xmlparser.java

linux下建立JAVA虚拟机的源码KAFFE

                parseNotationType();
                return typeString;
              }
            else if ("CDATA" == typeString
                     || "ID" == typeString
                     || "IDREF" == typeString
                     || "IDREFS" == typeString
                     || "ENTITY" == typeString
                     || "ENTITIES" == typeString
                     || "NMTOKEN" == typeString
                     || "NMTOKENS" == typeString)
              {
                return typeString;
              }
          }
        else
          {
            if ("NOTATION".equals(typeString))
              {
                parseNotationType();
                return typeString;
              }
            else if ("CDATA".equals(typeString)
                     || "ID".equals(typeString)
                     || "IDREF".equals(typeString)
                     || "IDREFS".equals(typeString)
                     || "ENTITY".equals(typeString)
                     || "ENTITIES".equals(typeString)
                     || "NMTOKEN".equals(typeString)
                     || "NMTOKENS".equals(typeString))
              {
                return typeString;
              }
          }
        error("illegal attribute type", typeString, null);
        return null;
      }
  }
  
  /**
   * Parse an enumeration.
   * <pre>
   * [59] Enumeration ::= '(' S? Nmtoken (S? '|' S? Nmtoken)* S? ')'
   * </pre>
   * <p>NOTE: the '(' has already been read.
   */
  private void parseEnumeration(boolean isNames)
    throws Exception
  {
    dataBufferAppend('(');

    // Read the first token.
    skipWhitespace();
    dataBufferAppend(readNmtoken(isNames));
    // Read the remaining tokens.
    skipWhitespace();
    while (!tryRead(')'))
      {
        require('|');
        dataBufferAppend('|');
        skipWhitespace();
        dataBufferAppend(readNmtoken (isNames));
        skipWhitespace();
      }
    dataBufferAppend(')');
  }

  /**
   * Parse a notation type for an attribute.
   * <pre>
   * [58] NotationType ::= 'NOTATION' S '(' S? NameNtoks
   *    (S? '|' S? name)* S? ')'
   * </pre>
   * <p>NOTE: the 'NOTATION' has already been read
   */
  private void parseNotationType()
    throws Exception
  {
    requireWhitespace();
    require('(');
    
    parseEnumeration(true);
  }

  /**
   * Parse the default value for an attribute.
   * <pre>
   * [60] DefaultDecl ::= '#REQUIRED' | '#IMPLIED'
   *    | (('#FIXED' S)? AttValue)
   * </pre>
   */
  private void parseDefault(String elementName, String name,
                            String type, String enumer)
    throws Exception
  {
    int valueType = ATTRIBUTE_DEFAULT_SPECIFIED;
    String value = null;
    int flags = LIT_ATTRIBUTE;
    boolean saved = expandPE;
    String defaultType = null;
    
    // LIT_ATTRIBUTE forces '<' checks now (ASAP) and turns whitespace
    // chars to spaces (doesn't matter when that's done if it doesn't
    // interfere with char refs expanding to whitespace).
    
    if (!skippedPE)
      {
        flags |= LIT_ENTITY_REF;
        if (handler.stringInterning)
          {
            if ("CDATA" != type)
              {
                flags |= LIT_NORMALIZE;
              }
          }
        else
          {
            if (!"CDATA".equals(type))
              {
                flags |= LIT_NORMALIZE;
              }
          }
      }
    
    expandPE = false;
    if (tryRead('#'))
      {
        if (tryRead("FIXED"))
          {
            defaultType = "#FIXED";
            valueType = ATTRIBUTE_DEFAULT_FIXED;
            requireWhitespace();
            value = readLiteral(flags);
          }
        else if (tryRead("REQUIRED"))
          {
            defaultType = "#REQUIRED";
            valueType = ATTRIBUTE_DEFAULT_REQUIRED;
          }
        else if (tryRead("IMPLIED"))
          {
            defaultType = "#IMPLIED";
            valueType = ATTRIBUTE_DEFAULT_IMPLIED;
          }
        else
          {
            error("illegal keyword for attribute default value");
          }
      }
    else
      {
        value = readLiteral(flags);
      }
    expandPE = saved;
    setAttribute(elementName, name, type, enumer, value, valueType);
    if (handler.stringInterning)
      {
        if ("ENUMERATION" == type)
          {
            type = enumer;
          }
        else if ("NOTATION" == type)
          {
            type = "NOTATION " + enumer;
          }
      }
    else
      {
        if ("ENUMERATION".equals(type))
          {
            type = enumer;
          }
        else if ("NOTATION".equals(type))
          {
            type = "NOTATION " + enumer;
          }
      }
    if (!skippedPE)
      {
        handler.getDeclHandler().attributeDecl(elementName, name, type,
                                               defaultType, value);
      }
  }
  
  /**
   * Parse a conditional section.
   * <pre>
   * [61] conditionalSect ::= includeSect || ignoreSect
   * [62] includeSect ::= '&lt;![' S? 'INCLUDE' S? '['
   *    extSubsetDecl ']]&gt;'
   * [63] ignoreSect ::= '&lt;![' S? 'IGNORE' S? '['
   *    ignoreSectContents* ']]&gt;'
   * [64] ignoreSectContents ::= Ignore
   *    ('&lt;![' ignoreSectContents* ']]&gt;' Ignore )*
   * [65] Ignore ::= Char* - (Char* ( '&lt;![' | ']]&gt;') Char* )
   * </pre>
   * <p> NOTE: the '&gt;![' has already been read.
   */
  private void parseConditionalSect(char[] saved)
    throws Exception
  {
    skipWhitespace();
    if (tryRead("INCLUDE"))
      {
        skipWhitespace();
        require('[');
        // VC: Proper Conditional Section/PE Nesting
        if (readBuffer != saved)
          {
            handler.verror("Illegal Conditional Section/PE nesting");
          }
        skipWhitespace();
        while (!tryRead("]]>"))
          {
            parseMarkupdecl();
            skipWhitespace();
          }
      }
    else if (tryRead("IGNORE"))
      {
        skipWhitespace();
        require('[');
        // VC: Proper Conditional Section/PE Nesting
        if (readBuffer != saved)
          {
            handler.verror("Illegal Conditional Section/PE nesting");
          }
        int nesting = 1;
        char c;
        expandPE = false;
        for (int nest = 1; nest > 0; )
          {
            c = readCh();
            switch (c)
              {
              case '<':
                if (tryRead("!["))
                  {
                    nest++;
                  }
                break;
              case ']':
                if (tryRead("]>"))
                  {
                    nest--;
                  }
              }
          }
        expandPE = true;
      }
    else
      {
        error("conditional section must begin with INCLUDE or IGNORE");
      }
  }
  
  private void parseCharRef()
    throws SAXException, IOException
  {
    parseCharRef(true /* do flushDataBuffer by default */);
  }

  /**
   * Try to read a character reference without consuming data from buffer.
   * <pre>
   * [66] CharRef ::= '&#' [0-9]+ ';' | '&#x' [0-9a-fA-F]+ ';'
   * </pre>
   * <p>NOTE: the '&#' has already been read.
   */
  private void tryReadCharRef()
    throws SAXException, IOException
  {
    int value = 0;
    char c;
    
    if (tryRead('x'))
      {
loop1:
        while (true)
          {
            c = readCh();
            if (c == ';')
              {
                break loop1;
              }
            else
              {
                int n = Character.digit(c, 16);
                if (n == -1)
                  {
                    error("illegal character in character reference", c, null);
                    break loop1;
                  }
                value *= 16;
                value += n;
              }
          }
      }
    else
      {
loop2:
        while (true)
          {
            c = readCh();
            if (c == ';')
              {
                break loop2;
              }
            else
              {
                int n = Character.digit(c, 10);
                if (n == -1)
                  {
                    error("illegal character in character reference", c, null);
                    break loop2;
                  }
                value *= 10;
                value += n;
              }
          }
      }
    
    // check for character refs being legal XML
    if ((value < 0x0020
         && ! (value == '\n' || value == '\t' || value == '\r'))
        || (value >= 0xD800 && value <= 0xDFFF)
        || value == 0xFFFE || value == 0xFFFF
        || value > 0x0010ffff)
      {
        error("illegal XML character reference U+"
              + Integer.toHexString(value));
      }
    
    // Check for surrogates: 00000000 0000xxxx yyyyyyyy zzzzzzzz
    //  (1101|10xx|xxyy|yyyy + 1101|11yy|zzzz|zzzz:
    if (value > 0x0010ffff)
      {
        // too big for surrogate
        error("character reference " + value + " is too large for UTF-16",
              new Integer(value).toString(), null);
      }
    
  }
  
  /**
   * Read and interpret a character reference.
   * <pre>
   * [66] CharRef ::= '&#' [0-9]+ ';' | '&#x' [0-9a-fA-F]+ ';'
   * </pre>
   * <p>NOTE: the '&#' has already been read.
   */
  private void parseCharRef(boolean doFlush)
    throws SAXException, IOException
  {
    int value = 0;
    char c;
    
    if (tryRead('x'))
      {
loop1:
        while (true)
          {
            c = readCh();
            if (c == ';')
              {
                break loop1;
              }
            else
              {
                int n = Character.digit(c, 16);
                if (n == -1)
                  {
                    error("illegal character in character reference", c, null);
                    break loop1;
                  }
                value *= 16;
                value += n;
              }
          }
      }
    else
      {
loop2:
        while (true)
          {
            c = readCh();
            if (c == ';')
              {
                break loop2;
              }
            else
              {
                int n = Character.digit(c, 10);
                if (n == -1)
                  {
                    error("illegal character in character reference", c, null);
                    break loop2;
                  }
                value *= 10;
                value += c - '0';
              }
          }
      }
    
    // check for character refs being legal XML
    if ((value < 0x0020
         && ! (value == '\n' || value == '\t' || value == '\r'))
        || (value >= 0xD800 && value <= 0xDFFF)
        || value == 0xFFFE || value == 0xFFFF
        || value > 0x0010ffff)
      {
        error("illegal XML character reference U+"
              + Integer.toHexString(value));
      }
    
    // 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 <= 0x0010ffff)
      {
        value -= 0x10000;
        // > 16 bits, surrogate needed
        dataBufferAppend((char) (0xd800 | (value >> 10)));
        dataBufferAppend((char) (0xdc00 | (value & 0x0003ff)));
      }
    else
      {
        // too big for surrogate
        error("character reference " + value + " is too large for UTF-16",
              new Integer(value).toString(), null);
      }
    if (doFlush)
      {
        dataBufferFlush();
      }
  }
  
  /**
   * Parse and expand an entity reference.
   * <pre>
   * [68] EntityRef ::= '&' Name ';'
   * </pre>
   * <p>NOTE: the '&amp;' has already been read.
   * @param externalAllowed External entities are allowed here.
   */
  private void parseEntityRef(boolean externalAllowed)
    throws SAXException, IOException
  {
    String name;
    
    name = readNmtoken(true);
    require(';');
    switch (getEntityType(name))
      {
      case ENTITY_UNDECLARED:
        // NOTE:  XML REC describes amazingly convoluted handling for
        // this case.  Nothing as meaningful as being a WFness error
        // unless the processor might _legitimately_ not have seen a
        // declaration ... which is what this implements.
        String message;
        
        message = "reference to undeclared general entity " + name;
        if (skippedPE && !docIsStandalone)
          {
            handler.verror(message);
            // we don't know this entity, and it might be external...
            if (externalAllowed)
              {
                handler.skippedEntity(name);
              }
          }
        else
          {
            error(message);
          }
        break;
      case ENTITY_INTERNAL:
          pushString(name, getEntityValue(name));
          
          //workaround for possible input pop before marking
          //the buffer reading position  
          cha