InternetAddress.java

/*
 * Copyright (c) 1997, 2020 Oracle and/or its affiliates. All rights reserved.
 *
 * This program and the accompanying materials are made available under the
 * terms of the Eclipse Public License v. 2.0, which is available at
 * http://www.eclipse.org/legal/epl-2.0.
 *
 * This Source Code may also be made available under the following Secondary
 * Licenses when the conditions for such availability set forth in the
 * Eclipse Public License v. 2.0 are satisfied: GNU General Public License,
 * version 2 with the GNU Classpath Exception, which is available at
 * https://www.gnu.org/software/classpath/license.html.
 *
 * SPDX-License-Identifier: EPL-2.0 OR GPL-2.0 WITH Classpath-exception-2.0
 */

package jakarta.mail.internet;

import java.io.UnsupportedEncodingException;
import java.net.InetAddress;
import java.net.UnknownHostException;
import java.util.List;
import java.util.ArrayList;
import java.util.StringTokenizer;
import java.util.Locale;
import java.nio.charset.StandardCharsets;
import jakarta.mail.*;
import com.sun.mail.util.PropUtil;

/**
 * This class represents an Internet email address using the syntax
 * of <a href="http://www.ietf.org/rfc/rfc822.txt" target="_top">RFC822</a>.
 * Typical address syntax is of the form "user@host.domain" or
 * "Personal Name &lt;user@host.domain&gt;".
 *
 * @author Bill Shannon
 * @author John Mani
 */

public class InternetAddress extends Address implements Cloneable {

    protected String address; // email address

    /**
     * The personal name.
     */
    protected String personal;

    /**
     * The RFC 2047 encoded version of the personal name. <p>
     *
     * This field and the <code>personal</code> field track each
     * other, so if a subclass sets one of these fields directly, it
     * should set the other to <code>null</code>, so that it is
     * suitably recomputed.
     */
    protected String encodedPersonal;

    private static final long serialVersionUID = -7507595530758302903L;

    private static final boolean ignoreBogusGroupName =
	PropUtil.getBooleanSystemProperty(
			    "mail.mime.address.ignorebogusgroupname", true);

    private static final boolean useCanonicalHostName =
	PropUtil.getBooleanSystemProperty(
			    "mail.mime.address.usecanonicalhostname", true);

    private static final boolean allowUtf8 =
	PropUtil.getBooleanSystemProperty("mail.mime.allowutf8", false);

    /**
     * Default constructor.
     */
    public InternetAddress() { }

    /**
     * Constructor. <p>
     *
     * Parse the given string and create an InternetAddress.
     * See the <code>parse</code> method for details of the parsing.
     * The address is parsed using "strict" parsing.
     * This constructor does <b>not</b> perform the additional
     * syntax checks that the
     * <code>InternetAddress(String address, boolean strict)</code>
     * constructor does when <code>strict</code> is <code>true</code>.
     * This constructor is equivalent to
     * <code>InternetAddress(address, false)</code>.
     *
     * @param address	the address in RFC822 format
     * @exception	AddressException if the parse failed
     */
    public InternetAddress(String address) throws AddressException {
	// use our address parsing utility routine to parse the string
	InternetAddress a[] = parse(address, true);
	// if we got back anything other than a single address, it's an error
	if (a.length != 1)
	    throw new AddressException("Illegal address", address);

	/*
	 * Now copy the contents of the single address we parsed
	 * into the current object, which will be returned from the
	 * constructor.
	 * XXX - this sure is a round-about way of getting this done.
	 */
	this.address = a[0].address;
	this.personal = a[0].personal;
	this.encodedPersonal = a[0].encodedPersonal;
    }

    /**
     * Parse the given string and create an InternetAddress.
     * If <code>strict</code> is false, the detailed syntax of the
     * address isn't checked.
     *
     * @param	address		the address in RFC822 format
     * @param	strict		enforce RFC822 syntax
     * @exception		AddressException if the parse failed
     * @since			JavaMail 1.3
     */
    public InternetAddress(String address, boolean strict)
						throws AddressException {
	this(address);
	if (strict) {
	    if (isGroup())
		getGroup(true);	// throw away the result
	    else
		checkAddress(this.address, true, true);
	}
    }

    /**
     * Construct an InternetAddress given the address and personal name.
     * The address is assumed to be a syntactically valid RFC822 address.
     *
     * @param address	the address in RFC822 format
     * @param personal	the personal name
     * @exception	UnsupportedEncodingException if the personal name
     *			can't be encoded in the given charset
     */
    public InternetAddress(String address, String personal)
				throws UnsupportedEncodingException {
	this(address, personal, null);
    }

    /**
     * Construct an InternetAddress given the address and personal name.
     * The address is assumed to be a syntactically valid RFC822 address.
     *
     * @param address	the address in RFC822 format
     * @param personal	the personal name
     * @param charset	the MIME charset for the name
     * @exception	UnsupportedEncodingException if the personal name
     *			can't be encoded in the given charset
     */
    public InternetAddress(String address, String personal, String charset)
				throws UnsupportedEncodingException {
	this.address = address;
	setPersonal(personal, charset);
    }

    /**
     * Return a copy of this InternetAddress object.
     * @since		JavaMail 1.2
     */
    @Override
    public Object clone() {
	InternetAddress a = null;
	try {
	    a = (InternetAddress)super.clone();
	} catch (CloneNotSupportedException e) {} // Won't happen
	return a;
    }

    /**
     * Return the type of this address. The type of an InternetAddress
     * is "rfc822".
     */
    @Override
    public String getType() {
	return "rfc822";
    }

    /**
     * Set the email address.
     *
     * @param	address email address
     */
    public void setAddress(String address) {
	this.address = address;
    }

    /**
     * Set the personal name. If the name contains non US-ASCII
     * characters, then the name will be encoded using the specified
     * charset as per RFC 2047. If the name contains only US-ASCII
     * characters, no encoding is done and the name is used as is. <p>
     *
     * @param	name 	personal name
     * @param	charset	MIME charset to be used to encode the name as 
     *			per RFC 2047
     * @see 	#setPersonal(String)
     * @exception UnsupportedEncodingException if the charset encoding
     *		  fails.
     */
    public void setPersonal(String name, String charset)
				throws UnsupportedEncodingException {
	personal = name;
	if (name != null)
	    encodedPersonal = MimeUtility.encodeWord(name, charset, null);
	else
	    encodedPersonal = null;
    }

    /**
     * Set the personal name. If the name contains non US-ASCII
     * characters, then the name will be encoded using the platform's 
     * default charset. If the name contains only US-ASCII characters,
     * no encoding is done and the name is used as is. <p>
     *
     * @param	name 	personal name
     * @see 	#setPersonal(String name, String charset)
     * @exception UnsupportedEncodingException if the charset encoding
     *		  fails.
     */
    public void setPersonal(String name) 
		throws UnsupportedEncodingException {
	personal = name;
	if (name != null)
	    encodedPersonal = MimeUtility.encodeWord(name);
	else
	    encodedPersonal = null;
    }

    /**
     * Get the email address.
     * @return	email address
     */
    public String getAddress() {
	return address;
    }

    /**
     * Get the personal name. If the name is encoded as per RFC 2047,
     * it is decoded and converted into Unicode. If the decoding or
     * conversion fails, the raw data is returned as is.
     *
     * @return	personal name
     */
    public String getPersonal() {
	if (personal != null)
	    return personal;
	
	if (encodedPersonal != null) {
	    try {
		personal = MimeUtility.decodeText(encodedPersonal);
		return personal;
	    } catch (Exception ex) {
		// 1. ParseException: either its an unencoded string or
		//	it can't be parsed
		// 2. UnsupportedEncodingException: can't decode it.
		return encodedPersonal;
	    }
	}
	// No personal or encodedPersonal, return null
	return null;
    }

    /**
     * Convert this address into a RFC 822 / RFC 2047 encoded address.
     * The resulting string contains only US-ASCII characters, and
     * hence is mail-safe.
     *
     * @return		possibly encoded address string
     */
    @Override
    public String toString() {
	String a = address == null ? "" : address;
	if (encodedPersonal == null && personal != null)
	    try {
		encodedPersonal = MimeUtility.encodeWord(personal);
	    } catch (UnsupportedEncodingException ex) { }
	
	if (encodedPersonal != null)
	    return quotePhrase(encodedPersonal) + " <" + a + ">";
	else if (isGroup() || isSimple())
	    return a;
	else
	    return "<" + a + ">";
    }

    /**
     * Returns a properly formatted address (RFC 822 syntax) of
     * Unicode characters.
     *   
     * @return          Unicode address string
     * @since           JavaMail 1.2
     */  
    public String toUnicodeString() {
	String p = getPersonal();
        if (p != null)
            return quotePhrase(p) + " <" + address + ">";
        else if (isGroup() || isSimple())
            return address;
        else
            return "<" + address + ">";
    }

    /*
     * quotePhrase() quotes the words within a RFC822 phrase.
     *
     * This is tricky, since a phrase is defined as 1 or more
     * RFC822 words, separated by LWSP. Now, a word that contains
     * LWSP is supposed to be quoted, and this is exactly what the 
     * MimeUtility.quote() method does. However, when dealing with
     * a phrase, any LWSP encountered can be construed to be the
     * separator between words, and not part of the words themselves.
     * To deal with this funkiness, we have the below variant of
     * MimeUtility.quote(), which essentially ignores LWSP when
     * deciding whether to quote a word.
     *
     * It aint pretty, but it gets the job done :)
     */

    private static final String rfc822phrase =
	HeaderTokenizer.RFC822.replace(' ', '\0').replace('\t', '\0');

    private static String quotePhrase(String phrase) {
        int len = phrase.length();
        boolean needQuoting = false;

        for (int i = 0; i < len; i++) {
            char c = phrase.charAt(i);
            if (c == '"' || c == '\\') { 
                // need to escape them and then quote the whole string
                StringBuilder sb = new StringBuilder(len + 3);
                sb.append('"');
                for (int j = 0; j < len; j++) {
                    char cc = phrase.charAt(j);
                    if (cc == '"' || cc == '\\')
                        // Escape the character
                        sb.append('\\');
                    sb.append(cc);
                }
                sb.append('"');
                return sb.toString();
            } else if ((c < 040 && c != '\r' && c != '\n' && c != '\t') || 
		    (c >= 0177 && !allowUtf8) || rfc822phrase.indexOf(c) >= 0)
		// These characters cause the string to be quoted
                needQuoting = true;
        }

        if (needQuoting) {
            StringBuilder sb = new StringBuilder(len + 2);
            sb.append('"').append(phrase).append('"');
            return sb.toString();
        } else
            return phrase;
    }

    private static String unquote(String s) {
	if (s.startsWith("\"") && s.endsWith("\"") && s.length() > 1) {
	    s = s.substring(1, s.length() - 1);
	    // check for any escaped characters
	    if (s.indexOf('\\') >= 0) {
		StringBuilder sb = new StringBuilder(s.length());	// approx
		for (int i = 0; i < s.length(); i++) {
		    char c = s.charAt(i);
		    if (c == '\\' && i < s.length() - 1)
			c = s.charAt(++i);
		    sb.append(c);
		}
		s = sb.toString();
	    }
	}
	return s;
    }

    /**
     * The equality operator.
     */
    @Override
    public boolean equals(Object a) {
	if (!(a instanceof InternetAddress))
	    return false;

	String s = ((InternetAddress)a).getAddress();
	if (s == address)
	    return true;
	if (address != null && address.equalsIgnoreCase(s))
	    return true;

	return false;
    }

    /**
     * Compute a hash code for the address.
     */
    @Override
    public int hashCode() {
	if (address == null)
	    return 0;
	else
	    return address.toLowerCase(Locale.ENGLISH).hashCode();
    }

    /**
     * Convert the given array of InternetAddress objects into
     * a comma separated sequence of address strings. The
     * resulting string contains only US-ASCII characters, and
     * hence is mail-safe. <p>
     *
     * @param addresses	array of InternetAddress objects
     * @exception 	ClassCastException if any address object in the 
     *			given array is not an InternetAddress object. Note
     *			that this is a RuntimeException.
     * @return		comma separated string of addresses
     */
    public static String toString(Address[] addresses) {
	return toString(addresses, 0);
    }

    /**
     * Convert the given array of InternetAddress objects into
     * a comma separated sequence of address strings. The
     * resulting string contains Unicode characters. <p>
     *
     * @param addresses	array of InternetAddress objects
     * @exception 	ClassCastException if any address object in the 
     *			given array is not an InternetAddress object. Note
     *			that this is a RuntimeException.
     * @return		comma separated string of addresses
     * @since		JavaMail 1.6
     */
    public static String toUnicodeString(Address[] addresses) {
	return toUnicodeString(addresses, 0);
    }

    /**
     * Convert the given array of InternetAddress objects into
     * a comma separated sequence of address strings. The
     * resulting string contains only US-ASCII characters, and
     * hence is mail-safe. <p>
     *
     * The 'used' parameter specifies the number of character positions
     * already taken up in the field into which the resulting address 
     * sequence string is to be inserted. It is used to determine the 
     * line-break positions in the resulting address sequence string.
     *
     * @param addresses	array of InternetAddress objects
     * @param used	number of character positions already used, in
     *			the field into which the address string is to
     *			be inserted.
     * @exception 	ClassCastException if any address object in the 
     *			given array is not an InternetAddress object. Note
     *			that this is a RuntimeException.
     * @return		comma separated string of addresses
     */
    public static String toString(Address[] addresses, int used) {
	if (addresses == null || addresses.length == 0)
	    return null;

	StringBuilder sb = new StringBuilder();

	for (int i = 0; i < addresses.length; i++) {
	    if (i != 0) { // need to append comma
		sb.append(", ");
		used += 2;
	    }

	    // prefer not to split a single address across lines so used=0 below
	    String s = MimeUtility.fold(0, addresses[i].toString());
	    int len = lengthOfFirstSegment(s); // length till CRLF
	    if (used + len > 76) { // overflows ...
		// smash trailing space from ", " above
		int curlen = sb.length();
		if (curlen > 0 && sb.charAt(curlen - 1) == ' ')
		    sb.setLength(curlen - 1);
		sb.append("\r\n\t"); // .. start new continuation line
		used = 8; // account for the starting <tab> char
	    }
	    sb.append(s);
	    used = lengthOfLastSegment(s, used);
	}

	return sb.toString();
    }

    /**
     * Convert the given array of InternetAddress objects into
     * a comma separated sequence of address strings. The
     * resulting string contains Unicode characters. <p>
     *
     * The 'used' parameter specifies the number of character positions
     * already taken up in the field into which the resulting address 
     * sequence string is to be inserted. It is used to determine the 
     * line-break positions in the resulting address sequence string.
     *
     * @param addresses	array of InternetAddress objects
     * @param used	number of character positions already used, in
     *			the field into which the address string is to
     *			be inserted.
     * @exception 	ClassCastException if any address object in the 
     *			given array is not an InternetAddress object. Note
     *			that this is a RuntimeException.
     * @return		comma separated string of addresses
     * @since		JavaMail 1.6
     */
    /*
     * XXX - This is exactly the same as the above, except it uses
     *	     toUnicodeString instead of toString.
     * XXX - Since the line length restrictions are in bytes, not characters,
     *	     we convert all non-ASCII addresses to UTF-8 byte strings,
     *	     which we then convert to ISO-8859-1 Strings where every
     *	     character respresents one UTF-8 byte.  At the end we reverse
     *	     the conversion to get back to a correct Unicode string.
     *	     This is a hack to allow all the other character-based methods
     *	     to work properly with UTF-8 bytes.
     */
    public static String toUnicodeString(Address[] addresses, int used) {
	if (addresses == null || addresses.length == 0)
	    return null;

	StringBuilder sb = new StringBuilder();

	boolean sawNonAscii = false;
	for (int i = 0; i < addresses.length; i++) {
	    if (i != 0) { // need to append comma
		sb.append(", ");
		used += 2;
	    }

	    // prefer not to split a single address across lines so used=0 below
	    String as = ((InternetAddress)addresses[i]).toUnicodeString();
	    if (MimeUtility.checkAscii(as) != MimeUtility.ALL_ASCII) {
		sawNonAscii = true;
		as = new String(as.getBytes(StandardCharsets.UTF_8),
	    			StandardCharsets.ISO_8859_1);
	    }
	    String s = MimeUtility.fold(0, as);
	    int len = lengthOfFirstSegment(s); // length till CRLF
	    if (used + len > 76) { // overflows ...
		// smash trailing space from ", " above
		int curlen = sb.length();
		if (curlen > 0 && sb.charAt(curlen - 1) == ' ')
		    sb.setLength(curlen - 1);
		sb.append("\r\n\t"); // .. start new continuation line
		used = 8; // account for the starting <tab> char
	    }
	    sb.append(s);
	    used = lengthOfLastSegment(s, used);
	}

	String ret = sb.toString();
	if (sawNonAscii)
	    ret = new String(ret.getBytes(StandardCharsets.ISO_8859_1),
				StandardCharsets.UTF_8);
    	return ret;
    }

    /*
     * Return the length of the first segment within this string.
     * If no segments exist, the length of the whole line is returned.
     */
    private static int lengthOfFirstSegment(String s) {
	int pos;
	if ((pos = s.indexOf("\r\n")) != -1)
	    return pos;
	else
	    return s.length();
    }

    /*
     * Return the length of the last segment within this string.
     * If no segments exist, the length of the whole line plus
     * <code>used</code> is returned.
     */
    private static int lengthOfLastSegment(String s, int used) {
	int pos;
	if ((pos = s.lastIndexOf("\r\n")) != -1)
	    return s.length() - pos - 2;
	else 
	    return s.length() + used;
    }

    /**
     * Return an InternetAddress object representing the current user.
     * The entire email address may be specified in the "mail.from"
     * property.  If not set, the "mail.user" and "mail.host" properties
     * are tried.  If those are not set, the "user.name" property and
     * <code>InetAddress.getLocalHost</code> method are tried.
     * Security exceptions that may occur while accessing this information
     * are ignored.  If it is not possible to determine an email address,
     * null is returned.
     *
     * @param	session		Session object used for property lookup
     * @return			current user's email address
     */
    public static InternetAddress getLocalAddress(Session session) {
	try {
	    return _getLocalAddress(session);
	} catch (SecurityException sex) {	// ignore it
	} catch (AddressException ex) {		// ignore it
	} catch (UnknownHostException ex) { }	// ignore it
	return null;
    }

    /**
     * A package-private version of getLocalAddress that doesn't swallow
     * the exception.  Used by MimeMessage.setFrom() to report the reason
     * for the failure.
     */
    // package-private
    static InternetAddress _getLocalAddress(Session session)
	    throws SecurityException, AddressException, UnknownHostException {
	String user = null, host = null, address = null;
	if (session == null) {
	    user = System.getProperty("user.name");
	    host = getLocalHostName();
	} else {
	    address = session.getProperty("mail.from");
	    if (address == null) {
		user = session.getProperty("mail.user");
		if (user == null || user.length() == 0)
		    user = session.getProperty("user.name");
		if (user == null || user.length() == 0)
		    user = System.getProperty("user.name");
		host = session.getProperty("mail.host");
		if (host == null || host.length() == 0)
		    host = getLocalHostName();
	    }
	}

	if (address == null && user != null && user.length() != 0 &&
		host != null && host.length() != 0)
	    address = MimeUtility.quote(user.trim(), specialsNoDot + "\t ") +
							    "@" + host;

	if (address == null)
	    return null;

	return new InternetAddress(address);
    }

    /**
     * Get the local host name from InetAddress and return it in a form
     * suitable for use in an email address.
     */
    private static String getLocalHostName() throws UnknownHostException {
	String host = null;
	InetAddress me = InetAddress.getLocalHost();
	if (me != null) {
	    // try canonical host name first
	    if (useCanonicalHostName)
		host = me.getCanonicalHostName();
	    if (host == null)
		host = me.getHostName();
	    // if we can't get our name, use local address literal
	    if (host == null)
		host = me.getHostAddress();
	    if (host != null && host.length() > 0 && isInetAddressLiteral(host))
		host = '[' + host + ']';
	}
	return host;
    }

    /**
     * Is the address an IPv4 or IPv6 address literal, which needs to
     * be enclosed in "[]" in an email address?  IPv4 literals contain
     * decimal digits and dots, IPv6 literals contain hex digits, dots,
     * and colons.  We're lazy and don't check the exact syntax, just
     * the allowed characters; strings that have only the allowed
     * characters in a literal but don't meet the syntax requirements
     * for a literal definitely can't be a host name and thus will fail
     * later when used as an address literal.
     */
    private static boolean isInetAddressLiteral(String addr) {
	boolean sawHex = false, sawColon = false;
	for (int i = 0; i < addr.length(); i++) {
	    char c = addr.charAt(i);
	    if (c >= '0' && c <= '9')
		;	// digits always ok
	    else if (c == '.')
		;	// dot always ok
	    else if ((c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z'))
		sawHex = true;	// need to see a colon too
	    else if (c == ':')
		sawColon = true;
	    else
		return false;	// anything else, definitely not a literal
	}
	return !sawHex || sawColon;
    }

    /**
     * Parse the given comma separated sequence of addresses into
     * InternetAddress objects.  Addresses must follow RFC822 syntax.
     *
     * @param addresslist	comma separated address strings
     * @return			array of InternetAddress objects
     * @exception		AddressException if the parse failed
     */
    public static InternetAddress[] parse(String addresslist) 
				throws AddressException {
	return parse(addresslist, true);
    }

    /**
     * Parse the given sequence of addresses into InternetAddress
     * objects.  If <code>strict</code> is false, simple email addresses
     * separated by spaces are also allowed.  If <code>strict</code> is
     * true, many (but not all) of the RFC822 syntax rules are enforced.
     * In particular, even if <code>strict</code> is true, addresses
     * composed of simple names (with no "@domain" part) are allowed.
     * Such "illegal" addresses are not uncommon in real messages. <p>
     *
     * Non-strict parsing is typically used when parsing a list of
     * mail addresses entered by a human.  Strict parsing is typically
     * used when parsing address headers in mail messages.
     *
     * @param	addresslist	comma separated address strings
     * @param	strict		enforce RFC822 syntax
     * @return			array of InternetAddress objects
     * @exception		AddressException if the parse failed
     */
    public static InternetAddress[] parse(String addresslist, boolean strict)
					    throws AddressException {
	return parse(addresslist, strict, false);
    }

    /**
     * Parse the given sequence of addresses into InternetAddress
     * objects.  If <code>strict</code> is false, the full syntax rules for
     * individual addresses are not enforced.  If <code>strict</code> is
     * true, many (but not all) of the RFC822 syntax rules are enforced. <p>
     *
     * To better support the range of "invalid" addresses seen in real
     * messages, this method enforces fewer syntax rules than the
     * <code>parse</code> method when the strict flag is false
     * and enforces more rules when the strict flag is true.  If the
     * strict flag is false and the parse is successful in separating out an
     * email address or addresses, the syntax of the addresses themselves
     * is not checked.
     *
     * @param	addresslist	comma separated address strings
     * @param	strict		enforce RFC822 syntax
     * @return			array of InternetAddress objects
     * @exception		AddressException if the parse failed
     * @since			JavaMail 1.3
     */
    public static InternetAddress[] parseHeader(String addresslist,
				boolean strict) throws AddressException {
	return parse(MimeUtility.unfold(addresslist), strict, true);
    }

    /*
     * RFC822 Address parser.
     *
     * XXX - This is complex enough that it ought to be a real parser,
     *       not this ad-hoc mess, and because of that, this is not perfect.
     *
     * XXX - Deal with encoded Headers too.
     */
    @SuppressWarnings("fallthrough")
    private static InternetAddress[] parse(String s, boolean strict,
				    boolean parseHdr) throws AddressException {
	int start, end, index, nesting;
	int start_personal = -1, end_personal = -1;
	int length = s.length();
	boolean ignoreErrors = parseHdr && !strict;
	boolean in_group = false;	// we're processing a group term
	boolean route_addr = false;	// address came from route-addr term
	boolean rfc822 = false;		// looks like an RFC822 address
	char c;
	List<InternetAddress> v = new ArrayList<>();
	InternetAddress ma;

	for (start = end = -1, index = 0; index < length; index++) {
    	    c = s.charAt(index);

	    switch (c) {
	    case '(': // We are parsing a Comment. Ignore everything inside.
		// XXX - comment fields should be parsed as whitespace,
		//	 more than one allowed per address
		rfc822 = true;
		if (start >= 0 && end == -1)
		    end = index;
		int pindex = index;
		for (index++, nesting = 1; index < length && nesting > 0;
		  index++) {
		    c = s.charAt(index);
		    switch (c) {
		    case '\\':
			index++; // skip both '\' and the escaped char
			break;
		    case '(':
			nesting++;
			break;
		    case ')':
			nesting--;
			break;
		    default:
			break;
		    }
		}
		if (nesting > 0) {
		    if (!ignoreErrors)
			throw new AddressException("Missing ')'", s, index);
		    // pretend the first paren was a regular character and
		    // continue parsing after it
		    index = pindex + 1;
		    break;
		}
		index--;	// point to closing paren
		if (start_personal == -1)
		    start_personal = pindex + 1;
		if (end_personal == -1)
		    end_personal = index;
		break;

	    case ')':
		if (!ignoreErrors)
		    throw new AddressException("Missing '('", s, index);
		// pretend the left paren was a regular character and
		// continue parsing
		if (start == -1)
		    start = index;
		break;

	    case '<':
		rfc822 = true;
		if (route_addr) {
		    if (!ignoreErrors)
			throw new AddressException(
						"Extra route-addr", s, index);

		    // assume missing comma between addresses
		    if (start == -1) {
			route_addr = false;
			rfc822 = false;
			start = end = -1;
			break;	// nope, nothing there
		    }
		    if (!in_group) {
			// got a token, add this to our InternetAddress list
			if (end == -1)	// should never happen
			    end = index;
			String addr = s.substring(start, end).trim();

			ma = new InternetAddress();
			ma.setAddress(addr);
			if (start_personal >= 0) {
			    ma.encodedPersonal = unquote(
				s.substring(start_personal, end_personal).
								trim());
			}
			v.add(ma);

			route_addr = false;
			rfc822 = false;
			start = end = -1;
			start_personal = end_personal = -1;
			// continue processing this new address...
		    }
		}

		int rindex = index;
		boolean inquote = false;
	      outf:
		for (index++; index < length; index++) {
		    c = s.charAt(index);
		    switch (c) {
		    case '\\':	// XXX - is this needed?
			index++; // skip both '\' and the escaped char
			break;
		    case '"':
			inquote = !inquote;
			break;
		    case '>':
			if (inquote)
			    continue;
			break outf; // out of for loop
		    default:
			break;
		    }
		}

		// did we find a matching quote?
		if (inquote) {
		    if (!ignoreErrors)
			throw new AddressException("Missing '\"'", s, index);
		    // didn't find matching quote, try again ignoring quotes
		    // (e.g., ``<"@foo.com>'')
		  outq:
		    for (index = rindex + 1; index < length; index++) {
			c = s.charAt(index);
			if (c == '\\')	// XXX - is this needed?
			    index++;	// skip both '\' and the escaped char
			else if (c == '>')
			    break;
		    }
		}

		// did we find a terminating '>'?
		if (index >= length) {
		    if (!ignoreErrors)
			throw new AddressException("Missing '>'", s, index);
		    // pretend the "<" was a regular character and
		    // continue parsing after it (e.g., ``<@foo.com'')
		    index = rindex + 1;
		    if (start == -1)
			start = rindex;	// back up to include "<"
		    break;
		}

		if (!in_group) {
		    if (start >= 0) {
			// seen some characters?  use them as the personal name
			start_personal = start;
			end_personal = rindex;
		    }
		    start = rindex + 1;
		}
		route_addr = true;
		end = index;
		break;

	    case '>':
		if (!ignoreErrors)
		    throw new AddressException("Missing '<'", s, index);
		// pretend the ">" was a regular character and
		// continue parsing (e.g., ``>@foo.com'')
		if (start == -1)
		    start = index;
		break;

	    case '"':	// parse quoted string
		int qindex = index;
		rfc822 = true;
		if (start == -1)
		    start = index;
	      outq:
		for (index++; index < length; index++) {
		    c = s.charAt(index);
		    switch (c) {
		    case '\\':
			index++; // skip both '\' and the escaped char
			break;
		    case '"':
			break outq; // out of for loop
		    default:
			break;
		    }
		}
		if (index >= length) {
		    if (!ignoreErrors)
			throw new AddressException("Missing '\"'", s, index);
		    // pretend the quote was a regular character and
		    // continue parsing after it (e.g., ``"@foo.com'')
		    index = qindex + 1;
		}
		break;

	    case '[':	// a domain-literal, probably
		int lindex = index;
		rfc822 = true;
		if (start == -1)
		    start = index;
	      outb:
		for (index++; index < length; index++) {
		    c = s.charAt(index);
		    switch (c) {
		    case '\\':
			index++; // skip both '\' and the escaped char
			break;
		    case ']':
			break outb; // out of for loop
		    default:
			break;
		    }
		}
		if (index >= length) {
		    if (!ignoreErrors)
			throw new AddressException("Missing ']'", s, index);
		    // pretend the "[" was a regular character and
		    // continue parsing after it (e.g., ``[@foo.com'')
		    index = lindex + 1;
		}
		break;

	    case ';':
		if (start == -1) {
		    route_addr = false;
		    rfc822 = false;
		    start = end = -1;
		    break;	// nope, nothing there
		}
		if (in_group) {
		    in_group = false;
		    /*
		     * If parsing headers, but not strictly, peek ahead.
		     * If next char is "@", treat the group name
		     * like the local part of the address, e.g.,
		     * "Undisclosed-Recipient:;@java.sun.com".
		     */
		    if (parseHdr && !strict &&
			    index + 1 < length && s.charAt(index + 1) == '@')
			break;
		    ma = new InternetAddress();
		    end = index + 1;
		    ma.setAddress(s.substring(start, end).trim());
		    v.add(ma);

		    route_addr = false;
		    rfc822 = false;
		    start = end = -1;
		    start_personal = end_personal = -1;
		    break;
		}
		if (!ignoreErrors)
		    throw new AddressException(
			    "Illegal semicolon, not in group", s, index);

		// otherwise, parsing a header; treat semicolon like comma
		// fall through to comma case...

	    case ',':	// end of an address, probably
		if (start == -1) {
		    route_addr = false;
		    rfc822 = false;
		    start = end = -1;
		    break;	// nope, nothing there
		}
		if (in_group) {
		    route_addr = false;
		    break;
		}
		// got a token, add this to our InternetAddress list
		if (end == -1)
		    end = index;

		String addr = s.substring(start, end).trim();
		String pers = null;
		if (rfc822 && start_personal >= 0) {
		    pers = unquote(
			    s.substring(start_personal, end_personal).trim());
		    if (pers.trim().length() == 0)
			pers = null;
		}

		/*
		 * If the personal name field has an "@" and the address
		 * field does not, assume they were reversed, e.g.,
		 * ``"joe doe" (john.doe@example.com)''.
		 */
		if (parseHdr && !strict && pers != null &&
			pers.indexOf('@') >= 0 &&
			addr.indexOf('@') < 0 && addr.indexOf('!') < 0) {
		    String tmp = addr;
		    addr = pers;
		    pers = tmp;
		}
		if (rfc822 || strict || parseHdr) {
		    if (!ignoreErrors)
			checkAddress(addr, route_addr, false);
		    ma = new InternetAddress();
		    ma.setAddress(addr);
		    if (pers != null)
			ma.encodedPersonal = pers;
		    v.add(ma);
		} else {
		    // maybe we passed over more than one space-separated addr
		    StringTokenizer st = new StringTokenizer(addr);
		    while (st.hasMoreTokens()) {
			String a = st.nextToken();
			checkAddress(a, false, false);
			ma = new InternetAddress();
			ma.setAddress(a);
			v.add(ma);
		    }
		}

		route_addr = false;
		rfc822 = false;
		start = end = -1;
		start_personal = end_personal = -1;
		break;

	    case ':':
		rfc822 = true;
		if (in_group)
		    if (!ignoreErrors)
			throw new AddressException("Nested group", s, index);
		if (start == -1)
		    start = index;
		if (parseHdr && !strict) {
		    /*
		     * If next char is a special character that can't occur at
		     * the start of a valid address, treat the group name
		     * as the entire address, e.g., "Date:, Tue", "Re:@foo".
		     */
		    if (index + 1 < length) {
			String addressSpecials = ")>[]:@\\,.";
			char nc = s.charAt(index + 1);
			if (addressSpecials.indexOf(nc) >= 0) {
			    if (nc != '@')
				break;	// don't change in_group
			    /*
			     * Handle a common error:
			     * ``Undisclosed-Recipient:@example.com;''
			     *
			     * Scan ahead.  If we find a semicolon before
			     * one of these other special characters,
			     * consider it to be a group after all.
			     */
			    for (int i = index + 2; i < length; i++) {
				nc = s.charAt(i);
				if (nc == ';')
				    break;
				if (addressSpecials.indexOf(nc) >= 0)
				    break;
			    }
			    if (nc == ';')
				break;	// don't change in_group
			}
		    }

		    // ignore bogus "mailto:" prefix in front of an address,
		    // or bogus mail header name included in the address field
		    String gname = s.substring(start, index);
		    if (ignoreBogusGroupName &&
			(gname.equalsIgnoreCase("mailto") ||
			gname.equalsIgnoreCase("From") ||
			gname.equalsIgnoreCase("To") ||
			gname.equalsIgnoreCase("Cc") ||
			gname.equalsIgnoreCase("Subject") ||
			gname.equalsIgnoreCase("Re")))
			start = -1;	// we're not really in a group
		    else
			in_group = true;
		} else
		    in_group = true;
		break;

	    // Ignore whitespace
	    case ' ':
	    case '\t':
	    case '\r':
	    case '\n':
		break;

	    default:
		if (start == -1)
		    start = index;
		break;
	     }
	}

	if (start >= 0) {
	    /*
	     * The last token, add this to our InternetAddress list.
	     * Note that this block of code should be identical to the
	     * block above for "case ','".
	     */
	    if (end == -1)
		end = length;

	    String addr = s.substring(start, end).trim();
	    String pers = null;
	    if (rfc822 && start_personal >= 0) {
		pers = unquote(
			s.substring(start_personal, end_personal).trim());
		if (pers.trim().length() == 0)
		    pers = null;
	    }

	    /*
	     * If the personal name field has an "@" and the address
	     * field does not, assume they were reversed, e.g.,
	     * ``"joe doe" (john.doe@example.com)''.
	     */
	    if (parseHdr && !strict &&
		    pers != null && pers.indexOf('@') >= 0 &&
		    addr.indexOf('@') < 0 && addr.indexOf('!') < 0) {
		String tmp = addr;
		addr = pers;
		pers = tmp;
	    }
	    if (rfc822 || strict || parseHdr) {
		if (!ignoreErrors)
		    checkAddress(addr, route_addr, false);
		ma = new InternetAddress();
		ma.setAddress(addr);
		if (pers != null)
		    ma.encodedPersonal = pers;
		v.add(ma);
	    } else {
		// maybe we passed over more than one space-separated addr
		StringTokenizer st = new StringTokenizer(addr);
		while (st.hasMoreTokens()) {
		    String a = st.nextToken();
		    checkAddress(a, false, false);
		    ma = new InternetAddress();
		    ma.setAddress(a);
		    v.add(ma);
		}
	    }
	}

	InternetAddress[] a = new InternetAddress[v.size()];
	v.toArray(a);
	return a;
    }

    /**
     * Validate that this address conforms to the syntax rules of
     * RFC 822.  The current implementation checks many, but not
     * all, syntax rules.  Note that even though the syntax of
     * the address may be correct, there's no guarantee that a
     * mailbox of that name exists.
     *
     * @exception	AddressException if the address isn't valid.
     * @since		JavaMail 1.3
     */
    public void validate() throws AddressException {
	if (isGroup())
	    getGroup(true);	// throw away the result
	else
	    checkAddress(getAddress(), true, true);
    }

    private static final String specialsNoDotNoAt = "()<>,;:\\\"[]";
    private static final String specialsNoDot = specialsNoDotNoAt + "@";

    /**
     * Check that the address is a valid "mailbox" per RFC822.
     * (We also allow simple names.)
     *
     * XXX - much more to check
     * XXX - doesn't handle domain-literals properly (but no one uses them)
     */
    private static void checkAddress(String addr,
				boolean routeAddr, boolean validate)
				throws AddressException {
	int i, start = 0;

	if (addr == null)
	    throw new AddressException("Address is null");
	int len = addr.length();
	if (len == 0)
	    throw new AddressException("Empty address", addr);

	/*
	 * routeAddr indicates that the address is allowed
	 * to have an RFC 822 "route".
	 */
	if (routeAddr && addr.charAt(0) == '@') {
	    /*
	     * Check for a legal "route-addr":
	     *		[@domain[,@domain ...]:]local@domain
	     */
	    for (start = 0; (i = indexOfAny(addr, ",:", start)) >= 0;
		    start = i+1) {
		if (addr.charAt(start) != '@')
		    throw new AddressException("Illegal route-addr", addr);
		if (addr.charAt(i) == ':') {
		    // end of route-addr
		    start = i + 1;
		    break;
		}
	    }
	}

	/*
	 * The rest should be "local@domain", but we allow simply "local"
	 * unless called from validate.
	 *
	 * local-part must follow RFC 822 - no specials except '.'
	 * unless quoted.
	 */

	char c = (char)-1;
	char lastc = (char)-1;
	boolean inquote = false;
	for (i = start; i < len; i++) {
	    lastc = c;
	    c = addr.charAt(i);
	    // a quoted-pair is only supposed to occur inside a quoted string,
	    // but some people use it outside so we're more lenient
	    if (c == '\\' || lastc == '\\')
		continue;
	    if (c == '"') {
		if (inquote) {
		    // peek ahead, next char must be "@"
		    if (validate && i + 1 < len && addr.charAt(i + 1) != '@')
			throw new AddressException(
				"Quote not at end of local address", addr);
		    inquote = false;
		} else {
		    if (validate && i != 0)
			throw new AddressException(
				"Quote not at start of local address", addr);
		    inquote = true;
		}
		continue;
	    } else if (c == '\r') {
		// peek ahead, next char must be LF
		if (i + 1 < len && addr.charAt(i + 1) != '\n')
		    throw new AddressException(
			"Quoted local address contains CR without LF", addr);
	    } else if (c == '\n') {
		/*
		 * CRLF followed by whitespace is allowed in a quoted string.
		 * We allowed naked LF, but ensure LF is always followed by
		 * whitespace to prevent spoofing the end of the header.
		 */
		if (i + 1 < len && addr.charAt(i + 1) != ' ' &&
				    addr.charAt(i + 1) != '\t')
		    throw new AddressException(
		     "Quoted local address contains newline without whitespace",
			addr);
	    }
	    if (inquote)
		continue;
	    // dot rules should not be applied to quoted-string
	    if (c == '.') {
		if (i == start)
		    throw new AddressException(
			"Local address starts with dot", addr);
		if (lastc == '.')
		    throw new AddressException(
			"Local address contains dot-dot", addr);
	    }
	    if (c == '@') {
		if (i == 0)
		    throw new AddressException("Missing local name", addr);
		if (lastc == '.')
		    throw new AddressException(
			"Local address ends with dot", addr);
		break;		// done with local part
	    }
	    if (c <= 040 || c == 0177)
		throw new AddressException(
			"Local address contains control or whitespace", addr);
	    if (specialsNoDot.indexOf(c) >= 0)
		throw new AddressException(
			"Local address contains illegal character", addr);
	}
	if (inquote)
	    throw new AddressException("Unterminated quote", addr);

	/*
	 * Done with local part, now check domain.
	 *
	 * Note that the MimeMessage class doesn't remember addresses
	 * as separate objects; it writes them out as headers and then
	 * parses the headers when the addresses are requested.
	 * In order to support the case where a "simple" address is used,
	 * but the address also has a personal name and thus looks like
	 * it should be a valid RFC822 address when parsed, we only check
	 * this if we're explicitly called from the validate method.
	 */

	if (c != '@') {
	    if (validate)
		throw new AddressException("Missing final '@domain'", addr);
	    return;
	}

	// check for illegal chars in the domain, but ignore domain literals

	start = i + 1;
	if (start >= len)
	    throw new AddressException("Missing domain", addr);

	if (addr.charAt(start) == '.')
	    throw new AddressException("Domain starts with dot", addr);
	boolean inliteral = false;
	for (i = start; i < len; i++) {
	    c = addr.charAt(i);
	    if (c == '[') {
		if (i != start)
		    throw new AddressException(
				"Domain literal not at start of domain", addr);
		inliteral = true;	// domain literal, don't validate
	    } else if (c == ']') {
		if (i != len - 1)
		    throw new AddressException(
			    "Domain literal end not at end of domain", addr);
		inliteral = false;
	    } else if (c <= 040 || c == 0177) {
		throw new AddressException(
				"Domain contains control or whitespace", addr);
	    } else {
		// RFC 2822 rule
		//if (specialsNoDot.indexOf(c) >= 0)
		/*
		 * RFC 1034 rule is more strict
		 * the full rule is:
		 * 
		 * <domain> ::= <subdomain> | " "
		 * <subdomain> ::= <label> | <subdomain> "." <label>
		 * <label> ::= <letter> [ [ <ldh-str> ] <let-dig> ]
		 * <ldh-str> ::= <let-dig-hyp> | <let-dig-hyp> <ldh-str>
		 * <let-dig-hyp> ::= <let-dig> | "-"
		 * <let-dig> ::= <letter> | <digit>
		 */
		if (!inliteral) {
		    if (!(Character.isLetterOrDigit(c) || c == '-' || c == '.'))
			throw new AddressException(
				    "Domain contains illegal character", addr);
		    if (c == '.' && lastc == '.')
			throw new AddressException(
					"Domain contains dot-dot", addr);
		}
	    }
	    lastc = c;
	}
	if (lastc == '.')
	    throw new AddressException("Domain ends with dot", addr);
    }

    /**
     * Is this a "simple" address?  Simple addresses don't contain quotes
     * or any RFC822 special characters other than '@' and '.'.
     */
    private boolean isSimple() {
	return address == null || indexOfAny(address, specialsNoDotNoAt) < 0;
    }

    /**
     * Indicates whether this address is an RFC 822 group address.
     * Note that a group address is different than the mailing
     * list addresses supported by most mail servers.  Group addresses
     * are rarely used; see RFC 822 for details.
     *
     * @return		true if this address represents a group
     * @since		JavaMail 1.3
     */
    public boolean isGroup() {
	// quick and dirty check
	return address != null &&
	    address.endsWith(";") && address.indexOf(':') > 0;
    }

    /**
     * Return the members of a group address.  A group may have zero,
     * one, or more members.  If this address is not a group, null
     * is returned.  The <code>strict</code> parameter controls whether
     * the group list is parsed using strict RFC 822 rules or not.
     * The parsing is done using the <code>parseHeader</code> method.
     *
     * @param	strict	use strict RFC 822 rules?
     * @return		array of InternetAddress objects, or null
     * @exception	AddressException if the group list can't be parsed
     * @since		JavaMail 1.3
     */
    public InternetAddress[] getGroup(boolean strict) throws AddressException {
	String addr = getAddress();
	if (addr == null)
	    return null;
	// groups are of the form "name:addr,addr,...;"
	if (!addr.endsWith(";"))
	    return null;
	int ix = addr.indexOf(':');
	if (ix < 0)
	    return null;
	// extract the list
	String list = addr.substring(ix + 1, addr.length() - 1);
	// parse it and return the individual addresses
	return InternetAddress.parseHeader(list, strict);
    }

    /**
     * Return the first index of any of the characters in "any" in "s",
     * or -1 if none are found.
     *
     * This should be a method on String.
     */
    private static int indexOfAny(String s, String any) {
	return indexOfAny(s, any, 0);
    }

    private static int indexOfAny(String s, String any, int start) {
	try {
	    int len = s.length();
	    for (int i = start; i < len; i++) {
		if (any.indexOf(s.charAt(i)) >= 0)
		    return i;
	    }
	    return -1;
	} catch (StringIndexOutOfBoundsException e) {
	    return -1;
	}
    }

    /*
    public static void main(String argv[]) throws Exception {
	for (int i = 0; i < argv.length; i++) {
	    InternetAddress[] a = InternetAddress.parse(argv[i]);
	    for (int j = 0; j < a.length; j++) {
		System.out.println("arg " + i + " address " + j + ": " + a[j]);
		System.out.println("\tAddress: " + a[j].getAddress() +
				    "\tPersonal: " + a[j].getPersonal());
	    }
	    if (a.length > 1) {
		System.out.println("address 0 hash code: " + a[0].hashCode());
		System.out.println("address 1 hash code: " + a[1].hashCode());
		if (a[0].hashCode() == a[1].hashCode())
		    System.out.println("success, hashcodes equal");
		else
		    System.out.println("fail, hashcodes not equal");
		if (a[0].equals(a[1]))
		    System.out.println("success, addresses equal");
		else
		    System.out.println("fail, addresses not equal");
		if (a[1].equals(a[0]))
		    System.out.println("success, addresses equal");
		else
		    System.out.println("fail, addresses not equal");
	    }
	}
    }
    */
}