xmltok.c

GNU/Linux程序开发指南书的所有例子源代码

static
int initScan(const ENCODING *enc, int state, const char *ptr, const char *end,
	     const char **nextTokPtr)
{
  const ENCODING **encPtr;

  if (ptr == end)
    return XML_TOK_NONE;
  encPtr = ((const INIT_ENCODING *)enc)->encPtr;
  if (ptr + 1 == end) {
    switch ((unsigned char)*ptr) {
    case 0xFE:
    case 0xFF:
    case 0x00:
    case 0x3C:
      return XML_TOK_PARTIAL;
    }
  }
  else {
    switch (((unsigned char)ptr[0] << 8) | (unsigned char)ptr[1]) {
    case 0x003C:
      *encPtr = &big2_encoding.enc;
      return XmlTok(*encPtr, state, ptr, end, nextTokPtr);
    case 0xFEFF:
      *nextTokPtr = ptr + 2;
      *encPtr = &big2_encoding.enc;
      return XML_TOK_BOM;
    case 0x3C00:
      *encPtr = &little2_encoding.enc;
      return XmlTok(*encPtr, state, ptr, end, nextTokPtr);
    case 0xFFFE:
      *nextTokPtr = ptr + 2;
      *encPtr = &little2_encoding.enc;
      return XML_TOK_BOM;
    }
  }
  *encPtr = &utf8_encoding.enc;
  return XmlTok(*encPtr, state, ptr, end, nextTokPtr);
}

static
int initScanProlog(const ENCODING *enc, const char *ptr, const char *end,
		   const char **nextTokPtr)
{
  return initScan(enc, XML_PROLOG_STATE, ptr, end, nextTokPtr);
}

static
int initScanContent(const ENCODING *enc, const char *ptr, const char *end,
		    const char **nextTokPtr)
{
  return initScan(enc, XML_CONTENT_STATE, ptr, end, nextTokPtr);
}

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

const ENCODING *XmlGetUtf8InternalEncoding()
{
  return &internal_utf8_encoding.enc;
}

const ENCODING *XmlGetUtf16InternalEncoding()
{
#if BYTE_ORDER == 12
  return &internal_little2_encoding.enc;
#elif BYTE_ORDER == 21
  return &internal_big2_encoding.enc;
#else
  const short n = 1;
  return *(const char *)&n ? &internal_little2_encoding.enc : &internal_big2_encoding.enc;
#endif
}

int XmlInitEncoding(INIT_ENCODING *p, const ENCODING **encPtr, const char *name)
{
  if (name) {
    if (streqci(name, "ISO-8859-1")) {
      *encPtr = &latin1_encoding.enc;
      return 1;
    }
    if (streqci(name, "UTF-8")) {
      *encPtr = &utf8_encoding.enc;
      return 1;
    }
    if (streqci(name, "US-ASCII")) {
      *encPtr = &ascii_encoding.enc;
      return 1;
    }
    if (!streqci(name, "UTF-16"))
      return 0;
  }
  p->initEnc.scanners[XML_PROLOG_STATE] = initScanProlog;
  p->initEnc.scanners[XML_CONTENT_STATE] = initScanContent;
  p->initEnc.updatePosition = initUpdatePosition;
  p->initEnc.minBytesPerChar = 1;
  p->encPtr = encPtr;
  *encPtr = &(p->initEnc);
  return 1;
}

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 isSpace(int c)
{
  switch (c) {
  case ' ':
  case '\r':
  case '\n':
  case '\t':
    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 **valPtr,
			 const char **nextTokPtr)
{
  int c;
  char open;
  if (ptr == end) {
    *namePtr = 0;
    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 = 0;
    return 1;
  }
  *namePtr = ptr;
  for (;;) {
    c = toAscii(enc, ptr, end);
    if (c == -1) {
      *nextTokPtr = ptr;
      return 0;
    }
    if (c == '=')
      break;
    if (isSpace(c)) {
      do {
	ptr += enc->minBytesPerChar;
      } while (isSpace(c = toAscii(enc, ptr, end)));
      if (c != '=') {
	*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 != '"' && c != '\'') {
    *nextTokPtr = ptr;
    return 0;
  }
  open = c;
  ptr += enc->minBytesPerChar;
  *valPtr = ptr;
  for (;; ptr += enc->minBytesPerChar) {
    c = toAscii(enc, ptr, end);
    if (c == open)
      break;
    if (!('a' <= c && c <= 'z')
	&& !('A' <= c && c <= 'Z')
	&& !('0' <= c && c <= '9')
	&& c != '.'
	&& c != '-'
	&& c != '_') {
      *nextTokPtr = ptr;
      return 0;
    }
  }
  *nextTokPtr = ptr + enc->minBytesPerChar;
  return 1;
}

static
const ENCODING *findEncoding(const ENCODING *enc, const char *ptr, const char *end)
{
#define ENCODING_MAX 128
  char buf[ENCODING_MAX];
  char *p = buf;
  int i;
  XmlUtf8Convert(enc, &ptr, end, &p, p + ENCODING_MAX - 1);
  if (ptr != end)
    return 0;
  *p = 0;
  for (i = 0; buf[i]; i++) {
    if ('a' <= buf[i] && buf[i] <= 'z')
      buf[i] +=  'A' - 'a';
  }
  if (streqci(buf, "UTF-8"))
    return &utf8_encoding.enc;
  if (streqci(buf, "ISO-8859-1"))
    return &latin1_encoding.enc;
  if (streqci(buf, "US-ASCII"))
    return &ascii_encoding.enc;
  if (streqci(buf, "UTF-16")) {
    static const unsigned short n = 1;
    if (enc->minBytesPerChar == 2)
      return enc;
    return &big2_encoding.enc;
  }
  return 0;  
}

int XmlParseXmlDecl(int isGeneralTextEntity,
		    const ENCODING *enc,
		    const char *ptr,
		    const char *end,
		    const char **badPtr,
		    const char **versionPtr,
		    const char **encodingName,
		    const ENCODING **encoding,
		    int *standalone)
{
  const char *val = 0;
  const char *name = 0;
  ptr += 5 * enc->minBytesPerChar;
  end -= 2 * enc->minBytesPerChar;
  if (!parsePseudoAttribute(enc, ptr, end, &name, &val, &ptr) || !name) {
    *badPtr = ptr;
    return 0;
  }
  if (!XmlNameMatchesAscii(enc, name, "version")) {
    if (!isGeneralTextEntity) {
      *badPtr = name;
      return 0;
    }
  }
  else {
    if (versionPtr)
      *versionPtr = val;
    if (!parsePseudoAttribute(enc, ptr, end, &name, &val, &ptr)) {
      *badPtr = ptr;
      return 0;
    }
    if (!name) {
      if (isGeneralTextEntity) {
	/* a TextDecl must have an EncodingDecl */
	*badPtr = ptr;
	return 0;
      }
      return 1;
    }
  }
  if (XmlNameMatchesAscii(enc, name, "encoding")) {
    int c = toAscii(enc, val, end);
    if (!('a' <= c && c <= 'z') && !('A' <= c && c <= 'Z')) {
      *badPtr = val;
      return 0;
    }
    if (encodingName)
      *encodingName = val;
    if (encoding)
      *encoding = findEncoding(enc, val, ptr - enc->minBytesPerChar);
    if (!parsePseudoAttribute(enc, ptr, end, &name, &val, &ptr)) {
      *badPtr = ptr;
      return 0;
    }
    if (!name)
      return 1;
  }
  if (!XmlNameMatchesAscii(enc, name, "standalone") || isGeneralTextEntity) {
    *badPtr = name;
    return 0;
  }
  if (XmlNameMatchesAscii(enc, val, "yes")) {
    if (standalone)
      *standalone = 1;
  }
  else if (XmlNameMatchesAscii(enc, val, "no")) {
    if (standalone)
      *standalone = 0;
  }
  else {
    *badPtr = val;
    return 0;
  }
  while (isSpace(toAscii(enc, ptr, end)))
    ptr += enc->minBytesPerChar;
  if (ptr != end) {
    *badPtr = ptr;
    return 0;
  }
  return 1;
}

static
int checkCharRefNumber(int result)
{
  switch (result >> 8) {
  case 0xD8: case 0xD9: case 0xDA: case 0xDB:
  case 0xDC: case 0xDD: case 0xDE: case 0xDF:
    return -1;
  case 0:
    if (latin1_encoding.type[result] == BT_NONXML)
      return -1;
    break;
  case 0xFF:
    if (result == 0xFFFE || result == 0xFFFF)
      return -1;
    break;
  }
  return result;
}

int XmlUtf8Encode(int c, char *buf)
{
  enum {
    /* minN is minimum legal resulting value for N byte sequence */
    min2 = 0x80,
    min3 = 0x800,
    min4 = 0x10000
  };

  if (c < 0)
    return 0;
  if (c < min2) {
    buf[0] = (c | UTF8_cval1);
    return 1;
  }
  if (c < min3) {
    buf[0] = ((c >> 6) | UTF8_cval2);
    buf[1] = ((c & 0x3f) | 0x80);
    return 2;
  }
  if (c < min4) {
    buf[0] = ((c >> 12) | UTF8_cval3);
    buf[1] = (((c >> 6) & 0x3f) | 0x80);
    buf[2] = ((c & 0x3f) | 0x80);
    return 3;
  }
  if (c < 0x110000) {
    buf[0] = ((c >> 18) | UTF8_cval4);
    buf[1] = (((c >> 12) & 0x3f) | 0x80);
    buf[2] = (((c >> 6) & 0x3f) | 0x80);
    buf[3] = ((c & 0x3f) | 0x80);
    return 4;
  }
  return 0;
}

int XmlUtf16Encode(int charNum, unsigned short *buf)
{
  if (charNum < 0)
    return 0;
  if (charNum < 0x10000) {
    buf[0] = charNum;
    return 1;
  }
  if (charNum < 0x110000) {
    charNum -= 0x10000;
    buf[0] = (charNum >> 10) + 0xD800;
    buf[1] = (charNum & 0x3FF) + 0xDC00;
    return 2;
  }
  return 0;
}

struct unknown_encoding {
  struct normal_encoding normal;
  int (*convert)(void *userData, const char *p);
  void *userData;
  unsigned short utf16[256];
  char utf8[256][4];
};

int XmlSizeOfUnknownEncoding()
{
  return sizeof(struct unknown_encoding);
}

static
int unknown_isName(const ENCODING *enc, const char *p)
{
  int c = ((const struct unknown_encoding *)enc)
	  ->convert(((const struct unknown_encoding *)enc)->userData, p);
  if (c & ~0xFFFF)
    return 0;
  return UCS2_GET_NAMING(namePages, c >> 8, c & 0xFF);
}

static
int unknown_isNmstrt(const ENCODING *enc, const char *p)
{
  int c = ((const struct unknown_encoding *)enc)
	  ->convert(((const struct unknown_encoding *)enc)->userData, p);
  if (c & ~0xFFFF)
    return 0;
  return UCS2_GET_NAMING(nmstrtPages, c >> 8, c & 0xFF);
}

static
int unknown_isInvalid(const ENCODING *enc, const char *p)
{
  int c = ((const struct unknown_encoding *)enc)
	   ->convert(((const struct unknown_encoding *)enc)->userData, p);
  return (c & ~0xFFFF) || checkCharRefNumber(c) < 0;
}

static
void unknown_toUtf8(const ENCODING *enc,
		    const char **fromP, const char *fromLim,
		    char **toP, const char *toLim)
{
  char buf[XML_UTF8_ENCODE_MAX];
  for (;;) {
    const char *utf8;
    int n;
    if (*fromP == fromLim)
      break;
    utf8 = ((const struct unknown_encoding *)enc)->utf8[(unsigned char)**fromP];
    n = *utf8++;
    if (n == 0) {
      int c = ((const struct unknown_encoding *)enc)
	      ->convert(((const struct unknown_encoding *)enc)->userData, *fromP);
      n = XmlUtf8Encode(c, buf);
      if (n > toLim - *toP)
	break;
      utf8 = buf;
      *fromP += ((const struct normal_encoding *)enc)->type[(unsigned char)**fromP]
	         - (BT_LEAD2 - 2);
    }
    else {
      if (n > toLim - *toP)
	break;
      (*fromP)++;
    }
    do {
      *(*toP)++ = *utf8++;
    } while (--n != 0);
  }
}

static
void unknown_toUtf16(const ENCODING *enc,
		     const char **fromP, const char *fromLim,
		     unsigned short **toP, const unsigned short *toLim)
{
  while (*fromP != fromLim && *toP != toLim) {
    unsigned short c
      = ((const struct unknown_encoding *)enc)->utf16[(unsigned char)**fromP];
    if (c == 0) {
      c = (unsigned short)((const struct unknown_encoding *)enc)
	   ->convert(((const struct unknown_encoding *)enc)->userData, *fromP);
      *fromP += ((const struct normal_encoding *)enc)->type[(unsigned char)**fromP]
	         - (BT_LEAD2 - 2);
    }
    else
      (*fromP)++;
    *(*toP)++ = c;
  }
}

ENCODING *
XmlInitUnknownEncoding(void *mem,
		       int *table,
		       int (*convert)(void *userData, const char *p),
		       void *userData)
{
  int i;
  struct unknown_encoding *e = mem;
  for (i = 0; i < sizeof(struct normal_encoding); i++)
    ((char *)mem)[i] = ((char *)&latin1_encoding)[i];
  for (i = 0; i < 128; i++)
    if (latin1_encoding.type[i] != BT_OTHER
        && latin1_encoding.type[i] != BT_NONXML
	&& table[i] != i)
      return 0;
  for (i = 0; i < 256; i++) {
    int c = table[i];
    if (c == -1) {
      e->normal.type[i] = BT_MALFORM;
      /* This shouldn't really get used. */
      e->utf16[i] = 0xFFFF;
      e->utf8[i][0] = 1;
      e->utf8[i][1] = 0;
    }
    else if (c < 0) {
      if (c < -4)
	return 0;
      e->normal.type[i] = BT_LEAD2 - (c + 2);
      e->utf8[i][0] = 0;
      e->utf16[i] = 0;
    }
    else if (c < 0x80) {
      if (latin1_encoding.type[c] != BT_OTHER
	  && latin1_encoding.type[c] != BT_NONXML
	  && c != i)
	return 0;
      e->normal.type[i] = latin1_encoding.type[c];
      e->utf8[i][0] = 1;
      e->utf8[i][1] = (char)c;
      e->utf16[i] = c == 0 ? 0xFFFF : c;
    }
    else if (checkCharRefNumber(c) < 0) {
      e->normal.type[i] = BT_NONXML;
      /* This shouldn't really get used. */
      e->utf16[i] = 0xFFFF;
      e->utf8[i][0] = 1;
      e->utf8[i][1] = 0;
    }
    else {
      if (c > 0xFFFF)
	return 0;
      if (UCS2_GET_NAMING(nmstrtPages, c >> 8, c & 0xff))
	e->normal.type[i] = BT_NMSTRT;
      else if (UCS2_GET_NAMING(namePages, c >> 8, c & 0xff))
	e->normal.type[i] = BT_NAME;
      else
	e->normal.type[i] = BT_OTHER;
      e->utf8[i][0] = (char)XmlUtf8Encode(c, e->utf8[i] + 1);
      e->utf16[i] = c;
    }
  }
  e->userData = userData;
  e->convert = convert;
  if (convert) {
    e->normal.isName2 = unknown_isName;
    e->normal.isName3 = unknown_isName;
    e->normal.isName4 = unknown_isName;
    e->normal.isNmstrt2 = unknown_isNmstrt;
    e->normal.isNmstrt3 = unknown_isNmstrt;
    e->normal.isNmstrt4 = unknown_isNmstrt;
    e->normal.isInvalid2 = unknown_isInvalid;
    e->normal.isInvalid3 = unknown_isInvalid;
    e->normal.isInvalid4 = unknown_isInvalid;
  }
  e->normal.enc.utf8Convert = unknown_toUtf8;
  e->normal.enc.utf16Convert = unknown_toUtf16;
  return &(e->normal.enc);
}