grammar: split %union to group together related aspects

[bison.git] / src / parse-gram.y

diff --git a/src/parse-gram.y b/src/parse-gram.y
index 518969948f0586a22d438f41606175240e32e19d..50b90cf0b732c5279b11a3d48d4f9663457395d0 100644 (file)
--- a/src/parse-gram.y
+++ b/src/parse-gram.y
@@ -1,6 +1,6 @@
 %{/* Bison Grammar Parser                             -*- C -*-
 
 %{/* Bison Grammar Parser                             -*- C -*-
 

-   Copyright (C) 2002-2012 Free Software Foundation, Inc.
+   Copyright (C) 2002-2013 Free Software Foundation, Inc.

 
    This file is part of Bison, the GNU Compiler Compiler.
 
 
    This file is part of Bison, the GNU Compiler Compiler.
 


@@ -20,6 +20,7 @@
 #include <config.h>
 #include "system.h"
 
 #include <config.h>
 #include "system.h"
 

+#include "c-ctype.h"

 #include "complain.h"
 #include "conflicts.h"
 #include "files.h"
 #include "complain.h"
 #include "conflicts.h"
 #include "files.h"


@@ -39,17 +40,13 @@
 static YYLTYPE lloc_default (YYLTYPE const *, int);
 
 #define YY_LOCATION_PRINT(File, Loc) \
 static YYLTYPE lloc_default (YYLTYPE const *, int);
 
 #define YY_LOCATION_PRINT(File, Loc) \

-          location_print (File, Loc)
+          location_print (Loc, File)

 
 static void version_check (location const *loc, char const *version);
 
 
 static void version_check (location const *loc, char const *version);
 

-/* Request detailed syntax error messages, and pass them to GRAM_ERROR.
-   FIXME: depends on the undocumented availability of YYLLOC.  */
-#undef  yyerror
-#define yyerror(Msg) \
-        gram_error (&yylloc, Msg)

 static void gram_error (location const *, char const *);
 
 static void gram_error (location const *, char const *);
 

+/* A string that describes a char (e.g., 'a' -> "'a'").  */

 static char const *char_name (char);
 %}
 
 static char const *char_name (char);
 %}
 


@@ -67,7 +64,7 @@ static char const *char_name (char);
    */
   static
   void
    */
   static
   void

-  current_lhs(symbol *sym, location loc, named_ref *ref)
+  current_lhs (symbol *sym, location loc, named_ref *ref)

   {
     current_lhs_symbol = sym;
     current_lhs_location = loc;
   {
     current_lhs_symbol = sym;
     current_lhs_location = loc;


@@ -86,15 +83,15 @@ static char const *char_name (char);
   #define YYTYPE_UINT8 uint_fast8_t
 }
 
   #define YYTYPE_UINT8 uint_fast8_t
 }
 

-%debug
-%verbose
-%defines
-%locations
-%pure-parser
-%define parse.error "verbose"
+%define api.prefix "gram_"
+%define api.pure full
+%define locations
+%define parse.error verbose

 %define parse.lac full
 %define parse.lac full

-%name-prefix="gram_"
+%define parse.trace
+%defines

 %expect 0
 %expect 0

+%verbose

 
 %initial-action
 {
 
 %initial-action
 {


@@ -104,23 +101,9 @@ static char const *char_name (char);
   boundary_set (&@$.end, current_file, 1, 1);
 }
 
   boundary_set (&@$.end, current_file, 1, 1);
 }
 

-%union
-{
-  assoc assoc;
-  char *code;
-  char const *chars;
-  int integer;
-  named_ref *named_ref;
-  symbol *symbol;
-  symbol_list *list;
-  uniqstr uniqstr;
-  unsigned char character;
-};
-

 /* Define the tokens together with their human representation.  */
 %token GRAM_EOF 0 "end of file"
 %token STRING     "string"
 /* Define the tokens together with their human representation.  */
 %token GRAM_EOF 0 "end of file"
 %token STRING     "string"

-%token INT        "integer"

 
 %token PERCENT_TOKEN       "%token"
 %token PERCENT_NTERM       "%nterm"
 
 %token PERCENT_TOKEN       "%token"
 %token PERCENT_NTERM       "%nterm"


@@ -185,11 +168,17 @@ static char const *char_name (char);
 %token TAG_ANY         "<*>"
 %token TAG_NONE        "<>"
 
 %token TAG_ANY         "<*>"
 %token TAG_NONE        "<>"
 

+%union {unsigned char character;}

 %type <character> CHAR
 %printer { fputs (char_name ($$), yyo); } CHAR
 
 /* braceless is not to be used for rule or symbol actions, as it
    calls code_props_plain_init.  */
 %type <character> CHAR
 %printer { fputs (char_name ($$), yyo); } CHAR
 
 /* braceless is not to be used for rule or symbol actions, as it
    calls code_props_plain_init.  */

+%union
+{
+  char *code;
+  char const *chars;
+};

 %type <chars> STRING "%{...%}" EPILOGUE braceless content.opt
 %type <code> "{...}" "%?{...}"
 %printer { fputs (quotearg_style (c_quoting_style, $$), yyo); }
 %type <chars> STRING "%{...%}" EPILOGUE braceless content.opt
 %type <code> "{...}" "%?{...}"
 %printer { fputs (quotearg_style (c_quoting_style, $$), yyo); }


@@ -197,6 +186,7 @@ static char const *char_name (char);
 %printer { fprintf (yyo, "{\n%s\n}", $$); }
          braceless content.opt "{...}" "%{...%}" EPILOGUE
 
 %printer { fprintf (yyo, "{\n%s\n}", $$); }
          braceless content.opt "{...}" "%{...%}" EPILOGUE
 

+%union {uniqstr uniqstr;}

 %type <uniqstr> BRACKETED_ID ID ID_COLON PERCENT_FLAG TAG tag variable
 %printer { fputs ($$, yyo); } <uniqstr>
 %printer { fprintf (yyo, "[%s]", $$); } BRACKETED_ID
 %type <uniqstr> BRACKETED_ID ID ID_COLON PERCENT_FLAG TAG tag variable
 %printer { fputs ($$, yyo); } <uniqstr>
 %printer { fprintf (yyo, "[%s]", $$); } BRACKETED_ID


@@ -204,15 +194,22 @@ static char const *char_name (char);
 %printer { fprintf (yyo, "%%%s", $$); } PERCENT_FLAG
 %printer { fprintf (yyo, "<%s>", $$); } TAG tag
 
 %printer { fprintf (yyo, "%%%s", $$); } PERCENT_FLAG
 %printer { fprintf (yyo, "<%s>", $$); } TAG tag
 

-%type <integer> INT
+%union {int integer;};
+%token <integer> INT "integer"

 %printer { fprintf (yyo, "%d", $$); } <integer>
 
 %printer { fprintf (yyo, "%d", $$); } <integer>
 

+%union {symbol *symbol;}

 %type <symbol> id id_colon string_as_id symbol symbol.prec
 %printer { fprintf (yyo, "%s", $$->tag); } <symbol>
 %printer { fprintf (yyo, "%s:", $$->tag); } id_colon
 
 %type <symbol> id id_colon string_as_id symbol symbol.prec
 %printer { fprintf (yyo, "%s", $$->tag); } <symbol>
 %printer { fprintf (yyo, "%s:", $$->tag); } id_colon
 

+%union {assoc assoc;};

 %type <assoc> precedence_declarator
 %type <assoc> precedence_declarator

+
+%union {symbol_list *list;}

 %type <list>  symbols.1 symbols.prec generic_symlist generic_symlist_item
 %type <list>  symbols.1 symbols.prec generic_symlist generic_symlist_item

+
+%union {named_ref *named_ref;}

 %type <named_ref> named_ref.opt
 
 /*---------.
 %type <named_ref> named_ref.opt
 
 /*---------.


@@ -253,9 +250,9 @@ static char const *char_name (char);
     {
 #define CASE(In, Out)                                           \
       case param_ ## In: fputs ("%" #Out, stderr); break
     {
 #define CASE(In, Out)                                           \
       case param_ ## In: fputs ("%" #Out, stderr); break

-      CASE(lex,   lex-param);
-      CASE(parse, parse-param);
-      CASE(both,  param);
+      CASE (lex,   lex-param);
+      CASE (parse, parse-param);
+      CASE (both,  param);

 #undef CASE
       case param_none: aver (false); break;
     }
 #undef CASE
       case param_none: aver (false); break;
     }


@@ -277,7 +274,7 @@ input:
         `------------------------------------*/
 
 prologue_declarations:
         `------------------------------------*/
 
 prologue_declarations:

-  /* Nothing */
+  %empty

 | prologue_declarations prologue_declaration
 ;
 
 | prologue_declarations prologue_declaration
 ;
 


@@ -316,7 +313,6 @@ prologue_declaration:
 | "%expect" INT                    { expected_sr_conflicts = $2; }
 | "%expect-rr" INT                 { expected_rr_conflicts = $2; }
 | "%file-prefix" STRING            { spec_file_prefix = $2; }
 | "%expect" INT                    { expected_sr_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"
     {
       nondeterministic_parser = true;
 | "%glr-parser"
     {
       nondeterministic_parser = true;


@@ -333,11 +329,9 @@ prologue_declaration:
     }
 | "%language" STRING            { language_argmatch ($2, grammar_prio, @1); }
 | "%name-prefix" STRING         { spec_name_prefix = $2; }
     }
 | "%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; }
 | "%output" STRING              { spec_outfile = $2; }
 | "%no-lines"                   { no_lines_flag = true; }
 | "%nondeterministic-parser"    { nondeterministic_parser = true; }
 | "%output" STRING              { spec_outfile = $2; }

-| "%output" "=" STRING          { spec_outfile = $3; }  /* deprecated */

 | "%param" { current_param = $1; } params { current_param = param_none; }
 | "%require" STRING             { version_check (&@2, $2); }
 | "%skeleton" STRING
 | "%param" { current_param = $1; } params { current_param = param_none; }
 | "%require" STRING             { version_check (&@2, $2); }
 | "%skeleton" STRING


@@ -436,15 +430,15 @@ code_props_type:
 %token PERCENT_UNION "%union";
 
 union_name:
 %token PERCENT_UNION "%union";
 
 union_name:

-  /* Nothing. */ {}
-| ID             { muscle_code_grow ("union_name", $1, @1); }
+  %empty {}
+| ID     { muscle_code_grow ("union_name", $1, @1); }

 ;
 
 grammar_declaration:
   "%union" union_name braceless
     {
       union_seen = true;
 ;
 
 grammar_declaration:
   "%union" union_name braceless
     {
       union_seen = true;

-      muscle_code_grow ("stype", $3, @3);
+      muscle_code_grow ("union_members", $3, @3);

       code_scanner_last_string_free ();
     }
 ;
       code_scanner_last_string_free ();
     }
 ;


@@ -496,8 +490,8 @@ precedence_declarator:
 ;
 
 tag.opt:
 ;
 
 tag.opt:

-  /* Nothing. */ { current_type = NULL; }
-| TAG            { current_type = $1; tag_seen = true; }
+  %empty { current_type = NULL; }
+| TAG    { current_type = $1; tag_seen = true; }

 ;
 
 /* Just like symbols.1 but accept INT for the sake of POSIX.  */
 ;
 
 /* Just like symbols.1 but accept INT for the sake of POSIX.  */


@@ -505,12 +499,21 @@ symbols.prec:
   symbol.prec
     { $$ = symbol_list_sym_new ($1, @1); }
 | symbols.prec symbol.prec
   symbol.prec
     { $$ = symbol_list_sym_new ($1, @1); }
 | symbols.prec symbol.prec

-    { $$ = symbol_list_prepend ($1, symbol_list_sym_new ($2, @2)); }
+    { $$ = symbol_list_append ($1, symbol_list_sym_new ($2, @2)); }

 ;
 
 symbol.prec:
 ;
 
 symbol.prec:

-  symbol     { $$ = $1; }
-| symbol INT { $$ = $1; symbol_user_token_number_set ($1, $2, @2); }
+  symbol
+    {
+      $$ = $1;
+      symbol_class_set ($1, token_sym, @1, false);
+    }
+| symbol INT
+    {
+      $$ = $1;
+      symbol_user_token_number_set ($1, $2, @2);
+      symbol_class_set ($1, token_sym, @1, false);
+    }

 ;
 
 /* One or more symbols to be %typed. */
 ;
 
 /* One or more symbols to be %typed. */


@@ -518,12 +521,12 @@ symbols.1:
   symbol
     { $$ = symbol_list_sym_new ($1, @1); }
 | symbols.1 symbol
   symbol
     { $$ = symbol_list_sym_new ($1, @1); }
 | symbols.1 symbol

-    { $$ = symbol_list_prepend ($1, symbol_list_sym_new ($2, @2)); }
+    { $$ = symbol_list_append ($1, symbol_list_sym_new ($2, @2)); }

 ;
 
 generic_symlist:
   generic_symlist_item { $$ = $1; }
 ;
 
 generic_symlist:
   generic_symlist_item { $$ = $1; }

-| generic_symlist generic_symlist_item { $$ = symbol_list_prepend ($1, $2); }
+| generic_symlist generic_symlist_item { $$ = symbol_list_append ($1, $2); }

 ;
 
 generic_symlist_item:
 ;
 
 generic_symlist_item:


@@ -540,15 +543,15 @@ tag:
 /* One token definition.  */
 symbol_def:
   TAG
 /* One token definition.  */
 symbol_def:
   TAG

-     {
-       current_type = $1;
-       tag_seen = true;
-     }
+    {
+      current_type = $1;
+      tag_seen = true;
+    }

 | id
 | id

-     {
-       symbol_class_set ($1, current_class, @1, true);
-       symbol_type_set ($1, current_type, @1);
-     }
+    {
+      symbol_class_set ($1, current_class, @1, true);
+      symbol_type_set ($1, current_type, @1);
+    }

 | id INT
     {
       symbol_class_set ($1, current_class, @1, true);
 | id INT
     {
       symbol_class_set ($1, current_class, @1, true);


@@ -611,8 +614,9 @@ rhses.1:
 | rhses.1 ";"
 ;
 
 | rhses.1 ";"
 ;
 

+%token PERCENT_EMPTY "%empty";

 rhs:
 rhs:

-  /* Nothing.  */
+  %empty

     { grammar_current_rule_begin (current_lhs_symbol, current_lhs_location,
                                   current_lhs_named_ref); }
 | rhs symbol named_ref.opt
     { grammar_current_rule_begin (current_lhs_symbol, current_lhs_location,
                                   current_lhs_named_ref); }
 | rhs symbol named_ref.opt


@@ -621,6 +625,8 @@ rhs:
     { grammar_current_rule_action_append ($2, @2, $3, false); }
 | rhs "%?{...}"
     { grammar_current_rule_action_append ($2, @2, NULL, true); }
     { grammar_current_rule_action_append ($2, @2, $3, false); }
 | rhs "%?{...}"
     { grammar_current_rule_action_append ($2, @2, NULL, true); }

+| rhs "%empty"
+    { grammar_current_rule_empty_set (@2); }

 | rhs "%prec" symbol
     { grammar_current_rule_prec_set ($3, @3); }
 | rhs "%dprec" INT
 | rhs "%prec" symbol
     { grammar_current_rule_prec_set ($3, @3); }
 | rhs "%dprec" INT


@@ -630,9 +636,8 @@ rhs:
 ;
 
 named_ref.opt:
 ;
 
 named_ref.opt:

-  /* Nothing. */ { $$ = 0; }
-|
-  BRACKETED_ID   { $$ = named_ref_new($1, @1); }
+  %empty         { $$ = 0; }
+| BRACKETED_ID   { $$ = named_ref_new($1, @1); }

 ;
 
 /*---------------------------.
 ;
 
 /*---------------------------.


@@ -640,7 +645,7 @@ named_ref.opt:
 `---------------------------*/
 
 /* The STRING form of variable is deprecated and is not M4-friendly.
 `---------------------------*/
 
 /* The STRING form of variable is deprecated and is not M4-friendly.

-   For example, M4 fails for `%define "[" "value"'.  */
+   For example, M4 fails for '%define "[" "value"'.  */

 variable:
   ID
 | STRING { $$ = uniqstr_new ($1); }
 variable:
   ID
 | STRING { $$ = uniqstr_new ($1); }


@@ -648,9 +653,9 @@ variable:
 
 /* Some content or empty by default. */
 content.opt:
 
 /* Some content or empty by default. */
 content.opt:

-  /* Nothing. */   { $$ = ""; }
-| ID { $$ = $1; }
-| STRING
+  %empty    { $$ = ""; }
+| ID        { $$ = $1; }
+| STRING    { $$ = $1; }

 ;
 
 
 ;
 
 


@@ -709,7 +714,7 @@ string_as_id:
 ;
 
 epilogue.opt:
 ;
 
 epilogue.opt:

-  /* Nothing.  */
+  %empty

 | "%%" EPILOGUE
     {
       code_props plain_code;
 | "%%" EPILOGUE
     {
       code_props plain_code;


@@ -723,7 +728,6 @@ epilogue.opt:
 
 %%
 
 
 %%
 

-

 /* Return the location of the left-hand side of a rule whose
    right-hand side is RHS[1] ... RHS[N].  Ignore empty nonterminals in
    the right-hand side, and return an empty location equal to the end
 /* Return the location of the left-hand side of a rule whose
    right-hand side is RHS[1] ... RHS[N].  Ignore empty nonterminals in
    the right-hand side, and return an empty location equal to the end


@@ -775,16 +779,17 @@ add_param (param_type type, char *decl, location loc)
 
     /* Strip the surrounding '{' and '}', and any blanks just inside
        the braces.  */
 
     /* Strip the surrounding '{' and '}', and any blanks just inside
        the braces.  */

-    while (*--p == ' ' || *p == '\t')
-      continue;
+    --p;
+  while (c_isspace ((unsigned char) *p))
+      --p;

     p[1] = '\0';
     p[1] = '\0';

-    while (*++decl == ' ' || *decl == '\t')
-      continue;
+    ++decl;
+  while (c_isspace ((unsigned char) *decl))
+      ++decl;

   }
 
   if (! name_start)
   }
 
   if (! name_start)

-    complain_at (loc, complaint,
-                 _("missing identifier in parameter declaration"));
+    complain (&loc, complaint, _("missing identifier in parameter declaration"));

   else
     {
       char *name = xmemdup0 (name_start, strspn (name_start, alphanum));
   else
     {
       char *name = xmemdup0 (name_start, strspn (name_start, alphanum));


@@ -804,8 +809,8 @@ version_check (location const *loc, char const *version)
 {
   if (strverscmp (version, PACKAGE_VERSION) > 0)
     {
 {
   if (strverscmp (version, PACKAGE_VERSION) > 0)
     {

-      complain_at (*loc, complaint, "require bison %s, but have %s",
-                   version, PACKAGE_VERSION);
+      complain (loc, complaint, "require bison %s, but have %s",
+                version, PACKAGE_VERSION);

       exit (EX_MISMATCH);
     }
 }
       exit (EX_MISMATCH);
     }
 }


@@ -813,7 +818,7 @@ version_check (location const *loc, char const *version)
 static void
 gram_error (location const *loc, char const *msg)
 {
 static void
 gram_error (location const *loc, char const *msg)
 {

-  complain_at (*loc, complaint, "%s", msg);
+  complain (loc, complaint, "%s", msg);

 }
 
 char const *
 }
 
 char const *