| 1 |
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> |
| 2 |
<html xmlns="http://www.w3.org/1999/xhtml"> |
| 3 |
<!-- |
| 4 |
|
| 5 |
Copyright (c) 2006, Sun Microsystems, Inc. |
| 6 |
All rights reserved. |
| 7 |
|
| 8 |
Redistribution and use in source and binary forms, with or without |
| 9 |
modification, are permitted provided that the following conditions are met: |
| 10 |
|
| 11 |
* Redistributions of source code must retain the above copyright notice, |
| 12 |
this list of conditions and the following disclaimer. |
| 13 |
* Redistributions in binary form must reproduce the above copyright |
| 14 |
notice, this list of conditions and the following disclaimer in the |
| 15 |
documentation and/or other materials provided with the distribution. |
| 16 |
* Neither the name of the Sun Microsystems, Inc. nor the names of its |
| 17 |
contributors may be used to endorse or promote products derived from |
| 18 |
this software without specific prior written permission. |
| 19 |
|
| 20 |
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" |
| 21 |
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE |
| 22 |
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE |
| 23 |
ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE |
| 24 |
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR |
| 25 |
CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF |
| 26 |
SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS |
| 27 |
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN |
| 28 |
CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) |
| 29 |
ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF |
| 30 |
THE POSSIBILITY OF SUCH DAMAGE. |
| 31 |
|
| 32 |
--> |
| 33 |
<head> |
| 34 |
<title>JavaCC Grammar Files</title> |
| 35 |
<!-- Changed by: Michael Van De Vanter, 14-Jan-2003 --> |
| 36 |
</head> |
| 37 |
<body bgcolor="#FFFFFF" > |
| 38 |
|
| 39 |
<h1>JavaCC [tm]: Grammar Files</h1> |
| 40 |
<p> |
| 41 |
This page contains the complete syntax of Java Compiler Compiler [tm] |
| 42 |
grammar files with detailed explanations of each construct. |
| 43 |
</p> |
| 44 |
<p> |
| 45 |
Tokens in the grammar files follow the same conventions as for the Java programming language. |
| 46 |
Hence identifiers, strings, characters, etc. used in the grammars are |
| 47 |
the same as Java identifiers, Java strings, Java characters, etc. |
| 48 |
</p> |
| 49 |
<p> |
| 50 |
<em>White space</em> in the grammar files also follows the same conventions as |
| 51 |
for the Java programming language. This includes the syntax for comments. Most comments present in |
| 52 |
the grammar files are generated into the generated parser/lexical analyzer. |
| 53 |
</p> |
| 54 |
<p> |
| 55 |
Grammar files are preprocessed for Unicode escapes just as Java files |
| 56 |
are (i.e., occurrences of strings such as <code>\uxxxx</code> - where <code>xxxx</code> is a hex value - |
| 57 |
are converted to the corresponding Unicode character before lexical analysis). |
| 58 |
</p> |
| 59 |
<p> |
| 60 |
<em>Exceptions to the above rules:</em> |
| 61 |
The Java operators "<code><<</code>", "<code>>></code>", "<code>>>></code>", "<code><<=</code>", |
| 62 |
"<code>>>=</code>", and "<code>>>>=</code>" are left out of JavaCC's input token list |
| 63 |
in order to allow convenient nested use of token specifications. |
| 64 |
Finally, the following are the additional reserved words in the Java Compiler |
| 65 |
Compiler [tm] grammar files. |
| 66 |
</p> |
| 67 |
|
| 68 |
<table cellpadding="3"> |
| 69 |
<tr> |
| 70 |
<td align="left"><strong>EOF</strong></td> |
| 71 |
<td align="left"><strong><a href="#IGNORE_CASE">IGNORE_CASE</a></strong></td> |
| 72 |
<td align="left"><strong><a href="#JAVACODE">JAVACODE</a></strong></td> |
| 73 |
<td align="left"><strong><a href="#LOOKAHEAD">LOOKAHEAD</a></strong></td> |
| 74 |
</tr> |
| 75 |
<tr> |
| 76 |
<td align="left"><strong><a href="#MORE">MORE</a></strong></td> |
| 77 |
<td align="left"><strong><a href="#PARSER_BEGIN">PARSER_BEGIN</a></strong></td> |
| 78 |
<td align="left"><strong><a href="#PARSER_END">PARSER_END</a></strong></td> |
| 79 |
<td align="left"><strong><a href="#SKIP">SKIP</a></strong></td> |
| 80 |
</tr> |
| 81 |
<tr> |
| 82 |
<td align="left"><strong><a href="#SPECIAL_TOKEN">SPECIAL_TOKEN</a></strong></td> |
| 83 |
<td align="left"><strong><a href="#TOKEN">TOKEN</a></strong></td> |
| 84 |
<td align="left"><strong><a href="#TOKEN_MGR_DECLS">TOKEN_MGR_DECLS</a></strong></td> |
| 85 |
</tr> |
| 86 |
</table> |
| 87 |
|
| 88 |
<p> |
| 89 |
Any Java entities used in the grammar rules that follow appear italicized |
| 90 |
with the prefix <em>java_</em> (<em>e.g.</em>, <em>java_compilation_unit</em>). |
| 91 |
</p> |
| 92 |
|
| 93 |
<hr /> |
| 94 |
|
| 95 |
<a name="PARSER_BEGIN"></a><a name="PARSER_END"></a> |
| 96 |
<table> |
| 97 |
<tr> |
| 98 |
<td align="right" valign="baseline"><a name="prod1">javacc_input</a></td> |
| 99 |
<td align="center" valign="baseline">::=</td> |
| 100 |
<td align="left" valign="baseline"><a href="#prod2">javacc_options</a></td> |
| 101 |
</tr> |
| 102 |
<tr> |
| 103 |
<td></td><td></td> |
| 104 |
<td align="left" valign="baseline">"PARSER_BEGIN" "(" <IDENTIFIER> ")"</td> |
| 105 |
</tr> |
| 106 |
<tr> |
| 107 |
<td></td><td></td> |
| 108 |
<td align="left" valign="baseline"><em>java_compilation_unit</em></td> |
| 109 |
</tr> |
| 110 |
<tr> |
| 111 |
<td></td><td></td> |
| 112 |
<td align="left" valign="baseline">"PARSER_END" "(" <IDENTIFIER> ")"</td> |
| 113 |
</tr> |
| 114 |
<tr> |
| 115 |
<td></td><td></td> |
| 116 |
<td align="left" valign="baseline">( <a href="#prod5">production</a> )*</td> |
| 117 |
</tr> |
| 118 |
<tr> |
| 119 |
<td></td><td></td> |
| 120 |
<td align="left" valign="baseline"><EOF></td> |
| 121 |
</tr> |
| 122 |
</table> |
| 123 |
|
| 124 |
<p> |
| 125 |
The grammar file starts with a list of options (which is optional). |
| 126 |
This is then followed by a Java compilation unit enclosed between |
| 127 |
"PARSER_BEGIN(name)" and "PARSER_END(name)". After this is a list |
| 128 |
of grammar productions. <a href="#prod2">Options</a> and |
| 129 |
<a href="#prod5">productions</a> are described later. |
| 130 |
</p> |
| 131 |
<p> |
| 132 |
The <em>name</em> that follows "PARSER_BEGIN" and "PARSER_END" must |
| 133 |
be the same and this identifies the name of the generated parser. |
| 134 |
For example, if <em>name</em> is "MyParser", then the following files |
| 135 |
are generated: |
| 136 |
</p> |
| 137 |
<p> |
| 138 |
<strong>MyParser.java:</strong> |
| 139 |
The generate parser. |
| 140 |
<br /> |
| 141 |
<strong>MyParserTokenManager.java:</strong> |
| 142 |
The generated token manager (or scanner/lexical analyzer). |
| 143 |
<br /> |
| 144 |
<strong>MyParserConstants.java:</strong> |
| 145 |
A bunch of useful constants. |
| 146 |
</p> |
| 147 |
<p> |
| 148 |
Other files such as "Token.java", "ParseException.java", etc. are also |
| 149 |
generated. However, these files contain boilerplate code and are |
| 150 |
the same for any grammar and may be reused across grammars (provided the |
| 151 |
grammars use compatible options). |
| 152 |
</p> |
| 153 |
<p> |
| 154 |
Between the PARSER_BEGIN and PARSER_END constructs is a regular |
| 155 |
Java compilation unit (a compilation unit in Java lingo is the entire |
| 156 |
contents of a Java file). This may be any arbitrary |
| 157 |
Java compilation unit so long as it contains a class declaration |
| 158 |
whose name is the same as the name of the generated parser ("MyParser" |
| 159 |
in the above example). Hence, in general, this part of the grammar |
| 160 |
file looks like: |
| 161 |
</p> |
| 162 |
<pre> |
| 163 |
PARSER_BEGIN(parser_name) |
| 164 |
. . . |
| 165 |
class parser_name . . . { |
| 166 |
. . . |
| 167 |
} |
| 168 |
. . . |
| 169 |
PARSER_END(parser_name) |
| 170 |
</pre> |
| 171 |
<p> |
| 172 |
JavaCC does not perform detailed checks on the compilation unit, so |
| 173 |
it is possible for a grammar file to pass through JavaCC and generate |
| 174 |
Java files that produce errors when they are compiled. |
| 175 |
</p> |
| 176 |
<p> |
| 177 |
If the compilation unit includes a package declaration, this is |
| 178 |
included in all the generated files. If the compilation unit includes |
| 179 |
imports declarations, this is included in the generated parser and |
| 180 |
token manager files. |
| 181 |
</p> |
| 182 |
<p> |
| 183 |
The generated parser file contains everything in the compilation unit |
| 184 |
and, in addition, contains the generated parser code that is included at |
| 185 |
the end of the parser class. For the above example, the generated |
| 186 |
parser will look like: |
| 187 |
</p> |
| 188 |
<pre> |
| 189 |
. . . |
| 190 |
class parser_name . . . { |
| 191 |
. . . |
| 192 |
// generated parser is inserted here. |
| 193 |
} |
| 194 |
. . . |
| 195 |
</pre> |
| 196 |
<p> |
| 197 |
The generated parser includes a public method declaration corresponding |
| 198 |
to each non-terminal (see <a href="#prod9">javacode_production</a> and |
| 199 |
<a href="#prod11">bnf_production</a>) in the grammar file. Parsing with |
| 200 |
respect to a non-terminal is achieved by calling the method corresponding |
| 201 |
to that non-terminal. Unlike yacc, there is no single start symbol in |
| 202 |
JavaCC - one can parse with respect to any non-terminal in the grammar. |
| 203 |
</p> |
| 204 |
<p> |
| 205 |
The generated token manager provides one public method: |
| 206 |
</p> |
| 207 |
<pre> |
| 208 |
Token getNextToken() throws ParseError; |
| 209 |
</pre> |
| 210 |
<p> |
| 211 |
For more details on how this method may be used, please read |
| 212 |
<a href="apiroutines.html">the description of the Java Compiler Compiler |
| 213 |
API</a>. |
| 214 |
</p> |
| 215 |
|
| 216 |
<hr /> |
| 217 |
|
| 218 |
<table> |
| 219 |
<tr> |
| 220 |
<td align="right" valign="baseline"><a name="prod2">javacc_options</a></td> |
| 221 |
<td align="center" valign="baseline">::=</td> |
| 222 |
<td align="left" valign="baseline">[ "<a name="options">options</a>" "{" ( <a href="#prod6">option_binding</a> )* "}" ]</td> |
| 223 |
</tr> |
| 224 |
</table> |
| 225 |
|
| 226 |
<p> |
| 227 |
The options if present, starts with the reserved word "options" followed |
| 228 |
by a list of one or more option bindings within braces. Each option |
| 229 |
binding specifies the setting of one option. The same option may not be |
| 230 |
set multiple times. |
| 231 |
</p> |
| 232 |
<p> |
| 233 |
Options may be specified either here in the grammar file, or from |
| 234 |
<a href="commandline.html">the command line</a>. If the option is set |
| 235 |
from <a href="commandline.html">the command line</a>, that takes precedence. |
| 236 |
</p> |
| 237 |
<p> |
| 238 |
Option names are not case-sensitive. |
| 239 |
</p> |
| 240 |
|
| 241 |
<hr /> |
| 242 |
|
| 243 |
<table> |
| 244 |
<tr> |
| 245 |
<td align="right" valign="baseline"><a name="prod6">option_binding</a></td> |
| 246 |
<td align="center" valign="baseline">::=</td> |
| 247 |
<td align="left" valign="baseline">"LOOKAHEAD" "=" <em>java_integer_literal</em> ";"</td> |
| 248 |
</tr> |
| 249 |
<tr> |
| 250 |
<td></td><td align="center" valign="baseline">|</td> |
| 251 |
<td align="left" valign="baseline">"CHOICE_AMBIGUITY_CHECK" "=" <em>java_integer_literal</em> ";"</td> |
| 252 |
</tr> |
| 253 |
<tr> |
| 254 |
<td></td><td align="center" valign="baseline">|</td> |
| 255 |
<td align="left" valign="baseline">"OTHER_AMBIGUITY_CHECK" "=" <em>java_integer_literal</em> ";"</td> |
| 256 |
</tr> |
| 257 |
<tr> |
| 258 |
<td></td><td align="center" valign="baseline">|</td> |
| 259 |
<td align="left" valign="baseline">"STATIC" "=" <em>java_boolean_literal</em> ";"</td> |
| 260 |
</tr> |
| 261 |
<tr> |
| 262 |
<td></td><td align="center" valign="baseline">|</td> |
| 263 |
<td align="left" valign="baseline">"SUPPORT_CLASS_VISIBILITY_PUBLIC" "=" <em>java_boolean_literal</em> ";"</td> |
| 264 |
</tr> |
| 265 |
<tr> |
| 266 |
<td></td><td align="center" valign="baseline">|</td> |
| 267 |
<td align="left" valign="baseline">"DEBUG_PARSER" "=" <em>java_boolean_literal</em> ";"</td> |
| 268 |
</tr> |
| 269 |
<tr> |
| 270 |
<td></td><td align="center" valign="baseline">|</td> |
| 271 |
<td align="left" valign="baseline">"DEBUG_LOOKAHEAD" "=" <em>java_boolean_literal</em> ";"</td> |
| 272 |
</tr> |
| 273 |
<tr> |
| 274 |
<td></td><td align="center" valign="baseline">|</td> |
| 275 |
<td align="left" valign="baseline">"DEBUG_TOKEN_MANAGER" "=" <em>java_boolean_literal</em> ";"</td> |
| 276 |
</tr> |
| 277 |
<tr> |
| 278 |
<td></td><td align="center" valign="baseline">|</td> |
| 279 |
<td align="left" valign="baseline">"ERROR_REPORTING" "=" <em>java_boolean_literal</em> ";"</td> |
| 280 |
</tr> |
| 281 |
<tr> |
| 282 |
<td></td><td align="center" valign="baseline">|</td> |
| 283 |
<td align="left" valign="baseline">"JAVA_UNICODE_ESCAPE" "=" <em>java_boolean_literal</em> ";"</td> |
| 284 |
</tr> |
| 285 |
<tr> |
| 286 |
<td></td><td align="center" valign="baseline">|</td> |
| 287 |
<td align="left" valign="baseline">"UNICODE_INPUT" "=" <em>java_boolean_literal</em> ";"</td> |
| 288 |
</tr> |
| 289 |
<tr> |
| 290 |
<td></td><td align="center" valign="baseline">|</td> |
| 291 |
<td align="left" valign="baseline">"IGNORE_CASE" "=" <em>java_boolean_literal</em> ";"</td> |
| 292 |
</tr> |
| 293 |
<tr> |
| 294 |
<td></td><td align="center" valign="baseline">|</td> |
| 295 |
<td align="left" valign="baseline">"USER_TOKEN_MANAGER" "=" <em>java_boolean_literal</em> ";"</td> |
| 296 |
</tr> |
| 297 |
<tr> |
| 298 |
<td></td><td align="center" valign="baseline">|</td> |
| 299 |
<td align="left" valign="baseline">"USER_CHAR_STREAM" "=" <em>java_boolean_literal</em> ";"</td> |
| 300 |
</tr> |
| 301 |
<tr> |
| 302 |
<td></td><td align="center" valign="baseline">|</td> |
| 303 |
<td align="left" valign="baseline">"BUILD_PARSER" "=" <em>java_boolean_literal</em> ";"</td> |
| 304 |
</tr> |
| 305 |
<tr> |
| 306 |
<td></td><td align="center" valign="baseline">|</td> |
| 307 |
<td align="left" valign="baseline">"BUILD_TOKEN_MANAGER" "=" <em>java_boolean_literal</em> ";"</td> |
| 308 |
</tr> |
| 309 |
<tr> |
| 310 |
<td></td><td align="center" valign="baseline">|</td> |
| 311 |
<td align="left" valign="baseline">"TOKEN_EXTENDS" "=" <em>java_string_literal</em> ";"</td> |
| 312 |
</tr> |
| 313 |
<tr> |
| 314 |
<td></td><td align="center" valign="baseline">|</td> |
| 315 |
<td align="left" valign="baseline">"TOKEN_FACTORY" "=" <em>java_string_literal</em> ";"</td> |
| 316 |
</tr> |
| 317 |
<tr> |
| 318 |
<td></td><td align="center" valign="baseline">|</td> |
| 319 |
<td align="left" valign="baseline">"TOKEN_MANAGER_USES_PARSER" "=" <em>java_boolean_literal</em> ";"</td> |
| 320 |
</tr> |
| 321 |
<tr> |
| 322 |
<td></td><td align="center" valign="baseline">|</td> |
| 323 |
<td align="left" valign="baseline">"SANITY_CHECK" "=" <em>java_boolean_literal</em> ";"</td> |
| 324 |
</tr> |
| 325 |
<tr> |
| 326 |
<td></td><td align="center" valign="baseline">|</td> |
| 327 |
<td align="left" valign="baseline">"FORCE_LA_CHECK" "=" <em>java_boolean_literal</em> ";"</td> |
| 328 |
</tr> |
| 329 |
<tr> |
| 330 |
<td></td><td align="center" valign="baseline">|</td> |
| 331 |
<td align="left" valign="baseline">"COMMON_TOKEN_ACTION" "=" <em>java_boolean_literal</em> ";"</td> |
| 332 |
</tr> |
| 333 |
<tr> |
| 334 |
<td></td><td align="center" valign="baseline">|</td> |
| 335 |
<td align="left" valign="baseline">"CACHE_TOKENS" "=" <em>java_boolean_literal</em> ";"</td> |
| 336 |
</tr> |
| 337 |
<tr> |
| 338 |
<td></td><td align="center" valign="baseline">|</td> |
| 339 |
<td align="left" valign="baseline">"OUTPUT_DIRECTORY" "=" <em>java_string_literal</em> ";"</td> |
| 340 |
</tr> |
| 341 |
</table> |
| 342 |
|
| 343 |
<ul> |
| 344 |
<li> |
| 345 |
<strong><a name="LOOKAHEAD">LOOKAHEAD</a>:</strong> |
| 346 |
The number of tokens to look ahead before making a |
| 347 |
decision at a choice point during parsing. The default value is 1. |
| 348 |
The smaller this number, the faster the parser. This number may be |
| 349 |
overridden for specific productions within the grammar as described |
| 350 |
later. See the description of |
| 351 |
<a href="lookahead.html">the lookahead algorithm</a> for complete |
| 352 |
details on how lookahead works. |
| 353 |
</li> |
| 354 |
<li> |
| 355 |
<strong>CHOICE_AMBIGUITY_CHECK:</strong> |
| 356 |
This is an integer option whose default value is 2. |
| 357 |
This is the number of tokens considered in checking choices of the |
| 358 |
form "A | B | ..." for ambiguity. For example, if there is a common |
| 359 |
two token prefix for both A and B, but no common three token prefix, |
| 360 |
(assume this option is set to 3) then JavaCC can tell you to use a |
| 361 |
lookahead of 3 for disambiguation purposes. And if A and B have a |
| 362 |
common three token prefix, then JavaCC only tell you that you need to |
| 363 |
have a lookahead of 3 <em>or more</em>. Increasing this can give you more |
| 364 |
comprehensive ambiguity information at the cost of more processing |
| 365 |
time. For large grammars such as the Java grammar, increasing this number |
| 366 |
any further causes the checking to take too much time. |
| 367 |
</li> |
| 368 |
<li> |
| 369 |
<strong>OTHER_AMBIGUITY_CHECK:</strong> |
| 370 |
This is an integer option whose default value is 1. |
| 371 |
This is the number of tokens considered in checking all other kinds of |
| 372 |
choices (i.e., of the forms "(A)*", "(A)+", and "(A)?") for ambiguity. |
| 373 |
This takes more time to do than the choice checking, and hence the |
| 374 |
default value is set to 1 rather than 2. |
| 375 |
</li> |
| 376 |
<li> |
| 377 |
<strong>STATIC:</strong> |
| 378 |
This is a boolean option whose default value is true. If |
| 379 |
true, all methods and class variables are specified as static in the |
| 380 |
generated parser and token manager. This allows only one parser object to be present, |
| 381 |
but it improves the performance of the parser. To perform multiple |
| 382 |
parses during one run of your Java program, you will have to call the |
| 383 |
<a href="apiroutines.html">ReInit()</a> |
| 384 |
method to reinitialize your parser if it is static. |
| 385 |
If the parser is non-static, you may use the "new" operator to |
| 386 |
construct as many parsers as you wish. These can all be used |
| 387 |
simultaneously from different threads. |
| 388 |
</li> |
| 389 |
<li> |
| 390 |
<strong>DEBUG_PARSER:</strong> |
| 391 |
This is a boolean option whose default value is false. This |
| 392 |
option is used to obtain debugging information from the generated |
| 393 |
parser. Setting this option to true causes the parser to generate |
| 394 |
a trace of its actions. Tracing may be disabled by |
| 395 |
calling the method <a href="apiroutines.html">disable_tracing()</a> |
| 396 |
in the generated parser class. Tracing may be subsequently enabled |
| 397 |
by calling the method <a href="apiroutines.html">enable_tracing()</a> |
| 398 |
in the generated parser class. |
| 399 |
</li> |
| 400 |
<li> |
| 401 |
<strong>DEBUG_LOOKAHEAD:</strong> |
| 402 |
This is a boolean option whose default value is false. Setting this |
| 403 |
option to true causes the parser to generate all the tracing information |
| 404 |
it does when the option DEBUG_PARSER is true, and in addition, also |
| 405 |
causes it to generated a trace of actions performed during |
| 406 |
<a href="lookahead.html">lookahead operation</a>. |
| 407 |
</li> |
| 408 |
<li> |
| 409 |
<strong>DEBUG_TOKEN_MANAGER:</strong> |
| 410 |
This is a boolean option whose default value is false. This |
| 411 |
option is used to obtain debugging information from the generated |
| 412 |
token manager. Setting this option to true causes the token manager to generate |
| 413 |
a trace of its actions. This trace is rather large and should only |
| 414 |
be used when you have a lexical error that has been reported to you |
| 415 |
and you cannot understand why. Typically, in this situation, you |
| 416 |
can determine the problem by looking at the last few lines of this trace. |
| 417 |
</li> |
| 418 |
<li> |
| 419 |
<strong>ERROR_REPORTING:</strong> |
| 420 |
This is a boolean option whose default value is |
| 421 |
true. Setting it to false causes errors due to parse errors to be |
| 422 |
reported in somewhat less detail. The only reason to set this |
| 423 |
option to false is to improve performance. |
| 424 |
</li> |
| 425 |
<li> |
| 426 |
<strong>JAVA_UNICODE_ESCAPE:</strong> |
| 427 |
This is a boolean option whose default value is |
| 428 |
false. When set to true, the generated parser uses |
| 429 |
an input stream object that processes Java Unicode escapes |
| 430 |
(\u...) before sending characters to the token manager. By |
| 431 |
default, Java Unicode escapes are not processed. |
| 432 |
<br /> |
| 433 |
This option is ignored if either of options USER_TOKEN_MANAGER, |
| 434 |
USER_CHAR_STREAM is set to true. |
| 435 |
</li> |
| 436 |
<li> |
| 437 |
<strong>UNICODE_INPUT:</strong> |
| 438 |
This is a boolean option whose default value is |
| 439 |
false. When set to true, the generated parser uses |
| 440 |
uses an input stream object that reads Unicode files. By default, |
| 441 |
ASCII files are assumed. |
| 442 |
<br /> |
| 443 |
This option is ignored if either of |
| 444 |
options USER_TOKEN_MANAGER, USER_CHAR_STREAM is set to true. |
| 445 |
</li> |
| 446 |
<li> |
| 447 |
<strong><a name="IGNORE_CASE">IGNORE_CASE:</a></strong> |
| 448 |
This is a boolean option whose default value is false. |
| 449 |
Setting this option to true causes the generated token manager to ignore |
| 450 |
case in the token specifications and the input files. This is useful |
| 451 |
for writing grammars for languages such as HTML. It is also possible |
| 452 |
to localize the effect of IGNORE_CASE by using |
| 453 |
<a href="#prod10">an alternate mechanism described later</a>. |
| 454 |
</li> |
| 455 |
<li> |
| 456 |
<strong>USER_TOKEN_MANAGER:</strong> |
| 457 |
This is a boolean option whose default value is |
| 458 |
false. The default action is to generate a token manager |
| 459 |
that works on the specified grammar tokens. If this |
| 460 |
option is set to true, then the parser is generated to accept tokens |
| 461 |
from any token manager of type "TokenManager" - this interface |
| 462 |
is generated into the generated parser directory. |
| 463 |
</li> |
| 464 |
<li> |
| 465 |
<strong>SUPPORT_CLASS_VISIBILITY_PUBLIC:</strong> |
| 466 |
This is a boolean option whose default value is |
| 467 |
true. The default action is to generate support classes (such as |
| 468 |
Token.java, ParseException.java etc) with <em>Public</em> visibility. If |
| 469 |
set to false, the classes will be generated with package-private |
| 470 |
visibility. |
| 471 |
</li> |
| 472 |
<li> |
| 473 |
<strong>USER_CHAR_STREAM:</strong> |
| 474 |
This is a boolean option whose default value is |
| 475 |
false. The default action is to generate a character stream reader |
| 476 |
as specified by the options JAVA_UNICODE_ESCAPE and UNICODE_INPUT. |
| 477 |
The generated token manager receives characters |
| 478 |
from this stream reader. If this option is set to true, then the |
| 479 |
token manager is generated to read characters from any character |
| 480 |
stream reader of type "CharStream.java". This file is generated |
| 481 |
into the generated parser directory. |
| 482 |
<br /> |
| 483 |
This option is ignored if USER_TOKEN_MANAGER is set to true. |
| 484 |
</li> |
| 485 |
<li> |
| 486 |
<strong>BUILD_PARSER:</strong> |
| 487 |
This is a boolean option whose default value is true. |
| 488 |
The default action is to generate the parser file ("MyParser.java" |
| 489 |
in the above example). When set to false, the parser file is |
| 490 |
not generated. Typically, this option is set to false when |
| 491 |
you wish to generate only the token manager and use it without |
| 492 |
the associated parser. |
| 493 |
</li> |
| 494 |
<li> |
| 495 |
<strong>BUILD_TOKEN_MANAGER:</strong> |
| 496 |
This is a boolean option whose default value is true. |
| 497 |
The default action is to generate the token manager file |
| 498 |
("MyParserTokenManager.java" in the above example). When set to |
| 499 |
false the token manager file is not generated. The only reason |
| 500 |
to set this option to false is to save some time during parser |
| 501 |
generation when you fix problems in the parser part of the grammar |
| 502 |
file and leave the lexical specifications untouched. |
| 503 |
</li> |
| 504 |
<li> |
| 505 |
<strong>TOKEN_MANAGER_USES_PARSER:</strong> |
| 506 |
This is a boolean option whose default value is false. |
| 507 |
When set to true, the generated token manager will include a field |
| 508 |
called <CODE>parser</CODE> that references the instantiating parser |
| 509 |
instance (of type <CODE>MyParser</CODE> in the above example). |
| 510 |
The main reason for having a parser in a token manager is using |
| 511 |
some of its logic in lexical actions. |
| 512 |
This option has no effect if the STATIC option is set to true. |
| 513 |
</li> |
| 514 |
<li> |
| 515 |
<strong>TOKEN_EXTENDS:</strong> |
| 516 |
This is a string option whose default value is |
| 517 |
"", meaning that the generated Token class will extend |
| 518 |
java.lang.Object. This option may be set to the name of a |
| 519 |
class that will be used as the base class for the generated |
| 520 |
<code>Token</code> class. |
| 521 |
</li> |
| 522 |
<li> |
| 523 |
<strong>TOKEN_FACTORY:</strong> |
| 524 |
This is a string option whose default value is |
| 525 |
"", meaning that Tokens will be created by calling |
| 526 |
<code>Token.newToken()</code>. If set the option names a |
| 527 |
Token factory class containing a |
| 528 |
<code>public static Token newToken(int ofKind, String image)</code> |
| 529 |
method. |
| 530 |
</li> |
| 531 |
<li> |
| 532 |
<strong>SANITY_CHECK:</strong> |
| 533 |
This is a boolean option whose default value is true. |
| 534 |
JavaCC performs many syntactic and semantic checks on the grammar |
| 535 |
file during parser generation. Some checks such as detection of |
| 536 |
left recursion, detection of ambiguity, and bad usage of empty |
| 537 |
expansions may be suppressed for faster parser generation by |
| 538 |
setting this option to false. Note that the presence of these |
| 539 |
errors (even if they are not detected and reported by setting this |
| 540 |
option to false) can cause unexpected behavior from the generated |
| 541 |
parser. |
| 542 |
</li> |
| 543 |
<li> |
| 544 |
<strong>FORCE_LA_CHECK:</strong> |
| 545 |
This is a boolean option whose default value is false. |
| 546 |
This option setting controls lookahead ambiguity checking performed |
| 547 |
by JavaCC. By default (when this option is false), lookahead |
| 548 |
ambiguity checking is performed for all choice points where the |
| 549 |
default lookahead of 1 is used. Lookahead ambiguity checking is |
| 550 |
not performed at choice points where there is an |
| 551 |
<a href="lookahead.html">explicit lookahead specification</a>, |
| 552 |
or if the option LOOKAHEAD is set to something other than 1. |
| 553 |
Setting this option to true performs lookahead ambiguity checking |
| 554 |
at <em>all</em> choice points regardless of the lookahead specifications |
| 555 |
in the grammar file. |
| 556 |
</li> |
| 557 |
<li> |
| 558 |
<strong>COMMON_TOKEN_ACTION:</strong> |
| 559 |
This is a boolean option whose default value is false. |
| 560 |
When set to true, every call to the token manager's method |
| 561 |
"getNextToken" (<a href="apiroutines.html">see the description of the |
| 562 |
Java Compiler Compiler API</a>) will cause a call to a used defined |
| 563 |
method "CommonTokenAction" after the token has been scanned in by the |
| 564 |
token manager. The user must define this method within the |
| 565 |
<a href="#prod12">TOKEN_MGR_DECLS</a> section. |
| 566 |
The signature of this method is: |
| 567 |
<pre> |
| 568 |
void CommonTokenAction(Token t) |
| 569 |
</pre> |
| 570 |
</li> |
| 571 |
<li> |
| 572 |
<strong>CACHE_TOKENS:</strong> |
| 573 |
This is a boolean option whose default value is false. |
| 574 |
Setting this option to true causes the generated parser to lookahead for |
| 575 |
extra tokens ahead of time. This facilitates some performance improvements. |
| 576 |
However, in this case (when the option is true), interactive |
| 577 |
applications may not work since the parser needs to work synchronously |
| 578 |
with the availability of tokens from the input stream. In such cases, |
| 579 |
it's best to leave this option at its default value. |
| 580 |
</li> |
| 581 |
<li> |
| 582 |
<strong>OUTPUT_DIRECTORY:</strong> |
| 583 |
This is a string valued option whose default value is the current |
| 584 |
directory. This controls where output files are generated. |
| 585 |
</li> |
| 586 |
</ul> |
| 587 |
|
| 588 |
<hr /> |
| 589 |
|
| 590 |
<table> |
| 591 |
<tr> |
| 592 |
<td align="right" valign="baseline"><a name="prod5">production</a></td> |
| 593 |
<td align="center" valign="baseline">::=</td> |
| 594 |
<td align="left" valign="baseline"><a href="#prod9">javacode_production</a></td> |
| 595 |
</tr> |
| 596 |
<tr> |
| 597 |
<td align="right" valign="baseline"></td> |
| 598 |
<td align="center" valign="baseline">|</td> |
| 599 |
<td align="left" valign="baseline"><a href="#prod10">regular_expr_production</a></td> |
| 600 |
</tr> |
| 601 |
<tr> |
| 602 |
<td align="right" valign="baseline"></td> |
| 603 |
<td align="center" valign="baseline">|</td> |
| 604 |
<td align="left" valign="baseline"><a href="#prod11">bnf_production</a></td> |
| 605 |
</tr> |
| 606 |
<tr> |
| 607 |
<td align="right" valign="baseline"></td> |
| 608 |
<td align="center" valign="baseline">|</td> |
| 609 |
<td align="left" valign="baseline"><a href="#prod12">token_manager_decls</a></td> |
| 610 |
</tr> |
| 611 |
</table> |
| 612 |
|
| 613 |
<p> |
| 614 |
There are four kinds of productions in JavaCC. |
| 615 |
<a href="#prod9">javacode_production</a> and <a href="#prod11">bnf_production</a> |
| 616 |
are used to define the grammar from which the parser is generated. |
| 617 |
<a href="#prod10">regular_expr_production</a> is used to define the grammar |
| 618 |
tokens - the token manager is generated from this information (as well as from |
| 619 |
inline token specifications in the parser grammar). |
| 620 |
<a href="#prod12">token_manager_decls</a> is used to introduce declarations |
| 621 |
that get inserted into the generated token manager. |
| 622 |
</p> |
| 623 |
|
| 624 |
<hr /> |
| 625 |
<table> |
| 626 |
<tr> |
| 627 |
<td align="right" valign="baseline"><a name="prod9">javacode_production</a></td> |
| 628 |
<td align="center" valign="baseline">::=</td> |
| 629 |
<td align="left" valign="baseline">"<a name="JAVACODE">JAVACODE</a>"</td> |
| 630 |
</tr> |
| 631 |
<tr> |
| 632 |
<td></td><td></td> |
| 633 |
<td align="left" valign="baseline"><em>java_access_modifier</em> <em>java_return_type</em> <em>java_identifier</em> "(" <em>java_parameter_list</em> ")"</td> |
| 634 |
</tr> |
| 635 |
<tr> |
| 636 |
<td></td><td></td> |
| 637 |
<td align="left" valign="baseline"><em>java_block</em></td> |
| 638 |
</tr> |
| 639 |
</table> |
| 640 |
|
| 641 |
<p> |
| 642 |
The JAVACODE production is a way to write Java code for some |
| 643 |
productions instead of the usual EBNF expansion. This is useful when |
| 644 |
there is the need to recognize something that is not context-free |
| 645 |
or for whatever reason is very difficult to write a grammar for. |
| 646 |
An example of the use of JAVACODE is shown below. In this example, |
| 647 |
the non-terminal "skip_to_matching_brace" consumes tokens in the input |
| 648 |
stream all the way up to a matching closing brace (the opening brace |
| 649 |
is assumed to have been just scanned): |
| 650 |
</p> |
| 651 |
<pre> |
| 652 |
JAVACODE |
| 653 |
void skip_to_matching_brace() { |
| 654 |
<a href="apiroutines.html">Token</a> tok; |
| 655 |
int nesting = 1; |
| 656 |
while (true) { |
| 657 |
tok = <a href="apiroutines.html">getToken</a>(1); |
| 658 |
if (tok.kind == LBRACE) nesting++; |
| 659 |
if (tok.kind == RBRACE) { |
| 660 |
nesting--; |
| 661 |
if (nesting == 0) break; |
| 662 |
} |
| 663 |
tok = <a href="apiroutines.html">getNextToken</a>(); |
| 664 |
} |
| 665 |
} |
| 666 |
</pre> |
| 667 |
<p> |
| 668 |
Care must be taken when using JAVACODE productions. While you can |
| 669 |
say pretty much what you want with these productions, JavaCC simply |
| 670 |
considers it a black box (that somehow performs its parsing task). |
| 671 |
This becomes a problem when JAVACODE productions appear at |
| 672 |
<a href="lookahead.html">choice points</a>. For example, if the |
| 673 |
above JAVACODE production was referred to from the following production: |
| 674 |
</p> |
| 675 |
<pre> |
| 676 |
void NT() : |
| 677 |
{} |
| 678 |
{ |
| 679 |
skip_to_matching_brace() |
| 680 |
| |
| 681 |
some_other_production() |
| 682 |
} |
| 683 |
</pre> |
| 684 |
<p> |
| 685 |
Then JavaCC would not know how to choose between the two choices. |
| 686 |
On the other hand, if the JAVACODE production is used at a non-choice |
| 687 |
point as in the following example, there is no problem: |
| 688 |
</p> |
| 689 |
<pre> |
| 690 |
void NT() : |
| 691 |
{} |
| 692 |
{ |
| 693 |
"{" skip_to_matching_brace() |
| 694 |
| |
| 695 |
"(" parameter_list() ")" |
| 696 |
} |
| 697 |
</pre> |
| 698 |
<p> |
| 699 |
JAVACODE productions at choice points may also be preceded by syntactic or |
| 700 |
semantic LOOKAHEAD, as in this example: |
| 701 |
</p> |
| 702 |
<pre> |
| 703 |
void NT() : |
| 704 |
{} |
| 705 |
{ |
| 706 |
LOOKAHEAD( {errorOccurred} ) skip_to_matching_brace() |
| 707 |
| |
| 708 |
"(" parameter_list() ")" |
| 709 |
} |
| 710 |
</pre> |
| 711 |
<!-- JavaCC *should* print a warning message, but currently doesn't (see issue 166). |
| 712 |
<p> |
| 713 |
When this issue is fixed ww should re-instate this paragraph. |
| 714 |
When JAVACODE productions are used at choice points, JavaCC will |
| 715 |
print a warning message stating this fact. You will then have to |
| 716 |
insert some explicit LOOKAHEAD specifications to help JavaCC. See |
| 717 |
<a href="lookahead.html">the minitutorial on LOOKAHEAD</a> for a |
| 718 |
detailed guide on such issues. |
| 719 |
</p> |
| 720 |
--> |
| 721 |
<p> |
| 722 |
The default access modifier for JAVACODE productions is package private. |
| 723 |
</p> |
| 724 |
|
| 725 |
<hr /> |
| 726 |
|
| 727 |
<table> |
| 728 |
<tr> |
| 729 |
<td align="right" valign="baseline"><a name="prod11">bnf_production</a></td> |
| 730 |
<td align="center" valign="baseline">::=</td> |
| 731 |
<td align="left" valign="baseline"><em>java_access_modifier</em> <em>java_return_type</em> <em>java_identifier</em> "(" <em>java_parameter_list</em> ")" ":"</td> |
| 732 |
</tr> |
| 733 |
<tr> |
| 734 |
<td></td><td></td> |
| 735 |
<td align="left" valign="baseline"><em>java_block</em></td> |
| 736 |
</tr> |
| 737 |
<tr> |
| 738 |
<td></td><td></td> |
| 739 |
<td align="left" valign="baseline">"{" <a href="#prod16">expansion_choices</a> "}"</td> |
| 740 |
</tr> |
| 741 |
</table> |
| 742 |
|
| 743 |
<p> |
| 744 |
The BNF production is the standard production used |
| 745 |
in specifying JavaCC grammars. Each BNF production has a left hand |
| 746 |
side which is a non-terminal specification. The BNF production then |
| 747 |
defines this non-terminal in terms of BNF expansions on the right hand |
| 748 |
side. The non-terminal is written exactly like a declared Java method. |
| 749 |
Since each non-terminal is translated into a method |
| 750 |
in the generated parser, this style of writing the non-terminal makes |
| 751 |
this association obvious. The name of the non-terminal is the name of |
| 752 |
the method, and the parameters and return value declared are the means |
| 753 |
to pass values up and down the parse tree. As will be seen later, |
| 754 |
non-terminals on the right hand sides of productions are written as |
| 755 |
method calls, so the passing of values up and down the tree are done |
| 756 |
using exactly the same paradigm as method call and return. |
| 757 |
The default access modifier for BNF productions is public. |
| 758 |
</p> |
| 759 |
<p> |
| 760 |
There are two parts on the right hand side of an BNF production. The |
| 761 |
first part is a set of arbitrary Java declarations and code (the Java |
| 762 |
block). This code is generated at the beginning |
| 763 |
of the method generated for the Java non-terminal. Hence, every time |
| 764 |
this non-terminal is used in the parsing process, these declarations and |
| 765 |
code are executed. The declarations in this part are visible to all Java |
| 766 |
code in actions in the BNF expansions. JavaCC does not do any processing |
| 767 |
of these declarations and code, except to skip to the matching ending |
| 768 |
brace, collecting all text encountered on the way. Hence, a Java compiler |
| 769 |
can detect errors in this code that has been processed by JavaCC. |
| 770 |
</p> |
| 771 |
<p> |
| 772 |
The second part of the right hand side are the BNF expansions. This |
| 773 |
is described <a href="#prod16">later</a>. |
| 774 |
</p> |
| 775 |
|
| 776 |
<hr /> |
| 777 |
|
| 778 |
<table> |
| 779 |
<tr> |
| 780 |
<td align="right" valign="baseline"><a name="prod10">regular_expr_production</a></td> |
| 781 |
<td align="center" valign="baseline">::=</td> |
| 782 |
<td align="left" valign="baseline">[ <a href="#newprod1">lexical_state_list</a> ]</td> |
| 783 |
</tr> |
| 784 |
<tr> |
| 785 |
<td></td><td></td> |
| 786 |
<td align="left" valign="baseline"><a href="#prod17">regexpr_kind</a> [ "[" "IGNORE_CASE" "]" ] ":"</td> |
| 787 |
</tr> |
| 788 |
<tr> |
| 789 |
<td></td><td></td> |
| 790 |
<td align="left" valign="baseline">"{" <a href="#prod18">regexpr_spec</a> ( "|" <a href="#prod18">regexpr_spec</a> )* "}"</td> |
| 791 |
</tr> |
| 792 |
</table> |
| 793 |
<p> |
| 794 |
A regular expression production is used to define lexical entities |
| 795 |
that get processed by the generated token manager. A detailed description |
| 796 |
of how the token manager works is provided in |
| 797 |
<a href="tokenmanager.html">this minitutorial (click here)</a>. This |
| 798 |
page describes the syntactic aspects of specifying lexical entities, |
| 799 |
while <a href="tokenmanager.html">the minitutorial</a> describes how |
| 800 |
these syntactic constructs tie in with how the token manager actually |
| 801 |
works. |
| 802 |
</p> |
| 803 |
<p> |
| 804 |
A regular expression production starts with a specification of the |
| 805 |
lexical states for which it applies (the |
| 806 |
<a href="#newprod1">lexical state list</a>). |
| 807 |
There is a standard lexical state called "DEFAULT". If the |
| 808 |
<a href="#newprod1">lexical state list</a> is omitted, the regular |
| 809 |
expression production applies to the lexical state "DEFAULT". |
| 810 |
</p> |
| 811 |
<p> |
| 812 |
Following this is a description of what kind of regular expression |
| 813 |
production this is (<a href="#prod17">see below for what this means</a>). |
| 814 |
</p> |
| 815 |
<p> |
| 816 |
After this is an optional "[IGNORE_CASE]". If this is present, the |
| 817 |
regular expression production is case insensitive - it has the same |
| 818 |
effect as the |
| 819 |
<a href="#prod6">IGNORE_CASE</a> |
| 820 |
option, except that in this case it applies locally to this regular |
| 821 |
expression production. |
| 822 |
</p> |
| 823 |
<p> |
| 824 |
This is then followed by a list of regular expression specifications |
| 825 |
that describe in more detail the lexical entities of this regular |
| 826 |
expression production. |
| 827 |
</p> |
| 828 |
|
| 829 |
<hr /> |
| 830 |
|
| 831 |
<table> |
| 832 |
<tr> |
| 833 |
<td align="right" valign="baseline"><a name="prod12">token_manager_decls</a></td> |
| 834 |
<td align="center" valign="baseline">::=</td> |
| 835 |
<td align="left" valign="baseline">"<a name="TOKEN_MGR_DECLS">TOKEN_MGR_DECLS</a>" ":" <em>java_block</em></td> |
| 836 |
</tr> |
| 837 |
</table> |
| 838 |
|
| 839 |
<p> |
| 840 |
The token manager declarations starts with the reserved word |
| 841 |
"TOKEN_MGR_DECLS" followed by a ":" and then a set of Java declarations |
| 842 |
and statements (the Java block). These declarations and statements are |
| 843 |
written into the generated token manager and are accessible from within |
| 844 |
<a href="#prod18">lexical actions</a>. See |
| 845 |
<a href="tokenmanager.html">the minitutorial on the token manager</a> |
| 846 |
for more details. |
| 847 |
</p> |
| 848 |
<p> |
| 849 |
There can only be one token manager declaration in a JavaCC grammar file. |
| 850 |
</p> |
| 851 |
|
| 852 |
<hr /> |
| 853 |
|
| 854 |
<table> |
| 855 |
<tr> |
| 856 |
<td align="right" valign="baseline"><a name="newprod1">lexical_state_list</a></td> |
| 857 |
<td align="center" valign="baseline">::=</td> |
| 858 |
<td align="left" valign="baseline">"<" "*" ">"</td> |
| 859 |
</tr> |
| 860 |
<tr> |
| 861 |
<td></td><td>|</td> |
| 862 |
<td align="left" valign="baseline">"<" <em>java_identifier</em> ( "," <em>java_identifier</em> )* ">"</td> |
| 863 |
</tr> |
| 864 |
</table> |
| 865 |
|
| 866 |
<p> |
| 867 |
The lexical state list describes the set of lexical states for which |
| 868 |
the corresponding <a href="#prod10">regular expression production</a> |
| 869 |
applies. If this is written as "<*>", the regular expression production |
| 870 |
applies to all lexical states. Otherwise, it applies to all the lexical |
| 871 |
states in the identifier list within the angular brackets. |
| 872 |
</p> |
| 873 |
|
| 874 |
|
| 875 |
<hr /> |
| 876 |
|
| 877 |
<table> |
| 878 |
<tr> |
| 879 |
<td align="right" valign="baseline"><a name="prod17">regexpr_kind</a></td> |
| 880 |
<td align="center" valign="baseline">::=</td> |
| 881 |
<td align="left" valign="baseline">"TOKEN"</td> |
| 882 |
</tr> |
| 883 |
<tr> |
| 884 |
<td align="right" valign="baseline"></td> |
| 885 |
<td align="center" valign="baseline">|</td> |
| 886 |
<td align="left" valign="baseline">"SPECIAL_TOKEN"</td> |
| 887 |
</tr> |
| 888 |
<tr> |
| 889 |
<td align="right" valign="baseline"></td> |
| 890 |
<td align="center" valign="baseline">|</td> |
| 891 |
<td align="left" valign="baseline">"SKIP"</td> |
| 892 |
</tr> |
| 893 |
<tr> |
| 894 |
<td align="right" valign="baseline"></td> |
| 895 |
<td align="center" valign="baseline">|</td> |
| 896 |
<td align="left" valign="baseline">"MORE"</td> |
| 897 |
</tr> |
| 898 |
</table> |
| 899 |
|
| 900 |
<p> |
| 901 |
This specifies the kind of |
| 902 |
<a href="#prod10">regular expression production</a>. |
| 903 |
There are four kinds: |
| 904 |
</p> |
| 905 |
<ul> |
| 906 |
<li> |
| 907 |
<strong><a name="TOKEN">TOKEN</a></strong>: |
| 908 |
The regular expressions in this regular expression production describe |
| 909 |
<em>tokens</em> in the grammar. The token manager creates a |
| 910 |
<a href="apiroutines.html">Token</a> object for each match of such |
| 911 |
a regular expression and returns it to the parser. |
| 912 |
</li> |
| 913 |
<li> |
| 914 |
<strong><a name="SPECIAL_TOKEN">SPECIAL_TOKEN</a></strong>: |
| 915 |
The regular expressions in this regular expression production describe |
| 916 |
<em>special tokens</em>. Special tokens are like tokens, except that |
| 917 |
they do not have significance during parsing - that is the BNF productions |
| 918 |
ignore them. Special tokens are, however, still passed on to the parser |
| 919 |
so that parser actions can access them. Special tokens are passed |
| 920 |
to the parser by linking them to neighboring real tokens using the |
| 921 |
field "specialToken" in the <a href="apiroutines.html">Token</a> |
| 922 |
class. Special tokens are useful in the processing of lexical entities |
| 923 |
such as comments which have no significance to parsing, but still |
| 924 |
are an important part of the input file. See |
| 925 |
<a href="tokenmanager.html">the minitutorial on the token manager</a> |
| 926 |
for more details of special token handling. |
| 927 |
</li> |
| 928 |
<li> |
| 929 |
<strong><a name="SKIP">SKIP</a></strong>: |
| 930 |
Matches to regular expressions in this regular expression production |
| 931 |
are simply skipped (ignored) by the token manager. |
| 932 |
</li> |
| 933 |
<li> |
| 934 |
<strong><a name="MORE">MORE</a></strong>: |
| 935 |
Sometimes it is useful to gradually build up a token to be passed on |
| 936 |
to the parser. Matches to this kind of regular expression are stored |
| 937 |
in a buffer until the next TOKEN or SPECIAL_TOKEN match. Then all |
| 938 |
the matches in the buffer and the final TOKEN/SPECIAL_TOKEN match |
| 939 |
are concatenated together to form one TOKEN/SPECIAL_TOKEN that is |
| 940 |
passed on to the parser. If a match to a SKIP regular expression |
| 941 |
follows a sequence of MORE matches, the contents of the buffer is |
| 942 |
discarded. |
| 943 |
</li> |
| 944 |
</ul> |
| 945 |
|
| 946 |
<hr /> |
| 947 |
|
| 948 |
<table> |
| 949 |
<tr> |
| 950 |
<td align="right" valign="baseline"><a name="prod18">regexpr_spec</a></td> |
| 951 |
<td align="center" valign="baseline">::=</td> |
| 952 |
<td align="left" valign="baseline"><a href="#prod19">regular_expression</a> [ <em>java_block</em> ] [ ":" <em>java_identifier</em> ]</td> |
| 953 |
</tr> |
| 954 |
</table> |
| 955 |
<p> |
| 956 |
The regular expression specification begins the actual description |
| 957 |
of the lexical entities that are part of this |
| 958 |
<a href="#prod10">regular expression production</a>. |
| 959 |
Each regular expression production may contain any number of |
| 960 |
regular expression specifications. |
| 961 |
</p> |
| 962 |
<p> |
| 963 |
Each regular expression specification contains a regular expression |
| 964 |
followed by a Java block (the lexical action) which is optional. |
| 965 |
This is then followed by an identifier of a lexical state (which |
| 966 |
is also optional). Whenever this regular expression is matched, |
| 967 |
the lexical action (if any) gets executed, followed by any |
| 968 |
<a href="#prod6">common token actions</a>. Then the action depending |
| 969 |
on the |
| 970 |
<a href="#prod17">regular expression production kind</a> |
| 971 |
is taken. Finally, if a lexical state is specified, the token |
| 972 |
manager moves to that lexical state for further processing (the |
| 973 |
token manager starts initially in the state "DEFAULT"). |
| 974 |
</p> |
| 975 |
|
| 976 |
<hr /> |
| 977 |
|
| 978 |
<table> |
| 979 |
<tr> |
| 980 |
<td align="right" valign="baseline"><a name="prod16">expansion_choices</a></td> |
| 981 |
<td align="center" valign="baseline">::=</td> |
| 982 |
<td align="left" valign="baseline"><a href="#prod20">expansion</a> ( "|" <a href="#prod20">expansion</a> )*</td> |
| 983 |
</tr> |
| 984 |
</table> |
| 985 |
|
| 986 |
<p> |
| 987 |
Expansion choices are written as a list of one or more expansions |
| 988 |
separated by "|"s. The set of legal parses allowed by an expansion |
| 989 |
choice is a legal parse of any one of the contained expansions. |
| 990 |
</p> |
| 991 |
|
| 992 |
<hr /> |
| 993 |
|
| 994 |
<table> |
| 995 |
<tr> |
| 996 |
<td align="right" valign="baseline"><a name="prod20">expansion</a></td> |
| 997 |
<td align="center" valign="baseline">::=</td> |
| 998 |
<td align="left" valign="baseline">( <a href="#prod22">expansion_unit</a> )*</td> |
| 999 |
</tr> |
| 1000 |
</table> |
| 1001 |
|
| 1002 |
<p> |
| 1003 |
An expansion is written as a sequence of expansion units. |
| 1004 |
A concatenation of legal |
| 1005 |
parses of the expansion units is a legal parse of the expansion. |
| 1006 |
</p> |
| 1007 |
<p> |
| 1008 |
For example, the expansion "{" decls() "}" consists of three expansion |
| 1009 |
units - "{", decls(), and "}". A match for the expansion is a concatenation |
| 1010 |
of the matches for the individual expansion units - in this case, that would |
| 1011 |
be any string that begins with a "{", ends with a "}", and contains a match |
| 1012 |
for decls() in between. |
| 1013 |
</p> |
| 1014 |
|
| 1015 |
<hr /> |
| 1016 |
|
| 1017 |
<table> |
| 1018 |
<tr> |
| 1019 |
<td align="right" valign="baseline"><a name="prod22">expansion_unit</a></td> |
| 1020 |
<td align="center" valign="baseline">::=</td> |
| 1021 |
<td align="left" valign="baseline"><a href="#prod21">local_lookahead</a></td> |
| 1022 |
</tr> |
| 1023 |
<tr> |
| 1024 |
<td align="right" valign="baseline"></td> |
| 1025 |
<td align="center" valign="baseline">|</td> |
| 1026 |
<td align="left" valign="baseline"><em>java_block</em></td> |
| 1027 |
</tr> |
| 1028 |
<tr> |
| 1029 |
<td align="right" valign="baseline"></td> |
| 1030 |
<td align="center" valign="baseline">|</td> |
| 1031 |
<td align="left" valign="baseline">"(" <a href="#prod16">expansion_choices</a> ")" [ "+" | "*" | "?" ]</td> |
| 1032 |
</tr> |
| 1033 |
<tr> |
| 1034 |
<td align="right" valign="baseline"></td> |
| 1035 |
<td align="center" valign="baseline">|</td> |
| 1036 |
<td align="left" valign="baseline">"[" <a href="#prod16">expansion_choices</a> "]"</td> |
| 1037 |
</tr> |
| 1038 |
<tr> |
| 1039 |
<td align="right" valign="baseline"></td> |
| 1040 |
<td align="center" valign="baseline">|</td> |
| 1041 |
<td align="left" valign="baseline">[ <em>java_assignment_lhs</em> "=" ] <a href="#prod19">regular_expression</a></td> |
| 1042 |
</tr> |
| 1043 |
<tr> |
| 1044 |
<td align="right" valign="baseline"></td> |
| 1045 |
<td align="center" valign="baseline">|</td> |
| 1046 |
<td align="left" valign="baseline">[ <em>java_assignment_lhs</em> "=" ] <em>java_identifier</em> "(" <em>java_expression_list</em> ")"</td> |
| 1047 |
</tr> |
| 1048 |
</table> |
| 1049 |
|
| 1050 |
<p> |
| 1051 |
An expansion unit can be a <a href="#prod21">local LOOKAHEAD specification</a>. |
| 1052 |
This instructs the |
| 1053 |
generated parser on how to make choices at choice points. For details |
| 1054 |
on how LOOKAHEAD specifications work and how to write LOOKAHEAD specifications, |
| 1055 |
<a href="lookahead.html">click here to visit the minitutorial on LOOKAHEAD</a>. |
| 1056 |
</p> |
| 1057 |
<p> |
| 1058 |
An expansion unit can be a set of Java declarations and code enclosed |
| 1059 |
within braces (the Java block). These are also called <em>parser |
| 1060 |
actions</em>. This is generated into the method parsing the |
| 1061 |
non-terminal at the appropriate location. This block is executed |
| 1062 |
whenever the parsing process crosses this point successfully. |
| 1063 |
When JavaCC processes the Java block, it does not perform any detailed |
| 1064 |
syntax or semantic checking. Hence it is possible that the Java compiler |
| 1065 |
will find errors in your actions that have been processed by JavaCC. |
| 1066 |
<em>Actions are not executed during |
| 1067 |
<a href="lookahead.html">lookahead evaluation</a>.</em> |
| 1068 |
</p> |
| 1069 |
<p> |
| 1070 |
An expansion unit can be a parenthesized set of one or more |
| 1071 |
<a href="#prod16">expansion choices</a>. In which case, a legal parse of the expansion |
| 1072 |
unit is any legal parse of the nested expansion choices. |
| 1073 |
The parenthesized set of expansion choices can be suffixed (optionally) by: |
| 1074 |
</p> |
| 1075 |
<ul> |
| 1076 |
<li> |
| 1077 |
<strong>"+":</strong> |
| 1078 |
Then any legal parse of the expansion unit is one or more |
| 1079 |
repetitions of a legal parse of the parenthesized set of |
| 1080 |
expansion choices. |
| 1081 |
</li> |
| 1082 |
<li> |
| 1083 |
<strong>"*":</strong> |
| 1084 |
Then any legal parse of the expansion unit is zero or more |
| 1085 |
repetitions of a legal parse of the parenthesized set of |
| 1086 |
expansion choices. |
| 1087 |
</li> |
| 1088 |
<li> |
| 1089 |
<strong>"?":</strong> |
| 1090 |
Then a legal parse of the expansion unit is either the |
| 1091 |
empty token sequence or any legal parse of the nested expansion choices. |
| 1092 |
An alternate syntax for this construct is to enclose the |
| 1093 |
expansion choices within brackets "[...]". |
| 1094 |
</li> |
| 1095 |
</ul> |
| 1096 |
<p> |
| 1097 |
An expansion unit can be a <a href="#prod19">regular expression</a>. Then a legal parse |
| 1098 |
of the expansion unit is any token that matches this regular |
| 1099 |
expression. When a regular expression is matched, it creates an |
| 1100 |
object of type <a href="apiroutines.html">Token</a>. This object |
| 1101 |
can be accessed by assigning it to a variable by prefixing the |
| 1102 |
regular expression with "variable =". In general, you may have any |
| 1103 |
valid Java assignment left-hand side to the left of the "=". |
| 1104 |
<em>This assignment is not performed during |
| 1105 |
<a href="lookahead.html">lookahead evaluation</a>.</em> |
| 1106 |
</p> |
| 1107 |
<p> |
| 1108 |
An expansion unit can be a non-terminal (the last choice in the syntax |
| 1109 |
above). In which case, it takes |
| 1110 |
the form of a method call with the non-terminal name used as the |
| 1111 |
name of the method. A successful parse of the non-terminal causes |
| 1112 |
the parameters placed in the method call to be operated on and a |
| 1113 |
value returned (in case the non-terminal was not declared to be |
| 1114 |
of type "void"). The return value can be assigned (optionally) to |
| 1115 |
a variable by prefixing the regular expression with "variable =". |
| 1116 |
In general, you may have any |
| 1117 |
valid Java assignment left-hand side to the left of the "=". |
| 1118 |
<em>This assignment is not performed during |
| 1119 |
<a href="lookahead.html">lookahead evaluation</a>.</em> |
| 1120 |
Non-terminals may not be used in an expansion in a manner that introduces |
| 1121 |
left-recursion. JavaCC checks this for you. |
| 1122 |
</p> |
| 1123 |
|
| 1124 |
<hr /> |
| 1125 |
|
| 1126 |
<table> |
| 1127 |
<tr> |
| 1128 |
<td align="right" valign="baseline"><a name="prod21">local_lookahead</a></td> |
| 1129 |
<td align="center" valign="baseline">::=</td> |
| 1130 |
<td align="left" valign="baseline">"LOOKAHEAD" "(" [ <em>java_integer_literal</em> ] [ "," ] [ <a href="#prod16">expansion_choices</a> ] [ "," ] [ "{" <em>java_expression</em> "}" ] ")"</td> |
| 1131 |
</tr> |
| 1132 |
</table> |
| 1133 |
|
| 1134 |
<p> |
| 1135 |
A local lookahead specification is used to influence the way the generated |
| 1136 |
parser makes choices at the various |
| 1137 |
<a href="lookahead.html">choice points</a> |
| 1138 |
in the grammar. A local lookahead specification starts with the reserved |
| 1139 |
word "LOOKAHEAD" followed by a set of lookahead constraints within parentheses. |
| 1140 |
There are three different kinds of lookahead constraints - a lookahead limit |
| 1141 |
(the integer literal), a syntactic lookahead (the expansion choices), and |
| 1142 |
a semantic lookahead (the expression within braces). At least one lookahead |
| 1143 |
constraint must be present. If more than one lookahead constraint is present, |
| 1144 |
they must be separated by commas. |
| 1145 |
</p> |
| 1146 |
<p> |
| 1147 |
For a detailed description of how lookahead works, please |
| 1148 |
<a href="lookahead.html">click here to visit the minitutorial on LOOKAHEAD</a>. |
| 1149 |
A brief description of each kind of lookahead constraint is given below: |
| 1150 |
</p> |
| 1151 |
<ul> |
| 1152 |
<li> |
| 1153 |
<strong>Lookahead Limit:</strong> |
| 1154 |
This is the maximum number of tokens of lookahead that may be used for choice |
| 1155 |
determination purposes. This overrides the default value which is specified |
| 1156 |
by the <a href="#prod2">LOOKAHEAD option</a>. This lookahead limit applies |
| 1157 |
only to the <a href="lookahead.html">choice point</a> |
| 1158 |
at the location of the local lookahead specification. |
| 1159 |
If the local lookahead specification is not at a choice point, the lookahead |
| 1160 |
limit (if any) is ignored. |
| 1161 |
</li> |
| 1162 |
<li> |
| 1163 |
<strong>Syntactic Lookahead:</strong> |
| 1164 |
This is an expansion (or expansion choices) that is used for the purpose of |
| 1165 |
determining whether or not the particular choice that this local lookahead |
| 1166 |
specification applies to is to be taken. If this was not provided, the parser |
| 1167 |
uses the expansion to be selected during lookahead determination. |
| 1168 |
If the local lookahead specification is not at a |
| 1169 |
<a href="lookahead.html">choice point</a>, the syntactic |
| 1170 |
lookahead (if any) is ignored. |
| 1171 |
</li> |
| 1172 |
<li> |
| 1173 |
<strong>Semantic Lookahead:</strong> |
| 1174 |
This is a boolean expression that is evaluated whenever the parser crosses this |
| 1175 |
point during parsing. If the expression evaluates to true, the parsing |
| 1176 |
continues normally. If the expression evaluates to false and the local |
| 1177 |
lookahead specification is at a <a href="lookahead.html">choice point</a>, |
| 1178 |
the current choice is not taken and the next choice is considered. |
| 1179 |
If the expression evaluates to false and the local lookahead specification |
| 1180 |
is <em>not</em> at a choice point, then parsing aborts with a parse error. |
| 1181 |
Unlike the other two lookahead constraints that are ignored at non-choice |
| 1182 |
points, semantic lookahead is always evaluated. In fact, semantic lookahead |
| 1183 |
is even evaluated if it is encountered during the evaluation of some other |
| 1184 |
syntactic lookahead check (for more details |
| 1185 |
<a href="lookahead.html">click here to visit the minitutorial on LOOKAHEAD</a>). |
| 1186 |
</li> |
| 1187 |
</ul> |
| 1188 |
<p> |
| 1189 |
<strong>Default values for lookahead constraints:</strong> |
| 1190 |
If a local lookahead specification has been provided, but not all lookahead |
| 1191 |
constraints have been included, then the missing ones are assigned default |
| 1192 |
values as follows: |
| 1193 |
</p> |
| 1194 |
<ul> |
| 1195 |
<li> |
| 1196 |
If the lookahead limit is not provided and if the syntactic lookahead is |
| 1197 |
provided, then the lookahead limit defaults to the largest integer value |
| 1198 |
(2147483647). This essentially implements "infinite lookahead" - namely, |
| 1199 |
look ahead as many tokens as necessary to match the syntactic lookahead that |
| 1200 |
has been provided. |
| 1201 |
</li> |
| 1202 |
<li> |
| 1203 |
If neither the lookahead limit nor the syntactic lookahead has been |
| 1204 |
provided (which means the semantic lookahead is provided), the lookahead |
| 1205 |
limit defaults to 0. This means that syntactic lookahead is not performed |
| 1206 |
(it passes trivially), and only semantic lookahead is performed. |
| 1207 |
</li> |
| 1208 |
<li> |
| 1209 |
If the syntactic lookahead is not provided, it defaults to the choice |
| 1210 |
to which the local lookahead specification applies. If the local lookahead |
| 1211 |
specification is not at a choice point, then the syntactic lookahead is |
| 1212 |
ignored - hence a default value is not relevant. |
| 1213 |
</li> |
| 1214 |
<li> |
| 1215 |
If the semantic lookahead is not provided, it defaults to the boolean |
| 1216 |
expression "true". That is, it trivially passes. |
| 1217 |
</li> |
| 1218 |
</ul> |
| 1219 |
|
| 1220 |
<hr /> |
| 1221 |
|
| 1222 |
|
| 1223 |
<table> |
| 1224 |
<tr> |
| 1225 |
<td align="right" valign="baseline"><a name="prod19">regular_expression</a></td> |
| 1226 |
<td align="center" valign="baseline">::=</td> |
| 1227 |
<td align="left" valign="baseline"><em>java_string_literal</em></td> |
| 1228 |
</tr> |
| 1229 |
<tr> |
| 1230 |
<td align="right" valign="baseline"></td> |
| 1231 |
<td align="center" valign="baseline">|</td> |
| 1232 |
<td align="left" valign="baseline">"<" [ [ "#" ] <em>java_identifier</em> ":" ] <a href="#prod29">complex_regular_expression_choices</a> ">"</td> |
| 1233 |
</tr> |
| 1234 |
<tr> |
| 1235 |
<td align="right" valign="baseline"></td> |
| 1236 |
<td align="center" valign="baseline">|</td> |
| 1237 |
<td align="left" valign="baseline">"<" <em>java_identifier</em> ">"</td> |
| 1238 |
</tr> |
| 1239 |
<tr> |
| 1240 |
<td align="right" valign="baseline"></td> |
| 1241 |
<td align="center" valign="baseline">|</td> |
| 1242 |
<td align="left" valign="baseline">"<" "EOF" ">"</td> |
| 1243 |
</tr> |
| 1244 |
</table> |
| 1245 |
|
| 1246 |
<p> |
| 1247 |
There are two places in a grammar files where regular expressions may be |
| 1248 |
written: |
| 1249 |
</p> |
| 1250 |
<ul> |
| 1251 |
<li> |
| 1252 |
Within a <a href="#prod18">regular expression specification</a> |
| 1253 |
(part of a <a href="#prod10">regular expression production</a>), |
| 1254 |
</li> |
| 1255 |
<li> |
| 1256 |
As an <a href="#prod22">expansion unit</a> with an <a href="#prod20">expansion</a>. |
| 1257 |
When a regular expression is used in this manner, it is as if the regular expression |
| 1258 |
were defined in the following manner at this location and then referred to by its |
| 1259 |
label from the expansion unit: |
| 1260 |
<pre> |
| 1261 |
<DEFAULT> TOKEN : |
| 1262 |
{ |
| 1263 |
regular expression |
| 1264 |
} |
| 1265 |
</pre> |
| 1266 |
That is, this usage of regular expression can be rewritten using the other |
| 1267 |
kind of usage. |
| 1268 |
</li> |
| 1269 |
</ul> |
| 1270 |
<p> |
| 1271 |
The complete details of regular expression matching by the token manager is |
| 1272 |
available in |
| 1273 |
<a href="tokenmanager.html">the minitutorial on the token manager</a>. The |
| 1274 |
description of the syntactic constructs follows. |
| 1275 |
</p> |
| 1276 |
<p> |
| 1277 |
The first kind of regular expression is a string literal. The input being |
| 1278 |
parsed matches this regular expression if the token manager is in a |
| 1279 |
<a href="#prod10">lexical state</a> for which this regular expression applies |
| 1280 |
and the next set of characters in the input stream is the same (possibly with |
| 1281 |
case ignored) as this string literal. |
| 1282 |
</p> |
| 1283 |
<p> |
| 1284 |
A regular expression may also be a more <a href="#prod29">complex regular expression</a> |
| 1285 |
using which more involved regular expression (than string literals can be defined). |
| 1286 |
Such a regular expression is placed within angular brackets "<...>", and |
| 1287 |
may be labeled optionally with an identifier. This label may be used to refer |
| 1288 |
to this regular expression from |
| 1289 |
<a href="#prod22">expansion units</a> |
| 1290 |
or from within other regular expressions. |
| 1291 |
If the label is preceded by a "#", then this regular expression may not be |
| 1292 |
referred to from expansion units, but only from within other regular expressions. |
| 1293 |
When the "#" is present, the regular expression is referred to as a |
| 1294 |
"private regular expression". |
| 1295 |
</p> |
| 1296 |
<p> |
| 1297 |
A regular expression may be a reference to some other labeled regular expression |
| 1298 |
in which case it is written as the label enclosed in angular brackets "<...>". |
| 1299 |
</p> |
| 1300 |
<p> |
| 1301 |
Finally, a regular expression may be a reference to the predefined regular |
| 1302 |
expression "<EOF>" which is matched by the end of file. |
| 1303 |
</p> |
| 1304 |
<p> |
| 1305 |
Private regular expressions are not matched as tokens by the token manager. |
| 1306 |
Their purpose is solely to facilitate the definition of other more complex |
| 1307 |
regular expressions. |
| 1308 |
</p> |
| 1309 |
<p> |
| 1310 |
Consider the following example defining Java floating point literals: |
| 1311 |
</p> |
| 1312 |
<pre> |
| 1313 |
TOKEN : |
| 1314 |
{ |
| 1315 |
< FLOATING_POINT_LITERAL: |
| 1316 |
(["0"-"9"])+ "." (["0"-"9"])* (<EXPONENT>)? (["f","F","d","D"])? |
| 1317 |
| "." (["0"-"9"])+ (<EXPONENT>)? (["f","F","d","D"])? |
| 1318 |
| (["0"-"9"])+ <EXPONENT> (["f","F","d","D"])? |
| 1319 |
| (["0"-"9"])+ (<EXPONENT>)? ["f","F","d","D"] |
| 1320 |
> |
| 1321 |
| |
| 1322 |
< #EXPONENT: ["e","E"] (["+","-"])? (["0"-"9"])+ > |
| 1323 |
} |
| 1324 |
</pre> |
| 1325 |
<p> |
| 1326 |
In this example, the token FLOATING_POINT_LITERAL is defined using the |
| 1327 |
definition of another token, namely, EXPONENT. The "#" before the label |
| 1328 |
EXPONENT indicates that this exists solely for the purpose of defining other |
| 1329 |
tokens (FLOATING_POINT_LITERAL in this case). The definition of |
| 1330 |
FLOATING_POINT_LITERAL is not affected by the presence or absence of the "#". |
| 1331 |
However, the token manager's behavior is. If the "#" is omitted, the |
| 1332 |
token manager will |
| 1333 |
erroneously recognize a string like E123 as a legal token of kind EXPONENT |
| 1334 |
(instead of IDENTIFIER in the Java grammar). |
| 1335 |
</p> |
| 1336 |
|
| 1337 |
<hr /> |
| 1338 |
|
| 1339 |
<table> |
| 1340 |
<tr> |
| 1341 |
<td align="right" valign="baseline"><a name="prod29">complex_regular_expression_choices</a></td> |
| 1342 |
<td align="center" valign="baseline">::=</td> |
| 1343 |
<td align="left" valign="baseline"><a href="#prod30">complex_regular_expression</a> ( "|" <a href="#prod30">complex_regular_expression</a> )*</td> |
| 1344 |
</tr> |
| 1345 |
</table> |
| 1346 |
|
| 1347 |
<p> |
| 1348 |
Complex regular expression choices is made up of a list of one or more |
| 1349 |
<a href="#prod30">complex regular expressions</a> separated by "|"s. |
| 1350 |
A match for a complex regular expression choice is a match of any of its |
| 1351 |
constituent complex regular expressions. |
| 1352 |
</p> |
| 1353 |
|
| 1354 |
<hr /> |
| 1355 |
|
| 1356 |
<table> |
| 1357 |
<tr> |
| 1358 |
<td align="right" valign="baseline"><a name="prod30">complex_regular_expression</a></td> |
| 1359 |
<td align="center" valign="baseline">::=</td> |
| 1360 |
<td align="left" valign="baseline">( <a href="#prod31">complex_regular_expression_unit</a> )*</td> |
| 1361 |
</tr> |
| 1362 |
</table> |
| 1363 |
|
| 1364 |
<p> |
| 1365 |
A complex regular expression is a sequence of complex regular expression units. |
| 1366 |
A match for a complex regular expression is a concatenation of matches to |
| 1367 |
the complex regular expression units. |
| 1368 |
</p> |
| 1369 |
|
| 1370 |
<hr /> |
| 1371 |
|
| 1372 |
<table> |
| 1373 |
<tr> |
| 1374 |
<td align="right" valign="baseline"><a name="prod31">complex_regular_expression_unit</a></td> |
| 1375 |
<td align="center" valign="baseline">::=</td> |
| 1376 |
<td align="left" valign="baseline"><em>java_string_literal</em></td> |
| 1377 |
</tr> |
| 1378 |
<tr> |
| 1379 |
<td align="right" valign="baseline"></td> |
| 1380 |
<td align="center" valign="baseline">|</td> |
| 1381 |
<td align="left" valign="baseline">"<" <em>java_identifier</em> ">"</td> |
| 1382 |
</tr> |
| 1383 |
<tr> |
| 1384 |
<td align="right" valign="baseline"></td> |
| 1385 |
<td align="center" valign="baseline">|</td> |
| 1386 |
<td align="left" valign="baseline"><a href="#prod32">character_list</a></td> |
| 1387 |
</tr> |
| 1388 |
<tr> |
| 1389 |
<td align="right" valign="baseline"></td> |
| 1390 |
<td align="center" valign="baseline">|</td> |
| 1391 |
<td align="left" valign="baseline">"(" <a href="#prod29">complex_regular_expression_choices</a> ")" [ "+" | "*" | "?" ]</td> |
| 1392 |
</tr> |
| 1393 |
</table> |
| 1394 |
|
| 1395 |
<p> |
| 1396 |
A complex regular expression unit can be a string literal, in which case |
| 1397 |
there is exactly one match for this unit, namely, the string literal itself. |
| 1398 |
</p> |
| 1399 |
<p> |
| 1400 |
A complex regular expression unit can be a reference to another regular |
| 1401 |
expression. The other regular expression has to be labeled so that it |
| 1402 |
can be referenced. The matches of this unit are all the matches of this |
| 1403 |
other regular expression. Such references in regular expressions cannot |
| 1404 |
introduce loops in the dependency between tokens. |
| 1405 |
</p> |
| 1406 |
<p> |
| 1407 |
A complex regular expression unit can be a <a href="#prod32">character list</a>. |
| 1408 |
A character list is a way of defining a set of characters. A match for this |
| 1409 |
kind of complex regular expression unit is any character that is allowed |
| 1410 |
by the character list. |
| 1411 |
</p> |
| 1412 |
<p> |
| 1413 |
A complex regular expression unit can be a parenthesized set of |
| 1414 |
complex regular expression choices. In this case, a legal match of |
| 1415 |
the unit is any legal match of the nested choices. The parenthesized |
| 1416 |
set of choices can be suffixed (optionally) by: |
| 1417 |
</p> |
| 1418 |
<ul> |
| 1419 |
<li> |
| 1420 |
<strong>"+":</strong> |
| 1421 |
Then any legal match of the unit is one or more |
| 1422 |
repetitions of a legal match of the parenthesized set of |
| 1423 |
choices. |
| 1424 |
</li> |
| 1425 |
<li> |
| 1426 |
<strong>"*":</strong> |
| 1427 |
Then any legal match of the unit is zero or more |
| 1428 |
repetitions of a legal match of the parenthesized set of |
| 1429 |
choices. |
| 1430 |
</li> |
| 1431 |
<li> |
| 1432 |
<strong>"?":</strong> |
| 1433 |
Then a legal match of the unit is either the |
| 1434 |
empty string or any legal match of the nested choices. |
| 1435 |
</li> |
| 1436 |
</ul> |
| 1437 |
<p> |
| 1438 |
Note that unlike the BNF <a href="#prod20">expansions</a>, |
| 1439 |
the regular expression "[...]" is not equivalent |
| 1440 |
to the regular expression "(...)?". This is because the [...] |
| 1441 |
construct is used to describe <a href="#prod32">character lists</a> |
| 1442 |
in regular expressions. |
| 1443 |
</p> |
| 1444 |
|
| 1445 |
<hr /> |
| 1446 |
|
| 1447 |
<table> |
| 1448 |
<tr> |
| 1449 |
<td align="right" valign="baseline"><a name="prod32">character_list</a></td> |
| 1450 |
<td align="center" valign="baseline">::=</td> |
| 1451 |
<td align="left" valign="baseline">[ "~" ] "[" [ <a href="#prod33">character_descriptor</a> ( "," <a href="#prod33">character_descriptor</a> )* ] "]"</td> |
| 1452 |
</tr> |
| 1453 |
</table> |
| 1454 |
|
| 1455 |
<p> |
| 1456 |
A character list describes a set of characters. A legal match for a |
| 1457 |
character list is any character in this set. A character list is a list |
| 1458 |
of character descriptors separated by commas within square brackets. |
| 1459 |
Each character descriptor describes a single character or a range of characters |
| 1460 |
(see <a href="#prod33">character descriptor</a> below), |
| 1461 |
and this is added to the set of characters of the character |
| 1462 |
list. If the character list is prefixed by the "~" symbol, the set of |
| 1463 |
characters it represents is any UNICODE character not in the specified set. |
| 1464 |
</p> |
| 1465 |
|
| 1466 |
<hr /> |
| 1467 |
|
| 1468 |
<table> |
| 1469 |
<tr> |
| 1470 |
<td align="right" valign="baseline"><a name="prod33">character_descriptor</a></td> |
| 1471 |
<td align="center" valign="baseline">::=</td> |
| 1472 |
<td align="left" valign="baseline"><em>java_string_literal</em> [ "-" <em>java_string_literal</em> ]</td> |
| 1473 |
</tr> |
| 1474 |
</table> |
| 1475 |
|
| 1476 |
<p> |
| 1477 |
A character descriptor can be a single character string literal, in which |
| 1478 |
case it describes a singleton set containing that character; or it is |
| 1479 |
two single character string literals separated by a "-", in which case, it |
| 1480 |
describes the set of all characters in the range between and including these |
| 1481 |
two characters. |
| 1482 |
</p> |
| 1483 |
|
| 1484 |
</body> |
| 1485 |
</html> |