdef XML_MIN_SIZE

static int PTRFASTCALL
big2_byteType(const ENCODING *enc, const char *p)
{
  return BIG2_BYTE_TYPE(enc, p);
}

static int PTRFASTCALL
big2_byteToAscii(const ENCODING *enc, const char *p)
{
  return BIG2_BYTE_TO_ASCII(enc, p);
}

static int PTRCALL
big2_charMatches(const ENCODING *enc, const char *p, int c)
{
  return BIG2_CHAR_MATCHES(enc, p, c);
}

static int PTRFASTCALL
big2_isNameMin(const ENCODING *enc, const char *p)
{
  return BIG2_IS_NAME_CHAR_MINBPC(enc, p);
}

static int PTRFASTCALL
big2_isNmstrtMin(const ENCODING *enc, const char *p)
{
  return BIG2_IS_NMSTRT_CHAR_MINBPC(enc, p);
}

#undef VTABLE
#define VTABLE VTABLE1, big2_toUtf8, big2_toUtf16

#else /* not XML_MIN_SIZE */

#undef PREFIX
#define PREFIX(ident) big2_ ## ident
#define MINBPC(enc) 2
/* CHAR_MATCHES is guaranteed to have MINBPC bytes available. */
#define BYTE_TYPE(enc, p) BIG2_BYTE_TYPE(enc, p)
#define BYTE_TO_ASCII(enc, p) BIG2_BYTE_TO_ASCII(enc, p)
#define CHAR_MATCHES(enc, p, c) BIG2_CHAR_MATCHES(enc, p, c)
#define IS_NAME_CHAR(enc, p, n) 0
#define IS_NAME_CHAR_MINBPC(enc, p) BIG2_IS_NAME_CHAR_MINBPC(enc, p)
#define IS_NMSTRT_CHAR(enc, p, n) (0)
#define IS_NMSTRT_CHAR_MINBPC(enc, p) BIG2_IS_NMSTRT_CHAR_MINBPC(enc, p)

#include "xmltok_impl.c"

#undef MINBPC
#undef BYTE_TYPE
#undef BYTE_TO_ASCII
#undef CHAR_MATCHES
#undef IS_NAME_CHAR
#undef IS_NAME_CHAR_MINBPC
#undef IS_NMSTRT_CHAR
#undef IS_NMSTRT_CHAR_MINBPC
#undef IS_INVALID_CHAR

#endif /* not XML_MIN_SIZE */

#ifdef XML_NS

static const struct normal_encoding big2_encoding_ns = {
  { VTABLE, 2, 0,
#if BYTEORDER == 4321
  1
#else
  0
#endif
  },
  {
#include "asciitab.h"
#include "latin1tab.h"
  },
  STANDARD_VTABLE(big2_)
};

#endif

static const struct normal_encoding big2_encoding = {
  { VTABLE, 2, 0,
#if BYTEORDER == 4321
  1
#else
  0
#endif
  },
  {
#define BT_COLON BT_NMSTRT
#include "asciitab.h"
#undef BT_COLON
#include "latin1tab.h"
  },
  STANDARD_VTABLE(big2_)
};

#if BYTEORDER != 1234

#ifdef XML_NS

static const struct normal_encoding internal_big2_encoding_ns = {
  { VTABLE, 2, 0, 1 },
  {
#include "iasciitab.h"
#include "latin1tab.h"
  },
  STANDARD_VTABLE(big2_)
};

#endif

static const struct normal_encoding internal_big2_encoding = {
  { VTABLE, 2, 0, 1 },
  {
#define BT_COLON BT_NMSTRT
#include "iasciitab.h"
#undef BT_COLON
#include "latin1tab.h"
  },
  STANDARD_VTABLE(big2_)
};

#endif

#undef PREFIX

static int FASTCALL
streqci(const char *s1, const char *s2)
{
  for (;;) {
    char c1 = *s1++;
    char c2 = *s2++;
    if (ASCII_a <= c1 && c1 <= ASCII_z)
      c1 += ASCII_A - ASCII_a;
    if (ASCII_a <= c2 && c2 <= ASCII_z)
      c2 += ASCII_A - ASCII_a;
    if (c1 != c2)
      return 0;
    if (!c1)
      break;
  }
  return 1;
}

static void PTRCALL
initUpdatePosition(const ENCODING *enc, const char *ptr,
                   const char *end, POSITION *pos)
{
  normal_updatePosition(&utf8_encoding.enc, ptr, end, pos);
}

static int
toAscii(const ENCODING *enc, const char *ptr, const char *end)
{
  char buf[1];
  char *p = buf;
  XmlUtf8Convert(enc, &ptr, end, &p, p + 1);
  if (p == buf)
    return -1;
  else
    return buf[0];
}

static int FASTCALL
isSpace(int c)
{
  switch (c) {
  case 0x20:
  case 0xD:
  case 0xA:
  case 0x9:
    return 1;
  }
  return 0;
}

/* Return 1 if there's just optional white space or there's an S
   followed by name=val.
*/
static int
parsePseudoAttribute(const ENCODING *enc,
                     const char *ptr,
                     const char *end,
                     const char **namePtr,
                     const char **nameEndPtr,
                     const char **valPtr,
                     const char **nextTokPtr)
{
  int c;
  char open;
  if (ptr == end) {
    *namePtr = NULL;
    return 1;
  }
  if (!isSpace(toAscii(enc, ptr, end))) {
    *nextTokPtr = ptr;
    return 0;
  }
  do {
    ptr += enc->minBytesPerChar;
  } while (isSpace(toAscii(enc, ptr, end)));
  if (ptr == end) {
    *namePtr = NULL;
    return 1;
  }
  *namePtr = ptr;
  for (;;) {
    c = toAscii(enc, ptr, end);
    if (c == -1) {
      *nextTokPtr = ptr;
      return 0;
    }
    if (c == ASCII_EQUALS) {
      *nameEndPtr = ptr;
      break;
    }
    if (isSpace(c)) {
      *nameEndPtr = ptr;
      do {
        ptr += enc->minBytesPerChar;
      } while (isSpace(c = toAscii(enc, ptr, end)));
      if (c != ASCII_EQUALS) {
        *nextTokPtr = ptr;
        return 0;
      }
      break;
    }
    ptr += enc->minBytesPerChar;
  }
  if (ptr == *namePtr) {
    *nextTokPtr = ptr;
    return 0;
  }
  ptr += enc->minBytesPerChar;
  c = toAscii(enc, ptr, end);
  while (isSpace(c)) {
    ptr += enc->minBytesPerChar;
    c = toAscii(enc, ptr, end);
  }
  if (c != ASCII_QUOT && c != ASCII_APOS) {
    *nextTokPtr = ptr;
    return 0;
  }
  open = (char)c;
  ptr += enc->minBytesPerChar;
  *valPtr = ptr;
  for (;; ptr += enc->minBytesPerChar) {
    c = toAscii(enc, ptr, end);
    if (c == open)
      break;
    if (!(ASCII_a <= c && c <= ASCII_z)
        && !(ASCII_A <= c && c <= ASCII_Z)
        && !(ASCII_0 <= c && c <= ASCII_9)
        && c != ASCII_PERIOD
        && c != ASCII_MINUS
        && c != ASCII_UNDERSCORE) {
      *nextTokPtr = ptr;
      return 0;
    }
  }
  *nextTokPtr = ptr + enc->minBytesPerChar;
  return 1;
}

static const char KW_version[] = {
  ASCII_v, ASCII_e, ASCII_r, ASCII_s, ASCII_i, ASCII_o, ASCII_n, '\0'
};

static const char KW_encoding[] = {
  ASCII_e, ASCII_n, ASCII_c, ASCII_o, ASCII_d, ASCII_i, ASCII_n, ASCII_g, '\0'
};

static const char KW_standalone[] = {
  ASCII_s, ASCII_t, ASCII_a, ASCII_n, ASCII_d, ASCII_a, ASCII_l, ASCII_o,
  ASCII_n, ASCII_e, '\0'
};

static const char KW_yes[] = {
  ASCII_y, ASCII_e, ASCII_s,  '\0'
};

static const char KW_no[] = {
  ASCII_n, ASCII_o,  '\0'
};

static int
doParseXmlDecl(const ENCODING *(*encodingFinder)(const ENCODING *,
                                                 const char *,
                                                 const char *),
               int isGeneralTextEntity,
               const ENCODING *enc,
               const char *ptr,
               const char *end,
               const char **badPtr,
               const char **versionPtr,
               const char **versionEndPtr,
               const char **encodingName,
               const ENCODING **encoding,
               int *standalone)
{
  const char *val = NULL;
  const char *name = NULL;
  const char *nameEnd = NULL;
  ptr += 5 * enc->minBytesPerChar;
  end -= 2 * enc->minBytesPerChar;
  if (!parsePseudoAttribute(enc, ptr, end, &name, &nameEnd, &val, &ptr)
      || !name) {
    *badPtr = ptr;
    return 0;
  }
  if (!XmlNameMatchesAscii(enc, name, nameEnd, KW_version)) {