/*
 * @(#)MimeType.java	1.4 03/01/23
 *
 * Copyright 2003 Sun Microsystems, Inc. All rights reserved.
 * SUN PROPRIETARY/CONFIDENTIAL. Use is subject to license terms.
 */

package javax.print;

import java.io.Serializable;

import java.util.AbstractMap;
import java.util.AbstractSet;
import java.util.Iterator;
import java.util.Map;
import java.util.NoSuchElementException;
import java.util.Set;
import java.util.Vector;

/**
 * Class MimeType encapsulates a Multipurpose Internet Mail Extensions (MIME) 
 * media type as defined in <A HREF="http://www.ietf.org/rfc/rfc2045.txt">RFC 
 * 2045</A> and <A HREF="http://www.ietf.org/rfc/rfc2046.txt">RFC 2046</A>. A 
 * MIME type object is part of a {@link DocFlavor DocFlavor} object and 
 * specifies the format of the print data. 
 * <P>
 * Class MimeType is similar to the like-named 
 * class in package {@link java.awt.datatransfer java.awt.datatransfer}. Class 
 * java.awt.datatransfer.MimeType is not used in the Jini Print Service API  
 * for two reasons: 
 * <OL TYPE=1>
 * <LI>
 * Since not all Java profiles include the AWT, the Jini Print Service should 
 * not depend on an AWT class. 
 * <P>
 * <LI>
 * The implementation of class java.awt.datatransfer.MimeType does not
 * guarantee 
 * that equivalent MIME types will have the same serialized representation. 
 * Thus, since the Jini Lookup Service (JLUS) matches service attributes based 
 * on equality of serialized representations, JLUS searches involving MIME
 * types encapsulated in class java.awt.datatransfer.MimeType may incorrectly
 * fail to match. 
 * </OL>
 * <P>
 * Class MimeType's serialized representation is based on the following 
 * canonical form of a MIME type string. Thus, two MIME types that are not 
 * identical but that are equivalent (that have the same canonical form) will  
 * be considered equal by the JLUS's matching algorithm. 
 * <UL>
 * <LI> The media type, media subtype, and parameters are retained, but all
 *      comments and whitespace characters are discarded. 
 * <LI> The media type, media subtype, and parameter names are converted to 
 *      lowercase.
 * <LI> The parameter values retain their original case, except a charset 
 *      parameter value for a text media type is converted to lowercase. 
 * <LI> Quote characters surrounding parameter values are removed.
 * <LI> Quoting backslash characters inside parameter values are removed.
 * <LI> The parameters are arranged in ascending order of parameter name. 
 * </UL>
 * <P>
 *
 * @author  Alan Kaminsky
 */
class MimeType implements Serializable, Cloneable {

    /**
     * Array of strings that hold pieces of this MIME type's canonical form.  
     * If the MIME type has <I>n</I> parameters, <I>n</I> >= 0, then the  
     * strings in the array are: 
     * <BR>Index 0 -- Media type.
     * <BR>Index 1 -- Media subtype.
     * <BR>Index 2<I>i</I>+2 -- Name of parameter <I>i</I>, 
     * <I>i</I>=0,1,...,<I>n</I>-1. 
     * <BR>Index 2<I>i</I>+3 -- Value of parameter <I>i</I>, 
     * <I>i</I>=0,1,...,<I>n</I>-1. 
     * <BR>Parameters are arranged in ascending order of parameter name.
     * @serial
     */
    private String[] myPieces;

    /**
     * String value for this MIME type. Computed when needed and cached. 
     */
    private transient String myStringValue = null;
    
    /**
     * Parameter map entry set. Computed when needed and cached.
     */
    private transient ParameterMapEntrySet myEntrySet = null;
    
    /**
     * Parameter map. Computed when needed and cached.
     */
    private transient ParameterMap myParameterMap = null;
    
    /**
     * Parameter map entry.
     */
    private class ParameterMapEntry implements Map.Entry {
	private int myIndex;
	public ParameterMapEntry(int theIndex) {
	    myIndex = theIndex;
	}
	public Object getKey(){
	    return myPieces[myIndex];
	}
	public Object getValue(){ 
	    return myPieces[myIndex+1];
	}
	public Object setValue (Object value) {
	    throw new UnsupportedOperationException();
	}
	public boolean equals(Object o) {
	    return (o != null &&
		    o instanceof Map.Entry &&
		    getKey().equals (((Map.Entry) o).getKey()) &&
		    getValue().equals(((Map.Entry) o).getValue()));
	}
	public int hashCode() {
	    return getKey().hashCode() ^ getValue().hashCode();
	}
    }

    /**
     * Parameter map entry set iterator.
     */
    private class ParameterMapEntrySetIterator implements Iterator {
	private int myIndex = 2;
	public boolean hasNext() {
	    return myIndex < myPieces.length;
	}
	public Object next() {
	    if (hasNext()) {
		ParameterMapEntry result = new ParameterMapEntry (myIndex);
		myIndex += 2;
		return result;
	    } else {
		throw new NoSuchElementException();
	    }
	}
	public void remove() {
	    throw new UnsupportedOperationException();
	}
    }

    /**
     * Parameter map entry set.
     */
    private class ParameterMapEntrySet extends AbstractSet {
	public Iterator iterator() {
	    return new ParameterMapEntrySetIterator();
	}
	public int size() {
	    return (myPieces.length - 2) / 2;
	}
    }

    /**
     * Parameter map.
     */
    private class ParameterMap extends AbstractMap {
	public Set entrySet() {
	    if (myEntrySet == null) {
		myEntrySet = new ParameterMapEntrySet();
	    }
	    return myEntrySet;
	}
    }

    /**
     * Construct a new MIME type object from the given string. The given  
     * string is converted into canonical form and stored internally.
     *
     * @param  s  MIME media type string.
     *
     * @exception  NullPointerException
     *     (unchecked exception) Thrown if <CODE>s</CODE> is null. 
     * @exception  IllegalArgumentException
     *     (unchecked exception) Thrown if <CODE>s</CODE> does not obey the 
     *     syntax for a MIME media type string. 
     */
    public MimeType(String s) {
	parse (s);
    }

    /**
     * Returns this MIME type object's MIME type string based on the canonical 
     * form. Each parameter value is enclosed in quotes. 
     */
    public String getMimeType() {
	return getStringValue();
    }
    
    /**
     * Returns this MIME type object's media type.
     */
    public String getMediaType() {
	return myPieces[0];
    }

    /**
     * Returns this MIME type object's media subtype.
     */
    public String getMediaSubtype() {
	return myPieces[1];
    }

    /**
     * Returns an unmodifiable map view of the parameters in this MIME type 
     * object. Each entry in the parameter map view consists of a parameter  
     * name String (key) mapping to a parameter value String. If this MIME  
     * type object has no parameters, an empty map is returned. 
     *
     * @return  Parameter map for this MIME type object.
     */
    public Map getParameterMap() {
	if (myParameterMap == null) {
	    myParameterMap = new ParameterMap();
	}
	return myParameterMap;
    }

    /**
     * Converts this MIME type object to a string.
     *
     * @return  MIME type string based on the canonical form. Each parameter 
     *          value is enclosed in quotes. 
     */
    public String toString() {
	return getStringValue();
    }

    /**
     * Returns a hash code for this MIME type object.
     */
    public int hashCode() {
	return getStringValue().hashCode();
    }

    /**
     * Determine if this MIME type object is equal to the given object. The two 
     * are equal if the given object is not null, is an instance of class 
     * net.jini.print.data.MimeType, and has the same canonical form as this 
     * MIME type object (that is, has the same type, subtype, and parameters). 
     * Thus, if two MIME type objects are the same except for comments, they are 
     * considered equal. However, "text/plain" and "text/plain; 
     * charset=us-ascii" are not considered equal, even though they represent 
     * the same media type (because the default character set for plain text is 
     * US-ASCII). 
     *
     * @param  obj  Object to test.
     *
     * @return  True if this MIME type object equals <CODE>obj</CODE>, false
     *          otherwise.
     */
    public boolean equals (Object obj) {
	return(obj != null &&
	       obj instanceof MimeType &&
	       getStringValue().equals(((MimeType) obj).getStringValue()));
    }

    /**
     * Returns this MIME type's string value in canonical form.
     */
    private String getStringValue() {
	if (myStringValue == null) {
	    StringBuffer result = new StringBuffer();
	    result.append (myPieces[0]);
	    result.append ('/');
	    result.append (myPieces[1]);
	    int n = myPieces.length;
	    for (int i = 2; i < n; i += 2) {
		result.append(';');
		result.append(' ');
		result.append(myPieces[i]);
		result.append('=');
		result.append(addQuotes (myPieces[i+1]));
	    }
	    myStringValue = result.toString();
	}
	return myStringValue;
    }
    
// Hidden classes, constants, and operations for parsing a MIME media type 
// string. 
    
    // Lexeme types.
    private static final int TOKEN_LEXEME         = 0;
    private static final int QUOTED_STRING_LEXEME = 1;
    private static final int TSPECIAL_LEXEME      = 2;
    private static final int EOF_LEXEME           = 3;
    private static final int ILLEGAL_LEXEME       = 4;
    
    // Class for a lexical analyzer.
    private static class LexicalAnalyzer {
	protected String mySource;
	protected int mySourceLength;
	protected int myCurrentIndex;
	protected int myLexemeType;
	protected int myLexemeBeginIndex;
	protected int myLexemeEndIndex;
	
	public LexicalAnalyzer(String theSource) {
	    mySource = theSource;
	    mySourceLength = theSource.length();
	    myCurrentIndex = 0;
	    nextLexeme();
	}

	public int getLexemeType() {
	    return myLexemeType;
	}

	public String getLexeme() {
	    return(myLexemeBeginIndex >= mySourceLength ?
		   null :
		   mySource.substring(myLexemeBeginIndex, myLexemeEndIndex));
	}

	public char getLexemeFirstCharacter() {
	    return(myLexemeBeginIndex >= mySourceLength ?
		   '\u0000' :
		   mySource.charAt(myLexemeBeginIndex));
	}

	public void nextLexeme() {
	    int state = 0;
	    int commentLevel = 0;
	    char c;
	    while (state >= 0) {
		switch (state) {
		    // Looking for a token, quoted string, or tspecial
		case 0:
		    if (myCurrentIndex >= mySourceLength) {
			myLexemeType = EOF_LEXEME;
			myLexemeBeginIndex = mySourceLength;
			myLexemeEndIndex = mySourceLength;
			state = -1;
		    } else if (Character.isWhitespace
			       (c = mySource.charAt (myCurrentIndex ++))) {
			state = 0;
		    } else if (c == '\"') {
			myLexemeType = QUOTED_STRING_LEXEME;
			myLexemeBeginIndex = myCurrentIndex;
			state = 1;
		    } else if (c == '(') {
			++ commentLevel;
			state = 3;
		    } else if (c == '/'  || c == ';' || c == '=' ||
			       c == ')'  || c == '<' || c == '>' ||
			       c == '@'  || c == ',' || c == ':' ||
			       c == '\\' || c == '[' || c == ']' ||
			       c == '?') {
			myLexemeType = TSPECIAL_LEXEME;
			myLexemeBeginIndex = myCurrentIndex - 1;
			myLexemeEndIndex = myCurrentIndex;
			state = -1;
		    } else {
			myLexemeType = TOKEN_LEXEME;
			myLexemeBeginIndex = myCurrentIndex - 1;
			state = 5;
		    }
		    break;
		    // In a quoted string
		case 1:
		    if (myCurrentIndex >= mySourceLength) {
			myLexemeType = ILLEGAL_LEXEME;
			myLexemeBeginIndex = mySourceLength;
			myLexemeEndIndex = mySourceLength;
			state = -1;
		    } else if ((c = mySource.charAt (myCurrentIndex ++)) == '\"') {
			myLexemeEndIndex = myCurrentIndex - 1;
			state = -1;
		    } else if (c == '\\') {
			state = 2;
		    } else {
			state = 1;
		    }
		    break;
		    // In a quoted string, backslash seen
		case 2:
		    if (myCurrentIndex >= mySourceLength) {
			myLexemeType = ILLEGAL_LEXEME;
			myLexemeBeginIndex = mySourceLength;
			myLexemeEndIndex = mySourceLength;
			state = -1;
		    } else {
			++ myCurrentIndex;
			state = 1;
		    } break;
		    // In a comment
		case 3: if (myCurrentIndex >= mySourceLength) {
		    myLexemeType = ILLEGAL_LEXEME;
		    myLexemeBeginIndex = mySourceLength;
		    myLexemeEndIndex = mySourceLength;
		    state = -1;
		} else if ((c = mySource.charAt (myCurrentIndex ++)) == '(') {
		    ++ commentLevel;
		    state = 3;
		} else if (c == ')') {
		    -- commentLevel;
		    state = commentLevel == 0 ? 0 : 3;
		} else if (c == '\\') {
		    state = 4;
		} else { state = 3;
		}
		break;
		// In a comment, backslash seen
		case 4:
		    if (myCurrentIndex >= mySourceLength) {
			myLexemeType = ILLEGAL_LEXEME;
			myLexemeBeginIndex = mySourceLength;
			myLexemeEndIndex = mySourceLength;
			state = -1;
		    } else {
			++ myCurrentIndex;
			state = 3;
		    }
		    break;
		    // In a token
		case 5:
		    if (myCurrentIndex >= mySourceLength) {
			myLexemeEndIndex = myCurrentIndex;
			state = -1;
		    } else if (Character.isWhitespace
			       (c = mySource.charAt (myCurrentIndex ++))) {
			myLexemeEndIndex = myCurrentIndex - 1;
			state = -1;
		    } else if (c == '\"' || c == '(' || c == '/' ||
			       c == ';'  || c == '=' || c == ')' ||
			       c == '<' || c == '>'  || c == '@' ||
			       c == ',' || c == ':' || c == '\\' ||
			       c == '[' || c == ']' || c == '?') {
			-- myCurrentIndex;
			myLexemeEndIndex = myCurrentIndex;
			state = -1;
		    } else {
			state = 5;
		    }
		    break;
		}
	    }
	    
	}
	
    }

    /**
     * Returns a lowercase version of the given string. The lowercase version  
     * is constructed by applying Character.toLowerCase() to each character of
     * the given string, which maps characters to lowercase using the rules of 
     * Unicode. This mapping is the same regardless of locale, whereas the 
     * mapping of String.toLowerCase() may be different depending on the  
     * default locale. 
     */
    private static String toUnicodeLowerCase(String s) {
	int n = s.length();
	char[] result = new char [n];
	for (int i = 0; i < n; ++ i) {
	    result[i] = Character.toLowerCase (s.charAt (i));
	}
	return new String (result);
    }

    /**
     * Returns a version of the given string with backslashes removed.
     */
    private static String removeBackslashes(String s) {
	int n = s.length();
	char[] result = new char [n];
	int i;
	int j = 0;
	char c;
	for (i = 0; i < n; ++ i) {
	    c = s.charAt (i);
	    if (c == '\\') {
		c = s.charAt (++ i);
	    }
	    result[j++] = c;
	}
	return new String (result, 0, j);
    }

    /**
     * Returns a version of the string surrounded by quotes and with interior
     * quotes preceded by a backslash.
     */
    private static String addQuotes(String s) {
	int n = s.length();
	int i;
	char c;
	StringBuffer result = new StringBuffer (n+2);
	result.append ('\"');
	for (i = 0; i < n; ++ i) {
	    c = s.charAt (i);
	    if (c == '\"') {
		result.append ('\\');
	    }
	    result.append (c);
	}
	result.append ('\"');
	return result.toString();
    }

    /**
     * Parses the given string into canonical pieces and stores the pieces in
     * {@link #myPieces <CODE>myPieces</CODE>}.
     * <P>
     * Special rules applied:
     * <UL>
     * <LI> If the media type is text, the value of a charset parameter is
     *      converted to lowercase.
     * </UL>
     *
     * @param  s  MIME media type string.
     *
     * @exception  NullPointerException
     *     (unchecked exception) Thrown if <CODE>s</CODE> is null. 
     * @exception  IllegalArgumentException
     *     (unchecked exception) Thrown if <CODE>s</CODE> does not obey the 
     *     syntax for a MIME media type string. 
     */
    private void parse(String s) {
	// Initialize.
	if (s == null) {
	    throw new NullPointerException();
	}
	LexicalAnalyzer theLexer = new LexicalAnalyzer (s);
	int theLexemeType;
	Vector thePieces = new Vector();
	boolean mediaTypeIsText = false;
	boolean parameterNameIsCharset = false;

	// Parse media type.
	if (theLexer.getLexemeType() == TOKEN_LEXEME) {
	    String mt = toUnicodeLowerCase (theLexer.getLexeme());
	    thePieces.add (mt);
	    theLexer.nextLexeme();
	    mediaTypeIsText = mt.equals ("text");
	} else {
	    throw new IllegalArgumentException();
	}
	// Parse slash.
	if (theLexer.getLexemeType() == TSPECIAL_LEXEME &&
	      theLexer.getLexemeFirstCharacter() == '/') {
	    theLexer.nextLexeme();
	} else {
	    throw new IllegalArgumentException();
	}
	if (theLexer.getLexemeType() == TOKEN_LEXEME) {
	    thePieces.add (toUnicodeLowerCase (theLexer.getLexeme()));
	    theLexer.nextLexeme();
	} else {
	    throw new IllegalArgumentException();
	}
	// Parse zero or more parameters.
	while (theLexer.getLexemeType() == TSPECIAL_LEXEME &&
	       theLexer.getLexemeFirstCharacter() == ';') {
	    // Parse semicolon.
	    theLexer.nextLexeme();

	    // Parse parameter name.
	    if (theLexer.getLexemeType() == TOKEN_LEXEME) {
		String pn = toUnicodeLowerCase (theLexer.getLexeme());
		thePieces.add (pn);
		theLexer.nextLexeme();
		parameterNameIsCharset = pn.equals ("charset");
	    } else {
		throw new IllegalArgumentException();
	    }
	    
	    // Parse equals.
	    if (theLexer.getLexemeType() == TSPECIAL_LEXEME &&
		theLexer.getLexemeFirstCharacter() == '=') {
		theLexer.nextLexeme();
	    } else {
		throw new IllegalArgumentException();
	    }

	    // Parse parameter value.
	    if (theLexer.getLexemeType() == TOKEN_LEXEME) {
		String pv = theLexer.getLexeme();
		thePieces.add(mediaTypeIsText && parameterNameIsCharset ?
			      toUnicodeLowerCase (pv) :
			      pv);
		theLexer.nextLexeme();
	    } else if (theLexer.getLexemeType() == QUOTED_STRING_LEXEME) {
		String pv = removeBackslashes (theLexer.getLexeme());
		thePieces.add(mediaTypeIsText && parameterNameIsCharset ?
			      toUnicodeLowerCase (pv) :
			      pv);
		theLexer.nextLexeme();
	    } else {
		throw new IllegalArgumentException();
	    }
	}

	// Make sure we've consumed everything.
	if (theLexer.getLexemeType() != EOF_LEXEME) {
	    throw new IllegalArgumentException();
	}

	// Save the pieces. Parameters are not in ascending order yet.
	int n = thePieces.size();
	myPieces = (String[]) thePieces.toArray (new String [n]);

	// Sort the parameters into ascending order using an insertion sort.
	int i, j;
	String temp;
	for (i = 4; i < n; i += 2) {
	    j = 2;
	    while (j < i && myPieces[j].compareTo (myPieces[i]) <= 0) {
		j += 2;
	    }
	    while (j < i) {
		temp = myPieces[j];
		myPieces[j] = myPieces[i];
		myPieces[i] = temp;
		temp = myPieces[j+1];
		myPieces[j+1] = myPieces[i+1];
		myPieces[i+1] = temp;
		j += 2;
	    }
	}
    }
}