xmlparser.java

linux下建立JAVA虚拟机的源码KAFFE

        return;
      }
    
    // four byte encodings
    if (encoding == ENCODING_UCS_4_1234
        || encoding == ENCODING_UCS_4_4321
        || encoding == ENCODING_UCS_4_2143
        || encoding == ENCODING_UCS_4_3412)
      {
        // Strictly:  "UCS-4" == "UTF-32BE"; also, "UTF-32LE" exists
        if (!encodingName.equals("ISO-10646-UCS-4"))
          {
            error("unsupported 32-bit encoding", encodingName,
                  "ISO-10646-UCS-4");
          }
        return;
      }
    
    // assert encoding == ENCODING_EXTERNAL
    // if (encoding != ENCODING_EXTERNAL)
    //     throw new RuntimeException ("encoding = " + encoding);
    
    if (encodingName.equals("UTF-16BE"))
      {
        encoding = ENCODING_UCS_2_12;
        return;
      }
    if (encodingName.equals("UTF-16LE"))
      {
        encoding = ENCODING_UCS_2_21;
        return;
      }
    
    // We couldn't use the builtin decoders at all.  But we can try to
    // create a reader, since we haven't messed up buffering.  Tweak
    // the encoding name if necessary.
    
    if (encodingName.equals("UTF-16")
        || encodingName.equals("ISO-10646-UCS-2"))
      {
        encodingName = "Unicode";
      }
    // Ignoring all the EBCDIC aliases here
    
    reader = new InputStreamReader(is, encodingName);
    sourceType = INPUT_READER;
  }
  
  /**
   * Parse miscellaneous markup outside the document element and DOCTYPE
   * declaration.
   * <pre>
   * [27] Misc ::= Comment | PI | S
   * </pre>
   */
  private void parseMisc()
    throws Exception
  {
    while (true)
      {
        skipWhitespace();
        if (tryRead(startDelimPI))
          {
            parsePI();
          }
        else if (tryRead(startDelimComment))
          {
            parseComment();
          }
        else
          {
            return;
          }
      }
  }

  /**
   * Parse a document type declaration.
   * <pre>
   * [28] doctypedecl ::= '&lt;!DOCTYPE' S Name (S ExternalID)? S?
   *    ('[' (markupdecl | PEReference | S)* ']' S?)? '&gt;'
   * </pre>
   * <p> (The <code>&lt;!DOCTYPE</code> has already been read.)
   */
  private void parseDoctypedecl()
    throws Exception
  {
    String rootName;
    ExternalIdentifiers ids;

    // Read the document type name.
    requireWhitespace();
    rootName = readNmtoken(true);

    // Read the External subset's IDs
    skipWhitespace();
    ids = readExternalIds(false, true);

    // report (a) declaration of name, (b) lexical info (ids)
    handler.doctypeDecl(rootName, ids.publicId, ids.systemId);
    
    // Internal subset is parsed first, if present
    skipWhitespace();
    if (tryRead('['))
      {
        
        // loop until the subset ends
        while (true)
          {
            doReport = expandPE = true;
            skipWhitespace();
            doReport = expandPE = false;
            if (tryRead(']'))
              {
                break;     // end of subset
              }
            else
              {
                // WFC, PEs in internal subset (only between decls)
                peIsError = expandPE = true;
                parseMarkupdecl();
                peIsError = expandPE = false;
              }
          }
      }
    skipWhitespace();
    require('>');
    
    // Read the external subset, if any
    InputSource subset;
    
    if (ids.systemId == null)
      {
        subset = handler.getExternalSubset(rootName,
                                           handler.getSystemId());
      }
    else
      {
        subset = null;
      }
    if (ids.systemId != null || subset != null)
      {
        pushString(null, ">");
      
        // NOTE:  [dtd] is so we say what SAX2 expects,
        // though it's misleading (subset, not entire dtd)
        if (ids.systemId != null)
          {
            pushURL(true, "[dtd]", ids, null, null, null, true);
          }
        else
          {
            handler.warn("modifying document by adding external subset");
            pushURL(true, "[dtd]",
                    new ExternalIdentifiers(subset.getPublicId(),
                                            subset.getSystemId(),
                                            null),
                    subset.getCharacterStream(),
                    subset.getByteStream(),
                    subset.getEncoding(),
                    false);
          }
        
        // Loop until we end up back at '>'
        while (true)
          {
            doReport = expandPE = true;
            skipWhitespace();
            doReport = expandPE = false;
            if (tryRead('>'))
              {
                break;
              }
            else
              {
                expandPE = true;
                parseMarkupdecl();
                expandPE = false;
              }
          }
        
        // the ">" string isn't popped yet
        if (inputStack.size() != 1)
          {
            error("external subset has unmatched '>'");
          }
      }
    
    // done dtd
    handler.endDoctype();
    expandPE = false;
    doReport = true;
  }
  
  /**
   * Parse a markup declaration in the internal or external DTD subset.
   * <pre>
   * [29] markupdecl ::= elementdecl | Attlistdecl | EntityDecl
   *    | NotationDecl | PI | Comment
   * [30] extSubsetDecl ::= (markupdecl | conditionalSect
   *    | PEReference | S) *
   * </pre>
   * <p> Reading toplevel PE references is handled as a lexical issue
   * by the caller, as is whitespace.
   */
  private void parseMarkupdecl()
    throws Exception
  {
    char[] saved = null;
    boolean savedPE = expandPE;

    // prevent "<%foo;" and ensures saved entity is right
    require('<');
    unread('<');
    expandPE = false;
    
    if (tryRead("<!ELEMENT"))
      {
        saved = readBuffer;
        expandPE = savedPE;
        parseElementDecl();
      }
    else if (tryRead("<!ATTLIST"))
      {
        saved = readBuffer;
        expandPE = savedPE;
        parseAttlistDecl();
      }
    else if (tryRead("<!ENTITY"))
      {
        saved = readBuffer;
        expandPE = savedPE;
        parseEntityDecl();
      }
    else if (tryRead("<!NOTATION"))
      {
        saved = readBuffer;
        expandPE = savedPE;
        parseNotationDecl();
      }
    else if (tryRead(startDelimPI))
      {
        saved = readBuffer;
        expandPE = savedPE;
        parsePI();
      }
    else if (tryRead(startDelimComment))
      {
        saved = readBuffer;
        expandPE = savedPE;
        parseComment();
      }
    else if (tryRead("<!["))
      {
        saved = readBuffer;
        expandPE = savedPE;
        if (inputStack.size() > 0)
          {
            parseConditionalSect(saved);
          }
        else
          {
            error("conditional sections illegal in internal subset");
          }
      }
    else
      {
        error("expected markup declaration");
      }

    // VC: Proper Decl/PE Nesting
    if (readBuffer != saved)
      {
        handler.verror("Illegal Declaration/PE nesting");
      }
  }
  
  /**
   * Parse an element, with its tags.
   * <pre>
   * [39] element ::= EmptyElementTag | STag content ETag
   * [40] STag ::= '&lt;' Name (S Attribute)* S? '&gt;'
   * [44] EmptyElementTag ::= '&lt;' Name (S Attribute)* S? '/&gt;'
   * </pre>
   * <p> (The '&lt;' has already been read.)
   * <p>NOTE: this method actually chains onto parseContent (), if necessary,
   * and parseContent () will take care of calling parseETag ().
   */
  private void parseElement(boolean maybeGetSubset)
    throws Exception
  {
    String gi;
    char c;
    int oldElementContent = currentElementContent;
    String oldElement = currentElement;
    ElementDecl element;

    // This is the (global) counter for the
    // array of specified attributes.
    tagAttributePos = 0;
    
    // Read the element type name.
    gi = readNmtoken(true);
    
    // If we saw no DTD, and this is the document root element,
    // let the application modify the input stream by providing one.
    if (maybeGetSubset)
      {
        InputSource subset = handler.getExternalSubset(gi,
                                                       handler.getSystemId());
        if (subset != null)
          {
            String publicId = subset.getPublicId();
            String systemId = subset.getSystemId();
            
            handler.warn("modifying document by adding DTD");
            handler.doctypeDecl(gi, publicId, systemId);
            pushString(null, ">");
            
            // NOTE:  [dtd] is so we say what SAX2 expects,
            // though it's misleading (subset, not entire dtd)
            pushURL(true, "[dtd]",
                    new ExternalIdentifiers(publicId, systemId, null),
                    subset.getCharacterStream(),
                    subset.getByteStream(),
                    subset.getEncoding(),
                    false);
            
            // Loop until we end up back at '>'
            while (true)
              {
                doReport = expandPE = true;
                skipWhitespace();
                doReport = expandPE = false;
                if (tryRead('>'))
                  {
                    break;
                  }
                else
                  {
                    expandPE = true;
                    parseMarkupdecl();
                    expandPE = false;
                  }
              }
            
            // the ">" string isn't popped yet
            if (inputStack.size() != 1)
              {
                error("external subset has unmatched '>'");
              }
            
            handler.endDoctype();
          }
      }
    
    // Determine the current content type.
    currentElement = gi;
    element = (ElementDecl) elementInfo.get(gi);
    currentElementContent = getContentType(element, CONTENT_ANY);

    // Read the attributes, if any.
    // After this loop, "c" is the closing delimiter.
    boolean white = tryWhitespace();
    c = readCh();
    while (c != '/' && c != '>')
      {
        unread(c);
        if (!white)
          {
            error("need whitespace between attributes");
          }
        parseAttribute(gi);
        white = tryWhitespace();
        c = readCh();
      }
    
    // Supply any defaulted attributes.
    Iterator atts = declaredAttributes(element);
    if (atts != null)
      {
        String aname;
loop:
        while (atts.hasNext())
          {
            aname = (String) atts.next();
            // See if it was specified.
            for (int i = 0; i < tagAttributePos; i++)
              {
                if (tagAttributes[i] == aname)
                  {
                    continue loop;
                  }
              }
            // ... or has a default
            String value = getAttributeDefaultValue(gi, aname);
            
            if (value == null)
              {
                continue;
              }
            handler.attribute(aname, value, false);
          }
      }

    // Figure out if this is a start tag
    // or an empty element, and dispatch an
    // event accordingly.
    switch (c)
      {
      case '>':
        handler.startElement(gi);
        parseContent();
        break;
      case '/':
        require('>');
        handler.startElement(gi);
        handler.endElement(gi);
        break;
      }

    // Restore the previous state.
    currentElement = oldElement;
    currentElementContent = oldElementContent;
  }
    
  /**
   * Parse an attribute assignment.
   * <pre>
   * [41] Attribute ::= Name Eq AttValue
   * </pre>
   * @param name The name of the attribute's element.
   * @see SAXDriver#attribute
   */
  private void parseAttribute(String name)
    throws Exception
  {
    String aname;
    String type;
    String value;
    int flags = LIT_ATTRIBUTE |  LIT_ENTITY_REF;
    
    // Read the attribute name.
    aname = readNmtoken(true);
    type = getAttributeType(name, aname);
    
    // Parse '='
    parseEq();

    // Read the value, normalizing whitespace
    // unless it is CDATA.
    if (handler.stringInterning)
      {
        if (type == "CDATA" || type == null)
          {
            value = readLiteral(flags);
          }
        else
          {
            value = readLiteral(flags | LIT_NORMALIZE);
          }
      }
    else
      {
        if (type == null || type.equals("CDATA"))
          {
            value = readLiteral(flags);
          }
        else
          {
            value = readLiteral(flags | LIT_NORMALIZE);
          }
      }

    // WFC: no duplicate attributes
    for (int i = 0; i < tagAttributePos; i++)
      {
        if (aname.equals(tagAttributes [i]))
          {
            error("duplicate attribute", aname, null);
          }
      }

    // Inform the handler about the
    // attribute.
    handler.attribute(aname, value, true);