[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.io; |
---|
| 21 | |
---|
| 22 | import java.io.FilterInputStream; |
---|
| 23 | import java.io.IOException; |
---|
| 24 | import java.io.InputStream; |
---|
| 25 | |
---|
| 26 | import java.util.List; |
---|
| 27 | import java.util.ArrayList; |
---|
| 28 | import java.util.Collections; |
---|
| 29 | import java.util.regex.Pattern; |
---|
| 30 | import java.util.regex.Matcher; |
---|
| 31 | |
---|
| 32 | /** |
---|
| 33 | * <code>InputStream</code> used by the parser to wrap the original user |
---|
| 34 | * supplied stream. This stream keeps track of the current line number. |
---|
| 35 | */ |
---|
| 36 | public class LineNumberInputStream extends FilterInputStream implements LineNumberSource { |
---|
| 37 | |
---|
| 38 | /** |
---|
| 39 | * Creates a new <code>LineNumberInputStream</code>. |
---|
| 40 | * |
---|
| 41 | * @param is the stream to read from. |
---|
| 42 | */ |
---|
| 43 | public LineNumberInputStream(InputStream is) { |
---|
| 44 | super(is); |
---|
| 45 | |
---|
| 46 | currentEntity = new Entity(); |
---|
| 47 | currentEntity.startLine = 1; |
---|
| 48 | rootEntity = currentEntity; |
---|
| 49 | } |
---|
| 50 | |
---|
| 51 | @Override |
---|
| 52 | public int read() throws IOException { |
---|
| 53 | int b = in.read(); |
---|
| 54 | if(b >= 0) { |
---|
| 55 | process(b); |
---|
| 56 | } |
---|
| 57 | return b; |
---|
| 58 | } |
---|
| 59 | |
---|
| 60 | @Override |
---|
| 61 | public int read(byte[] b, int off, int len) throws IOException { |
---|
| 62 | int n = in.read(b, off, len); |
---|
| 63 | if(n >= 0) { |
---|
| 64 | for (int i = off; i < off + n; i++) { |
---|
| 65 | process(b[i]); |
---|
| 66 | } |
---|
| 67 | } |
---|
| 68 | return n; |
---|
| 69 | } |
---|
| 70 | |
---|
| 71 | |
---|
| 72 | //========================================================================= |
---|
| 73 | private static final int CR = 13; |
---|
| 74 | private static final int LF = 10; |
---|
| 75 | |
---|
| 76 | private int lineNumber = 1; |
---|
| 77 | private int lineOffset = 0; |
---|
| 78 | private int offset = 0; |
---|
| 79 | |
---|
| 80 | private boolean pendingCR = false; |
---|
| 81 | private boolean insideBody = false; |
---|
| 82 | |
---|
| 83 | private StringBuilder lineBuilder = new StringBuilder(4096); |
---|
| 84 | private StringBuilder fieldBuilder = new StringBuilder(8192); |
---|
| 85 | |
---|
| 86 | private Entity currentEntity; |
---|
| 87 | |
---|
| 88 | private final Entity rootEntity; |
---|
| 89 | |
---|
| 90 | public Entity getRootEntity() { |
---|
| 91 | return rootEntity; |
---|
| 92 | } |
---|
| 93 | |
---|
| 94 | public int getLineNumber() { |
---|
| 95 | return lineNumber; |
---|
| 96 | } |
---|
| 97 | |
---|
| 98 | |
---|
| 99 | private void process(int b) { |
---|
| 100 | if(pendingCR) { |
---|
| 101 | if(b == LF) { |
---|
| 102 | ++offset; |
---|
| 103 | processLine(); |
---|
| 104 | } |
---|
| 105 | else { |
---|
| 106 | processLine(); |
---|
| 107 | process(b); |
---|
| 108 | } |
---|
| 109 | } |
---|
| 110 | else { |
---|
| 111 | ++offset; |
---|
| 112 | if(b == CR) { |
---|
| 113 | pendingCR = true; |
---|
| 114 | } |
---|
| 115 | else if(b == LF) { |
---|
| 116 | processLine(); |
---|
| 117 | } |
---|
| 118 | else { |
---|
| 119 | lineBuilder.append((char)b); |
---|
| 120 | } |
---|
| 121 | } |
---|
| 122 | } |
---|
| 123 | |
---|
| 124 | |
---|
| 125 | private void processLine() { |
---|
| 126 | if(insideBody) { |
---|
| 127 | processBodyLine(); |
---|
| 128 | } |
---|
| 129 | else { |
---|
| 130 | processHeaderLine(); |
---|
| 131 | } |
---|
| 132 | pendingCR = false; |
---|
| 133 | lineBuilder.setLength(0); |
---|
| 134 | lineOffset = offset; |
---|
| 135 | ++lineNumber; |
---|
| 136 | } |
---|
| 137 | |
---|
| 138 | |
---|
| 139 | private void processHeaderLine() { |
---|
| 140 | |
---|
| 141 | String line = lineBuilder.toString(); |
---|
| 142 | if(line.trim().isEmpty()) { |
---|
| 143 | |
---|
| 144 | //reach the separator line between header and body |
---|
| 145 | |
---|
| 146 | processField(); |
---|
| 147 | |
---|
| 148 | currentEntity.separatorLine = lineNumber; |
---|
| 149 | currentEntity.bodyOffset = offset; |
---|
| 150 | |
---|
| 151 | if(currentEntity.isMessageMimeType()) { |
---|
| 152 | |
---|
| 153 | if(currentEntity.isBase64Encoded()) { |
---|
| 154 | insideBody = true; |
---|
| 155 | } |
---|
| 156 | |
---|
| 157 | currentEntity = currentEntity.getEmbeddedMessage(); |
---|
| 158 | currentEntity.startLine = lineNumber + 1; |
---|
| 159 | } |
---|
| 160 | else { |
---|
| 161 | insideBody = true; |
---|
| 162 | } |
---|
| 163 | } |
---|
| 164 | else if(Character.isWhitespace(line.charAt(0))) { |
---|
| 165 | fieldBuilder.append(line); |
---|
| 166 | } |
---|
| 167 | else { |
---|
| 168 | processField(); |
---|
| 169 | fieldBuilder.append(line); |
---|
| 170 | } |
---|
| 171 | } |
---|
| 172 | |
---|
| 173 | |
---|
| 174 | private static final Pattern base64EncodedPattern = Pattern.compile("^content-transfer-encoding(\\s*):(\\s*)(base64)", Pattern.CASE_INSENSITIVE); |
---|
| 175 | private static final Pattern compositeTypePattern = Pattern.compile("^content-type(\\s*):(\\s*)((message)|(multipart))/([\\w-]+)", Pattern.CASE_INSENSITIVE); |
---|
| 176 | private static final Pattern boundaryPattern = Pattern.compile(";(\\s*)boundary(\\s*)=(\\s*)([^\"\\s;]+|\"[^\"\\\\\\r\\n]*(?:\\\\.[^\"\\\\\\r\\n]*)*\")", Pattern.CASE_INSENSITIVE); |
---|
| 177 | |
---|
| 178 | private void processField() { |
---|
| 179 | |
---|
| 180 | if(fieldBuilder.length() == 0) { |
---|
| 181 | return; |
---|
| 182 | } |
---|
| 183 | |
---|
| 184 | Matcher base64Matcher = base64EncodedPattern.matcher(fieldBuilder); |
---|
| 185 | |
---|
| 186 | if(base64Matcher.lookingAt()) { |
---|
| 187 | currentEntity.base64Encoded = true; |
---|
| 188 | } |
---|
| 189 | |
---|
| 190 | Matcher compositeMatcher = compositeTypePattern.matcher(fieldBuilder); |
---|
| 191 | |
---|
| 192 | if(compositeMatcher.lookingAt()) { |
---|
| 193 | |
---|
| 194 | if(compositeMatcher.group(4) != null) { |
---|
| 195 | currentEntity.messageMimeType = true; |
---|
| 196 | } |
---|
| 197 | else { |
---|
| 198 | currentEntity.multipartMimeType = true; |
---|
| 199 | |
---|
| 200 | Matcher boundaryMatcher = boundaryPattern.matcher(fieldBuilder.subSequence(compositeMatcher.end(6), fieldBuilder.length())); |
---|
| 201 | |
---|
| 202 | if(boundaryMatcher.lookingAt()) { |
---|
| 203 | String boundary = boundaryMatcher.group(4); |
---|
| 204 | if(boundary.charAt(0) == '\"') { |
---|
| 205 | boundary = boundary.substring(1, boundary.length() - 1); |
---|
| 206 | } |
---|
| 207 | currentEntity.boundary = boundary; |
---|
| 208 | } |
---|
| 209 | else { |
---|
| 210 | throw new RuntimeException("boundary parameter not found"); |
---|
| 211 | } |
---|
| 212 | } |
---|
| 213 | } |
---|
| 214 | fieldBuilder.setLength(0); |
---|
| 215 | } |
---|
| 216 | |
---|
| 217 | private void processBodyLine() { |
---|
| 218 | |
---|
| 219 | if((lineBuilder.length() < 2) || (lineBuilder.charAt(0) != '-') || (lineBuilder.charAt(1) != '-')) { |
---|
| 220 | return; |
---|
| 221 | } |
---|
| 222 | |
---|
| 223 | Entity ancestorEntity; |
---|
| 224 | |
---|
| 225 | if(currentEntity.isMultipartMimeType() && (currentEntity.parts == null)) { |
---|
| 226 | ancestorEntity = currentEntity; |
---|
| 227 | } |
---|
| 228 | else { |
---|
| 229 | ancestorEntity = currentEntity.parent; |
---|
| 230 | |
---|
| 231 | while(ancestorEntity != null) { |
---|
| 232 | if(ancestorEntity.isMultipartMimeType()) { |
---|
| 233 | break; |
---|
| 234 | } |
---|
| 235 | ancestorEntity = ancestorEntity.parent; |
---|
| 236 | } |
---|
| 237 | } |
---|
| 238 | |
---|
| 239 | if((ancestorEntity == null) || ((ancestorEntity.boundary.length() + 2) > lineBuilder.length())) { |
---|
| 240 | return; |
---|
| 241 | } |
---|
| 242 | |
---|
| 243 | for(int i = 0, j = 2; i < ancestorEntity.boundary.length(); i++, j++) { |
---|
| 244 | if(ancestorEntity.boundary.charAt(i) != lineBuilder.charAt(j)) { |
---|
| 245 | return; |
---|
| 246 | } |
---|
| 247 | } |
---|
| 248 | |
---|
| 249 | int k = 0; |
---|
| 250 | |
---|
| 251 | for(int i = 0, j = ancestorEntity.boundary.length() + 2; j < lineBuilder.length(); i++, j++) { |
---|
| 252 | if((i == 0) && (lineBuilder.charAt(j) == '-')) { |
---|
| 253 | k = 1; |
---|
| 254 | continue; |
---|
| 255 | } |
---|
| 256 | else if((i == 1) && (k == 1)) { |
---|
| 257 | if(lineBuilder.charAt(j) == '-') { |
---|
| 258 | k = 2; |
---|
| 259 | continue; |
---|
| 260 | } |
---|
| 261 | return; |
---|
| 262 | } |
---|
| 263 | else if(Character.isWhitespace(lineBuilder.charAt(j))){ |
---|
| 264 | continue; |
---|
| 265 | } |
---|
| 266 | return; |
---|
| 267 | } |
---|
| 268 | |
---|
| 269 | if(k == 1) { |
---|
| 270 | return; |
---|
| 271 | } |
---|
| 272 | |
---|
| 273 | //it's a boundary line ! |
---|
| 274 | |
---|
| 275 | if(ancestorEntity.parts != null) { |
---|
| 276 | do { |
---|
| 277 | currentEntity.endLine = lineNumber - 1; |
---|
| 278 | currentEntity.bodyLength = lineOffset - currentEntity.bodyOffset; |
---|
| 279 | |
---|
| 280 | currentEntity = currentEntity.parent; |
---|
| 281 | } |
---|
| 282 | while(currentEntity != ancestorEntity); |
---|
| 283 | } |
---|
| 284 | |
---|
| 285 | if(k == 0) { |
---|
| 286 | currentEntity = currentEntity.getLocalNextPart(); |
---|
| 287 | currentEntity.startLine = lineNumber + 1; |
---|
| 288 | insideBody = false; |
---|
| 289 | } |
---|
| 290 | } |
---|
| 291 | |
---|
| 292 | public void endOfStream() { |
---|
| 293 | if(lineBuilder.length() > 0) { |
---|
| 294 | if(insideBody) { |
---|
| 295 | processBodyLine(); |
---|
| 296 | } |
---|
| 297 | else { |
---|
| 298 | processHeaderLine(); |
---|
| 299 | } |
---|
| 300 | } |
---|
| 301 | |
---|
| 302 | for(;;) { |
---|
| 303 | currentEntity.endLine = lineNumber - 1; |
---|
| 304 | currentEntity.bodyLength = lineOffset - currentEntity.bodyOffset; |
---|
| 305 | |
---|
| 306 | if(currentEntity.parent == null) { |
---|
| 307 | break; |
---|
| 308 | } |
---|
| 309 | |
---|
| 310 | currentEntity = currentEntity.parent; |
---|
| 311 | } |
---|
| 312 | } |
---|
| 313 | |
---|
| 314 | |
---|
| 315 | public class Entity { |
---|
| 316 | |
---|
| 317 | private int startLine = -1; |
---|
| 318 | private int separatorLine = -1; |
---|
| 319 | private int endLine = -1; |
---|
| 320 | |
---|
| 321 | private int bodyOffset = -1; |
---|
| 322 | private int bodyLength = -1; |
---|
| 323 | |
---|
| 324 | public int getStartLine() { return startLine; } |
---|
| 325 | public int getSeparatorLine() { return separatorLine; } |
---|
| 326 | public int getEndLine() { return endLine; } |
---|
| 327 | |
---|
| 328 | public int getBodyOffset() { return bodyOffset; } |
---|
| 329 | public int getBodyLength() { return bodyLength; } |
---|
| 330 | |
---|
| 331 | //----------------------------- |
---|
| 332 | private boolean base64Encoded; |
---|
| 333 | |
---|
| 334 | public boolean isBase64Encoded() { |
---|
| 335 | return base64Encoded; |
---|
| 336 | } |
---|
| 337 | |
---|
| 338 | //----------------------------- |
---|
| 339 | private Entity parent; |
---|
| 340 | |
---|
| 341 | public Entity getParent() { |
---|
| 342 | return parent; |
---|
| 343 | } |
---|
| 344 | |
---|
| 345 | //----------------------------- |
---|
| 346 | private boolean multipartMimeType; |
---|
| 347 | private String boundary; |
---|
| 348 | private List<Entity> parts; |
---|
| 349 | private int nextPart = 0; |
---|
| 350 | private int localNextPart = 0; |
---|
| 351 | |
---|
| 352 | public boolean isMultipartMimeType() { |
---|
| 353 | return multipartMimeType; |
---|
| 354 | } |
---|
| 355 | |
---|
| 356 | public String getBoundary() { |
---|
| 357 | return boundary; |
---|
| 358 | } |
---|
| 359 | |
---|
| 360 | public synchronized Entity getNextPart() { |
---|
| 361 | if(parts == null) { |
---|
| 362 | parts = new ArrayList<Entity>(); |
---|
| 363 | multipartMimeType = true; |
---|
| 364 | } |
---|
| 365 | while(parts.size() <= nextPart) { |
---|
| 366 | Entity bodyPart = new Entity(); |
---|
| 367 | bodyPart.parent = this; |
---|
| 368 | parts.add(bodyPart); |
---|
| 369 | } |
---|
| 370 | return parts.get(nextPart++); |
---|
| 371 | } |
---|
| 372 | |
---|
| 373 | private synchronized Entity getLocalNextPart() { |
---|
| 374 | if(parts == null) { |
---|
| 375 | parts = new ArrayList<Entity>(); |
---|
| 376 | multipartMimeType = true; |
---|
| 377 | } |
---|
| 378 | while(parts.size() <= localNextPart) { |
---|
| 379 | Entity bodyPart = new Entity(); |
---|
| 380 | bodyPart.parent = this; |
---|
| 381 | parts.add(bodyPart); |
---|
| 382 | } |
---|
| 383 | return parts.get(localNextPart++); |
---|
| 384 | } |
---|
| 385 | |
---|
| 386 | //----------------------------- |
---|
| 387 | private boolean messageMimeType; |
---|
| 388 | private Entity embeddedMessage; |
---|
| 389 | |
---|
| 390 | public boolean isMessageMimeType() { |
---|
| 391 | return messageMimeType; |
---|
| 392 | } |
---|
| 393 | |
---|
| 394 | public synchronized Entity getEmbeddedMessage() { |
---|
| 395 | if(embeddedMessage == null) { |
---|
| 396 | embeddedMessage = new Entity(); |
---|
| 397 | embeddedMessage.parent = this; |
---|
| 398 | messageMimeType = true; |
---|
| 399 | } |
---|
| 400 | return embeddedMessage; |
---|
| 401 | } |
---|
| 402 | |
---|
| 403 | //----------------------------- |
---|
| 404 | private void dump(String label, String indent, StringBuilder sb) { |
---|
| 405 | sb.append(indent).append(label); |
---|
| 406 | |
---|
| 407 | if(this == currentEntity) { |
---|
| 408 | sb.append(" #{").append(hashCode()).append("}#"); |
---|
| 409 | } |
---|
| 410 | else { |
---|
| 411 | sb.append(" {").append(hashCode()).append("}"); |
---|
| 412 | } |
---|
| 413 | sb.append(" (startLine:").append(startLine) |
---|
| 414 | .append(", separatorLine:").append(separatorLine) |
---|
| 415 | .append(", endLine:").append(endLine) |
---|
| 416 | .append(", bodyOffset:").append(bodyOffset) |
---|
| 417 | .append(", bodyLength:").append(bodyLength); |
---|
| 418 | if(boundary != null) { |
---|
| 419 | sb.append(", boundary:").append(boundary); |
---|
| 420 | } |
---|
| 421 | sb.append(")\n"); |
---|
| 422 | if(isMessageMimeType()) { |
---|
| 423 | if(embeddedMessage == null) { |
---|
| 424 | sb.append(indent + " Embedded: null"); |
---|
| 425 | } |
---|
| 426 | else { |
---|
| 427 | embeddedMessage.dump("Embedded", indent + " ", sb); |
---|
| 428 | } |
---|
| 429 | } |
---|
| 430 | else if(isMultipartMimeType()) { |
---|
| 431 | if(parts == null) { |
---|
| 432 | sb.append(indent + " Parts: null"); |
---|
| 433 | } |
---|
| 434 | else { |
---|
| 435 | for(int i = 0; i < parts.size(); i++) { |
---|
| 436 | parts.get(i).dump("Part[" + (i+1) + "]", indent + " ", sb); |
---|
| 437 | } |
---|
| 438 | } |
---|
| 439 | } |
---|
| 440 | } |
---|
| 441 | |
---|
| 442 | public String dump(String label) { |
---|
| 443 | StringBuilder sb = new StringBuilder(); |
---|
| 444 | dump(label, "", sb); |
---|
| 445 | return sb.toString(); |
---|
| 446 | } |
---|
| 447 | } |
---|
| 448 | } |
---|