1YACC(1) User Commands YACC(1)
2
6 Yacc - an LALR(1) parser generator
7
9 yacc [ -BdghilLPrtvVy ] [ -b file_prefix ] [ -H defines_file ] [ -o
10 output_file ] [ -p symbol_prefix ] filename
11
13 Yacc reads the grammar specification in the file filename and generates
14 an LALR(1) parser for it. The parsers consist of a set of LALR(1)
15 parsing tables and a driver routine written in the C programming lan‐
16 guage. Yacc normally writes the parse tables and the driver routine to
17 the file y.tab.c.
18 The following options are available:
20 -b file_prefix
22 The -b option changes the prefix prepended to the output file
23 names to the string denoted by file_prefix. The default prefix is
24 the character y.
25 -B create a backtracking parser (compile-time configuration for bty‐
27 acc).
28 -d causes the header file y.tab.h to be written. It contains #de‐
30 fine's for the token identifiers.
31 -h print a usage message to the standard error.
33 -H defines_file
35 causes #define's for the token identifiers to be written to the
36 given defines_file rather than the y.tab.h file used by the -d op‐
37 tion.
38 -g The -g option causes a graphical description of the generated
40 LALR(1) parser to be written to the file y.dot in graphviz format,
41 ready to be processed by dot(1).
42 -i The -i option causes a supplementary header file y.tab.i to be
44 written. It contains extern declarations and supplementary #de‐
45 fine's as needed to map the conventional yacc yy-prefixed names to
46 whatever the -p option may specify. The code file, e.g., y.tab.c
47 is modified to #include this file as well as the y.tab.h file, en‐
48 forcing consistent usage of the symbols defined in those files.
49 The supplementary header file makes it simpler to separate compi‐
51 lation of lex- and yacc-files.
52 -l If the -l option is not specified, yacc will insert #line direc‐
54 tives in the generated code. The #line directives let the C com‐
55 piler relate errors in the generated code to the user's original
56 code. If the -l option is specified, yacc will not insert the
57 #line directives. #line directives specified by the user will be
58 retained.
59 -L enable position processing, e.g., “%locations” (compile-time con‐
61 figuration for btyacc).
62 -o output_file
64 specify the filename for the parser file. If this option is not
65 given, the output filename is the file prefix concatenated with
66 the file suffix, e.g., y.tab.c. This overrides the -b option.
67 -p symbol_prefix
69 The -p option changes the prefix prepended to yacc-generated sym‐
70 bols to the string denoted by symbol_prefix. The default prefix
71 is the string yy.
72 -P create a reentrant parser, e.g., “%pure-parser”.
74 -r The -r option causes yacc to produce separate files for code and
76 tables. The code file is named y.code.c, and the tables file is
77 named y.tab.c. The prefix “y.” can be overridden using the -b op‐
78 tion.
79 -s suppress “#define” statements generated for string literals in a
81 “%token” statement, to more closely match original yacc behavior.
82 Normally when yacc sees a line such as
84 %token OP_ADD "ADD"
86 it notices that the quoted “ADD” is a valid C identifier, and gen‐
88 erates a #define not only for OP_ADD, but for ADD as well, e.g.,
89 #define OP_ADD 257
91 #define ADD 258
92 The original yacc does not generate the second “#define”. The -s
94 option suppresses this “#define”.
95 POSIX (IEEE 1003.1 2004) documents only names and numbers for
97 “%token”, though original yacc and bison also accept string liter‐
98 als.
99 -t The -t option changes the preprocessor directives generated by
101 yacc so that debugging statements will be incorporated in the com‐
102 piled code.
103 Yacc sends debugging output to the standard output (compatible
105 with both the original yacc and btyacc), while btyacc writes de‐
106 bugging output to the standard error (like bison).
107 -v The -v option causes a human-readable description of the generated
109 parser to be written to the file y.output.
110 -V print the version number to the standard output.
112 -y yacc ignores this option, which bison supports for ostensible
114 POSIX compatibility.
115 The filename parameter is not optional. However, yacc accepts a single
117 “-” to read the grammar from the standard input. A double “--” marker
118 denotes the end of options. A single filename parameter is expected
119 after a “--” marker.
120
122 Yacc provides some extensions for compatibility with bison and other
123 implementations of yacc. It accepts several long options which have
124 equivalents in yacc. The %destructor and %locations features are
125 available only if yacc has been configured and compiled to support the
126 back-tracking (btyacc) functionality. The remaining features are al‐
127 ways available:
128 %code keyword { code }
130 Adds the indicated source code at a given point in the output
131 file. The optional keyword tells yacc where to insert the code:
132 top just after the version-definition in the generated code-
134 file.
135 requires
137 just after the declaration of public parser variables. If
138 the -d option is given, the code is inserted at the begin‐
139 ning of the defines-file.
140 provides
142 just after the declaration of private parser variables. If
143 the -d option is given, the code is inserted at the end of
144 the defines-file.
145 If no keyword is given, the code is inserted at the beginning of
147 the section of code copied verbatim from the source file. Mul‐
148 tiple %code directives may be given; yacc inserts those into the
149 corresponding code- or defines-file in the order that they ap‐
150 pear in the source file.
151 %debug
153 This has the same effect as the “-t” command-line option.
154 %destructor { code } symbol+
156 defines code that is invoked when a symbol is automatically dis‐
157 carded during error recovery. This code can be used to reclaim
158 dynamically allocated memory associated with the corresponding
159 semantic value for cases where user actions cannot manage the
160 memory explicitly.
161 On encountering a parse error, the generated parser discards
163 symbols on the stack and input tokens until it reaches a state
164 that will allow parsing to continue. This error recovery ap‐
165 proach results in a memory leak if the YYSTYPE value is, or con‐
166 tains, pointers to dynamically allocated memory.
167 The bracketed code is invoked whenever the parser discards one
169 of the symbols. Within code, “$$” or “$<tag>$” designates the
170 semantic value associated with the discarded symbol, and “@$”
171 designates its location (see %locations directive).
172 A per-symbol destructor is defined by listing a grammar symbol
174 in symbol+. A per-type destructor is defined by listing a se‐
175 mantic type tag (e.g., “<some_tag>”) in symbol+; in this case,
176 the parser will invoke code whenever it discards any grammar
177 symbol that has that semantic type tag, unless that symbol has
178 its own per-symbol destructor.
179 Two categories of default destructor are supported that are in‐
181 voked when discarding any grammar symbol that has no per-symbol
182 and no per-type destructor:
183 • the code for “<*>” is used for grammar symbols that have an
185 explicitly declared semantic type tag (via “%type”);
186 • the code for “<>” is used for grammar symbols that have no
188 declared semantic type tag.
189 %empty
191 ignored by yacc.
192 %expect number
194 tells yacc the expected number of shift/reduce conflicts. That
195 makes it only report the number if it differs.
196 %expect-rr number
198 tell yacc the expected number of reduce/reduce conflicts. That
199 makes it only report the number if it differs. This is (unlike
200 bison) allowable in LALR parsers.
201 %locations
203 tells yacc to enable management of position information associ‐
204 ated with each token, provided by the lexer in the global vari‐
205 able yylloc, similar to management of semantic value information
206 provided in yylval.
207 As for semantic values, locations can be referenced within ac‐
209 tions using @$ to refer to the location of the left hand side
210 symbol, and @N (N an integer) to refer to the location of one of
211 the right hand side symbols. Also as for semantic values, when
212 a rule is matched, a default action is used the compute the lo‐
213 cation represented by @$ as the beginning of the first symbol
214 and the end of the last symbol in the right hand side of the
215 rule. This default computation can be overridden by explicit
216 assignment to @$ in a rule action.
217 The type of yylloc is YYLTYPE, which is defined by default as:
219 typedef struct YYLTYPE {
221 int first_line;
222 int first_column;
223 int last_line;
224 int last_column;
225 } YYLTYPE;
226 YYLTYPE can be redefined by the user (YYLTYPE_IS_DEFINED must be
228 defined, to inhibit the default) in the declarations section of
229 the specification file. As in bison, the macro YYLLOC_DEFAULT
230 is invoked each time a rule is matched to calculate a position
231 for the left hand side of the rule, before the associated action
232 is executed; this macro can be redefined by the user.
233 This directive adds a YYLTYPE parameter to yyerror(). If the
235 %pure-parser directive is present, a YYLTYPE parameter is added
236 to yylex() calls.
237 %lex-param { argument-declaration }
239 By default, the lexer accepts no parameters, e.g., yylex(). Use
240 this directive to add parameter declarations for your customized
241 lexer.
242 %parse-param { argument-declaration }
244 By default, the parser accepts no parameters, e.g., yyparse().
245 Use this directive to add parameter declarations for your cus‐
246 tomized parser.
247 %pure-parser
249 Most variables (other than yydebug and yynerrs) are allocated on
250 the stack within yyparse, making the parser reasonably reen‐
251 trant.
252 %token-table
254 Make the parser's names for tokens available in the yytname ar‐
255 ray. However, yacc does not predefine “$end”, “$error” or “$un‐
256 defined” in this array.
257
259 According to Robert Corbett,
260 Berkeley Yacc is an LALR(1) parser generator. Berkeley Yacc
262 has been made as compatible as possible with AT&T Yacc.
263 Berkeley Yacc can accept any input specification that
264 conforms to the AT&T Yacc documentation. Specifications
265 that take advantage of undocumented features of AT&T Yacc
266 will probably be rejected.
267 The rationale in
269 http://pubs.opengroup.org/onlinepubs/9699919799/utilities/yacc.html
271 documents some features of AT&T yacc which are no longer required for
273 POSIX compliance.
274 That said, you may be interested in reusing grammar files with some
276 other implementation which is not strictly compatible with AT&T yacc.
277 For instance, there is bison. Here are a few differences:
278 • Yacc accepts an equals mark preceding the left curly brace of an
280 action (as in the original grammar file ftp.y):
281 | STAT CRLF
283 = {
284 statcmd();
285 }
286 • Yacc and bison emit code in different order, and in particular bi‐
288 son makes forward reference to common functions such as yylex, yy‐
289 parse and yyerror without providing prototypes.
290 • Bison's support for “%expect” is broken in more than one release.
292 For best results using bison, delete that directive.
293 • Bison has no equivalent for some of yacc's command-line options,
295 relying on directives embedded in the grammar file.
296 • Bison's “-y” option does not affect bison's lack of support for
298 features of AT&T yacc which were deemed obsolescent.
299 • Yacc accepts multiple parameters with %lex-param and %parse-param
301 in two forms
302 {type1 name1} {type2 name2} ...
304 {type1 name1, type2 name2 ...}
305 Bison accepts the latter (though undocumented), but depending on
307 the release may generate bad code.
308 • Like bison, yacc will add parameters specified via %parse-param to
310 yyparse, yyerror and (if configured for back-tracking) to the de‐
311 structor declared using %destructor. Bison puts the additional pa‐
312 rameters first for yyparse and yyerror but last for destructors.
313 Yacc matches this behavior.
314
316 If there are rules that are never reduced, the number of such rules is
317 reported on standard error. If there are any LALR(1) conflicts, the
318 number of conflicts is reported on standard error.
319Berkeley Yacc November 6, 2022 YACC(1)