1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/src/share/classes/com/sun/tools/javac/parser/UnicodeReader.java Mon Oct 24 13:00:20 2011 +0100
1.3 @@ -0,0 +1,227 @@
1.4 +/*
1.5 + * Copyright (c) 2011, 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 +package com.sun.tools.javac.parser;
1.30 +
1.31 +import com.sun.tools.javac.file.JavacFileManager;
1.32 +import java.nio.CharBuffer;
1.33 +import com.sun.tools.javac.util.Log;
1.34 +import static com.sun.tools.javac.util.LayoutCharacters.*;
1.35 +
1.36 +/** The char reader used by the javac lexer/tokenizer. Returns the sequence of
1.37 + * characters contained in the input stream, handling unicode escape accordingly.
1.38 + * Additionally, it provide features for saving chars into a buffer and to retrieve
1.39 + * them at a later stage.
1.40 + *
1.41 + * <p><b>This is NOT part of any supported API.
1.42 + * If you write code that depends on this, you do so at your own risk.
1.43 + * This code and its internal interfaces are subject to change or
1.44 + * deletion without notice.</b>
1.45 + */
1.46 +public class UnicodeReader {
1.47 +
1.48 + /** The input buffer, index of next character to be read,
1.49 + * index of one past last character in buffer.
1.50 + */
1.51 + protected char[] buf;
1.52 + protected int bp;
1.53 + protected final int buflen;
1.54 +
1.55 + /** The current character.
1.56 + */
1.57 + protected char ch;
1.58 +
1.59 + /** The buffer index of the last converted unicode character
1.60 + */
1.61 + protected int unicodeConversionBp = -1;
1.62 +
1.63 + protected Log log;
1.64 +
1.65 + /**
1.66 + * Create a scanner from the input array. This method might
1.67 + * modify the array. To avoid copying the input array, ensure
1.68 + * that {@code inputLength < input.length} or
1.69 + * {@code input[input.length -1]} is a white space character.
1.70 + *
1.71 + * @param fac the factory which created this Scanner
1.72 + * @param input the input, might be modified
1.73 + * @param inputLength the size of the input.
1.74 + * Must be positive and less than or equal to input.length.
1.75 + */
1.76 + protected UnicodeReader(ScannerFactory sf, CharBuffer buffer) {
1.77 + this(sf, JavacFileManager.toArray(buffer), buffer.limit());
1.78 + }
1.79 +
1.80 + protected UnicodeReader(ScannerFactory sf, char[] input, int inputLength) {
1.81 + log = sf.log;
1.82 + if (inputLength == input.length) {
1.83 + if (input.length > 0 && Character.isWhitespace(input[input.length - 1])) {
1.84 + inputLength--;
1.85 + } else {
1.86 + char[] newInput = new char[inputLength + 1];
1.87 + System.arraycopy(input, 0, newInput, 0, input.length);
1.88 + input = newInput;
1.89 + }
1.90 + }
1.91 + buf = input;
1.92 + buflen = inputLength;
1.93 + buf[buflen] = EOI;
1.94 + bp = -1;
1.95 + scanChar();
1.96 + }
1.97 +
1.98 + /** Read next character.
1.99 + */
1.100 + protected void scanChar() {
1.101 + if (bp < buflen) {
1.102 + ch = buf[++bp];
1.103 + if (ch == '\\') {
1.104 + convertUnicode();
1.105 + }
1.106 + }
1.107 + }
1.108 +
1.109 + /** Convert unicode escape; bp points to initial '\' character
1.110 + * (Spec 3.3).
1.111 + */
1.112 + protected void convertUnicode() {
1.113 + if (ch == '\\' && unicodeConversionBp != bp) {
1.114 + bp++; ch = buf[bp];
1.115 + if (ch == 'u') {
1.116 + do {
1.117 + bp++; ch = buf[bp];
1.118 + } while (ch == 'u');
1.119 + int limit = bp + 3;
1.120 + if (limit < buflen) {
1.121 + int d = digit(bp, 16);
1.122 + int code = d;
1.123 + while (bp < limit && d >= 0) {
1.124 + bp++; ch = buf[bp];
1.125 + d = digit(bp, 16);
1.126 + code = (code << 4) + d;
1.127 + }
1.128 + if (d >= 0) {
1.129 + ch = (char)code;
1.130 + unicodeConversionBp = bp;
1.131 + return;
1.132 + }
1.133 + }
1.134 + log.error(bp, "illegal.unicode.esc");
1.135 + } else {
1.136 + bp--;
1.137 + ch = '\\';
1.138 + }
1.139 + }
1.140 + }
1.141 +
1.142 + /** Are surrogates supported?
1.143 + */
1.144 + final static boolean surrogatesSupported = surrogatesSupported();
1.145 + private static boolean surrogatesSupported() {
1.146 + try {
1.147 + Character.isHighSurrogate('a');
1.148 + return true;
1.149 + } catch (NoSuchMethodError ex) {
1.150 + return false;
1.151 + }
1.152 + }
1.153 +
1.154 + /** Scan surrogate pairs. If 'ch' is a high surrogate and
1.155 + * the next character is a low surrogate, then put the low
1.156 + * surrogate in 'ch', and return the high surrogate.
1.157 + * otherwise, just return 0.
1.158 + */
1.159 + protected char scanSurrogates() {
1.160 + if (surrogatesSupported && Character.isHighSurrogate(ch)) {
1.161 + char high = ch;
1.162 +
1.163 + scanChar();
1.164 +
1.165 + if (Character.isLowSurrogate(ch)) {
1.166 + return high;
1.167 + }
1.168 +
1.169 + ch = high;
1.170 + }
1.171 +
1.172 + return 0;
1.173 + }
1.174 +
1.175 + /** Convert an ASCII digit from its base (8, 10, or 16)
1.176 + * to its value.
1.177 + */
1.178 + protected int digit(int pos, int base) {
1.179 + char c = ch;
1.180 + int result = Character.digit(c, base);
1.181 + if (result >= 0 && c > 0x7f) {
1.182 + log.error(pos + 1, "illegal.nonascii.digit");
1.183 + ch = "0123456789abcdef".charAt(result);
1.184 + }
1.185 + return result;
1.186 + }
1.187 +
1.188 + protected boolean isUnicode() {
1.189 + return unicodeConversionBp == bp;
1.190 + }
1.191 +
1.192 + protected void skipChar() {
1.193 + bp++;
1.194 + }
1.195 +
1.196 + protected char peekChar() {
1.197 + return buf[bp + 1];
1.198 + }
1.199 +
1.200 + /**
1.201 + * Returns a copy of the input buffer, up to its inputLength.
1.202 + * Unicode escape sequences are not translated.
1.203 + */
1.204 + public char[] getRawCharacters() {
1.205 + char[] chars = new char[buflen];
1.206 + System.arraycopy(buf, 0, chars, 0, buflen);
1.207 + return chars;
1.208 + }
1.209 +
1.210 + /**
1.211 + * Returns a copy of a character array subset of the input buffer.
1.212 + * The returned array begins at the <code>beginIndex</code> and
1.213 + * extends to the character at index <code>endIndex - 1</code>.
1.214 + * Thus the length of the substring is <code>endIndex-beginIndex</code>.
1.215 + * This behavior is like
1.216 + * <code>String.substring(beginIndex, endIndex)</code>.
1.217 + * Unicode escape sequences are not translated.
1.218 + *
1.219 + * @param beginIndex the beginning index, inclusive.
1.220 + * @param endIndex the ending index, exclusive.
1.221 + * @throws IndexOutOfBounds if either offset is outside of the
1.222 + * array bounds
1.223 + */
1.224 + public char[] getRawCharacters(int beginIndex, int endIndex) {
1.225 + int length = endIndex - beginIndex;
1.226 + char[] chars = new char[length];
1.227 + System.arraycopy(buf, beginIndex, chars, 0, length);
1.228 + return chars;
1.229 + }
1.230 +}
