1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/src/jdk/internal/dynalink/support/NameCodec.java Thu Feb 14 13:22:26 2013 +0100
1.3 @@ -0,0 +1,218 @@
1.4 +/*
1.5 + * Copyright (c) 2010, 2013, Oracle and/or its affiliates. All rights reserved.
1.6 + * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
1.7 + *
1.8 + * This code is free software; you can redistribute it and/or modify it
1.9 + * under the terms of the GNU General Public License version 2 only, as
1.10 + * published by the Free Software Foundation. Oracle designates this
1.11 + * particular file as subject to the "Classpath" exception as provided
1.12 + * by Oracle in the LICENSE file that accompanied this code.
1.13 + *
1.14 + * This code is distributed in the hope that it will be useful, but WITHOUT
1.15 + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
1.16 + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
1.17 + * version 2 for more details (a copy is included in the LICENSE file that
1.18 + * accompanied this code).
1.19 + *
1.20 + * You should have received a copy of the GNU General Public License version
1.21 + * 2 along with this work; if not, write to the Free Software Foundation,
1.22 + * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
1.23 + *
1.24 + * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
1.25 + * or visit www.oracle.com if you need additional information or have any
1.26 + * questions.
1.27 + */
1.28 +
1.29 +/*
1.30 + * This file is available under and governed by the GNU General Public
1.31 + * License version 2 only, as published by the Free Software Foundation.
1.32 + * However, the following notice accompanied the original version of this
1.33 + * file, and Oracle licenses the original version of this file under the BSD
1.34 + * license:
1.35 + */
1.36 +/*
1.37 + Copyright 2009-2013 Attila Szegedi
1.38 +
1.39 + Licensed under both the Apache License, Version 2.0 (the "Apache License")
1.40 + and the BSD License (the "BSD License"), with licensee being free to
1.41 + choose either of the two at their discretion.
1.42 +
1.43 + You may not use this file except in compliance with either the Apache
1.44 + License or the BSD License.
1.45 +
1.46 + If you choose to use this file in compliance with the Apache License, the
1.47 + following notice applies to you:
1.48 +
1.49 + You may obtain a copy of the Apache License at
1.50 +
1.51 + http://www.apache.org/licenses/LICENSE-2.0
1.52 +
1.53 + Unless required by applicable law or agreed to in writing, software
1.54 + distributed under the License is distributed on an "AS IS" BASIS,
1.55 + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
1.56 + implied. See the License for the specific language governing
1.57 + permissions and limitations under the License.
1.58 +
1.59 + If you choose to use this file in compliance with the BSD License, the
1.60 + following notice applies to you:
1.61 +
1.62 + Redistribution and use in source and binary forms, with or without
1.63 + modification, are permitted provided that the following conditions are
1.64 + met:
1.65 + * Redistributions of source code must retain the above copyright
1.66 + notice, this list of conditions and the following disclaimer.
1.67 + * Redistributions in binary form must reproduce the above copyright
1.68 + notice, this list of conditions and the following disclaimer in the
1.69 + documentation and/or other materials provided with the distribution.
1.70 + * Neither the name of the copyright holder nor the names of
1.71 + contributors may be used to endorse or promote products derived from
1.72 + this software without specific prior written permission.
1.73 +
1.74 + THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS
1.75 + IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
1.76 + TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
1.77 + PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL COPYRIGHT HOLDER
1.78 + BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
1.79 + CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
1.80 + SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
1.81 + BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
1.82 + WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
1.83 + OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
1.84 + ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
1.85 +*/
1.86 +
1.87 +package jdk.internal.dynalink.support;
1.88 +
1.89 +import jdk.internal.dynalink.CallSiteDescriptor;
1.90 +
1.91 +/**
1.92 + * Implements the name mangling and demangling as specified by John Rose's
1.93 + * <a href="https://blogs.oracle.com/jrose/entry/symbolic_freedom_in_the_vm" target="_blank">"Symbolic Freedom in the
1.94 + * VM"</a> article. It is recommended that implementers of languages on the JVM uniformly adopt this for symbolic
1.95 + * interoperability between languages. Normally, you would mangle the names as you're generating bytecode, and then
1.96 + * demangle them when you're creating {@link CallSiteDescriptor} objects. Note that you are expected to mangle
1.97 + * individual tokens, and not the whole name at the call site, i.e. the colon character normally separating the tokens
1.98 + * is never mangled. I.e. you wouldn't mangle {@code dyn:getProp:color} into {@code dyn\!getProp\!color}, but you would
1.99 + * mangle {@code dyn:getProp:color$} into {@code dyn:getProp:\=color\%} (only mangling the individual token containing
1.100 + * the symbol {@code color$}). {@link CallSiteDescriptorFactory#tokenizeName(String)} (and by implication, all call site
1.101 + * descriptors it creates) will automatically perform demangling on the passed names. If you use this factory, or you
1.102 + * have your own way of creating call site descriptors, but you still delegate to this method of the default factory
1.103 + * (it is recommended that you do), then you have demangling handled for you already, and only need to ensure that you
1.104 + * mangle the names when you're emitting them in the bytecode.
1.105 + *
1.106 + * @author Attila Szegedi
1.107 + */
1.108 +public class NameCodec {
1.109 + private static final char ESCAPE_CHAR = '\\';
1.110 + private static final char EMPTY_ESCAPE = '=';
1.111 + private static final String EMPTY_NAME = new String(new char[] { ESCAPE_CHAR, EMPTY_ESCAPE });
1.112 + private static final char EMPTY_CHAR = 0xFEFF;
1.113 +
1.114 + private static final int MIN_ENCODING = '$';
1.115 + private static final int MAX_ENCODING = ']';
1.116 + private static final char[] ENCODING = new char[MAX_ENCODING - MIN_ENCODING + 1];
1.117 + private static final int MIN_DECODING = '!';
1.118 + private static final int MAX_DECODING = '}';
1.119 + private static final char[] DECODING = new char[MAX_DECODING - MIN_DECODING + 1];
1.120 +
1.121 + static {
1.122 + addEncoding('/', '|');
1.123 + addEncoding('.', ',');
1.124 + addEncoding(';', '?');
1.125 + addEncoding('$', '%');
1.126 + addEncoding('<', '^');
1.127 + addEncoding('>', '_');
1.128 + addEncoding('[', '{');
1.129 + addEncoding(']', '}');
1.130 + addEncoding(':', '!');
1.131 + addEncoding('\\', '-');
1.132 + DECODING[EMPTY_ESCAPE - MIN_DECODING] = EMPTY_CHAR;
1.133 + }
1.134 +
1.135 + private NameCodec() {
1.136 + }
1.137 +
1.138 + /**
1.139 + * Encodes ("mangles") an unencoded symbolic name.
1.140 + * @param name the symbolic name to mangle
1.141 + * @return the mangled form of the symbolic name.
1.142 + */
1.143 + public static String encode(String name) {
1.144 + final int l = name.length();
1.145 + if(l == 0) {
1.146 + return EMPTY_NAME;
1.147 + }
1.148 + StringBuilder b = null;
1.149 + int lastEscape = -1;
1.150 + for(int i = 0; i < l; ++i) {
1.151 + final int encodeIndex = name.charAt(i) - MIN_ENCODING;
1.152 + if(encodeIndex >= 0 && encodeIndex < ENCODING.length) {
1.153 + final char e = ENCODING[encodeIndex];
1.154 + if(e != 0) {
1.155 + if(b == null) {
1.156 + b = new StringBuilder(name.length() + 3);
1.157 + if(name.charAt(0) != ESCAPE_CHAR && i > 0) {
1.158 + b.append(EMPTY_NAME);
1.159 + }
1.160 + b.append(name, 0, i);
1.161 + } else {
1.162 + b.append(name, lastEscape + 1, i);
1.163 + }
1.164 + b.append(ESCAPE_CHAR).append(e);
1.165 + lastEscape = i;
1.166 + }
1.167 + }
1.168 + }
1.169 + if(b == null) {
1.170 + return name.toString();
1.171 + }
1.172 + assert lastEscape != -1;
1.173 + b.append(name, lastEscape + 1, l);
1.174 + return b.toString();
1.175 + }
1.176 +
1.177 + /**
1.178 + * Decodes ("demangles") an encoded symbolic name.
1.179 + * @param name the symbolic name to demangle
1.180 + * @return the demangled form of the symbolic name.
1.181 + */
1.182 + public static String decode(String name) {
1.183 + if(name.charAt(0) != ESCAPE_CHAR) {
1.184 + return name;
1.185 + }
1.186 + final int l = name.length();
1.187 + if(l == 2 && name.charAt(1) == EMPTY_CHAR) {
1.188 + return "";
1.189 + }
1.190 + StringBuilder b = new StringBuilder(name.length());
1.191 + int lastEscape = -2;
1.192 + int lastBackslash = -1;
1.193 + for(;;) {
1.194 + int nextBackslash = name.indexOf(ESCAPE_CHAR, lastBackslash + 1);
1.195 + if(nextBackslash == -1 || nextBackslash == l - 1) {
1.196 + break;
1.197 + }
1.198 + final int decodeIndex = name.charAt(nextBackslash + 1) - MIN_DECODING;
1.199 + if(decodeIndex >= 0 && decodeIndex < DECODING.length) {
1.200 + final char d = DECODING[decodeIndex];
1.201 + if(d == EMPTY_CHAR) {
1.202 + // "\=" is only valid at the beginning of a mangled string
1.203 + if(nextBackslash == 0) {
1.204 + lastEscape = 0;
1.205 + }
1.206 + } else if(d != 0) {
1.207 + b.append(name, lastEscape + 2, nextBackslash).append(d);
1.208 + lastEscape = nextBackslash;
1.209 + }
1.210 + }
1.211 + lastBackslash = nextBackslash;
1.212 + }
1.213 + b.append(name, lastEscape + 2, l);
1.214 + return b.toString();
1.215 + }
1.216 +
1.217 + private static void addEncoding(char from, char to) {
1.218 + ENCODING[from - MIN_ENCODING] = to;
1.219 + DECODING[to - MIN_DECODING] = from;
1.220 + }
1.221 +}
