xmlparser.java

来自「RESIN 3.2 最新源码」· Java 代码 · 共 2,651 行 · 第 1/5 页

/*
 * Copyright (c) 1998-2008 Caucho Technology -- all rights reserved
 *
 * This file is part of Resin(R) Open Source
 *
 * Each copy or derived work must preserve the copyright notice and this
 * notice unmodified.
 *
 * Resin Open Source is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * Resin Open Source is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, or any warranty
 * of NON-INFRINGEMENT.  See the GNU General Public License for more
 * details.
 *
 * You should have received a copy of the GNU General Public License
 * along with Resin Open Source; if not, write to the
 *   Free SoftwareFoundation, Inc.
 *   59 Temple Place, Suite 330
 *   Boston, MA 02111-1307  USA
 *
 * @author Scott Ferguson
 */

package com.caucho.xml;

import com.caucho.util.CharBuffer;
import com.caucho.vfs.Path;
import com.caucho.vfs.ReadStream;
import com.caucho.vfs.ReaderWriterStream;
import com.caucho.vfs.Vfs;
import com.caucho.vfs.WriteStream;
import com.caucho.xml.readers.MacroReader;
import com.caucho.xml.readers.Utf16Reader;
import com.caucho.xml.readers.Utf8Reader;
import com.caucho.xml.readers.XmlReader;

import org.w3c.dom.Document;
import org.w3c.dom.Node;
import org.xml.sax.InputSource;
import org.xml.sax.Locator;
import org.xml.sax.SAXException;
import org.xml.sax.SAXParseException;

import java.io.FileNotFoundException;
import java.io.IOException;
import java.io.InputStream;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.logging.Level;

/**
 * A configurable XML parser.  Loose versions of XML and HTML are supported
 * by changing the Policy object.
 *
 * <p>Normally, applications will use Xml, LooseXml, Html, or LooseHtml.
 */
public class XmlParser extends AbstractParser {
  // Xerces uses the following
  public static final String XMLNS = "http://www.w3.org/2000/xmlns/";
  public static final String XML = "http://www.w3.org/XML/1998/namespace";

  static final QName DOC_NAME = new QName(null, "#document", null);
  static final QName TEXT_NAME = new QName(null, "#text", null);
  static final QName JSP_NAME = new QName(null, "#jsp", null);
  static final QName WHITESPACE_NAME = new QName(null, "#whitespace", null);
  static final QName JSP_ATTRIBUTE_NAME = new QName("xtp", "jsp-attribute", null);
  
  QAttributes _attributes;
  QAttributes _nullAttributes;

  boolean _inDtd;
  
  CharBuffer _text;
  CharBuffer _eltName;
  CharBuffer _cb;
  CharBuffer _buf = new CharBuffer();
  String _textFilename;
  int _textLine;

  char []_textBuffer = new char[1024];
  int _textLength;
  int _textCapacity = _textBuffer.length;
  boolean _isIgnorableWhitespace;
  boolean _isJspText;
  
  CharBuffer _name = new CharBuffer();
  CharBuffer _nameBuffer = new CharBuffer();
  
  MacroReader _macro = new MacroReader();
  int _macroIndex = 0;
  int _macroLength = 0;
  char []_macroBuffer;

  QName []_elementNames = new QName[64];
  NamespaceMap []_namespaces = new NamespaceMap[64];
  int []_elementLines = new int[64];
  int _elementTop;

  NamespaceMap _namespaceMap;

  ArrayList<String> _attrNames = new ArrayList<String>();
  ArrayList<String> _attrValues = new ArrayList<String>();

  ReadStream _is;
  XmlReader _reader;
  
  String _extPublicId;
  String _extSystemId;
  
  QName _activeNode;
  QName _topNamespaceNode;
  boolean _isTagStart;
  boolean _stopOnIncludeEnd;
  boolean _hasTopElement;
  boolean _hasDoctype;
  boolean _isHtml;
  Locator _locator = new LocatorImpl(this);

  public XmlParser()
  {
    clear();
  }

  /**
   * Creates a new parser with a given parsing policy and dtd.
   *
   * @param policy the parsing policy, handling optional tags.
   * @param dtd the parser's dtd.
   */
  XmlParser(Policy policy, QDocumentType dtd)
  {
    super(policy, dtd);
    
    clear();
  }

  /**
   * Initialize the parser.
   */
  void init()
  {
    super.init();
    
    _attributes = new QAttributes();
    _nullAttributes = new QAttributes();
    _eltName = new CharBuffer();
    _text = new CharBuffer();

    _isHtml = _policy instanceof HtmlPolicy;

    // jsp/193b
    // _namespaceMap = null;
    
    _textLength = 0;
    _isIgnorableWhitespace = true;
    _elementTop = 0;
    _elementLines[0] = 1;

    _line = 1;

    _dtd = null;
    _inDtd = false;
    _isTagStart = false;
    _stopOnIncludeEnd = false;

    _extPublicId = null;
    _extSystemId = null;

    // _filename = null;
    _publicId = null;
    _systemId = null;
    
    _hasTopElement = false;
    _hasDoctype = false;

    _macroIndex = 0;
    _macroLength = 0;

    _reader = null;

    // _owner = null;
    
    _policy.init();
  }

  /**
   * Parse the document from a read stream.
   *
   * @param is read stream to parse from.
   *
   * @return The parsed document.
   */
  Document parseInt(ReadStream is)
    throws IOException, SAXException
  {
    _is = is;

    if (_filename == null && _systemId != null)
      _filename = _systemId;
    else if (_filename == null)
      _filename = _is.getUserPath();

    if (_systemId == null) {
      _systemId = _is.getPath().getURL();
      if ("null:".equals(_systemId) || "string:".equals(_systemId))
	_systemId = "stream";
    }

    /* xsl/0401
    if (_isNamespaceAware)
      _namespaceMap = new NamespaceMap(null, "", "");
    */
    _policy.setNamespaceAware(_isNamespaceAware);
    
    if (_filename == null)
      _filename = _systemId;

    if (_filename == null)
      _filename = "stream";

    if (_dtd != null)
      _dtd.setSystemId(_systemId);
    
    if (_builder != null) {
      if (! "string:".equals(_systemId) && ! "stream".equals(_systemId))
	_builder.setSystemId(_systemId);
      _builder.setFilename(_is.getPath().getURL());
    }

    if (_contentHandler == null)
      _contentHandler = new org.xml.sax.helpers.DefaultHandler();

    _contentHandler.setDocumentLocator(_locator);

    if (_owner == null)
      _owner = new QDocument();
    if (_defaultEncoding != null)
      _owner.setAttribute("encoding", _defaultEncoding);
    _owner.addDepend(is.getPath());
    
    _activeNode = DOC_NAME;
    
    _policy.setStream(is);
    _policy.setNamespace(_namespaceMap);

    _contentHandler.startDocument();
    
    int ch = parseXMLDeclaration(null);
    
    ch = skipWhitespace(ch);
    parseNode(ch, false);

    /*
    if (dbg.canWrite()) {
      printDebugNode(dbg, doc, 0);
      dbg.flush();
    }
    */

    if (_strictXml && ! _hasTopElement)
      throw error(L.l("XML file has no top-element.  All well-formed XML files have a single top-level element."));

    if (_contentHandler != null)
      _contentHandler.endDocument();

    QDocument owner = _owner;
    _owner = null;
      
    return owner;
  }

  /**
   * The main dispatch loop.
   *
   * @param node the current node
   * @param ch the next character
   * @param special true for the short form, &lt;foo/bar/>
   */
  private void parseNode(int ch, boolean special)
    throws IOException, SAXException
  {
    //boolean isTop = node instanceof QDocument;

    _text.clear();

  loop:
    while (true) {
      if (_textLength == 0) {
        _textFilename = getFilename();
        _textLine = getLine();
      }

      switch (ch) {
      case -1:
	if (_textLength != 0)
	  appendText();
        if (! _stopOnIncludeEnd && _reader.getNext() != null) {
          popInclude();
          if (_reader != null)
            parseNode(_reader.read(), special);
          return;
        }
	closeTag("");
	return;

      case ' ': case '\t': case '\n': case '\r':
	if (! _normalizeWhitespace)
	  addText((char) ch);
	else if (_textLength == 0) {
	  if (! _isTagStart)
	    addText(' ');
	}
	else if (_textBuffer[_textLength - 1] != ' ') {
	  addText(' ');
	}
	ch = _reader.read();
	break;

      case 0xffff:
	// marker for end of text for serialization
	return;

      default:
	addText((char) ch);
	ch = _reader.read();
	break;

      case '/':
	if (! special) {
	  addText((char) ch);
	  ch = _reader.read();
	  continue;
	}
	ch = _reader.read();
	if (ch == '>' || ch == -1) {
          appendText();
          popNode();
	  return;
	}
	addText('/');
	break;

      case '&':
        ch = parseEntityReference();
	break;

      case '<':
	boolean endTag = false;
	ch = _reader.read();

	if (ch == '/' && ! special) {
	  if (_normalizeWhitespace &&
	      _textLength > 0 && _textBuffer[_textLength - 1] == ' ') {
	    _textLength--;
	  }
	  appendText();

	  ch = _reader.parseName(_name, _reader.read());

          if (ch != '>') {
            // XXX: Hack for Java PetStore
            while (XmlChar.isWhitespace(ch))
              ch = _reader.read();

            if (ch != '>')
              throw error(L.l("`</{0}>' expected `>' at {1}.  Closing tags must close immediately after the tag name.", _name, badChar(ch)));
          }

	  closeTag(_policy.getName(_name).getName());
	  ch = _reader.read();
	} 
	// element: <tag attr=value ... attr=value> ...
	else if (XmlChar.isNameStart(ch)) {
	  appendText();
	  
	  parseElement(ch);
	  ch = _reader.read();
	}
	// <! ...
	else if (ch == '!') {
	  // <![CDATA[ ... ]]>
	  if ((ch = _reader.read()) == '[') {
	    parseCdata();
	    ch = _reader.read();
	  }
	  // <!-- ... -->
          else if (ch == '-') {
	    parseComment();

	    ch = _reader.read();
	  } 
	  else if (XmlChar.isNameStart(ch)) {
	    appendText();
	    ch = _reader.parseName(_name, ch);
	    String declName = _name.toString();
	    if (declName.equals("DOCTYPE")) {
	      parseDoctype(ch);
              if (_contentHandler instanceof DOMBuilder)
                ((DOMBuilder) _contentHandler).dtd(_dtd);

	      ch = _reader.read();
	    } else if (_forgiving && declName.equalsIgnoreCase("doctype")) {
	      parseDoctype(ch);
              if (_contentHandler instanceof DOMBuilder)
                ((DOMBuilder) _contentHandler).dtd(_dtd);
              
	      ch = _reader.read();
	    } else
	      throw error(L.l("expected `<!DOCTYPE' declaration at {0}", declName));
	  } else if (_forgiving) {
	    addText("<!");
	  } else
	    throw error(L.l("expected `<!DOCTYPE' declaration at {0}", badChar(ch)));
	} 
	// PI: <?tag attr=value ... attr=value?>
	else if (ch == '?') {
	  ch = parsePI();
	} 
	else if (_strictXml) {
	  throw error(L.l("expected tag name after `<' at {0}.  Open tag names must immediately follow the open brace like `<foo ...>'", badChar(ch)));
        }
        // implicit <![CDATA[ for <% ... %>
        else if (_isJsp && ch == '%') {
          ch = _reader.read();

          appendText();
          _isJspText = ch != '=';
          
          addText("<%");

          while (ch >= 0) {
            if (ch == '%') {
              ch = _reader.read();
              if (ch == '>') {
                addText("%>");
                ch = _reader.read();
                break;
              }
              else
                addText('%');
            }
            else {
              addText((char) ch);
              ch = _reader.read();
            }
          }

          appendText();
          _isJspText = false;
	}
	else {
	  addText('<');
	}
      }
    }
  }

  /**
   * Parses the &lt;!DOCTYPE> declaration.
   */
  private void parseDoctype(int ch)
    throws IOException, SAXException
  {
    if (_activeNode != DOC_NAME)
      throw error(L.l("<!DOCTYPE immediately follow the <?xml ...?> declaration."));
    
    _inDtd = true;

    ch = skipWhitespace(ch);
    ch = _reader.parseName(_nameBuffer, ch);
    String name = _nameBuffer.toString();
    ch = skipWhitespace(ch);

    if (_dtd == null)
      _dtd = new QDocumentType(name);

    _dtd.setName(name);

    if (XmlChar.isNameStart(ch)) {
      ch = parseExternalID(ch);
      ch = skipWhitespace(ch);

      _dtd._publicId = _extPublicId;
      _dtd._systemId = _extSystemId;
    }

    if (_dtd._systemId != null && ! _dtd._systemId.equals("")) {
      InputStream is = null;

      unread(ch);
      
      XmlReader oldReader = _reader;
      boolean hasInclude = false;

      try {
        pushInclude(_extPublicId, _extSystemId);
        hasInclude = true;
      } catch (Exception e) {
	if (log.isLoggable(Level.FINEST))
	  log.log(Level.FINER, e.toString(), e);
	else
	  log.finer(e.toString());
      }

      if (hasInclude) {
        _stopOnIncludeEnd = true;
	try {
	  ch = parseDoctypeDecl(_dtd);
	} catch (XmlParseException e) {
	  if (_extSystemId != null &&
	      _extSystemId.startsWith("http")) {
	    log.log(Level.FINE, e.toString(), e);
	  }
	  else
	    throw e;
	}
        _stopOnIncludeEnd = false;

        while (_reader != null && _reader != oldReader)
          popInclude();
      }

      if (_reader != null)
        ch = skipWhitespace(read());
    }