xmlparser.java

克里斯托弗编写的linux web编程的所有实验源代码

	}
      }
    }

    // Check for surrogates: 00000000 0000xxxx yyyyyyyy zzzzzzzz
    //  (1101|10xx|xxyy|yyyy + 1101|11yy|zzzz|zzzz: 
    if (value <= 0x0000ffff) {
				// no surrogates needed
      dataBufferAppend((char)value);
    } else if (value <= 0x000fffff) {
				// > 16 bits, surrogate needed
      dataBufferAppend((char)(0xd8 | ((value & 0x000ffc00) >> 10)));
      dataBufferAppend((char)(0xdc | (value & 0x0003ff)));
    } else {
				// too big for surrogate
      error("character reference " + value + " is too large for UTF-16",
	    new Integer(value).toString(), null);
    }
  }


  /**
    * Parse a reference.
    * [69] EntityRef ::= '&' Name ';'
    * *NOTE: the '&' has already been read.
    * @param externalAllowed External entities are allowed here.
    */
  void parseEntityRef (boolean externalAllowed)
    throws java.lang.Exception
  {
    String name;

    name = readNmtoken(true);
    require(';');
    switch (getEntityType(name)) {
    case ENTITY_UNDECLARED:
      error("reference to undeclared entity", name, null);
      break;
    case ENTITY_INTERNAL:
      pushString(name, getEntityValue(name));
      break;
    case ENTITY_TEXT:
      if (externalAllowed) {
	pushURL(name, getEntityPublicId(name),
		getEntitySystemId(name),
		null, null, null);
      } else {
	error("reference to external entity in attribute value.", name, null);
      }
      break;
    case ENTITY_NDATA:
      if (externalAllowed) {
	error("data entity reference in content", name, null);
      } else {
	error("reference to external entity in attribute value.", name, null);
      }
      break;
    }
  }


  /**
    * Parse a parameter entity reference.
    * [70] PEReference ::= '%' Name ';'
    * *NOTE: the '%' has already been read.
    */
  void parsePEReference (boolean isEntityValue)
    throws java.lang.Exception
  {
    String name;

    name = "%" + readNmtoken(true);
    require(';');
    switch (getEntityType(name)) {
    case ENTITY_UNDECLARED:
      error("reference to undeclared parameter entity", name, null);
      break;
    case ENTITY_INTERNAL:
      if (isEntityValue) {
	pushString(name, getEntityValue(name));
      } else {
	pushString(name, " " + getEntityValue(name) + ' ');
      }
      break;
    case ENTITY_TEXT:
      if (isEntityValue) {
	pushString(null, " ");
      }
      pushURL(name, getEntityPublicId(name),
	      getEntitySystemId(name),
	      null, null, null);
      if (isEntityValue) {
	pushString(null, " ");
      }
      break;
    }
  }


  /**
    * Parse an entity declaration.
    * [71] EntityDecl ::= '<!ENTITY' S %Name S %EntityDef S? '>'
    *                   | '<!ENTITY' S '%' S %Name S %EntityDef S? '>'
    * [72] EntityDef ::= EntityValue | ExternalDef
    * [73] ExternalDef ::= ExternalID %NDataDecl?
    * [74] ExternalID ::= 'SYSTEM' S SystemLiteral
    *                   | 'PUBLIC' S PubidLiteral S SystemLiteral
    * [75] NDataDecl ::= S %'NDATA' S %Name
    * *NOTE: the '<!ENTITY' has already been read.
    */
  void parseEntityDecl ()
    throws java.lang.Exception
  {
    char c;
    boolean peFlag = false;
    String name, value, notationName, ids[];

				// Check for a parameter entity.
    requireWhitespace();
    if (tryRead('%')) {
      peFlag = true;
      requireWhitespace();
    }

				// Read the entity name, and prepend
				// '%' if necessary.
    name = readNmtoken(true);
    if (peFlag) {
      name = "%" + name;
    }

				// Read the entity value.
    requireWhitespace();
    c = readCh();
    unread(c);
    if (c == '"' || c == '\'') {
				// Internal entity.
      context = CONTEXT_ENTITYVALUE;
      value = readLiteral(LIT_CHAR_REF|LIT_PE_REF);
      context = CONTEXT_DTD;
      setInternalEntity(name,value);
    } else {
				// Read the external IDs
      ids = readExternalIds(false);
      if (ids[1] == null) {
	error("system identifer missing", name, null);
      }

				// Check for NDATA declaration.
      skipWhitespace();
      if (tryRead("NDATA")) {
	requireWhitespace();
	notationName = readNmtoken(true);
	setExternalDataEntity(name, ids[0], ids[1], notationName);
      } else {
	setExternalTextEntity(name, ids[0], ids[1]);
      }
    }

				// Finish the declaration.
    skipWhitespace();
    require('>');
  }


  /**
    * Parse a notation declaration.
    * [81] NotationDecl ::= '<!NOTATION' S %Name S %ExternalID S? '>'
    * *NOTE: the '<!NOTATION' has already been read.
    */
  void parseNotationDecl ()
    throws java.lang.Exception
  {
    String nname, ids[];
    

    requireWhitespace();
    nname = readNmtoken(true);

    requireWhitespace();

				// Read the external identifiers.
    ids = readExternalIds(true);
    if (ids[0] == null && ids[1] == null) {
      error("external identifer missing", nname, null);
    }

				// Register the notation.
    setNotation(nname, ids[0], ids[1]);

    skipWhitespace();
    require('>');
  }


  /**
    * Parse PCDATA.
    * <pre>
    * [16] PCData ::= [^&lt;&amp;]*
    * </pre>
    * <p>The trick here is that the data stays in the dataBuffer without
    * necessarily being converted to a string right away.
    */
  void parsePCData ()
    throws java.lang.Exception
  {
    char c;

				// Start with a little cheat -- in most
				// cases, the entire sequence of
				// character data will already be in
				// the readBuffer; if not, fall through to
				// the normal approach.
    if (USE_CHEATS) {
      int lineAugment = 0;
      int columnAugment = 0;

      loop: for (int i = readBufferPos; i < readBufferLength; i++) {
	switch (readBuffer[i]) {
	case '\n':
	  lineAugment++;
	  columnAugment = 0;
	  break;
	case '&':
	case '<':
	  int start = readBufferPos;
	  columnAugment++;
	  readBufferPos = i;
	  if (lineAugment > 0) {
	    line += lineAugment;
	    column = columnAugment;
	  } else {
	    column += columnAugment;
	  }
	  dataBufferAppend(readBuffer, start, i-start);
	  return;
	default:
	  columnAugment++;
	}
      }
    }

				// OK, the cheat didn't work; start over
				// and do it by the book.
    while (true) {
      c = readCh();
      switch (c) {
      case '<':
      case '&':
	unread(c);
	return;
      default:
	dataBufferAppend(c);
	break;
      }
    }
  }



  //////////////////////////////////////////////////////////////////////
  // High-level reading and scanning methods.
  //////////////////////////////////////////////////////////////////////

  /**
    * Require whitespace characters.
    * [1] S ::= (#x20 | #x9 | #xd | #xa)+
    */
  void requireWhitespace ()
    throws java.lang.Exception
  {
    char c = readCh();
    if (isWhitespace(c)) {
      skipWhitespace();
    } else {
      error("whitespace expected", c, null);
    }
  }


  /**
    * Parse whitespace characters, and leave them in the data buffer.
    */
  void parseWhitespace ()
    throws java.lang.Exception
  {
    char c = readCh();
    while (isWhitespace(c)) {
      dataBufferAppend(c);
      c = readCh();
    }
    unread(c);
  }


  /**
    * Skip whitespace characters.
    * [1] S ::= (#x20 | #x9 | #xd | #xa)+
    */
  void skipWhitespace ()
    throws java.lang.Exception
  {
				// Start with a little cheat.  Most of
				// the time, the white space will fall
				// within the current read buffer; if
				// not, then fall through.
    if (USE_CHEATS) {
      int lineAugment = 0;
      int columnAugment = 0;

      loop: for (int i = readBufferPos; i < readBufferLength; i++) {
	switch (readBuffer[i]) {
	case ' ':
	case '\t':
	case '\r':
	  columnAugment++;
	  break;
	case '\n':
	  lineAugment++;
	  columnAugment = 0;
	  break;
	case '%':
	  if (context == CONTEXT_DTD || context == CONTEXT_ENTITYVALUE) {
	    break loop;
	  } // else fall through...
	default:
	  readBufferPos = i;
	  if (lineAugment > 0) {
	    line += lineAugment;
	    column = columnAugment;
	  } else {
	    column += columnAugment;
	  }
	  return;
	}
      }
    }

				// OK, do it by the book.
    char c = readCh();
    while (isWhitespace(c)) {
      c = readCh();
    }
    unread(c);
  }


  /**
    * Read a name or name token.
    * [5] Name ::= (Letter | '_' | ':') (NameChar)*
    * [7] Nmtoken ::= (NameChar)+
    * *NOTE: [6] is implemented implicitly where required.
    */
  String readNmtoken (boolean isName)
    throws java.lang.Exception
  {
    char c;

    if (USE_CHEATS) {
      loop: for (int i = readBufferPos; i < readBufferLength; i++) {
	switch (readBuffer[i]) {
	case '%':
	  if (context == CONTEXT_DTD || context == CONTEXT_ENTITYVALUE) {
	    break loop;
	  } // else fall through...
	case '<':
	case '>':
	case '&':
	case ',':
	case '|':
	case '*':
	case '+':
	case '?':
	case ')':
	case '=':
	case '\'':
	case '"':
	case '[':
	case ' ':
	case '\t':
	case '\r':
	case '\n':
	case ';':
	case '/':
	case '#':
	  int start = readBufferPos;
	  if (i == start) {
	    error("name expected", readBuffer[i], null);
	  }
	  readBufferPos = i;
	  return intern(readBuffer, start, i - start);
	}
      }
    }

    nameBufferPos = 0;

				// Read the first character.
    loop: while (true) {
      c = readCh();
      switch (c) {
      case '%':
      case '<':
      case '>':
      case '&':
      case ',':
      case '|':
      case '*':
      case '+':
      case '?':
      case ')':
      case '=':
      case '\'':
      case '"':
      case '[':
      case ' ':
      case '\t':
      case '\n':
      case '\r':
      case ';':
      case '/':
	unread(c);
	if (nameBufferPos == 0) {
	  error("name expected", null, null);
	}
	String s = intern(nameBuffer,0,nameBufferPos);
	nameBufferPos = 0;
	return s;
      default:
	nameBuffer =
	  (char[])extendArray(nameBuffer, nameBuffer.length, nameBufferPos);
	nameBuffer[nameBufferPos++] = c;
      }
    }
  }


  /**
    * Read a literal.
    * [10] AttValue ::= '"' ([^<&"] | Reference)* '"'
    *                 | "'" ([^<&'] | Reference)* "'"
    * [11] SystemLiteral ::= '"' URLchar* '"' | "'" (URLchar - "'")* "'"
    * [13] PubidLiteral ::= '"' PubidChar* '"' | "'" (PubidChar - "'")* "'"
    * [9] EntityValue ::= '"' ([^%&"] | PEReference | Reference)* '"'
    *                   | "'" ([^%&'] | PEReference | Reference)* "'"
    */
  String readLiteral (int flags)
    throws java.lang.Exception
  {
    char delim, c;
    int startLine = line;

				// Find the delimiter.
    delim = readCh();
    if (delim != '"' && delim != '\'' && delim != (char)0) {
      error("expected '\"' or \"'\"", delim, null);
      return null;
    }

				// Read the literal.
    try {
      c = readCh();

    loop: while (c != delim) {
      switch (c) {
				// Literals never have line ends
      case '\n':
      case '\r':
	c = ' ';
	break;
				// References may be allowed
      case '&':
	if ((flags & LIT_CHAR_REF) > 0) {
	  c = readCh();
	  if (c == '#') {
	    parseCharRef();
	    c = readCh();
	    continue loop;		// check the next character
	  } else if ((flags & LIT_ENTITY_REF) > 0) {
	    unread(c);
	    parseEntityRef(false);
	    c = readCh();
	    continue loop;
	  } else {
	    dataBufferAppend('&');
	  }
	}
	break;

      default:
	break;
      }
      dataBufferAppend(c);
      c = readCh();
    }
    } catch (EOFException e) {
      error("end of input while looking for delimiter (started on line "
	    + startLine + ')', null, new Character(delim).toString());
    }

				// Normalise whitespace if necessary.
    if ((flags & LIT_NORMALIZE) > 0) {
      dataBufferNormalize();
    }

				// Return the value.
    return dataBufferToString();
  }


  /**
    * Try reading external identifiers.
    * <p>The system identifier is not required for notations.
    * @param inNotation Are we in a notation?
    * @return A two-member String array containing the identifiers.
    */
  String[] readExternalIds (boolean inNotation)
    throws java.lang.Exception
  {
    char c;
    String ids[] = new String[2];

    if (tryRead("PUBLIC")) {
      requireWhitespace();
      ids[0] = readLiteral(LIT_NORMALIZE); // public id
      if (inNotation) {
	skipWhitespace();
	if (tryRead('"') || tryRead('\'')) {
	  ids[1] = readLiteral(0);
	}
      } else {
	requireWhitespace();
	ids[1] = readLiteral(0); // system id
      }
    } else if (tryRead("SYSTEM")) {