1 /* 2 * Licensed to the Apache Software Foundation (ASF) under one or more 3 * contributor license agreements. See the NOTICE file distributed with 4 * this work for additional information regarding copyright ownership. 5 * The ASF licenses this file to You under the Apache License, Version 2.0 6 * (the "License"); you may not use this file except in compliance with 7 * the License. You may obtain a copy of the License at 8 * 9 * http://www.apache.org/licenses/LICENSE-2.0 10 * 11 * Unless required by applicable law or agreed to in writing, software 12 * distributed under the License is distributed on an "AS IS" BASIS, 13 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 14 * See the License for the specific language governing permissions and 15 * limitations under the License. 16 */ 17 18 package org.apache.log4j; 19 20 import org.apache.log4j.spi.LoggingEvent; 21 import org.apache.log4j.helpers.PatternParser; 22 import org.apache.log4j.helpers.PatternConverter; 23 24 25 // Contributors: Nelson Minar <nelson@monkey.org> 26 // Anders Kristensen <akristensen@dynamicsoft.com> 27 28 /** 29 30 A flexible layout configurable with pattern string. 31 32 This code is known to have synchronization and other issues 33 which are not present in org.apache.log4j.EnhancedPatternLayout. 34 EnhancedPatternLayout should be used in preference to PatternLayout. 35 EnhancedPatternLayout is distributed in the log4j extras companion. 36 37 <p>The goal of this class is to {@link #format format} a {@link 38 LoggingEvent} and return the results as a String. The results 39 depend on the <em>conversion pattern</em>. 40 41 <p>The conversion pattern is closely related to the conversion 42 pattern of the printf function in C. A conversion pattern is 43 composed of literal text and format control expressions called 44 <em>conversion specifiers</em>. 45 46 <p><i>You are free to insert any literal text within the conversion 47 pattern.</i> 48 49 <p>Each conversion specifier starts with a percent sign (%) and is 50 followed by optional <em>format modifiers</em> and a <em>conversion 51 character</em>. The conversion character specifies the type of 52 data, e.g. category, priority, date, thread name. The format 53 modifiers control such things as field width, padding, left and 54 right justification. The following is a simple example. 55 56 <p>Let the conversion pattern be <b>"%-5p [%t]: %m%n"</b> and assume 57 that the log4j environment was set to use a PatternLayout. Then the 58 statements 59 <pre> 60 Category root = Category.getRoot(); 61 root.debug("Message 1"); 62 root.warn("Message 2"); 63 </pre> 64 would yield the output 65 <pre> 66 DEBUG [main]: Message 1 67 WARN [main]: Message 2 68 </pre> 69 70 <p>Note that there is no explicit separator between text and 71 conversion specifiers. The pattern parser knows when it has reached 72 the end of a conversion specifier when it reads a conversion 73 character. In the example above the conversion specifier 74 <b>%-5p</b> means the priority of the logging event should be left 75 justified to a width of five characters. 76 77 The recognized conversion characters are 78 79 <p> 80 <table border="1" CELLPADDING="8"> 81 <th>Conversion Character</th> 82 <th>Effect</th> 83 84 <tr> 85 <td align=center><b>c</b></td> 86 87 <td>Used to output the category of the logging event. The 88 category conversion specifier can be optionally followed by 89 <em>precision specifier</em>, that is a decimal constant in 90 brackets. 91 92 <p>If a precision specifier is given, then only the corresponding 93 number of right most components of the category name will be 94 printed. By default the category name is printed in full. 95 96 <p>For example, for the category name "a.b.c" the pattern 97 <b>%c{2}</b> will output "b.c". 98 99 </td> 100 </tr> 101 102 <tr> 103 <td align=center><b>C</b></td> 104 105 <td>Used to output the fully qualified class name of the caller 106 issuing the logging request. This conversion specifier 107 can be optionally followed by <em>precision specifier</em>, that 108 is a decimal constant in brackets. 109 110 <p>If a precision specifier is given, then only the corresponding 111 number of right most components of the class name will be 112 printed. By default the class name is output in fully qualified form. 113 114 <p>For example, for the class name "org.apache.xyz.SomeClass", the 115 pattern <b>%C{1}</b> will output "SomeClass". 116 117 <p><b>WARNING</b> Generating the caller class information is 118 slow. Thus, use should be avoided unless execution speed is 119 not an issue. 120 121 </td> 122 </tr> 123 124 <tr> <td align=center><b>d</b></td> <td>Used to output the date of 125 the logging event. The date conversion specifier may be 126 followed by a <em>date format specifier</em> enclosed between 127 braces. For example, <b>%d{HH:mm:ss,SSS}</b> or 128 <b>%d{dd MMM yyyy HH:mm:ss,SSS}</b>. If no 129 date format specifier is given then ISO8601 format is 130 assumed. 131 132 <p>The date format specifier admits the same syntax as the 133 time pattern string of the {@link 134 java.text.SimpleDateFormat}. Although part of the standard 135 JDK, the performance of <code>SimpleDateFormat</code> is 136 quite poor. 137 138 <p>For better results it is recommended to use the log4j date 139 formatters. These can be specified using one of the strings 140 "ABSOLUTE", "DATE" and "ISO8601" for specifying {@link 141 org.apache.log4j.helpers.AbsoluteTimeDateFormat 142 AbsoluteTimeDateFormat}, {@link 143 org.apache.log4j.helpers.DateTimeDateFormat DateTimeDateFormat} 144 and respectively {@link 145 org.apache.log4j.helpers.ISO8601DateFormat 146 ISO8601DateFormat}. For example, <b>%d{ISO8601}</b> or 147 <b>%d{ABSOLUTE}</b>. 148 149 <p>These dedicated date formatters perform significantly 150 better than {@link java.text.SimpleDateFormat}. 151 </td> 152 </tr> 153 154 <tr> 155 <td align=center><b>F</b></td> 156 157 <td>Used to output the file name where the logging request was 158 issued. 159 160 <p><b>WARNING</b> Generating caller location information is 161 extremely slow and should be avoided unless execution speed 162 is not an issue. 163 164 </tr> 165 166 <tr> 167 <td align=center><b>l</b></td> 168 169 <td>Used to output location information of the caller which generated 170 the logging event. 171 172 <p>The location information depends on the JVM implementation but 173 usually consists of the fully qualified name of the calling 174 method followed by the callers source the file name and line 175 number between parentheses. 176 177 <p>The location information can be very useful. However, its 178 generation is <em>extremely</em> slow and should be avoided 179 unless execution speed is not an issue. 180 181 </td> 182 </tr> 183 184 <tr> 185 <td align=center><b>L</b></td> 186 187 <td>Used to output the line number from where the logging request 188 was issued. 189 190 <p><b>WARNING</b> Generating caller location information is 191 extremely slow and should be avoided unless execution speed 192 is not an issue. 193 194 </tr> 195 196 197 <tr> 198 <td align=center><b>m</b></td> 199 <td>Used to output the application supplied message associated with 200 the logging event.</td> 201 </tr> 202 203 <tr> 204 <td align=center><b>M</b></td> 205 206 <td>Used to output the method name where the logging request was 207 issued. 208 209 <p><b>WARNING</b> Generating caller location information is 210 extremely slow and should be avoided unless execution speed 211 is not an issue. 212 213 </tr> 214 215 <tr> 216 <td align=center><b>n</b></td> 217 218 <td>Outputs the platform dependent line separator character or 219 characters. 220 221 <p>This conversion character offers practically the same 222 performance as using non-portable line separator strings such as 223 "\n", or "\r\n". Thus, it is the preferred way of specifying a 224 line separator. 225 226 227 </tr> 228 229 <tr> 230 <td align=center><b>p</b></td> 231 <td>Used to output the priority of the logging event.</td> 232 </tr> 233 234 <tr> 235 236 <td align=center><b>r</b></td> 237 238 <td>Used to output the number of milliseconds elapsed from the construction 239 of the layout until the creation of the logging event.</td> 240 </tr> 241 242 243 <tr> 244 <td align=center><b>t</b></td> 245 246 <td>Used to output the name of the thread that generated the 247 logging event.</td> 248 249 </tr> 250 251 <tr> 252 253 <td align=center><b>x</b></td> 254 255 <td>Used to output the NDC (nested diagnostic context) associated 256 with the thread that generated the logging event. 257 </td> 258 </tr> 259 260 261 <tr> 262 <td align=center><b>X</b></td> 263 264 <td> 265 266 <p>Used to output the MDC (mapped diagnostic context) associated 267 with the thread that generated the logging event. The <b>X</b> 268 conversion character <em>must</em> be followed by the key for the 269 map placed between braces, as in <b>%X{clientNumber}</b> where 270 <code>clientNumber</code> is the key. The value in the MDC 271 corresponding to the key will be output.</p> 272 273 <p>See {@link MDC} class for more details. 274 </p> 275 276 </td> 277 </tr> 278 279 <tr> 280 281 <td align=center><b>%</b></td> 282 283 <td>The sequence %% outputs a single percent sign. 284 </td> 285 </tr> 286 287 </table> 288 289 <p>By default the relevant information is output as is. However, 290 with the aid of format modifiers it is possible to change the 291 minimum field width, the maximum field width and justification. 292 293 <p>The optional format modifier is placed between the percent sign 294 and the conversion character. 295 296 <p>The first optional format modifier is the <em>left justification 297 flag</em> which is just the minus (-) character. Then comes the 298 optional <em>minimum field width</em> modifier. This is a decimal 299 constant that represents the minimum number of characters to 300 output. If the data item requires fewer characters, it is padded on 301 either the left or the right until the minimum width is 302 reached. The default is to pad on the left (right justify) but you 303 can specify right padding with the left justification flag. The 304 padding character is space. If the data item is larger than the 305 minimum field width, the field is expanded to accommodate the 306 data. The value is never truncated. 307 308 <p>This behavior can be changed using the <em>maximum field 309 width</em> modifier which is designated by a period followed by a 310 decimal constant. If the data item is longer than the maximum 311 field, then the extra characters are removed from the 312 <em>beginning</em> of the data item and not from the end. For 313 example, it the maximum field width is eight and the data item is 314 ten characters long, then the first two characters of the data item 315 are dropped. This behavior deviates from the printf function in C 316 where truncation is done from the end. 317 318 <p>Below are various format modifier examples for the category 319 conversion specifier. 320 321 <p> 322 <TABLE BORDER=1 CELLPADDING=8> 323 <th>Format modifier 324 <th>left justify 325 <th>minimum width 326 <th>maximum width 327 <th>comment 328 329 <tr> 330 <td align=center>%20c</td> 331 <td align=center>false</td> 332 <td align=center>20</td> 333 <td align=center>none</td> 334 335 <td>Left pad with spaces if the category name is less than 20 336 characters long. 337 338 <tr> <td align=center>%-20c</td> <td align=center>true</td> <td 339 align=center>20</td> <td align=center>none</td> <td>Right pad with 340 spaces if the category name is less than 20 characters long. 341 342 <tr> 343 <td align=center>%.30c</td> 344 <td align=center>NA</td> 345 <td align=center>none</td> 346 <td align=center>30</td> 347 348 <td>Truncate from the beginning if the category name is longer than 30 349 characters. 350 351 <tr> 352 <td align=center>%20.30c</td> 353 <td align=center>false</td> 354 <td align=center>20</td> 355 <td align=center>30</td> 356 357 <td>Left pad with spaces if the category name is shorter than 20 358 characters. However, if category name is longer than 30 characters, 359 then truncate from the beginning. 360 361 <tr> 362 <td align=center>%-20.30c</td> 363 <td align=center>true</td> 364 <td align=center>20</td> 365 <td align=center>30</td> 366 367 <td>Right pad with spaces if the category name is shorter than 20 368 characters. However, if category name is longer than 30 characters, 369 then truncate from the beginning. 370 371 </table> 372 373 <p>Below are some examples of conversion patterns. 374 375 <dl> 376 377 <p><dt><b>%r [%t] %-5p %c %x - %m%n</b> 378 <p><dd>This is essentially the TTCC layout. 379 380 <p><dt><b>%-6r [%15.15t] %-5p %30.30c %x - %m%n</b> 381 382 <p><dd>Similar to the TTCC layout except that the relative time is 383 right padded if less than 6 digits, thread name is right padded if 384 less than 15 characters and truncated if longer and the category 385 name is left padded if shorter than 30 characters and truncated if 386 longer. 387 388 </dl> 389 390 <p>The above text is largely inspired from Peter A. Darnell and 391 Philip E. Margolis' highly recommended book "C -- a Software 392 Engineering Approach", ISBN 0-387-97389-3. 393 394 @author <a href="mailto:cakalijp@Maritz.com">James P. Cakalic</a> 395 @author Ceki Gülcü 396 397 398 @since 0.8.2 */ 399 public class PatternLayout extends Layout { 400 401 402 /** Default pattern string for log output. Currently set to the 403 string <b>"%m%n"</b> which just prints the application supplied 404 message. */ 405 public final static String DEFAULT_CONVERSION_PATTERN ="%m%n"; 406 407 /** A conversion pattern equivalent to the TTCCCLayout. 408 Current value is <b>%r [%t] %p %c %x - %m%n</b>. */ 409 public final static String TTCC_CONVERSION_PATTERN 410 = "%r [%t] %p %c %x - %m%n"; 411 412 413 protected final int BUF_SIZE = 256; 414 protected final int MAX_CAPACITY = 1024; 415 416 417 // output buffer appended to when format() is invoked 418 private StringBuffer sbuf = new StringBuffer(BUF_SIZE); 419 420 private String pattern; 421 422 private PatternConverter head; 423 424 /** 425 Constructs a PatternLayout using the DEFAULT_LAYOUT_PATTERN. 426 427 The default pattern just produces the application supplied message. 428 */ 429 public PatternLayout() { 430 this(DEFAULT_CONVERSION_PATTERN); 431 } 432 433 /** 434 Constructs a PatternLayout using the supplied conversion pattern. 435 */ 436 public PatternLayout(String pattern) { 437 this.pattern = pattern; 438 head = createPatternParser((pattern == null) ? DEFAULT_CONVERSION_PATTERN : 439 pattern).parse(); 440 } 441 442 /** 443 Set the <b>ConversionPattern</b> option. This is the string which 444 controls formatting and consists of a mix of literal content and 445 conversion specifiers. 446 */ 447 public 448 void setConversionPattern(String conversionPattern) { 449 pattern = conversionPattern; 450 head = createPatternParser(conversionPattern).parse(); 451 } 452 453 /** 454 Returns the value of the <b>ConversionPattern</b> option. 455 */ 456 public 457 String getConversionPattern() { 458 return pattern; 459 } 460 461 /** 462 Does not do anything as options become effective 463 */ 464 public 465 void activateOptions() { 466 // nothing to do. 467 } 468 469 /** 470 The PatternLayout does not handle the throwable contained within 471 {@link LoggingEvent LoggingEvents}. Thus, it returns 472 <code>true</code>. 473 474 @since 0.8.4 */ 475 public 476 boolean ignoresThrowable() { 477 return true; 478 } 479 480 /** 481 Returns PatternParser used to parse the conversion string. Subclasses 482 may override this to return a subclass of PatternParser which recognize 483 custom conversion characters. 484 485 @since 0.9.0 486 */ 487 protected PatternParser createPatternParser(String pattern) { 488 return new PatternParser(pattern); 489 } 490 491 492 /** 493 Produces a formatted string as specified by the conversion pattern. 494 */ 495 public String format(LoggingEvent event) { 496 // Reset working stringbuffer 497 if(sbuf.capacity() > MAX_CAPACITY) { 498 sbuf = new StringBuffer(BUF_SIZE); 499 } else { 500 sbuf.setLength(0); 501 } 502 503 PatternConverter c = head; 504 505 while(c != null) { 506 c.format(sbuf, event); 507 c = c.next; 508 } 509 return sbuf.toString(); 510 } 511 }