Back to index...

	/*
	* Copyright (c) 1996, 2017, Oracle and/or its affiliates. All rights reserved.
	* DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
	*
	* This code is free software; you can redistribute it and/or modify it
	* under the terms of the GNU General Public License version 2 only, as
	* published by the Free Software Foundation. Oracle designates this
	* particular file as subject to the "Classpath" exception as provided
	* by Oracle in the LICENSE file that accompanied this code.
	*
	* This code 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. See the GNU General Public License
	* version 2 for more details (a copy is included in the LICENSE file that
	* accompanied this code).
	*
	* You should have received a copy of the GNU General Public License version
	* 2 along with this work; if not, write to the Free Software Foundation,
	* Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
	*
	* Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
	* or visit www.oracle.com if you need additional information or have any
	* questions.
	*/

	package sun.security.util;

	import java.io.*;
	import java.math.BigInteger;
	import java.util.Arrays;

	/**
	* Represent an ISO Object Identifier.
	*
	* <P>Object Identifiers are arbitrary length hierarchical identifiers.
	* The individual components are numbers, and they define paths from the
	* root of an ISO-managed identifier space. You will sometimes see a
	* string name used instead of (or in addition to) the numerical id.
	* These are synonyms for the numerical IDs, but are not widely used
	* since most sites do not know all the requisite strings, while all
	* sites can parse the numeric forms.
	*
	* <P>So for example, JavaSoft has the sole authority to assign the
	* meaning to identifiers below the 1.3.6.1.4.1.42.2.17 node in the
	* hierarchy, and other organizations can easily acquire the ability
	* to assign such unique identifiers.
	*
	* @author David Brownell
	* @author Amit Kapoor
	* @author Hemma Prafullchandra
	*/

	public final
	class ObjectIdentifier implements Serializable
	{
	/**
	* We use the DER value (no tag, no length) as the internal format
	* @serial
	*/
	private byte[] encoding = null;

	private transient volatile String stringForm;

	/*
	* IMPORTANT NOTES FOR CODE CHANGES (bug 4811968) IN JDK 1.7.0
	* ===========================================================
	*
	* (Almost) serialization compatibility with old versions:
	*
	* serialVersionUID is unchanged. Old field "component" is changed to
	* type Object so that "poison" (unknown object type for old versions)
	* can be put inside if there are huge components that cannot be saved
	* as integers.
	*
	* New version use the new filed "encoding" only.
	*
	* Below are all 4 cases in a serialization/deserialization process:
	*
	* 1. old -> old: Not covered here
	* 2. old -> new: There's no "encoding" field, new readObject() reads
	* "components" and "componentLen" instead and inits correctly.
	* 3. new -> new: "encoding" field exists, new readObject() uses it
	* (ignoring the other 2 fields) and inits correctly.
	* 4. new -> old: old readObject() only recognizes "components" and
	* "componentLen" fields. If no huge components are involved, they
	* are serialized as legal values and old object can init correctly.
	* Otherwise, old object cannot recognize the form (component not int[])
	* and throw a ClassNotFoundException at deserialization time.
	*
	* Therfore, for the first 3 cases, exact compatibility is preserved. In
	* the 4th case, non-huge OID is still supportable in old versions, while
	* huge OID is not.
	*/
	private static final long serialVersionUID = 8697030238860181294L;

	/**
	* Changed to Object
	* @serial
	*/
	private Object components = null; // path from root
	/**
	* @serial
	*/
	private int componentLen = -1; // how much is used.

	// Is the components field calculated?
	private transient boolean componentsCalculated = false;

	private void readObject(ObjectInputStream is)
	throws IOException, ClassNotFoundException {
	is.defaultReadObject();

	if (encoding == null) { // from an old version
	int[] comp = (int[])components;
	if (componentLen > comp.length) {
	componentLen = comp.length;
	}
	init(comp, componentLen);
	}
	}

	private void writeObject(ObjectOutputStream os)
	throws IOException {
	if (!componentsCalculated) {
	int[] comps = toIntArray();
	if (comps != null) { // every one understands this
	components = comps;
	componentLen = comps.length;
	} else {
	components = HugeOidNotSupportedByOldJDK.theOne;
	}
	componentsCalculated = true;
	}
	os.defaultWriteObject();
	}

	static class HugeOidNotSupportedByOldJDK implements Serializable {
	private static final long serialVersionUID = 1L;
	static HugeOidNotSupportedByOldJDK theOne = new HugeOidNotSupportedByOldJDK();
	}

	/**
	* Constructs, from a string. This string should be of the form 1.23.56.
	* Validity check included.
	*/
	public ObjectIdentifier (String oid) throws IOException
	{
	int ch = '.';
	int start = 0;
	int end = 0;

	int pos = 0;
	byte[] tmp = new byte[oid.length()];
	int first = 0, second;
	int count = 0;

	try {
	String comp = null;
	do {
	int length = 0; // length of one section
	end = oid.indexOf(ch,start);
	if (end == -1) {
	comp = oid.substring(start);
	length = oid.length() - start;
	} else {
	comp = oid.substring(start,end);
	length = end - start;
	}

	if (length > 9) {
	BigInteger bignum = new BigInteger(comp);
	if (count == 0) {
	checkFirstComponent(bignum);
	first = bignum.intValue();
	} else {
	if (count == 1) {
	checkSecondComponent(first, bignum);
	bignum = bignum.add(BigInteger.valueOf(40*first));
	} else {
	checkOtherComponent(count, bignum);
	}
	pos += pack7Oid(bignum, tmp, pos);
	}
	} else {
	int num = Integer.parseInt(comp);
	if (count == 0) {
	checkFirstComponent(num);
	first = num;
	} else {
	if (count == 1) {
	checkSecondComponent(first, num);
	num += 40 * first;
	} else {
	checkOtherComponent(count, num);
	}
	pos += pack7Oid(num, tmp, pos);
	}
	}
	start = end + 1;
	count++;
	} while (end != -1);

	checkCount(count);
	encoding = new byte[pos];
	System.arraycopy(tmp, 0, encoding, 0, pos);
	this.stringForm = oid;
	} catch (IOException ioe) { // already detected by checkXXX
	throw ioe;
	} catch (Exception e) {
	throw new IOException("ObjectIdentifier() -- Invalid format: "
	+ e.toString(), e);
	}
	}

	/**
	* Constructor, from an array of integers.
	* Validity check included.
	*/
	public ObjectIdentifier(int[] values) throws IOException
	{
	checkCount(values.length);
	checkFirstComponent(values[0]);
	checkSecondComponent(values[0], values[1]);
	for (int i=2; i<values.length; i++)
	checkOtherComponent(i, values[i]);
	init(values, values.length);
	}

	/**
	* Constructor, from an ASN.1 encoded input stream.
	* Validity check NOT included.
	* The encoding of the ID in the stream uses "DER", a BER/1 subset.
	* In this case, that means a triple { typeId, length, data }.
	*
	* <P><STRONG>NOTE:</STRONG> When an exception is thrown, the
	* input stream has not been returned to its "initial" state.
	*
	* @param in DER-encoded data holding an object ID
	* @exception IOException indicates a decoding error
	*/
	public ObjectIdentifier (DerInputStream in) throws IOException
	{
	byte type_id;
	int bufferEnd;

	/*
	* Object IDs are a "universal" type, and their tag needs only
	* one byte of encoding. Verify that the tag of this datum
	* is that of an object ID.
	*
	* Then get and check the length of the ID's encoding. We set
	* up so that we can use in.available() to check for the end of
	* this value in the data stream.
	*/
	type_id = (byte) in.getByte ();
	if (type_id != DerValue.tag_ObjectId)
	throw new IOException (
	"ObjectIdentifier() -- data isn't an object ID"
	+ " (tag = " + type_id + ")"
	);

	int len = in.getDefiniteLength();
	if (len > in.available()) {
	throw new IOException("ObjectIdentifier() -- length exceeds" +
	"data available. Length: " + len + ", Available: " +
	in.available());
	}
	encoding = new byte[len];
	in.getBytes(encoding);
	check(encoding);
	}

	/*
	* Constructor, from the rest of a DER input buffer;
	* the tag and length have been removed/verified
	* Validity check NOT included.
	*/
	ObjectIdentifier (DerInputBuffer buf) throws IOException
	{
	DerInputStream in = new DerInputStream(buf);
	encoding = new byte[in.available()];
	in.getBytes(encoding);
	check(encoding);
	}

	private void init(int[] components, int length) {
	int pos = 0;
	byte[] tmp = new byte[length*5+1]; // +1 for empty input

	if (components[1] < Integer.MAX_VALUE - components[0]*40)
	pos += pack7Oid(components[0]*40+components[1], tmp, pos);
	else {
	BigInteger big = BigInteger.valueOf(components[1]);
	big = big.add(BigInteger.valueOf(components[0]*40));
	pos += pack7Oid(big, tmp, pos);
	}

	for (int i=2; i<length; i++) {
	pos += pack7Oid(components[i], tmp, pos);
	}
	encoding = new byte[pos];
	System.arraycopy(tmp, 0, encoding, 0, pos);
	}

	/**
	* This method is kept for compatibility reasons. The new implementation
	* does the check and conversion. All around the JDK, the method is called
	* in static blocks to initialize pre-defined ObjectIdentifieies. No
	* obvious performance hurt will be made after this change.
	*
	* Old doc: Create a new ObjectIdentifier for internal use. The values are
	* neither checked nor cloned.
	*/
	public static ObjectIdentifier newInternal(int[] values) {
	try {
	return new ObjectIdentifier(values);
	} catch (IOException ex) {
	throw new RuntimeException(ex);
	// Should not happen, internal calls always uses legal values.
	}
	}

	/*
	* n.b. the only public interface is DerOutputStream.putOID()
	*/
	void encode (DerOutputStream out) throws IOException
	{
	out.write (DerValue.tag_ObjectId, encoding);
	}

	/**
	* Compares this identifier with another, for equality.
	*
	* @return true iff the names are identical.
	*/
	@Override
	public boolean equals(Object obj) {
	if (this == obj) {
	return true;
	}
	if (obj instanceof ObjectIdentifier == false) {
	return false;
	}
	ObjectIdentifier other = (ObjectIdentifier)obj;
	return Arrays.equals(encoding, other.encoding);
	}

	@Override
	public int hashCode() {
	return Arrays.hashCode(encoding);
	}

	/**
	* Private helper method for serialization. To be compatible with old
	* versions of JDK.
	* @return components in an int array, if all the components are less than
	* Integer.MAX_VALUE. Otherwise, null.
	*/
	private int[] toIntArray() {
	int length = encoding.length;
	int[] result = new int[20];
	int which = 0;
	int fromPos = 0;
	for (int i = 0; i < length; i++) {
	if ((encoding[i] & 0x80) == 0) {
	// one section [fromPos..i]
	if (i - fromPos + 1 > 4) {
	BigInteger big = new BigInteger(pack(encoding, fromPos, i-fromPos+1, 7, 8));
	if (fromPos == 0) {
	result[which++] = 2;
	BigInteger second = big.subtract(BigInteger.valueOf(80));
	if (second.compareTo(BigInteger.valueOf(Integer.MAX_VALUE)) == 1) {
	return null;
	} else {
	result[which++] = second.intValue();
	}
	} else {
	if (big.compareTo(BigInteger.valueOf(Integer.MAX_VALUE)) == 1) {
	return null;
	} else {
	result[which++] = big.intValue();
	}
	}
	} else {
	int retval = 0;
	for (int j = fromPos; j <= i; j++) {
	retval <<= 7;
	byte tmp = encoding[j];
	retval \|= (tmp & 0x07f);
	}
	if (fromPos == 0) {
	if (retval < 80) {
	result[which++] = retval / 40;
	result[which++] = retval % 40;
	} else {
	result[which++] = 2;
	result[which++] = retval - 80;
	}
	} else {
	result[which++] = retval;
	}
	}
	fromPos = i+1;
	}
	if (which >= result.length) {
	result = Arrays.copyOf(result, which + 10);
	}
	}
	return Arrays.copyOf(result, which);
	}

	/**
	* Returns a string form of the object ID. The format is the
	* conventional "dot" notation for such IDs, without any
	* user-friendly descriptive strings, since those strings
	* will not be understood everywhere.
	*/
	@Override
	public String toString() {
	String s = stringForm;
	if (s == null) {
	int length = encoding.length;
	StringBuilder sb = new StringBuilder(length * 4);

	int fromPos = 0;
	for (int i = 0; i < length; i++) {
	if ((encoding[i] & 0x80) == 0) {
	// one section [fromPos..i]
	if (fromPos != 0) { // not the first segment
	sb.append('.');
	}
	if (i - fromPos + 1 > 4) { // maybe big integer
	BigInteger big = new BigInteger(pack(encoding, fromPos, i-fromPos+1, 7, 8));
	if (fromPos == 0) {
	// first section encoded with more than 4 bytes,
	// must be 2.something
	sb.append("2.");
	sb.append(big.subtract(BigInteger.valueOf(80)));
	} else {
	sb.append(big);
	}
	} else { // small integer
	int retval = 0;
	for (int j = fromPos; j <= i; j++) {
	retval <<= 7;
	byte tmp = encoding[j];
	retval \|= (tmp & 0x07f);
	}
	if (fromPos == 0) {
	if (retval < 80) {
	sb.append(retval/40);
	sb.append('.');
	sb.append(retval%40);
	} else {
	sb.append("2.");
	sb.append(retval - 80);
	}
	} else {
	sb.append(retval);
	}
	}
	fromPos = i+1;
	}
	}
	s = sb.toString();
	stringForm = s;
	}
	return s;
	}

	/**
	* Repack all bits from input to output. On the both sides, only a portion
	* (from the least significant bit) of the 8 bits in a byte is used. This
	* number is defined as the number of useful bits (NUB) for the array. All the
	* used bits from the input byte array and repacked into the output in the
	* exactly same order. The output bits are aligned so that the final bit of
	* the input (the least significant bit in the last byte), when repacked as
	* the final bit of the output, is still at the least significant position.
	* Zeroes will be padded on the left side of the first output byte if
	* necessary. All unused bits in the output are also zeroed.
	*
	* For example: if the input is 01001100 with NUB 8, the output which
	* has a NUB 6 will look like:
	* 00000001 00001100
	* The first 2 bits of the output bytes are unused bits. The other bits
	* turn out to be 000001 001100. While the 8 bits on the right are from
	* the input, the left 4 zeroes are padded to fill the 6 bits space.
	*
	* @param in the input byte array
	* @param ioffset start point inside <code>in</code>
	* @param ilength number of bytes to repack
	* @param iw NUB for input
	* @param ow NUB for output
	* @return the repacked bytes
	*/
	private static byte[] pack(byte[] in, int ioffset, int ilength, int iw, int ow) {
	assert (iw > 0 && iw <= 8): "input NUB must be between 1 and 8";
	assert (ow > 0 && ow <= 8): "output NUB must be between 1 and 8";

	if (iw == ow) {
	return in.clone();
	}

	int bits = ilength * iw; // number of all used bits
	byte[] out = new byte[(bits+ow-1)/ow];

	// starting from the 0th bit in the input
	int ipos = 0;

	// the number of padding 0's needed in the output, skip them
	int opos = (bits+ow-1)/ow*ow-bits;

	while(ipos < bits) {
	int count = iw - ipos%iw; // unpacked bits in current input byte
	if (count > ow - opos%ow) { // free space available in output byte
	count = ow - opos%ow; // choose the smaller number
	}
	// and move them!
	out[opos/ow] \|= // paste!
	(((in[ioffset+ipos/iw]+256) // locate the byte (+256 so that it's never negative)
	>> (iw-ipos%iw-count)) // move to the end of a byte
	& ((1 << (count))-1)) // zero out all other bits
	<< (ow-opos%ow-count); // move to the output position
	ipos += count; // advance
	opos += count; // advance
	}
	return out;
	}

	/**
	* Repack from NUB 8 to a NUB 7 OID sub-identifier, remove all
	* unnecessary 0 headings, set the first bit of all non-tail
	* output bytes to 1 (as ITU-T Rec. X.690 8.19.2 says), and
	* paste it into an existing byte array.
	* @param out the existing array to be pasted into
	* @param ooffset the starting position to paste
	* @return the number of bytes pasted
	*/
	private static int pack7Oid(byte[] in, int ioffset, int ilength, byte[] out, int ooffset) {
	byte[] pack = pack(in, ioffset, ilength, 8, 7);
	int firstNonZero = pack.length-1; // paste at least one byte
	for (int i=pack.length-2; i>=0; i--) {
	if (pack[i] != 0) {
	firstNonZero = i;
	}
	pack[i] \|= 0x80;
	}
	System.arraycopy(pack, firstNonZero, out, ooffset, pack.length-firstNonZero);
	return pack.length-firstNonZero;
	}

	/**
	* Repack from NUB 7 to NUB 8, remove all unnecessary 0
	* headings, and paste it into an existing byte array.
	* @param out the existing array to be pasted into
	* @param ooffset the starting position to paste
	* @return the number of bytes pasted
	*/
	private static int pack8(byte[] in, int ioffset, int ilength, byte[] out, int ooffset) {
	byte[] pack = pack(in, ioffset, ilength, 7, 8);
	int firstNonZero = pack.length-1; // paste at least one byte
	for (int i=pack.length-2; i>=0; i--) {
	if (pack[i] != 0) {
	firstNonZero = i;
	}
	}
	System.arraycopy(pack, firstNonZero, out, ooffset, pack.length-firstNonZero);
	return pack.length-firstNonZero;
	}

	/**
	* Pack the int into a OID sub-identifier DER encoding
	*/
	private static int pack7Oid(int input, byte[] out, int ooffset) {
	byte[] b = new byte[4];
	b[0] = (byte)(input >> 24);
	b[1] = (byte)(input >> 16);
	b[2] = (byte)(input >> 8);
	b[3] = (byte)(input);
	return pack7Oid(b, 0, 4, out, ooffset);
	}

	/**
	* Pack the BigInteger into a OID subidentifier DER encoding
	*/
	private static int pack7Oid(BigInteger input, byte[] out, int ooffset) {
	byte[] b = input.toByteArray();
	return pack7Oid(b, 0, b.length, out, ooffset);
	}

	/**
	* Private methods to check validity of OID. They must be --
	* 1. at least 2 components
	* 2. all components must be non-negative
	* 3. the first must be 0, 1 or 2
	* 4. if the first is 0 or 1, the second must be <40
	*/

	/**
	* Check the DER encoding. Since DER encoding defines that the integer bits
	* are unsigned, so there's no need to check the MSB.
	*/
	private static void check(byte[] encoding) throws IOException {
	int length = encoding.length;
	if (length < 1 \|\| // too short
	(encoding[length - 1] & 0x80) != 0) { // not ended
	throw new IOException("ObjectIdentifier() -- " +
	"Invalid DER encoding, not ended");
	}
	for (int i=0; i<length; i++) {
	// 0x80 at the beginning of a subidentifier
	if (encoding[i] == (byte)0x80 &&
	(i==0 \|\| (encoding[i-1] & 0x80) == 0)) {
	throw new IOException("ObjectIdentifier() -- " +
	"Invalid DER encoding, useless extra octet detected");
	}
	}
	}
	private static void checkCount(int count) throws IOException {
	if (count < 2) {
	throw new IOException("ObjectIdentifier() -- " +
	"Must be at least two oid components ");
	}
	}
	private static void checkFirstComponent(int first) throws IOException {
	if (first < 0 \|\| first > 2) {
	throw new IOException("ObjectIdentifier() -- " +
	"First oid component is invalid ");
	}
	}
	private static void checkFirstComponent(BigInteger first) throws IOException {
	if (first.signum() == -1 \|\| first.compareTo(BigInteger.TWO) > 0) {
	throw new IOException("ObjectIdentifier() -- " +
	"First oid component is invalid ");
	}
	}
	private static void checkSecondComponent(int first, int second) throws IOException {
	if (second < 0 \|\| first != 2 && second > 39) {
	throw new IOException("ObjectIdentifier() -- " +
	"Second oid component is invalid ");
	}
	}
	private static void checkSecondComponent(int first, BigInteger second) throws IOException {
	if (second.signum() == -1 \|\|
	first != 2 &&
	second.compareTo(BigInteger.valueOf(39)) == 1) {
	throw new IOException("ObjectIdentifier() -- " +
	"Second oid component is invalid ");
	}
	}
	private static void checkOtherComponent(int i, int num) throws IOException {
	if (num < 0) {
	throw new IOException("ObjectIdentifier() -- " +
	"oid component #" + (i+1) + " must be non-negative ");
	}
	}
	private static void checkOtherComponent(int i, BigInteger num) throws IOException {
	if (num.signum() == -1) {
	throw new IOException("ObjectIdentifier() -- " +
	"oid component #" + (i+1) + " must be non-negative ");
	}
	}
	}

Back to index...