| /* | |
| * Copyright (C) 2011 The Android Open Source Project | |
| * | |
| * Licensed under the Apache License, Version 2.0 (the "License"); | |
| * you may not use this file except in compliance with the License. | |
| * You may obtain a copy of the License at | |
| * | |
| * http://www.apache.org/licenses/LICENSE-2.0 | |
| * | |
| * Unless required by applicable law or agreed to in writing, software | |
| * distributed under the License is distributed on an "AS IS" BASIS, | |
| * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | |
| * See the License for the specific language governing permissions and | |
| * limitations under the License. | |
| */ | |
| package com.android.dex; | |
| import com.android.dex.util.ByteInput; | |
| import java.io.UTFDataFormatException; | |
| /** | |
| * Modified UTF-8 as described in the dex file format spec. | |
| */ | |
| public final class Mutf8 { | |
| private Mutf8() {} | |
| /** | |
| * Decodes bytes from {@code in} into {@code out} until a delimiter 0x00 is | |
| * encountered. Returns a new string containing the decoded characters. | |
| */ | |
| public static String decode(ByteInput in, char[] out) throws UTFDataFormatException { | |
| int s = 0; | |
| while (true) { | |
| char a = (char) (in.readByte() & 0xff); | |
| if (a == 0) { | |
| return new String(out, 0, s); | |
| } | |
| out[s] = a; | |
| if (a < '\u0080') { | |
| s++; | |
| } else if ((a & 0xe0) == 0xc0) { | |
| int b = in.readByte() & 0xff; | |
| if ((b & 0xC0) != 0x80) { | |
| throw new UTFDataFormatException("bad second byte"); | |
| } | |
| out[s++] = (char) (((a & 0x1F) << 6) | (b & 0x3F)); | |
| } else if ((a & 0xf0) == 0xe0) { | |
| int b = in.readByte() & 0xff; | |
| int c = in.readByte() & 0xff; | |
| if (((b & 0xC0) != 0x80) || ((c & 0xC0) != 0x80)) { | |
| throw new UTFDataFormatException("bad second or third byte"); | |
| } | |
| out[s++] = (char) (((a & 0x0F) << 12) | ((b & 0x3F) << 6) | (c & 0x3F)); | |
| } else { | |
| throw new UTFDataFormatException("bad byte"); | |
| } | |
| } | |
| } | |
| /** | |
| * Returns the number of bytes the modified UTF8 representation of 's' would take. | |
| */ | |
| private static long countBytes(String s, boolean shortLength) throws UTFDataFormatException { | |
| long result = 0; | |
| final int length = s.length(); | |
| for (int i = 0; i < length; ++i) { | |
| char ch = s.charAt(i); | |
| if (ch != 0 && ch <= 127) { // U+0000 uses two bytes. | |
| ++result; | |
| } else if (ch <= 2047) { | |
| result += 2; | |
| } else { | |
| result += 3; | |
| } | |
| if (shortLength && result > 65535) { | |
| throw new UTFDataFormatException("String more than 65535 UTF bytes long"); | |
| } | |
| } | |
| return result; | |
| } | |
| /** | |
| * Encodes the modified UTF-8 bytes corresponding to {@code s} into {@code | |
| * dst}, starting at {@code offset}. | |
| */ | |
| public static void encode(byte[] dst, int offset, String s) { | |
| final int length = s.length(); | |
| for (int i = 0; i < length; i++) { | |
| char ch = s.charAt(i); | |
| if (ch != 0 && ch <= 127) { // U+0000 uses two bytes. | |
| dst[offset++] = (byte) ch; | |
| } else if (ch <= 2047) { | |
| dst[offset++] = (byte) (0xc0 | (0x1f & (ch >> 6))); | |
| dst[offset++] = (byte) (0x80 | (0x3f & ch)); | |
| } else { | |
| dst[offset++] = (byte) (0xe0 | (0x0f & (ch >> 12))); | |
| dst[offset++] = (byte) (0x80 | (0x3f & (ch >> 6))); | |
| dst[offset++] = (byte) (0x80 | (0x3f & ch)); | |
| } | |
| } | |
| } | |
| /** | |
| * Returns an array containing the <i>modified UTF-8</i> form of {@code s}. | |
| */ | |
| public static byte[] encode(String s) throws UTFDataFormatException { | |
| int utfCount = (int) countBytes(s, true); | |
| byte[] result = new byte[utfCount]; | |
| encode(result, 0, s); | |
| return result; | |
| } | |
| } |