Gitiles
Code Review Sign In
gerrit.openfyde.cn / chromium.googlesource.com / chromium / deps / sqlite / dfe4e6bb33bc58a5d72e635c83d0895e412bdde6 / . / doc / lemon.html
blob: 114526f372ca8b870eb5cf3d1d6fa8d38268c6b1 [file] [log] [blame]
drh75897232000-05-29 14:26:00 +0000[diff] [blame]1<html>
2<head>
3<title>The Lemon Parser Generator</title>
4</head>
5<body bgcolor=white>
6<h1 align=center>The Lemon Parser Generator</h1>
7
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]8<p>Lemon is an LALR(1) parser generator for C.
9It does the same job as "bison" and "yacc".
10But lemon is not a bison or yacc clone. Lemon
drh75897232000-05-29 14:26:00 +0000[diff] [blame]11uses a different grammar syntax which is designed to
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]12reduce the number of coding errors. Lemon also uses a
13parsing engine that is faster than yacc and
14bison and which is both reentrant and threadsafe.
15(Update: Since the previous sentence was written, bison
16has also been updated so that it too can generate a
17reentrant and threadsafe parser.)
18Lemon also implements features that can be used
drh75897232000-05-29 14:26:00 +0000[diff] [blame]19to eliminate resource leaks, making is suitable for use
20in long-running programs such as graphical user interfaces
21or embedded controllers.</p>
22
23<p>This document is an introduction to the Lemon
24parser generator.</p>
25
26<h2>Theory of Operation</h2>
27
28<p>The main goal of Lemon is to translate a context free grammar (CFG)
29for a particular language into C code that implements a parser for
30that language.
31The program has two inputs:
32<ul>
33<li>The grammar specification.
34<li>A parser template file.
35</ul>
36Typically, only the grammar specification is supplied by the programmer.
37Lemon comes with a default parser template which works fine for most
38applications. But the user is free to substitute a different parser
39template if desired.</p>
40
41<p>Depending on command-line options, Lemon will generate between
42one and three files of outputs.
43<ul>
44<li>C code to implement the parser.
45<li>A header file defining an integer ID for each terminal symbol.
46<li>An information file that describes the states of the generated parser
47 automaton.
48</ul>
49By default, all three of these output files are generated.
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]50The header file is suppressed if the "-m" command-line option is
51used and the report file is omitted when "-q" is selected.</p>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]52
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]53<p>The grammar specification file uses a ".y" suffix, by convention.
drh75897232000-05-29 14:26:00 +0000[diff] [blame]54In the examples used in this document, we'll assume the name of the
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]55grammar file is "gram.y". A typical use of Lemon would be the
drh75897232000-05-29 14:26:00 +0000[diff] [blame]56following command:
57<pre>
58 lemon gram.y
59</pre>
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]60This command will generate three output files named "gram.c",
61"gram.h" and "gram.out".
drh75897232000-05-29 14:26:00 +0000[diff] [blame]62The first is C code to implement the parser. The second
63is the header file that defines numerical values for all
64terminal symbols, and the last is the report that explains
65the states used by the parser automaton.</p>
66
67<h3>Command Line Options</h3>
68
69<p>The behavior of Lemon can be modified using command-line options.
70You can obtain a list of the available command-line options together
71with a brief explanation of what each does by typing
72<pre>
73 lemon -?
74</pre>
75As of this writing, the following command-line options are supported:
76<ul>
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]77<li><b>-b</b>
78Show only the basis for each parser state in the report file.
79<li><b>-c</b>
80Do not compress the generated action tables.
81<li><b>-D<i>name</i></b>
82Define C preprocessor macro <i>name</i>. This macro is useable by
83"%ifdef" lines in the grammar file.
84<li><b>-g</b>
85Do not generate a parser. Instead write the input grammar to standard
86output with all comments, actions, and other extraneous text removed.
87<li><b>-l</b>
drhdfe4e6b2016-10-08 13:34:08 +0000[diff] [blame^]88Omit "#line" directives in the generated parser C code.
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]89<li><b>-m</b>
90Cause the output C source code to be compatible with the "makeheaders"
91program.
92<li><b>-p</b>
93Display all conflicts that are resolved by
94<a href='#precrules'>precedence rules</a>.
95<li><b>-q</b>
96Suppress generation of the report file.
97<li><b>-r</b>
98Do not sort or renumber the parser states as part of optimization.
99<li><b>-s</b>
100Show parser statistics before existing.
101<li><b>-T<i>file</i></b>
102Use <i>file</i> as the template for the generated C-code parser implementation.
103<li><b>-x</b>
104Print the Lemon version number.
drh75897232000-05-29 14:26:00 +0000[diff] [blame]105</ul>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]106
107<h3>The Parser Interface</h3>
108
109<p>Lemon doesn't generate a complete, working program. It only generates
110a few subroutines that implement a parser. This section describes
111the interface to those subroutines. It is up to the programmer to
112call these subroutines in an appropriate way in order to produce a
113complete system.</p>
114
115<p>Before a program begins using a Lemon-generated parser, the program
116must first create the parser.
117A new parser is created as follows:
118<pre>
119 void *pParser = ParseAlloc( malloc );
120</pre>
121The ParseAlloc() routine allocates and initializes a new parser and
122returns a pointer to it.
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]123The actual data structure used to represent a parser is opaque &mdash;
drh75897232000-05-29 14:26:00 +0000[diff] [blame]124its internal structure is not visible or usable by the calling routine.
125For this reason, the ParseAlloc() routine returns a pointer to void
126rather than a pointer to some particular structure.
127The sole argument to the ParseAlloc() routine is a pointer to the
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]128subroutine used to allocate memory. Typically this means malloc().</p>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]129
130<p>After a program is finished using a parser, it can reclaim all
131memory allocated by that parser by calling
132<pre>
133 ParseFree(pParser, free);
134</pre>
135The first argument is the same pointer returned by ParseAlloc(). The
136second argument is a pointer to the function used to release bulk
137memory back to the system.</p>
138
139<p>After a parser has been allocated using ParseAlloc(), the programmer
140must supply the parser with a sequence of tokens (terminal symbols) to
141be parsed. This is accomplished by calling the following function
142once for each token:
143<pre>
144 Parse(pParser, hTokenID, sTokenData, pArg);
145</pre>
146The first argument to the Parse() routine is the pointer returned by
147ParseAlloc().
148The second argument is a small positive integer that tells the parse the
149type of the next token in the data stream.
150There is one token type for each terminal symbol in the grammar.
151The gram.h file generated by Lemon contains #define statements that
152map symbolic terminal symbol names into appropriate integer values.
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]153A value of 0 for the second argument is a special flag to the
154parser to indicate that the end of input has been reached.
drh75897232000-05-29 14:26:00 +0000[diff] [blame]155The third argument is the value of the given token. By default,
156the type of the third argument is integer, but the grammar will
157usually redefine this type to be some kind of structure.
158Typically the second argument will be a broad category of tokens
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]159such as "identifier" or "number" and the third argument will
drh75897232000-05-29 14:26:00 +0000[diff] [blame]160be the name of the identifier or the value of the number.</p>
161
162<p>The Parse() function may have either three or four arguments,
drh45f31be2016-02-16 21:19:49 +0000[diff] [blame]163depending on the grammar. If the grammar specification file requests
164it (via the <a href='#extraarg'><tt>extra_argument</tt> directive</a>),
165the Parse() function will have a fourth parameter that can be
drh75897232000-05-29 14:26:00 +0000[diff] [blame]166of any type chosen by the programmer. The parser doesn't do anything
167with this argument except to pass it through to action routines.
168This is a convenient mechanism for passing state information down
169to the action routines without having to use global variables.</p>
170
171<p>A typical use of a Lemon parser might look something like the
172following:
173<pre>
174 01 ParseTree *ParseFile(const char *zFilename){
175 02 Tokenizer *pTokenizer;
176 03 void *pParser;
177 04 Token sToken;
178 05 int hTokenId;
179 06 ParserState sState;
180 07
181 08 pTokenizer = TokenizerCreate(zFilename);
182 09 pParser = ParseAlloc( malloc );
183 10 InitParserState(&sState);
184 11 while( GetNextToken(pTokenizer, &hTokenId, &sToken) ){
185 12 Parse(pParser, hTokenId, sToken, &sState);
186 13 }
187 14 Parse(pParser, 0, sToken, &sState);
188 15 ParseFree(pParser, free );
189 16 TokenizerFree(pTokenizer);
190 17 return sState.treeRoot;
191 18 }
192</pre>
193This example shows a user-written routine that parses a file of
194text and returns a pointer to the parse tree.
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]195(All error-handling code is omitted from this example to keep it
drh75897232000-05-29 14:26:00 +0000[diff] [blame]196simple.)
197We assume the existence of some kind of tokenizer which is created
198using TokenizerCreate() on line 8 and deleted by TokenizerFree()
199on line 16. The GetNextToken() function on line 11 retrieves the
200next token from the input file and puts its type in the
201integer variable hTokenId. The sToken variable is assumed to be
202some kind of structure that contains details about each token,
203such as its complete text, what line it occurs on, etc. </p>
204
205<p>This example also assumes the existence of structure of type
206ParserState that holds state information about a particular parse.
207An instance of such a structure is created on line 6 and initialized
208on line 10. A pointer to this structure is passed into the Parse()
209routine as the optional 4th argument.
210The action routine specified by the grammar for the parser can use
211the ParserState structure to hold whatever information is useful and
212appropriate. In the example, we note that the treeRoot field of
213the ParserState structure is left pointing to the root of the parse
214tree.</p>
215
216<p>The core of this example as it relates to Lemon is as follows:
217<pre>
218 ParseFile(){
219 pParser = ParseAlloc( malloc );
220 while( GetNextToken(pTokenizer,&hTokenId, &sToken) ){
221 Parse(pParser, hTokenId, sToken);
222 }
223 Parse(pParser, 0, sToken);
224 ParseFree(pParser, free );
225 }
226</pre>
227Basically, what a program has to do to use a Lemon-generated parser
228is first create the parser, then send it lots of tokens obtained by
229tokenizing an input source. When the end of input is reached, the
230Parse() routine should be called one last time with a token type
231of 0. This step is necessary to inform the parser that the end of
232input has been reached. Finally, we reclaim memory used by the
233parser by calling ParseFree().</p>
234
235<p>There is one other interface routine that should be mentioned
236before we move on.
237The ParseTrace() function can be used to generate debugging output
238from the parser. A prototype for this routine is as follows:
239<pre>
240 ParseTrace(FILE *stream, char *zPrefix);
241</pre>
242After this routine is called, a short (one-line) message is written
243to the designated output stream every time the parser changes states
244or calls an action routine. Each such message is prefaced using
245the text given by zPrefix. This debugging output can be turned off
246by calling ParseTrace() again with a first argument of NULL (0).</p>
247
248<h3>Differences With YACC and BISON</h3>
249
250<p>Programmers who have previously used the yacc or bison parser
251generator will notice several important differences between yacc and/or
252bison and Lemon.
253<ul>
254<li>In yacc and bison, the parser calls the tokenizer. In Lemon,
255 the tokenizer calls the parser.
256<li>Lemon uses no global variables. Yacc and bison use global variables
257 to pass information between the tokenizer and parser.
258<li>Lemon allows multiple parsers to be running simultaneously. Yacc
259 and bison do not.
260</ul>
261These differences may cause some initial confusion for programmers
262with prior yacc and bison experience.
263But after years of experience using Lemon, I firmly
264believe that the Lemon way of doing things is better.</p>
265
drh45f31be2016-02-16 21:19:49 +0000[diff] [blame]266<p><i>Updated as of 2016-02-16:</i>
267The text above was written in the 1990s.
268We are told that Bison has lately been enhanced to support the
269tokenizer-calls-parser paradigm used by Lemon, and to obviate the
270need for global variables.</p>
271
drh75897232000-05-29 14:26:00 +0000[diff] [blame]272<h2>Input File Syntax</h2>
273
274<p>The main purpose of the grammar specification file for Lemon is
275to define the grammar for the parser. But the input file also
276specifies additional information Lemon requires to do its job.
277Most of the work in using Lemon is in writing an appropriate
278grammar file.</p>
279
280<p>The grammar file for lemon is, for the most part, free format.
281It does not have sections or divisions like yacc or bison. Any
282declaration can occur at any point in the file.
283Lemon ignores whitespace (except where it is needed to separate
284tokens) and it honors the same commenting conventions as C and C++.</p>
285
286<h3>Terminals and Nonterminals</h3>
287
288<p>A terminal symbol (token) is any string of alphanumeric
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]289and/or underscore characters
drh75897232000-05-29 14:26:00 +0000[diff] [blame]290that begins with an upper case letter.
drhc8eee5e2011-07-30 23:50:12 +0000[diff] [blame]291A terminal can contain lowercase letters after the first character,
drh75897232000-05-29 14:26:00 +0000[diff] [blame]292but the usual convention is to make terminals all upper case.
293A nonterminal, on the other hand, is any string of alphanumeric
294and underscore characters than begins with a lower case letter.
295Again, the usual convention is to make nonterminals use all lower
296case letters.</p>
297
298<p>In Lemon, terminal and nonterminal symbols do not need to
299be declared or identified in a separate section of the grammar file.
300Lemon is able to generate a list of all terminals and nonterminals
301by examining the grammar rules, and it can always distinguish a
302terminal from a nonterminal by checking the case of the first
303character of the name.</p>
304
305<p>Yacc and bison allow terminal symbols to have either alphanumeric
306names or to be individual characters included in single quotes, like
307this: ')' or '$'. Lemon does not allow this alternative form for
308terminal symbols. With Lemon, all symbols, terminals and nonterminals,
309must have alphanumeric names.</p>
310
311<h3>Grammar Rules</h3>
312
313<p>The main component of a Lemon grammar file is a sequence of grammar
314rules.
315Each grammar rule consists of a nonterminal symbol followed by
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]316the special symbol "::=" and then a list of terminals and/or nonterminals.
drh75897232000-05-29 14:26:00 +0000[diff] [blame]317The rule is terminated by a period.
318The list of terminals and nonterminals on the right-hand side of the
319rule can be empty.
320Rules can occur in any order, except that the left-hand side of the
321first rule is assumed to be the start symbol for the grammar (unless
322specified otherwise using the <tt>%start</tt> directive described below.)
323A typical sequence of grammar rules might look something like this:
324<pre>
325 expr ::= expr PLUS expr.
326 expr ::= expr TIMES expr.
327 expr ::= LPAREN expr RPAREN.
328 expr ::= VALUE.
329</pre>
330</p>
331
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]332<p>There is one non-terminal in this example, "expr", and five
333terminal symbols or tokens: "PLUS", "TIMES", "LPAREN",
334"RPAREN" and "VALUE".</p>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]335
336<p>Like yacc and bison, Lemon allows the grammar to specify a block
337of C code that will be executed whenever a grammar rule is reduced
338by the parser.
339In Lemon, this action is specified by putting the C code (contained
340within curly braces <tt>{...}</tt>) immediately after the
341period that closes the rule.
342For example:
343<pre>
344 expr ::= expr PLUS expr. { printf("Doing an addition...\n"); }
345</pre>
346</p>
347
348<p>In order to be useful, grammar actions must normally be linked to
349their associated grammar rules.
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]350In yacc and bison, this is accomplished by embedding a "$$" in the
drh75897232000-05-29 14:26:00 +0000[diff] [blame]351action to stand for the value of the left-hand side of the rule and
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]352symbols "$1", "$2", and so forth to stand for the value of
drh75897232000-05-29 14:26:00 +0000[diff] [blame]353the terminal or nonterminal at position 1, 2 and so forth on the
354right-hand side of the rule.
355This idea is very powerful, but it is also very error-prone. The
356single most common source of errors in a yacc or bison grammar is
357to miscount the number of symbols on the right-hand side of a grammar
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]358rule and say "$7" when you really mean "$8".</p>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]359
360<p>Lemon avoids the need to count grammar symbols by assigning symbolic
361names to each symbol in a grammar rule and then using those symbolic
362names in the action.
363In yacc or bison, one would write this:
364<pre>
365 expr -> expr PLUS expr { $$ = $1 + $3; };
366</pre>
367But in Lemon, the same rule becomes the following:
368<pre>
369 expr(A) ::= expr(B) PLUS expr(C). { A = B+C; }
370</pre>
371In the Lemon rule, any symbol in parentheses after a grammar rule
372symbol becomes a place holder for that symbol in the grammar rule.
373This place holder can then be used in the associated C action to
374stand for the value of that symbol.<p>
375
376<p>The Lemon notation for linking a grammar rule with its reduce
377action is superior to yacc/bison on several counts.
378First, as mentioned above, the Lemon method avoids the need to
379count grammar symbols.
380Secondly, if a terminal or nonterminal in a Lemon grammar rule
381includes a linking symbol in parentheses but that linking symbol
382is not actually used in the reduce action, then an error message
383is generated.
384For example, the rule
385<pre>
386 expr(A) ::= expr(B) PLUS expr(C). { A = B; }
387</pre>
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]388will generate an error because the linking symbol "C" is used
drh75897232000-05-29 14:26:00 +0000[diff] [blame]389in the grammar rule but not in the reduce action.</p>
390
391<p>The Lemon notation for linking grammar rules to reduce actions
392also facilitates the use of destructors for reclaiming memory
393allocated by the values of terminals and nonterminals on the
394right-hand side of a rule.</p>
395
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]396<a name='precrules'></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]397<h3>Precedence Rules</h3>
398
399<p>Lemon resolves parsing ambiguities in exactly the same way as
400yacc and bison. A shift-reduce conflict is resolved in favor
401of the shift, and a reduce-reduce conflict is resolved by reducing
402whichever rule comes first in the grammar file.</p>
403
404<p>Just like in
405yacc and bison, Lemon allows a measure of control
406over the resolution of paring conflicts using precedence rules.
407A precedence value can be assigned to any terminal symbol
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]408using the
409<a href='#pleft'>%left</a>,
410<a href='#pright'>%right</a> or
411<a href='#pnonassoc'>%nonassoc</a> directives. Terminal symbols
drh75897232000-05-29 14:26:00 +0000[diff] [blame]412mentioned in earlier directives have a lower precedence that
413terminal symbols mentioned in later directives. For example:</p>
414
415<p><pre>
416 %left AND.
417 %left OR.
418 %nonassoc EQ NE GT GE LT LE.
419 %left PLUS MINUS.
420 %left TIMES DIVIDE MOD.
421 %right EXP NOT.
422</pre></p>
423
424<p>In the preceding sequence of directives, the AND operator is
425defined to have the lowest precedence. The OR operator is one
426precedence level higher. And so forth. Hence, the grammar would
427attempt to group the ambiguous expression
428<pre>
429 a AND b OR c
430</pre>
431like this
432<pre>
433 a AND (b OR c).
434</pre>
435The associativity (left, right or nonassoc) is used to determine
436the grouping when the precedence is the same. AND is left-associative
437in our example, so
438<pre>
439 a AND b AND c
440</pre>
441is parsed like this
442<pre>
443 (a AND b) AND c.
444</pre>
445The EXP operator is right-associative, though, so
446<pre>
447 a EXP b EXP c
448</pre>
449is parsed like this
450<pre>
451 a EXP (b EXP c).
452</pre>
453The nonassoc precedence is used for non-associative operators.
454So
455<pre>
456 a EQ b EQ c
457</pre>
458is an error.</p>
459
460<p>The precedence of non-terminals is transferred to rules as follows:
461The precedence of a grammar rule is equal to the precedence of the
462left-most terminal symbol in the rule for which a precedence is
463defined. This is normally what you want, but in those cases where
464you want to precedence of a grammar rule to be something different,
465you can specify an alternative precedence symbol by putting the
466symbol in square braces after the period at the end of the rule and
467before any C-code. For example:</p>
468
469<p><pre>
470 expr = MINUS expr. [NOT]
471</pre></p>
472
473<p>This rule has a precedence equal to that of the NOT symbol, not the
474MINUS symbol as would have been the case by default.</p>
475
476<p>With the knowledge of how precedence is assigned to terminal
477symbols and individual
478grammar rules, we can now explain precisely how parsing conflicts
479are resolved in Lemon. Shift-reduce conflicts are resolved
480as follows:
481<ul>
482<li> If either the token to be shifted or the rule to be reduced
483 lacks precedence information, then resolve in favor of the
484 shift, but report a parsing conflict.
485<li> If the precedence of the token to be shifted is greater than
486 the precedence of the rule to reduce, then resolve in favor
487 of the shift. No parsing conflict is reported.
488<li> If the precedence of the token it be shifted is less than the
489 precedence of the rule to reduce, then resolve in favor of the
490 reduce action. No parsing conflict is reported.
491<li> If the precedences are the same and the shift token is
492 right-associative, then resolve in favor of the shift.
493 No parsing conflict is reported.
mistachkind5578432012-08-25 10:01:29 +0000[diff] [blame]494<li> If the precedences are the same the shift token is
drh75897232000-05-29 14:26:00 +0000[diff] [blame]495 left-associative, then resolve in favor of the reduce.
496 No parsing conflict is reported.
497<li> Otherwise, resolve the conflict by doing the shift and
498 report the parsing conflict.
499</ul>
500Reduce-reduce conflicts are resolved this way:
501<ul>
502<li> If either reduce rule
503 lacks precedence information, then resolve in favor of the
504 rule that appears first in the grammar and report a parsing
505 conflict.
506<li> If both rules have precedence and the precedence is different
507 then resolve the dispute in favor of the rule with the highest
508 precedence and do not report a conflict.
509<li> Otherwise, resolve the conflict by reducing by the rule that
510 appears first in the grammar and report a parsing conflict.
511</ul>
512
513<h3>Special Directives</h3>
514
515<p>The input grammar to Lemon consists of grammar rules and special
516directives. We've described all the grammar rules, so now we'll
517talk about the special directives.</p>
518
519<p>Directives in lemon can occur in any order. You can put them before
520the grammar rules, or after the grammar rules, or in the mist of the
521grammar rules. It doesn't matter. The relative order of
522directives used to assign precedence to terminals is important, but
523other than that, the order of directives in Lemon is arbitrary.</p>
524
525<p>Lemon supports the following special directives:
526<ul>
drhf2340fc2001-06-08 00:25:18 +0000[diff] [blame]527<li><tt>%code</tt>
528<li><tt>%default_destructor</tt>
529<li><tt>%default_type</tt>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]530<li><tt>%destructor</tt>
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]531<li><tt>%endif</tt>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]532<li><tt>%extra_argument</tt>
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]533<li><tt>%fallback</tt>
534<li><tt>%ifdef</tt>
535<li><tt>%ifndef</tt>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]536<li><tt>%include</tt>
537<li><tt>%left</tt>
538<li><tt>%name</tt>
539<li><tt>%nonassoc</tt>
540<li><tt>%parse_accept</tt>
541<li><tt>%parse_failure </tt>
542<li><tt>%right</tt>
543<li><tt>%stack_overflow</tt>
544<li><tt>%stack_size</tt>
545<li><tt>%start_symbol</tt>
546<li><tt>%syntax_error</tt>
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]547<li><tt>%token_class</tt>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]548<li><tt>%token_destructor</tt>
549<li><tt>%token_prefix</tt>
550<li><tt>%token_type</tt>
551<li><tt>%type</tt>
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]552<li><tt>%wildcard</tt>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]553</ul>
554Each of these directives will be described separately in the
555following sections:</p>
556
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]557<a name='pcode'></a>
drhf2340fc2001-06-08 00:25:18 +0000[diff] [blame]558<h4>The <tt>%code</tt> directive</h4>
559
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]560<p>The %code directive is used to specify addition C code that
drhf2340fc2001-06-08 00:25:18 +0000[diff] [blame]561is added to the end of the main output file. This is similar to
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]562the <a href='#pinclude'>%include</a> directive except that %include
563is inserted at the beginning of the main output file.</p>
drhf2340fc2001-06-08 00:25:18 +0000[diff] [blame]564
565<p>%code is typically used to include some action routines or perhaps
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]566a tokenizer or even the "main()" function
567as part of the output file.</p>
drhf2340fc2001-06-08 00:25:18 +0000[diff] [blame]568
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]569<a name='default_destructor'></a>
drhf2340fc2001-06-08 00:25:18 +0000[diff] [blame]570<h4>The <tt>%default_destructor</tt> directive</h4>
571
572<p>The %default_destructor directive specifies a destructor to
573use for non-terminals that do not have their own destructor
574specified by a separate %destructor directive. See the documentation
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]575on the <a name='#destructor'>%destructor</a> directive below for
576additional information.</p>
drhf2340fc2001-06-08 00:25:18 +0000[diff] [blame]577
578<p>In some grammers, many different non-terminal symbols have the
579same datatype and hence the same destructor. This directive is
580a convenience way to specify the same destructor for all those
581non-terminals using a single statement.</p>
582
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]583<a name='default_type'></a>
drhf2340fc2001-06-08 00:25:18 +0000[diff] [blame]584<h4>The <tt>%default_type</tt> directive</h4>
585
586<p>The %default_type directive specifies the datatype of non-terminal
587symbols that do no have their own datatype defined using a separate
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]588<a href='#ptype'>%type</a> directive.
589</p>
drhf2340fc2001-06-08 00:25:18 +0000[diff] [blame]590
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]591<a name='destructor'></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]592<h4>The <tt>%destructor</tt> directive</h4>
593
594<p>The %destructor directive is used to specify a destructor for
595a non-terminal symbol.
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]596(See also the <a href='#token_destructor'>%token_destructor</a>
597directive which is used to specify a destructor for terminal symbols.)</p>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]598
599<p>A non-terminal's destructor is called to dispose of the
600non-terminal's value whenever the non-terminal is popped from
601the stack. This includes all of the following circumstances:
602<ul>
603<li> When a rule reduces and the value of a non-terminal on
604 the right-hand side is not linked to C code.
605<li> When the stack is popped during error processing.
606<li> When the ParseFree() function runs.
607</ul>
608The destructor can do whatever it wants with the value of
609the non-terminal, but its design is to deallocate memory
610or other resources held by that non-terminal.</p>
611
612<p>Consider an example:
613<pre>
614 %type nt {void*}
615 %destructor nt { free($$); }
616 nt(A) ::= ID NUM. { A = malloc( 100 ); }
617</pre>
618This example is a bit contrived but it serves to illustrate how
619destructors work. The example shows a non-terminal named
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]620"nt" that holds values of type "void*". When the rule for
621an "nt" reduces, it sets the value of the non-terminal to
drh75897232000-05-29 14:26:00 +0000[diff] [blame]622space obtained from malloc(). Later, when the nt non-terminal
623is popped from the stack, the destructor will fire and call
624free() on this malloced space, thus avoiding a memory leak.
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]625(Note that the symbol "$$" in the destructor code is replaced
drh75897232000-05-29 14:26:00 +0000[diff] [blame]626by the value of the non-terminal.)</p>
627
628<p>It is important to note that the value of a non-terminal is passed
629to the destructor whenever the non-terminal is removed from the
630stack, unless the non-terminal is used in a C-code action. If
631the non-terminal is used by C-code, then it is assumed that the
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]632C-code will take care of destroying it.
633More commonly, the value is used to build some
drh75897232000-05-29 14:26:00 +0000[diff] [blame]634larger structure and we don't want to destroy it, which is why
635the destructor is not called in this circumstance.</p>
636
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]637<p>Destructors help avoid memory leaks by automatically freeing
638allocated objects when they go out of scope.
drh75897232000-05-29 14:26:00 +0000[diff] [blame]639To do the same using yacc or bison is much more difficult.</p>
640
drh45f31be2016-02-16 21:19:49 +0000[diff] [blame]641<a name="extraarg"></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]642<h4>The <tt>%extra_argument</tt> directive</h4>
643
644The %extra_argument directive instructs Lemon to add a 4th parameter
645to the parameter list of the Parse() function it generates. Lemon
646doesn't do anything itself with this extra argument, but it does
647make the argument available to C-code action routines, destructors,
648and so forth. For example, if the grammar file contains:</p>
649
650<p><pre>
651 %extra_argument { MyStruct *pAbc }
652</pre></p>
653
654<p>Then the Parse() function generated will have an 4th parameter
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]655of type "MyStruct*" and all action routines will have access to
656a variable named "pAbc" that is the value of the 4th parameter
drh75897232000-05-29 14:26:00 +0000[diff] [blame]657in the most recent call to Parse().</p>
658
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]659<a name='pfallback'></a>
660<h4>The <tt>%fallback</tt> directive</h4>
661
662<p>The %fallback directive specifies an alternative meaning for one
663or more tokens. The alternative meaning is tried if the original token
664would have generated a syntax error.
665
666<p>The %fallback directive was added to support robust parsing of SQL
667syntax in <a href="https://www.sqlite.org/">SQLite</a>.
668The SQL language contains a large assortment of keywords, each of which
669appears as a different token to the language parser. SQL contains so
670many keywords, that it can be difficult for programmers to keep up with
671them all. Programmers will, therefore, sometimes mistakenly use an
672obscure language keyword for an identifier. The %fallback directive
673provides a mechanism to tell the parser: "If you are unable to parse
674this keyword, try treating it as an identifier instead."
675
676<p>The syntax of %fallback is as follows:
677
678<blockquote>
679<tt>%fallback</tt> <i>ID</i> <i>TOKEN...</i> <b>.</b>
680</blockquote>
681
682<p>In words, the %fallback directive is followed by a list of token names
683terminated by a period. The first token name is the fallback token - the
684token to which all the other tokens fall back to. The second and subsequent
685arguments are tokens which fall back to the token identified by the first
686argument.
687
688<a name='pifdef'></a>
689<h4>The <tt>%ifdef</tt>, <tt>%ifndef</tt>, and <tt>%endif</tt> directives.</h4>
690
691<p>The %ifdef, %ifndef, and %endif directives are similar to
692#ifdef, #ifndef, and #endif in the C-preprocessor, just not as general.
693Each of these directives must begin at the left margin. No whitespace
694is allowed between the "%" and the directive name.
695
696<p>Grammar text in between "%ifdef MACRO" and the next nested "%endif" is
697ignored unless the "-DMACRO" command-line option is used. Grammar text
698betwen "%ifndef MACRO" and the next nested "%endif" is included except when
699the "-DMACRO" command-line option is used.
700
701<p>Note that the argument to %ifdef and %ifndef must be a single
702preprocessor symbol name, not a general expression. There is no "%else"
703directive.
704
705
706<a name='pinclude'></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]707<h4>The <tt>%include</tt> directive</h4>
708
709<p>The %include directive specifies C code that is included at the
710top of the generated parser. You can include any text you want --
drhf2340fc2001-06-08 00:25:18 +0000[diff] [blame]711the Lemon parser generator copies it blindly. If you have multiple
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]712%include directives in your grammar file, their values are concatenated
713so that all %include code ultimately appears near the top of the
714generated parser, in the same order as it appeared in the grammer.</p>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]715
716<p>The %include directive is very handy for getting some extra #include
717preprocessor statements at the beginning of the generated parser.
718For example:</p>
719
720<p><pre>
721 %include {#include &lt;unistd.h&gt;}
722</pre></p>
723
724<p>This might be needed, for example, if some of the C actions in the
725grammar call functions that are prototyed in unistd.h.</p>
726
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]727<a name='pleft'></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]728<h4>The <tt>%left</tt> directive</h4>
729
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]730The %left directive is used (along with the <a href='#pright'>%right</a> and
731<a href='#pnonassoc'>%nonassoc</a> directives) to declare precedences of
732terminal symbols. Every terminal symbol whose name appears after
733a %left directive but before the next period (".") is
drh75897232000-05-29 14:26:00 +0000[diff] [blame]734given the same left-associative precedence value. Subsequent
735%left directives have higher precedence. For example:</p>
736
737<p><pre>
738 %left AND.
739 %left OR.
740 %nonassoc EQ NE GT GE LT LE.
741 %left PLUS MINUS.
742 %left TIMES DIVIDE MOD.
743 %right EXP NOT.
744</pre></p>
745
746<p>Note the period that terminates each %left, %right or %nonassoc
747directive.</p>
748
749<p>LALR(1) grammars can get into a situation where they require
750a large amount of stack space if you make heavy use or right-associative
751operators. For this reason, it is recommended that you use %left
752rather than %right whenever possible.</p>
753
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]754<a name='pname'></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]755<h4>The <tt>%name</tt> directive</h4>
756
757<p>By default, the functions generated by Lemon all begin with the
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]758five-character string "Parse". You can change this string to something
drh75897232000-05-29 14:26:00 +0000[diff] [blame]759different using the %name directive. For instance:</p>
760
761<p><pre>
762 %name Abcde
763</pre></p>
764
765<p>Putting this directive in the grammar file will cause Lemon to generate
766functions named
767<ul>
768<li> AbcdeAlloc(),
769<li> AbcdeFree(),
770<li> AbcdeTrace(), and
771<li> Abcde().
772</ul>
773The %name directive allows you to generator two or more different
774parsers and link them all into the same executable.
775</p>
776
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]777<a name='pnonassoc'></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]778<h4>The <tt>%nonassoc</tt> directive</h4>
779
780<p>This directive is used to assign non-associative precedence to
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]781one or more terminal symbols. See the section on
782<a href='#precrules'>precedence rules</a>
783or on the <a href='#pleft'>%left</a> directive for additional information.</p>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]784
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]785<a name='parse_accept'></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]786<h4>The <tt>%parse_accept</tt> directive</h4>
787
788<p>The %parse_accept directive specifies a block of C code that is
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]789executed whenever the parser accepts its input string. To "accept"
drh75897232000-05-29 14:26:00 +0000[diff] [blame]790an input string means that the parser was able to process all tokens
791without error.</p>
792
793<p>For example:</p>
794
795<p><pre>
796 %parse_accept {
797 printf("parsing complete!\n");
798 }
799</pre></p>
800
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]801<a name='parse_failure'></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]802<h4>The <tt>%parse_failure</tt> directive</h4>
803
804<p>The %parse_failure directive specifies a block of C code that
805is executed whenever the parser fails complete. This code is not
806executed until the parser has tried and failed to resolve an input
807error using is usual error recovery strategy. The routine is
808only invoked when parsing is unable to continue.</p>
809
810<p><pre>
811 %parse_failure {
812 fprintf(stderr,"Giving up. Parser is hopelessly lost...\n");
813 }
814</pre></p>
815
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]816<a name='pright'></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]817<h4>The <tt>%right</tt> directive</h4>
818
819<p>This directive is used to assign right-associative precedence to
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]820one or more terminal symbols. See the section on
821<a href='#precrules'>precedence rules</a>
822or on the <a href='#pleft'>%left</a> directive for additional information.</p>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]823
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]824<a name='stack_overflow'></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]825<h4>The <tt>%stack_overflow</tt> directive</h4>
826
827<p>The %stack_overflow directive specifies a block of C code that
828is executed if the parser's internal stack ever overflows. Typically
829this just prints an error message. After a stack overflow, the parser
830will be unable to continue and must be reset.</p>
831
832<p><pre>
833 %stack_overflow {
834 fprintf(stderr,"Giving up. Parser stack overflow\n");
835 }
836</pre></p>
837
838<p>You can help prevent parser stack overflows by avoiding the use
839of right recursion and right-precedence operators in your grammar.
840Use left recursion and and left-precedence operators instead, to
841encourage rules to reduce sooner and keep the stack size down.
842For example, do rules like this:
843<pre>
844 list ::= list element. // left-recursion. Good!
845 list ::= .
846</pre>
847Not like this:
848<pre>
849 list ::= element list. // right-recursion. Bad!
850 list ::= .
851</pre>
852
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]853<a name='stack_size'></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]854<h4>The <tt>%stack_size</tt> directive</h4>
855
856<p>If stack overflow is a problem and you can't resolve the trouble
857by using left-recursion, then you might want to increase the size
858of the parser's stack using this directive. Put an positive integer
859after the %stack_size directive and Lemon will generate a parse
860with a stack of the requested size. The default value is 100.</p>
861
862<p><pre>
863 %stack_size 2000
864</pre></p>
865
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]866<a name='start_symbol'></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]867<h4>The <tt>%start_symbol</tt> directive</h4>
868
869<p>By default, the start-symbol for the grammar that Lemon generates
870is the first non-terminal that appears in the grammar file. But you
871can choose a different start-symbol using the %start_symbol directive.</p>
872
873<p><pre>
874 %start_symbol prog
875</pre></p>
876
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]877<a name='token_destructor'></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]878<h4>The <tt>%token_destructor</tt> directive</h4>
879
880<p>The %destructor directive assigns a destructor to a non-terminal
881symbol. (See the description of the %destructor directive above.)
882This directive does the same thing for all terminal symbols.</p>
883
884<p>Unlike non-terminal symbols which may each have a different data type
885for their values, terminals all use the same data type (defined by
886the %token_type directive) and so they use a common destructor. Other
887than that, the token destructor works just like the non-terminal
888destructors.</p>
889
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]890<a name='token_prefix'></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]891<h4>The <tt>%token_prefix</tt> directive</h4>
892
893<p>Lemon generates #defines that assign small integer constants
894to each terminal symbol in the grammar. If desired, Lemon will
895add a prefix specified by this directive
896to each of the #defines it generates.
897So if the default output of Lemon looked like this:
898<pre>
899 #define AND 1
900 #define MINUS 2
901 #define OR 3
902 #define PLUS 4
903</pre>
904You can insert a statement into the grammar like this:
905<pre>
906 %token_prefix TOKEN_
907</pre>
908to cause Lemon to produce these symbols instead:
909<pre>
910 #define TOKEN_AND 1
911 #define TOKEN_MINUS 2
912 #define TOKEN_OR 3
913 #define TOKEN_PLUS 4
914</pre>
915
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]916<a name='token_type'></a><a name='ptype'></a>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]917<h4>The <tt>%token_type</tt> and <tt>%type</tt> directives</h4>
918
919<p>These directives are used to specify the data types for values
920on the parser's stack associated with terminal and non-terminal
921symbols. The values of all terminal symbols must be of the same
922type. This turns out to be the same data type as the 3rd parameter
923to the Parse() function generated by Lemon. Typically, you will
924make the value of a terminal symbol by a pointer to some kind of
925token structure. Like this:</p>
926
927<p><pre>
928 %token_type {Token*}
929</pre></p>
930
931<p>If the data type of terminals is not specified, the default value
drhdfe4e6b2016-10-08 13:34:08 +0000[diff] [blame^]932is "void*".</p>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]933
934<p>Non-terminal symbols can each have their own data types. Typically
935the data type of a non-terminal is a pointer to the root of a parse-tree
936structure that contains all information about that non-terminal.
937For example:</p>
938
939<p><pre>
940 %type expr {Expr*}
941</pre></p>
942
943<p>Each entry on the parser's stack is actually a union containing
944instances of all data types for every non-terminal and terminal symbol.
945Lemon will automatically use the correct element of this union depending
946on what the corresponding non-terminal or terminal symbol is. But
947the grammar designer should keep in mind that the size of the union
948will be the size of its largest element. So if you have a single
949non-terminal whose data type requires 1K of storage, then your 100
950entry parser stack will require 100K of heap space. If you are willing
951and able to pay that price, fine. You just need to know.</p>
952
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]953<a name='pwildcard'></a>
954<h4>The <tt>%wildcard</tt> directive</h4>
955
956<p>The %wildcard directive is followed by a single token name and a
957period. This directive specifies that the identified token should
958match any input token.
959
960<p>When the generated parser has the choice of matching an input against
961the wildcard token and some other token, the other token is always used.
962The wildcard token is only matched if there are no other alternatives.
963
drh75897232000-05-29 14:26:00 +0000[diff] [blame]964<h3>Error Processing</h3>
965
966<p>After extensive experimentation over several years, it has been
967discovered that the error recovery strategy used by yacc is about
968as good as it gets. And so that is what Lemon uses.</p>
969
970<p>When a Lemon-generated parser encounters a syntax error, it
971first invokes the code specified by the %syntax_error directive, if
972any. It then enters its error recovery strategy. The error recovery
973strategy is to begin popping the parsers stack until it enters a
974state where it is permitted to shift a special non-terminal symbol
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]975named "error". It then shifts this non-terminal and continues
drh75897232000-05-29 14:26:00 +0000[diff] [blame]976parsing. But the %syntax_error routine will not be called again
977until at least three new tokens have been successfully shifted.</p>
978
979<p>If the parser pops its stack until the stack is empty, and it still
980is unable to shift the error symbol, then the %parse_failed routine
981is invoked and the parser resets itself to its start state, ready
982to begin parsing a new file. This is what will happen at the very
983first syntax error, of course, if there are no instances of the
drh9bccde32016-03-19 18:00:44 +0000[diff] [blame]984"error" non-terminal in your grammar.</p>
drh75897232000-05-29 14:26:00 +0000[diff] [blame]985
986</body>
987</html>