#include "symlist.h"
#include "scan-gram.h"
#include "scan-code.h"
+#include "xmemdup0.h"
#define YYLLOC_DEFAULT(Current, Rhs, N) (Current) = lloc_default (Rhs, N)
static YYLTYPE lloc_default (YYLTYPE const *, int);
#define YY_LOCATION_PRINT(File, Loc) \
- location_print (File, Loc)
+ location_print (File, Loc)
static void version_check (location const *loc, char const *version);
FIXME: depends on the undocumented availability of YYLLOC. */
#undef yyerror
#define yyerror(Msg) \
- gram_error (&yylloc, Msg)
+ gram_error (&yylloc, Msg)
static void gram_error (location const *, char const *);
static char const *char_name (char);
+%}
-/** Add a lex-param or a parse-param.
- *
- * \param type \a lex_param or \a parse_param
- * \param decl the formal argument
- * \param loc the location in the source.
- */
-static void add_param (char const *type, char *decl, location loc);
-
-
-static symbol_class current_class = unknown_sym;
-static uniqstr current_type = NULL;
-static symbol *current_lhs_symbol;
-static location current_lhs_location;
-static named_ref *current_lhs_named_ref;
-static int current_prec = 0;
-
-/** Set the new current left-hand side symbol, possibly common
- * to several right-hand side parts of rule.
- */
-static
-void
-current_lhs(symbol *sym, location loc, named_ref *ref)
+%code
{
- current_lhs_symbol = sym;
- current_lhs_location = loc;
- /* In order to simplify memory management, named references for lhs
- are always assigned by deep copy into the current symbol_list
- node. This is because a single named-ref in the grammar may
- result in several uses when the user factors lhs between several
- rules using "|". Therefore free the parser's original copy. */
- free (current_lhs_named_ref);
- current_lhs_named_ref = ref;
-}
-
+ static int current_prec = 0;
+ static location current_lhs_location;
+ static named_ref *current_lhs_named_ref;
+ static symbol *current_lhs_symbol;
+ static symbol_class current_class = unknown_sym;
+ static uniqstr current_type = NULL;
+
+ /** Set the new current left-hand side symbol, possibly common
+ * to several right-hand side parts of rule.
+ */
+ static
+ void
+ current_lhs(symbol *sym, location loc, named_ref *ref)
+ {
+ current_lhs_symbol = sym;
+ current_lhs_location = loc;
+ /* In order to simplify memory management, named references for lhs
+ are always assigned by deep copy into the current symbol_list
+ node. This is because a single named-ref in the grammar may
+ result in several uses when the user factors lhs between several
+ rules using "|". Therefore free the parser's original copy. */
+ free (current_lhs_named_ref);
+ current_lhs_named_ref = ref;
+ }
-#define YYTYPE_INT16 int_fast16_t
-#define YYTYPE_INT8 int_fast8_t
-#define YYTYPE_UINT16 uint_fast16_t
-#define YYTYPE_UINT8 uint_fast8_t
-%}
+ #define YYTYPE_INT16 int_fast16_t
+ #define YYTYPE_INT8 int_fast8_t
+ #define YYTYPE_UINT16 uint_fast16_t
+ #define YYTYPE_UINT8 uint_fast8_t
+}
%debug
%verbose
%defines
%locations
%pure-parser
-%error-verbose
+%define parse.error "verbose"
%define parse.lac full
%name-prefix="gram_"
%expect 0
%union
{
+ assoc assoc;
+ char *code;
+ char const *chars;
+ int integer;
+ named_ref *named_ref;
symbol *symbol;
symbol_list *list;
- int integer;
- char const *chars;
- char *code;
- assoc assoc;
uniqstr uniqstr;
unsigned char character;
- named_ref *named_ref;
};
/* Define the tokens together with their human representation. */
%token PERCENT_LEFT "%left"
%token PERCENT_RIGHT "%right"
%token PERCENT_NONASSOC "%nonassoc"
+%token PERCENT_PRECEDENCE "%precedence"
%token PERCENT_PREC "%prec"
%token PERCENT_DPREC "%dprec"
%token PERCENT_MERGE "%merge"
-
/*----------------------.
| Global Declarations. |
`----------------------*/
%token
PERCENT_CODE "%code"
- PERCENT_DEBUG "%debug"
PERCENT_DEFAULT_PREC "%default-prec"
PERCENT_DEFINE "%define"
PERCENT_DEFINES "%defines"
PERCENT_ERROR_VERBOSE "%error-verbose"
PERCENT_EXPECT "%expect"
- PERCENT_EXPECT_RR "%expect-rr"
+ PERCENT_EXPECT_RR "%expect-rr"
+ PERCENT_FLAG "%<flag>"
PERCENT_FILE_PREFIX "%file-prefix"
PERCENT_GLR_PARSER "%glr-parser"
PERCENT_INITIAL_ACTION "%initial-action"
PERCENT_LANGUAGE "%language"
- PERCENT_LEX_PARAM "%lex-param"
- PERCENT_LOCATIONS "%locations"
PERCENT_NAME_PREFIX "%name-prefix"
PERCENT_NO_DEFAULT_PREC "%no-default-prec"
PERCENT_NO_LINES "%no-lines"
PERCENT_NONDETERMINISTIC_PARSER
- "%nondeterministic-parser"
+ "%nondeterministic-parser"
PERCENT_OUTPUT "%output"
- PERCENT_PARSE_PARAM "%parse-param"
- PERCENT_PURE_PARSER "%pure-parser"
- PERCENT_REQUIRE "%require"
+ PERCENT_REQUIRE "%require"
PERCENT_SKELETON "%skeleton"
PERCENT_START "%start"
PERCENT_TOKEN_TABLE "%token-table"
;
%token BRACED_CODE "{...}"
+%token BRACED_PREDICATE "%?{...}"
%token BRACKETED_ID "[identifier]"
%token CHAR "char"
%token EPILOGUE "epilogue"
%token PIPE "|"
%token PROLOGUE "%{...%}"
%token SEMICOLON ";"
-%token TYPE "type"
-%token TYPE_TAG_ANY "<*>"
-%token TYPE_TAG_NONE "<>"
+%token TAG "<tag>"
+%token TAG_ANY "<*>"
+%token TAG_NONE "<>"
%type <character> CHAR
%printer { fputs (char_name ($$), stderr); } CHAR
/* braceless is not to be used for rule or symbol actions, as it
calls code_props_plain_init. */
%type <chars> STRING "%{...%}" EPILOGUE braceless content.opt
-%type <code> "{...}"
+%type <code> "{...}" "%?{...}"
%printer { fputs (quotearg_style (c_quoting_style, $$), stderr); }
- STRING
+ STRING
%printer { fprintf (stderr, "{\n%s\n}", $$); }
- braceless content.opt "{...}" "%{...%}" EPILOGUE
+ braceless content.opt "{...}" "%{...%}" EPILOGUE
-%type <uniqstr> BRACKETED_ID ID ID_COLON TYPE variable
+%type <uniqstr> BRACKETED_ID ID ID_COLON PERCENT_FLAG TAG variable
%printer { fputs ($$, stderr); } <uniqstr>
%printer { fprintf (stderr, "[%s]", $$); } BRACKETED_ID
%printer { fprintf (stderr, "%s:", $$); } ID_COLON
-%printer { fprintf (stderr, "<%s>", $$); } TYPE
+%printer { fprintf (stderr, "%%%s", $$); } PERCENT_FLAG
+%printer { fprintf (stderr, "<%s>", $$); } TAG
%type <integer> INT
%printer { fprintf (stderr, "%d", $$); } <integer>
%type <list> symbols.1 symbols.prec generic_symlist generic_symlist_item
%type <named_ref> named_ref.opt
+/*---------.
+| %param. |
+`---------*/
+%code requires
+{
+# ifndef PARAM_TYPE
+# define PARAM_TYPE
+ typedef enum
+ {
+ param_none = 0,
+ param_lex = 1 << 0,
+ param_parse = 1 << 1,
+ param_both = param_lex | param_parse
+ } param_type;
+# endif
+};
+%code
+{
+ /** Add a lex-param and/or a parse-param.
+ *
+ * \param type where to push this formal argument.
+ * \param decl the formal argument. Destroyed.
+ * \param loc the location in the source.
+ */
+ static void add_param (param_type type, char *decl, location loc);
+ static param_type current_param = param_none;
+};
+%union
+{
+ param_type param;
+}
+%token <param> PERCENT_PARAM "%param";
+%printer
+{
+ switch ($$)
+ {
+#define CASE(In, Out) \
+ case param_ ## In: fputs ("%" #Out, stderr); break
+ CASE(lex, lex-param);
+ CASE(parse, parse-param);
+ CASE(both, param);
+#undef CASE
+ case param_none: aver (false); break;
+ }
+} <param>;
+
+
+ /*==========\
+ | Grammar. |
+ \==========*/
%%
input:
;
- /*------------------------------------.
- | Declarations: before the first %%. |
- `------------------------------------*/
+ /*------------------------------------.
+ | Declarations: before the first %%. |
+ `------------------------------------*/
prologue_declarations:
/* Nothing */
plain_code.code, @1);
code_scanner_last_string_free ();
}
-| "%debug" { debug_flag = true; }
+| "%<flag>"
+ {
+ muscle_percent_define_ensure ($1, @1, true);
+ }
| "%define" variable content.opt
{
muscle_percent_define_insert ($2, @2, $3,
defines_flag = true;
spec_defines_file = xstrdup ($2);
}
-| "%error-verbose" { error_verbose = true; }
+| "%error-verbose"
+ {
+ muscle_percent_define_insert ("parse.error", @1, "verbose",
+ MUSCLE_PERCENT_DEFINE_GRAMMAR_FILE);
+ }
| "%expect" INT { expected_sr_conflicts = $2; }
-| "%expect-rr" INT { expected_rr_conflicts = $2; }
+| "%expect-rr" INT { expected_rr_conflicts = $2; }
| "%file-prefix" STRING { spec_file_prefix = $2; }
| "%file-prefix" "=" STRING { spec_file_prefix = $3; } /* deprecated */
| "%glr-parser"
muscle_code_grow ("initial_action", action.code, @2);
code_scanner_last_string_free ();
}
-| "%language" STRING { language_argmatch ($2, grammar_prio, @1); }
-| "%lex-param" "{...}" { add_param ("lex_param", $2, @2); }
-| "%locations" { locations_flag = true; }
+| "%language" STRING { language_argmatch ($2, grammar_prio, @1); }
| "%name-prefix" STRING { spec_name_prefix = $2; }
| "%name-prefix" "=" STRING { spec_name_prefix = $3; } /* deprecated */
| "%no-lines" { no_lines_flag = true; }
-| "%nondeterministic-parser" { nondeterministic_parser = true; }
+| "%nondeterministic-parser" { nondeterministic_parser = true; }
| "%output" STRING { spec_outfile = $2; }
| "%output" "=" STRING { spec_outfile = $3; } /* deprecated */
-| "%parse-param" "{...}" { add_param ("parse_param", $2, @2); }
-| "%pure-parser"
- {
- /* %pure-parser is deprecated in favor of `%define api.pure', so use
- `%define api.pure' in a backward-compatible manner here. First, don't
- complain if %pure-parser is specified multiple times. */
- if (!muscle_find_const ("percent_define(api.pure)"))
- muscle_percent_define_insert ("api.pure", @1, "",
- MUSCLE_PERCENT_DEFINE_GRAMMAR_FILE);
- /* In all cases, use api.pure now so that the backend doesn't complain if
- the skeleton ignores api.pure, but do warn now if there's a previous
- conflicting definition from an actual %define. */
- if (!muscle_percent_define_flag_if ("api.pure"))
- muscle_percent_define_insert ("api.pure", @1, "",
- MUSCLE_PERCENT_DEFINE_GRAMMAR_FILE);
- }
+| "%param" { current_param = $1; } params { current_param = param_none; }
| "%require" STRING { version_check (&@2, $2); }
| "%skeleton" STRING
{
char const *skeleton_user = $2;
- if (mbschr (skeleton_user, '/'))
+ if (strchr (skeleton_user, '/'))
{
size_t dir_length = strlen (current_file);
char *skeleton_build;
xmalloc (dir_length + 1 + strlen (skeleton_user) + 1);
if (dir_length > 0)
{
- strncpy (skeleton_build, current_file, dir_length);
+ memcpy (skeleton_build, current_file, dir_length);
skeleton_build[dir_length++] = '/';
}
strcpy (skeleton_build + dir_length, skeleton_user);
| /*FIXME: Err? What is this horror doing here? */ ";"
;
+params:
+ params "{...}" { add_param (current_param, $2, @2); }
+| "{...}" { add_param (current_param, $1, @1); }
+;
+
+
+/*----------------------.
+| grammar_declaration. |
+`----------------------*/
+
grammar_declaration:
precedence_declaration
| symbol_declaration
{
symbol_list *list;
for (list = $3; list; list = list->next)
- symbol_list_destructor_set (list, $2, @2);
+ symbol_list_code_props_set (list, destructor, @2, $2);
symbol_list_free ($3);
}
| "%printer" "{...}" generic_symlist
{
symbol_list *list;
for (list = $3; list; list = list->next)
- symbol_list_printer_set (list, $2, @2);
+ symbol_list_code_props_set (list, printer, @2, $2);
symbol_list_free ($3);
}
| "%default-prec"
;
-/*----------*
- | %union. |
- *----------*/
+/*---------.
+| %union. |
+`---------*/
%token PERCENT_UNION "%union";
current_class = unknown_sym;
current_type = NULL;
}
-| "%type" TYPE symbols.1
+| "%type" TAG symbols.1
{
symbol_list *list;
tag_seen = true;
for (list = $3; list; list = list->next)
- symbol_type_set (list->content.sym, $2, @2);
+ symbol_type_set (list->content.sym, $2, @2);
symbol_list_free ($3);
}
;
precedence_declaration:
- precedence_declarator type.opt symbols.prec
+ precedence_declarator tag.opt symbols.prec
{
symbol_list *list;
++current_prec;
for (list = $3; list; list = list->next)
- {
- symbol_type_set (list->content.sym, current_type, @2);
- symbol_precedence_set (list->content.sym, current_prec, $1, @1);
- }
+ {
+ symbol_type_set (list->content.sym, current_type, @2);
+ symbol_precedence_set (list->content.sym, current_prec, $1, @1);
+ }
symbol_list_free ($3);
current_type = NULL;
}
;
precedence_declarator:
- "%left" { $$ = left_assoc; }
-| "%right" { $$ = right_assoc; }
-| "%nonassoc" { $$ = non_assoc; }
+ "%left" { $$ = left_assoc; }
+| "%right" { $$ = right_assoc; }
+| "%nonassoc" { $$ = non_assoc; }
+| "%precedence" { $$ = precedence_assoc; }
;
-type.opt:
+tag.opt:
/* Nothing. */ { current_type = NULL; }
-| TYPE { current_type = $1; tag_seen = true; }
+| TAG { current_type = $1; tag_seen = true; }
;
/* Just like symbols.1 but accept INT for the sake of POSIX. */
;
generic_symlist_item:
- symbol { $$ = symbol_list_sym_new ($1, @1); }
-| TYPE { $$ = symbol_list_type_new ($1, @1); }
-| "<*>" { $$ = symbol_list_default_tagged_new (@1); }
-| "<>" { $$ = symbol_list_default_tagless_new (@1); }
+ symbol { $$ = symbol_list_sym_new ($1, @1); }
+| TAG { $$ = symbol_list_type_new ($1, @1); }
+| "<*>" { $$ = symbol_list_default_tagged_new (@1); }
+| "<>" { $$ = symbol_list_default_tagless_new (@1); }
;
/* One token definition. */
symbol_def:
- TYPE
+ TAG
{
current_type = $1;
tag_seen = true;
;
- /*------------------------------------------.
- | The grammar section: between the two %%. |
- `------------------------------------------*/
+ /*------------------------------------------.
+ | The grammar section: between the two %%. |
+ `------------------------------------------*/
grammar:
rules_or_grammar_declaration
rhs:
/* Nothing. */
{ grammar_current_rule_begin (current_lhs_symbol, current_lhs_location,
- current_lhs_named_ref); }
+ current_lhs_named_ref); }
| rhs symbol named_ref.opt
{ grammar_current_rule_symbol_append ($2, @2, $3); }
| rhs "{...}" named_ref.opt
- { grammar_current_rule_action_append ($2, @2, $3); }
+ { grammar_current_rule_action_append ($2, @2, $3, false); }
+| rhs "%?{...}"
+ { grammar_current_rule_action_append ($2, @2, NULL, true); }
| rhs "%prec" symbol
{ grammar_current_rule_prec_set ($3, @3); }
| rhs "%dprec" INT
{ grammar_current_rule_dprec_set ($3, @3); }
-| rhs "%merge" TYPE
+| rhs "%merge" TAG
{ grammar_current_rule_merge_set ($3, @3); }
;
BRACKETED_ID { $$ = named_ref_new($1, @1); }
;
-
-/*----------------------------*
- | variable and content.opt. |
- *---------------------------*/
+/*---------------------------.
+| variable and content.opt. |
+`---------------------------*/
/* The STRING form of variable is deprecated and is not M4-friendly.
For example, M4 fails for `%define "[" "value"'. */
;
-/*-------------*
- | braceless. |
- *-------------*/
+/*------------.
+| braceless. |
+`------------*/
braceless:
"{...}"
;
-/*---------------*
- | Identifiers. |
- *---------------*/
+/*--------------.
+| Identifiers. |
+`--------------*/
/* Identifiers are returned as uniqstr values by the scanner.
Depending on their use, we may need to make them genuine symbols. */
for (i = 1; i <= n; i++)
if (! equal_boundaries (rhs[i].start, rhs[i].end))
{
- loc.start = rhs[i].start;
- break;
+ loc.start = rhs[i].start;
+ break;
}
return loc;
}
-/* Add a lex-param or a parse-param (depending on TYPE) with
- declaration DECL and location LOC. */
-
static void
-add_param (char const *type, char *decl, location loc)
+add_param (param_type type, char *decl, location loc)
{
static char const alphanum[26 + 26 + 1 + 10] =
"abcdefghijklmnopqrstuvwxyz"
"ABCDEFGHIJKLMNOPQRSTUVWXYZ"
"_"
"0123456789";
+
char const *name_start = NULL;
- char *p;
-
- /* Stop on last actual character. */
- for (p = decl; p[1]; p++)
- if ((p == decl
- || ! memchr (alphanum, p[-1], sizeof alphanum))
- && memchr (alphanum, p[0], sizeof alphanum - 10))
- name_start = p;
-
- /* Strip the surrounding '{' and '}', and any blanks just inside
- the braces. */
- while (*--p == ' ' || *p == '\t')
- continue;
- p[1] = '\0';
- while (*++decl == ' ' || *decl == '\t')
- continue;
+ {
+ char *p;
+ /* Stop on last actual character. */
+ for (p = decl; p[1]; p++)
+ if ((p == decl
+ || ! memchr (alphanum, p[-1], sizeof alphanum))
+ && memchr (alphanum, p[0], sizeof alphanum - 10))
+ name_start = p;
+
+ /* Strip the surrounding '{' and '}', and any blanks just inside
+ the braces. */
+ while (*--p == ' ' || *p == '\t')
+ continue;
+ p[1] = '\0';
+ while (*++decl == ' ' || *decl == '\t')
+ continue;
+ }
if (! name_start)
- complain_at (loc, _("missing identifier in parameter declaration"));
+ complain_at (loc, complaint,
+ _("missing identifier in parameter declaration"));
else
{
- char *name;
- size_t name_len;
-
- for (name_len = 1;
- memchr (alphanum, name_start[name_len], sizeof alphanum);
- name_len++)
- continue;
-
- name = xmalloc (name_len + 1);
- memcpy (name, name_start, name_len);
- name[name_len] = '\0';
- muscle_pair_list_grow (type, decl, name);
+ char *name = xmemdup0 (name_start, strspn (name_start, alphanum));
+ if (type & param_lex)
+ muscle_pair_list_grow ("lex_param", decl, name);
+ if (type & param_parse)
+ muscle_pair_list_grow ("parse_param", decl, name);
free (name);
}
{
if (strverscmp (version, PACKAGE_VERSION) > 0)
{
- complain_at (*loc, "require bison %s, but have %s",
+ complain_at (*loc, complaint, "require bison %s, but have %s",
version, PACKAGE_VERSION);
- exit (EX_CONFIG);
+ exit (EX_MISMATCH);
}
}
static void
gram_error (location const *loc, char const *msg)
{
- complain_at (*loc, "%s", msg);
+ complain_at (*loc, complaint, "%s", msg);
}
char const *