[6785] | 1 | /**************************************************************** |
---|
| 2 | * Licensed to the Apache Software Foundation (ASF) under one * |
---|
| 3 | * or more contributor license agreements. See the NOTICE file * |
---|
| 4 | * distributed with this work for additional information * |
---|
| 5 | * regarding copyright ownership. The ASF licenses this file * |
---|
| 6 | * to you under the Apache License, Version 2.0 (the * |
---|
| 7 | * "License"); you may not use this file except in compliance * |
---|
| 8 | * with the License. You may obtain a copy of the License at * |
---|
| 9 | * * |
---|
| 10 | * http://www.apache.org/licenses/LICENSE-2.0 * |
---|
| 11 | * * |
---|
| 12 | * Unless required by applicable law or agreed to in writing, * |
---|
| 13 | * software distributed under the License is distributed on an * |
---|
| 14 | * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY * |
---|
| 15 | * KIND, either express or implied. See the License for the * |
---|
| 16 | * specific language governing permissions and limitations * |
---|
| 17 | * under the License. * |
---|
| 18 | ****************************************************************/ |
---|
| 19 | |
---|
| 20 | package org.apache.james.mime4j.codec; |
---|
| 21 | |
---|
| 22 | import java.io.IOException; |
---|
| 23 | import java.io.InputStream; |
---|
| 24 | |
---|
| 25 | import org.apache.james.mime4j.util.ByteArrayBuffer; |
---|
| 26 | |
---|
| 27 | /** |
---|
| 28 | * Performs Quoted-Printable decoding on an underlying stream. |
---|
| 29 | */ |
---|
| 30 | public class QuotedPrintableInputStream extends InputStream { |
---|
| 31 | |
---|
| 32 | private static final int DEFAULT_BUFFER_SIZE = 1024 * 2; |
---|
| 33 | |
---|
| 34 | private static final byte EQ = 0x3D; |
---|
| 35 | private static final byte CR = 0x0D; |
---|
| 36 | private static final byte LF = 0x0A; |
---|
| 37 | |
---|
| 38 | private final byte[] singleByte = new byte[1]; |
---|
| 39 | |
---|
| 40 | private final InputStream in; |
---|
| 41 | private final ByteArrayBuffer decodedBuf; |
---|
| 42 | private final ByteArrayBuffer blanks; |
---|
| 43 | |
---|
| 44 | private final byte[] encoded; |
---|
| 45 | private int pos = 0; // current index into encoded buffer |
---|
| 46 | private int limit = 0; // current size of encoded buffer |
---|
| 47 | |
---|
| 48 | private boolean closed; |
---|
| 49 | |
---|
| 50 | private final DecodeMonitor monitor; |
---|
| 51 | |
---|
| 52 | public QuotedPrintableInputStream(final InputStream in, DecodeMonitor monitor) { |
---|
| 53 | this(DEFAULT_BUFFER_SIZE, in, monitor); |
---|
| 54 | } |
---|
| 55 | |
---|
| 56 | protected QuotedPrintableInputStream(final int bufsize, final InputStream in, DecodeMonitor monitor) { |
---|
| 57 | super(); |
---|
| 58 | this.in = in; |
---|
| 59 | this.encoded = new byte[bufsize]; |
---|
| 60 | this.decodedBuf = new ByteArrayBuffer(512); |
---|
| 61 | this.blanks = new ByteArrayBuffer(512); |
---|
| 62 | this.closed = false; |
---|
| 63 | this.monitor = monitor; |
---|
| 64 | } |
---|
| 65 | |
---|
| 66 | protected QuotedPrintableInputStream(final int bufsize, final InputStream in, boolean strict) { |
---|
| 67 | this(bufsize, in, strict ? DecodeMonitor.STRICT : DecodeMonitor.SILENT); |
---|
| 68 | } |
---|
| 69 | |
---|
| 70 | public QuotedPrintableInputStream(final InputStream in, boolean strict) { |
---|
| 71 | this(DEFAULT_BUFFER_SIZE, in, strict); |
---|
| 72 | } |
---|
| 73 | |
---|
| 74 | public QuotedPrintableInputStream(final InputStream in) { |
---|
| 75 | this(in, false); |
---|
| 76 | } |
---|
| 77 | |
---|
| 78 | /** |
---|
| 79 | * Terminates Quoted-Printable coded content. This method does NOT close |
---|
| 80 | * the underlying input stream. |
---|
| 81 | * |
---|
| 82 | * @throws IOException on I/O errors. |
---|
| 83 | */ |
---|
| 84 | @Override |
---|
| 85 | public void close() throws IOException { |
---|
| 86 | closed = true; |
---|
| 87 | } |
---|
| 88 | |
---|
| 89 | private int fillBuffer() throws IOException { |
---|
| 90 | // Compact buffer if needed |
---|
| 91 | if (pos < limit) { |
---|
| 92 | System.arraycopy(encoded, pos, encoded, 0, limit - pos); |
---|
| 93 | limit -= pos; |
---|
| 94 | pos = 0; |
---|
| 95 | } else { |
---|
| 96 | limit = 0; |
---|
| 97 | pos = 0; |
---|
| 98 | } |
---|
| 99 | |
---|
| 100 | int capacity = encoded.length - limit; |
---|
| 101 | if (capacity > 0) { |
---|
| 102 | int bytesRead = in.read(encoded, limit, capacity); |
---|
| 103 | if (bytesRead > 0) { |
---|
| 104 | limit += bytesRead; |
---|
| 105 | } |
---|
| 106 | return bytesRead; |
---|
| 107 | } else { |
---|
| 108 | return 0; |
---|
| 109 | } |
---|
| 110 | } |
---|
| 111 | |
---|
| 112 | private int getnext() { |
---|
| 113 | if (pos < limit) { |
---|
| 114 | byte b = encoded[pos]; |
---|
| 115 | pos++; |
---|
| 116 | return b & 0xFF; |
---|
| 117 | } else { |
---|
| 118 | return -1; |
---|
| 119 | } |
---|
| 120 | } |
---|
| 121 | |
---|
| 122 | private int peek(int i) { |
---|
| 123 | if (pos + i < limit) { |
---|
| 124 | return encoded[pos + i] & 0xFF; |
---|
| 125 | } else { |
---|
| 126 | return -1; |
---|
| 127 | } |
---|
| 128 | } |
---|
| 129 | |
---|
| 130 | private int transfer( |
---|
| 131 | final int b, final byte[] buffer, final int from, final int to, boolean keepblanks) throws IOException { |
---|
| 132 | int index = from; |
---|
| 133 | if (keepblanks && blanks.length() > 0) { |
---|
| 134 | int chunk = Math.min(blanks.length(), to - index); |
---|
| 135 | System.arraycopy(blanks.buffer(), 0, buffer, index, chunk); |
---|
| 136 | index += chunk; |
---|
| 137 | int remaining = blanks.length() - chunk; |
---|
| 138 | if (remaining > 0) { |
---|
| 139 | decodedBuf.append(blanks.buffer(), chunk, remaining); |
---|
| 140 | } |
---|
| 141 | blanks.clear(); |
---|
| 142 | } else if (blanks.length() > 0 && !keepblanks) { |
---|
| 143 | StringBuilder sb = new StringBuilder(blanks.length() * 3); |
---|
| 144 | for (int i = 0; i < blanks.length(); i++) sb.append(" "+blanks.byteAt(i)); |
---|
| 145 | if (monitor.warn("ignored blanks", sb.toString())) |
---|
| 146 | throw new IOException("ignored blanks"); |
---|
| 147 | } |
---|
| 148 | if (b != -1) { |
---|
| 149 | if (index < to) { |
---|
| 150 | buffer[index++] = (byte) b; |
---|
| 151 | } else { |
---|
| 152 | decodedBuf.append(b); |
---|
| 153 | } |
---|
| 154 | } |
---|
| 155 | return index; |
---|
| 156 | } |
---|
| 157 | |
---|
| 158 | private int read0(final byte[] buffer, final int off, final int len) throws IOException { |
---|
| 159 | boolean eof = false; |
---|
| 160 | int from = off; |
---|
| 161 | int to = off + len; |
---|
| 162 | int index = off; |
---|
| 163 | |
---|
| 164 | // check if a previous invocation left decoded content |
---|
| 165 | if (decodedBuf.length() > 0) { |
---|
| 166 | int chunk = Math.min(decodedBuf.length(), to - index); |
---|
| 167 | System.arraycopy(decodedBuf.buffer(), 0, buffer, index, chunk); |
---|
| 168 | decodedBuf.remove(0, chunk); |
---|
| 169 | index += chunk; |
---|
| 170 | } |
---|
| 171 | |
---|
| 172 | while (index < to) { |
---|
| 173 | |
---|
| 174 | if (limit - pos < 3) { |
---|
| 175 | int bytesRead = fillBuffer(); |
---|
| 176 | eof = bytesRead == -1; |
---|
| 177 | } |
---|
| 178 | |
---|
| 179 | // end of stream? |
---|
| 180 | if (limit - pos == 0 && eof) { |
---|
| 181 | return index == from ? -1 : index - from; |
---|
| 182 | } |
---|
| 183 | |
---|
| 184 | boolean lastWasCR = false; |
---|
| 185 | while (pos < limit && index < to) { |
---|
| 186 | int b = encoded[pos++] & 0xFF; |
---|
| 187 | |
---|
| 188 | if (lastWasCR && b != LF) { |
---|
| 189 | if (monitor.warn("Found CR without LF", "Leaving it as is")) |
---|
| 190 | throw new IOException("Found CR without LF"); |
---|
| 191 | index = transfer(CR, buffer, index, to, false); |
---|
| 192 | } else if (!lastWasCR && b == LF) { |
---|
| 193 | if (monitor.warn("Found LF without CR", "Translating to CRLF")) |
---|
| 194 | throw new IOException("Found LF without CR"); |
---|
| 195 | } |
---|
| 196 | |
---|
| 197 | if (b == CR) { |
---|
| 198 | lastWasCR = true; |
---|
| 199 | continue; |
---|
| 200 | } else { |
---|
| 201 | lastWasCR = false; |
---|
| 202 | } |
---|
| 203 | |
---|
| 204 | if (b == LF) { |
---|
| 205 | // at end of line |
---|
| 206 | if (blanks.length() == 0) { |
---|
| 207 | index = transfer(CR, buffer, index, to, false); |
---|
| 208 | index = transfer(LF, buffer, index, to, false); |
---|
| 209 | } else { |
---|
| 210 | if (blanks.byteAt(0) != EQ) { |
---|
| 211 | // hard line break |
---|
| 212 | index = transfer(CR, buffer, index, to, false); |
---|
| 213 | index = transfer(LF, buffer, index, to, false); |
---|
| 214 | } |
---|
| 215 | } |
---|
| 216 | blanks.clear(); |
---|
| 217 | } else if (b == EQ) { |
---|
| 218 | if (limit - pos < 2 && !eof) { |
---|
| 219 | // not enough buffered data |
---|
| 220 | pos--; |
---|
| 221 | break; |
---|
| 222 | } |
---|
| 223 | |
---|
| 224 | // found special char '=' |
---|
| 225 | int b2 = getnext(); |
---|
| 226 | if (b2 == EQ) { |
---|
| 227 | index = transfer(b2, buffer, index, to, true); |
---|
| 228 | // deal with '==\r\n' brokenness |
---|
| 229 | int bb1 = peek(0); |
---|
| 230 | int bb2 = peek(1); |
---|
| 231 | if (bb1 == LF || (bb1 == CR && bb2 == LF)) { |
---|
| 232 | monitor.warn("Unexpected ==EOL encountered", "== 0x"+bb1+" 0x"+bb2); |
---|
| 233 | blanks.append(b2); |
---|
| 234 | } else { |
---|
| 235 | monitor.warn("Unexpected == encountered", "=="); |
---|
| 236 | } |
---|
| 237 | } else if (Character.isWhitespace((char) b2)) { |
---|
| 238 | // soft line break |
---|
| 239 | index = transfer(-1, buffer, index, to, true); |
---|
| 240 | if (b2 != LF) { |
---|
| 241 | blanks.append(b); |
---|
| 242 | blanks.append(b2); |
---|
| 243 | } |
---|
| 244 | } else { |
---|
| 245 | int b3 = getnext(); |
---|
| 246 | int upper = convert(b2); |
---|
| 247 | int lower = convert(b3); |
---|
| 248 | if (upper < 0 || lower < 0) { |
---|
| 249 | monitor.warn("Malformed encoded value encountered", "leaving "+((char) EQ)+((char) b2)+((char) b3)+" as is"); |
---|
| 250 | // TODO see MIME4J-160 |
---|
| 251 | index = transfer(EQ, buffer, index, to, true); |
---|
| 252 | index = transfer(b2, buffer, index, to, false); |
---|
| 253 | index = transfer(b3, buffer, index, to, false); |
---|
| 254 | } else { |
---|
| 255 | index = transfer((upper << 4) | lower, buffer, index, to, true); |
---|
| 256 | } |
---|
| 257 | } |
---|
| 258 | } else if (Character.isWhitespace(b)) { |
---|
| 259 | blanks.append(b); |
---|
| 260 | } else { |
---|
| 261 | index = transfer((int) b & 0xFF, buffer, index, to, true); |
---|
| 262 | } |
---|
| 263 | } |
---|
| 264 | } |
---|
| 265 | return to - from; |
---|
| 266 | } |
---|
| 267 | |
---|
| 268 | /** |
---|
| 269 | * Converts '0' => 0, 'A' => 10, etc. |
---|
| 270 | * @param c ASCII character value. |
---|
| 271 | * @return Numeric value of hexadecimal character. |
---|
| 272 | */ |
---|
| 273 | private int convert(int c) { |
---|
| 274 | if (c >= '0' && c <= '9') { |
---|
| 275 | return (c - '0'); |
---|
| 276 | } else if (c >= 'A' && c <= 'F') { |
---|
| 277 | return (0xA + (c - 'A')); |
---|
| 278 | } else if (c >= 'a' && c <= 'f') { |
---|
| 279 | return (0xA + (c - 'a')); |
---|
| 280 | } else { |
---|
| 281 | return -1; |
---|
| 282 | } |
---|
| 283 | } |
---|
| 284 | |
---|
| 285 | @Override |
---|
| 286 | public int read() throws IOException { |
---|
| 287 | if (closed) { |
---|
| 288 | throw new IOException("Stream has been closed"); |
---|
| 289 | } |
---|
| 290 | for (;;) { |
---|
| 291 | int bytes = read(singleByte, 0, 1); |
---|
| 292 | if (bytes == -1) { |
---|
| 293 | return -1; |
---|
| 294 | } |
---|
| 295 | if (bytes == 1) { |
---|
| 296 | return singleByte[0] & 0xff; |
---|
| 297 | } |
---|
| 298 | } |
---|
| 299 | } |
---|
| 300 | |
---|
| 301 | @Override |
---|
| 302 | public int read(byte[] b, int off, int len) throws IOException { |
---|
| 303 | if (closed) { |
---|
| 304 | throw new IOException("Stream has been closed"); |
---|
| 305 | } |
---|
| 306 | return read0(b, off, len); |
---|
| 307 | } |
---|
| 308 | |
---|
| 309 | } |
---|