xmlparse.c

GNU/Linux程序开发指南书的所有例子源代码

    case XML_TOK_DATA_CHARS:
      if (!poolAppend(pool, enc, ptr, next))
	return XML_ERROR_NO_MEMORY;
      break;
      break;
    case XML_TOK_TRAILING_CR:
      next = ptr + enc->minBytesPerChar;
      /* fall through */
    case XML_TOK_ATTRIBUTE_VALUE_S:
    case XML_TOK_DATA_NEWLINE:
      if (!isCdata && (poolLength(pool) == 0 || poolLastChar(pool) == XML_T(' ')))
	break;
      if (!poolAppendChar(pool, XML_T(' ')))
	return XML_ERROR_NO_MEMORY;
      break;
    case XML_TOK_ENTITY_REF:
      {
	const XML_Char *name;
	ENTITY *entity;
	XML_Char ch = XmlPredefinedEntityName(enc,
					      ptr + enc->minBytesPerChar,
					      next - enc->minBytesPerChar);
	if (ch) {
	  if (!poolAppendChar(pool, ch))
  	    return XML_ERROR_NO_MEMORY;
	  break;
	}
	name = poolStoreString(&temp2Pool, enc,
			       ptr + enc->minBytesPerChar,
			       next - enc->minBytesPerChar);
	if (!name)
	  return XML_ERROR_NO_MEMORY;
	entity = (ENTITY *)lookup(&dtd.generalEntities, name, 0);
	poolDiscard(&temp2Pool);
	if (!entity) {
	  if (dtd.complete) {
	    if (enc == encoding)
	      eventPtr = ptr;
	    return XML_ERROR_UNDEFINED_ENTITY;
	  }
	}
	else if (entity->open) {
	  if (enc == encoding)
	    eventPtr = ptr;
	  return XML_ERROR_RECURSIVE_ENTITY_REF;
	}
	else if (entity->notation) {
	  if (enc == encoding)
	    eventPtr = ptr;
	  return XML_ERROR_BINARY_ENTITY_REF;
	}
	else if (!entity->textPtr) {
	  if (enc == encoding)
	    eventPtr = ptr;
  	  return XML_ERROR_ATTRIBUTE_EXTERNAL_ENTITY_REF;
	}
	else {
	  enum XML_Error result;
	  const XML_Char *textEnd = entity->textPtr + entity->textLen;
	  entity->open = 1;
	  result = appendAttributeValue(parser, internalEnc, isCdata, (char *)entity->textPtr, (char *)textEnd, pool);
	  entity->open = 0;
	  if (result)
	    return result;
	}
      }
      break;
    default:
      abort();
    }
    ptr = next;
  }
  /* not reached */
}

static
enum XML_Error storeEntityValue(XML_Parser parser,
				const char *entityTextPtr,
				const char *entityTextEnd)
{
  const ENCODING *internalEnc = XmlGetInternalEncoding();
  STRING_POOL *pool = &(dtd.pool);
  entityTextPtr += encoding->minBytesPerChar;
  entityTextEnd -= encoding->minBytesPerChar;
  for (;;) {
    const char *next;
    int tok = XmlEntityValueTok(encoding, entityTextPtr, entityTextEnd, &next);
    switch (tok) {
    case XML_TOK_PARAM_ENTITY_REF:
      eventPtr = entityTextPtr;
      return XML_ERROR_SYNTAX;
    case XML_TOK_NONE:
      if (declEntity) {
	declEntity->textPtr = pool->start;
	declEntity->textLen = pool->ptr - pool->start;
	poolFinish(pool);
      }
      else
	poolDiscard(pool);
      return XML_ERROR_NONE;
    case XML_TOK_ENTITY_REF:
    case XML_TOK_DATA_CHARS:
      if (!poolAppend(pool, encoding, entityTextPtr, next))
	return XML_ERROR_NO_MEMORY;
      break;
    case XML_TOK_TRAILING_CR:
      next = entityTextPtr + encoding->minBytesPerChar;
      /* fall through */
    case XML_TOK_DATA_NEWLINE:
      if (pool->end == pool->ptr && !poolGrow(pool))
	return XML_ERROR_NO_MEMORY;
      *(pool->ptr)++ = XML_T('\n');
      break;
    case XML_TOK_CHAR_REF:
      {
	XML_Char buf[XML_ENCODE_MAX];
	int i;
	int n = XmlCharRefNumber(encoding, entityTextPtr);
	if (n < 0) {
	  eventPtr = entityTextPtr;
	  return XML_ERROR_BAD_CHAR_REF;
	}
	n = XmlEncode(n, (ICHAR *)buf);
	if (!n) {
	  eventPtr = entityTextPtr;
	  return XML_ERROR_BAD_CHAR_REF;
	}
	for (i = 0; i < n; i++) {
	  if (pool->end == pool->ptr && !poolGrow(pool))
	    return XML_ERROR_NO_MEMORY;
	  *(pool->ptr)++ = buf[i];
	}
      }
      break;
    case XML_TOK_PARTIAL:
      eventPtr = entityTextPtr;
      return XML_ERROR_INVALID_TOKEN;
    case XML_TOK_INVALID:
      eventPtr = next;
      return XML_ERROR_INVALID_TOKEN;
    default:
      abort();
    }
    entityTextPtr = next;
  }
  /* not reached */
}

static void
normalizeLines(XML_Char *s)
{
  XML_Char *p;
  for (;; s++) {
    if (*s == XML_T('\0'))
      return;
    if (*s == XML_T('\r'))
      break;
  }
  p = s;
  do {
    if (*s == XML_T('\r')) {
      *p++ = XML_T('\n');
      if (*++s == XML_T('\n'))
        s++;
    }
    else
      *p++ = *s++;
  } while (*s);
  *p = XML_T('\0');
}

static int
reportProcessingInstruction(XML_Parser parser, const ENCODING *enc, const char *start, const char *end)
{
  const XML_Char *target;
  XML_Char *data;
  const char *tem;
  if (!processingInstructionHandler) {
    if (defaultHandler)
      reportDefault(parser, enc, start, end);
    return 1;
  }
  start += enc->minBytesPerChar * 2;
  tem = start + XmlNameLength(enc, start);
  target = poolStoreString(&tempPool, enc, start, tem);
  if (!target)
    return 0;
  poolFinish(&tempPool);
  data = poolStoreString(&tempPool, enc,
			XmlSkipS(enc, tem),
			end - enc->minBytesPerChar*2);
  if (!data)
    return 0;
  normalizeLines(data);
  processingInstructionHandler(handlerArg, target, data);
  poolClear(&tempPool);
  return 1;
}

static void
reportDefault(XML_Parser parser, const ENCODING *enc, const char *s, const char *end)
{
  if (MUST_CONVERT(enc, s)) {
    for (;;) {
      ICHAR *dataPtr = (ICHAR *)dataBuf;
      XmlConvert(enc, &s, end, &dataPtr, (ICHAR *)dataBufEnd);
      if (s == end) {
	defaultHandler(handlerArg, dataBuf, dataPtr - (ICHAR *)dataBuf);
	break;
      }
      if (enc == encoding) {
	eventEndPtr = s;
	defaultHandler(handlerArg, dataBuf, dataPtr - (ICHAR *)dataBuf);
	eventPtr = s;
      }
      else
	defaultHandler(handlerArg, dataBuf, dataPtr - (ICHAR *)dataBuf);
    }
  }
  else
    defaultHandler(handlerArg, (XML_Char *)s, (XML_Char *)end - (XML_Char *)s);
}


static int
defineAttribute(ELEMENT_TYPE *type, ATTRIBUTE_ID *attId, int isCdata, const XML_Char *value)
{
  DEFAULT_ATTRIBUTE *att;
  if (type->nDefaultAtts == type->allocDefaultAtts) {
    if (type->allocDefaultAtts == 0) {
      type->allocDefaultAtts = 8;
      type->defaultAtts = malloc(type->allocDefaultAtts*sizeof(DEFAULT_ATTRIBUTE));
    }
    else {
      type->allocDefaultAtts *= 2;
      type->defaultAtts = realloc(type->defaultAtts,
				  type->allocDefaultAtts*sizeof(DEFAULT_ATTRIBUTE));
    }
    if (!type->defaultAtts)
      return 0;
  }
  att = type->defaultAtts + type->nDefaultAtts;
  att->id = attId;
  att->value = value;
  att->isCdata = isCdata;
  if (!isCdata)
    attId->maybeTokenized = 1;
  type->nDefaultAtts += 1;
  return 1;
}

static ATTRIBUTE_ID *
getAttributeId(XML_Parser parser, const ENCODING *enc, const char *start, const char *end)
{
  ATTRIBUTE_ID *id;
  const XML_Char *name;
  if (!poolAppendChar(&dtd.pool, XML_T('\0')))
    return 0;
  name = poolStoreString(&dtd.pool, enc, start, end);
  if (!name)
    return 0;
  ++name;
  id = (ATTRIBUTE_ID *)lookup(&dtd.attributeIds, name, sizeof(ATTRIBUTE_ID));
  if (!id)
    return 0;
  if (id->name != name)
    poolDiscard(&dtd.pool);
  else
    poolFinish(&dtd.pool);
  return id;
}

static
const XML_Char *getOpenEntityNames(XML_Parser parser)
{
  HASH_TABLE_ITER iter;

  hashTableIterInit(&iter, &(dtd.generalEntities));
  for (;;) {
    const XML_Char *s;
    ENTITY *e = (ENTITY *)hashTableIterNext(&iter);
    if (!e)
      break;
    if (!e->open)
      continue;
    if (poolLength(&tempPool) > 0 && !poolAppendChar(&tempPool, XML_T(' ')))
      return 0;
    for (s = e->name; *s; s++)
      if (!poolAppendChar(&tempPool, *s))
        return 0;
  }

  if (!poolAppendChar(&tempPool, XML_T('\0')))
    return 0;
  return tempPool.start;
}

static
int setOpenEntityNames(XML_Parser parser, const XML_Char *openEntityNames)
{
  const XML_Char *s = openEntityNames;
  while (*openEntityNames != XML_T('\0')) {
    if (*s == XML_T(' ') || *s == XML_T('\0')) {
      ENTITY *e;
      if (!poolAppendChar(&tempPool, XML_T('\0')))
	return 0;
      e = (ENTITY *)lookup(&dtd.generalEntities, poolStart(&tempPool), 0);
      if (e)
	e->open = 1;
      if (*s == XML_T(' '))
	s++;
      openEntityNames = s;
      poolDiscard(&tempPool);
    }
    else {
      if (!poolAppendChar(&tempPool, *s))
	return 0;
      s++;
    }
  }
  return 1;
}


static
void normalizePublicId(XML_Char *publicId)
{
  XML_Char *p = publicId;
  XML_Char *s;
  for (s = publicId; *s; s++) {
    switch (*s) {
    case XML_T(' '):
    case XML_T('\r'):
    case XML_T('\n'):
      if (p != publicId && p[-1] != XML_T(' '))
	*p++ = XML_T(' ');
      break;
    default:
      *p++ = *s;
    }
  }
  if (p != publicId && p[-1] == XML_T(' '))
    --p;
  *p = XML_T('\0');
}

static int dtdInit(DTD *p)
{
  poolInit(&(p->pool));
  hashTableInit(&(p->generalEntities));
  hashTableInit(&(p->elementTypes));
  hashTableInit(&(p->attributeIds));
  p->complete = 1;
  p->standalone = 0;
  p->base = 0;
  return 1;
}

static void dtdDestroy(DTD *p)
{
  HASH_TABLE_ITER iter;
  hashTableIterInit(&iter, &(p->elementTypes));
  for (;;) {
    ELEMENT_TYPE *e = (ELEMENT_TYPE *)hashTableIterNext(&iter);
    if (!e)
      break;
    if (e->allocDefaultAtts != 0)
      free(e->defaultAtts);
  }
  hashTableDestroy(&(p->generalEntities));
  hashTableDestroy(&(p->elementTypes));
  hashTableDestroy(&(p->attributeIds));
  poolDestroy(&(p->pool));
}

/* Do a deep copy of the DTD.  Return 0 for out of memory; non-zero otherwise.
The new DTD has already been initialized. */

static int dtdCopy(DTD *newDtd, const DTD *oldDtd)
{
  HASH_TABLE_ITER iter;

  if (oldDtd->base) {
    const XML_Char *tem = poolCopyString(&(newDtd->pool), oldDtd->base);
    if (!tem)
      return 0;
    newDtd->base = tem;
  }

  hashTableIterInit(&iter, &(oldDtd->attributeIds));

  /* Copy the attribute id table. */

  for (;;) {
    ATTRIBUTE_ID *newA;
    const XML_Char *name;
    const ATTRIBUTE_ID *oldA = (ATTRIBUTE_ID *)hashTableIterNext(&iter);

    if (!oldA)
      break;
    /* Remember to allocate the scratch byte before the name. */
    if (!poolAppendChar(&(newDtd->pool), XML_T('\0')))
      return 0;
    name = poolCopyString(&(newDtd->pool), oldA->name);
    if (!name)
      return 0;
    ++name;
    newA = (ATTRIBUTE_ID *)lookup(&(newDtd->attributeIds), name, sizeof(ATTRIBUTE_ID));
    if (!newA)
      return 0;
    newA->maybeTokenized = oldA->maybeTokenized;
  }

  /* Copy the element type table. */

  hashTableIterInit(&iter, &(oldDtd->elementTypes));

  for (;;) {
    int i;
    ELEMENT_TYPE *newE;
    const XML_Char *name;
    const ELEMENT_TYPE *oldE = (ELEMENT_TYPE *)hashTableIterNext(&iter);
    if (!oldE)
      break;
    name = poolCopyString(&(newDtd->pool), oldE->name);
    if (!name)
      return 0;
    newE = (ELEMENT_TYPE *)lookup(&(newDtd->elementTypes), name, sizeof(ELEMENT_TYPE));
    if (!newE)
      return 0;
    newE->defaultAtts = (DEFAULT_ATTRIBUTE *)malloc(oldE->nDefaultAtts * sizeof(DEFAULT_ATTRIBUTE));
    if (!newE->defaultAtts)
      return 0;
    newE->allocDefaultAtts = newE->nDefaultAtts = oldE->nDefaultAtts;
    for (i = 0; i < newE->nDefaultAtts; i++) {
      newE->defaultAtts[i].id = (ATTRIBUTE_ID *)lookup(&(newDtd->attributeIds), oldE->defaultAtts[i].id->name, 0);
      newE->defaultAtts[i].isCdata = oldE->defaultAtts[i].isCdata;
      if (oldE->defaultAtts[i].value) {
	newE->defaultAtts[i].value = poolCopyString(&(newDtd->pool), oldE->defaultAtts[i].value);
	if (!newE->defaultAtts[i].value)
  	  return 0;
      }
      else
	newE->defaultAtts[i].value = 0;
    }
  }

  /* Copy the entity table. */

  hashTableIterInit(&iter, &(oldDtd->generalEntities));

  for (;;) {
    ENTITY *newE;
    const XML_Char *name;
    const ENTITY *oldE = (ENTITY *)hashTableIterNext(&iter);
    if (!oldE)
      break;
    name = poolCopyString(&(newDtd->pool), oldE->name);
    if (!name)
      return 0;
    newE = (ENTITY *)lookup(&(newDtd->generalEntities), name, sizeof(ENTITY));
    if (!newE)
      return 0;
    if (oldE->systemId) {
      const XML_Char *tem = poolCopyString(&(newDtd->pool), oldE->systemId);
      if (!tem)
	return 0;
      newE->systemId = tem;
      if (oldE->base) {
	if (oldE->base == oldDtd->base)
	  newE->base = newDtd->base;
	tem = poolCopyString(&(newDtd->pool), oldE->base);
	if (!tem)
	  return 0;
	newE->base = tem;
      }
    }
    else {
      const XML_Char *tem = poolCopyStringN(&(newDtd->pool),