* data/glr.c (YYERROR): Update definition.

[bison.git] / src / scan-gram.l

diff --git a/src/scan-gram.l b/src/scan-gram.l
index eaf8d1762dfe226a020616e83dc6a93badb37bdc..2e7823175eb393e38f969a6c9b64c46e89812308 100644 (file)
--- a/src/scan-gram.l
+++ b/src/scan-gram.l
@@ -1,5 +1,6 @@
 /* Bison Grammar Scanner                             -*- C -*-
 /* Bison Grammar Scanner                             -*- C -*-

-   Copyright (C) 2002 Free Software Foundation, Inc.
+
+   Copyright (C) 2002, 2003 Free Software Foundation, Inc.

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


@@ -19,36 +20,44 @@
    02111-1307  USA
 */
 
    02111-1307  USA
 */
 

-%option debug nodefault noyywrap never-interactive
+%option debug nodefault nounput noyywrap never-interactive

 %option prefix="gram_" outfile="lex.yy.c"
 
 %{
 #include "system.h"
 %option prefix="gram_" outfile="lex.yy.c"
 
 %{
 #include "system.h"

-#include "mbswidth.h"
+
+#include <mbswidth.h>
+#include <get-errno.h>
+#include <quote.h>
+

 #include "complain.h"
 #include "complain.h"

-#include "quote.h"
-#include "struniq.h"
+#include "files.h"

 #include "getargs.h"
 #include "gram.h"
 #include "reader.h"
 #include "getargs.h"
 #include "gram.h"
 #include "reader.h"

+#include "uniqstr.h"

 
 

-#define YY_USER_INIT                           \
-do {                                           \
-  LOCATION_RESET (*loc);                       \
-  loc->file = current_file;                    \
-} while (0)
+#define YY_USER_INIT                                   \
+  do                                                   \
+    {                                                  \
+      scanner_cursor.file = current_file;              \
+      scanner_cursor.line = 1;                         \
+      scanner_cursor.column = 1;                       \
+      code_start = scanner_cursor;                     \
+    }                                                  \
+  while (0)

 
 

-/* Each time we match a string, move the end cursor to its end. */
-#define STEP  LOCATION_STEP (*loc)
+/* Location of scanner cursor.  */
+boundary scanner_cursor;

 
 

-static void extend_location (location_t *, char const *, int);
-#define YY_USER_ACTION  extend_location (loc, yytext, yyleng);
+static void adjust_location (location *, char const *, size_t);
+#define YY_USER_ACTION  adjust_location (loc, yytext, yyleng);

 
 static size_t no_cr_read (FILE *, char *, size_t);
 #define YY_INPUT(buf, result, size) ((result) = no_cr_read (yyin, buf, size))
 
 
 
 static size_t no_cr_read (FILE *, char *, size_t);
 #define YY_INPUT(buf, result, size) ((result) = no_cr_read (yyin, buf, size))
 
 

-/* STRING_OBSTACK -- Used to store all the characters that we need to
+/* OBSTACK_FOR_STRING -- Used to store all the characters that we need to

    keep (to construct ID, STRINGS etc.).  Use the following macros to
    use it.
 
    keep (to construct ID, STRINGS etc.).  Use the following macros to
    use it.
 


@@ -57,23 +66,23 @@ static size_t no_cr_read (FILE *, char *, size_t);
    STRING_FINISH also stores this string in LAST_STRING, which can be
    used, and which is used by STRING_FREE to free the last string.  */
 
    STRING_FINISH also stores this string in LAST_STRING, which can be
    used, and which is used by STRING_FREE to free the last string.  */
 

-static struct obstack string_obstack;
+static struct obstack obstack_for_string;

 
 /* A string representing the most recently saved token.  */
 static char *last_string;
 
 
 #define STRING_GROW   \
 
 /* A string representing the most recently saved token.  */
 static char *last_string;
 
 
 #define STRING_GROW   \

-  obstack_grow (&string_obstack, yytext, yyleng)
+  obstack_grow (&obstack_for_string, yytext, yyleng)

 
 #define STRING_FINISH                                  \
   do {                                                 \
 
 #define STRING_FINISH                                  \
   do {                                                 \

-    obstack_1grow (&string_obstack, '\0');             \
-    last_string = obstack_finish (&string_obstack);    \
+    obstack_1grow (&obstack_for_string, '\0');         \
+    last_string = obstack_finish (&obstack_for_string);        \

   } while (0)
 
 #define STRING_FREE \
   } while (0)
 
 #define STRING_FREE \

-  obstack_free (&string_obstack, last_string)
+  obstack_free (&obstack_for_string, last_string)

 
 void
 scanner_last_string_free (void)
 
 void
 scanner_last_string_free (void)


@@ -89,24 +98,23 @@ scanner_last_string_free (void)
    Outside of well-formed rules, RULE_LENGTH has an undefined value.  */
 static int rule_length;
 
    Outside of well-formed rules, RULE_LENGTH has an undefined value.  */
 static int rule_length;
 

-static void handle_dollar (braced_code_t code_kind,
-                          char *cp, location_t location);
-static void handle_at (braced_code_t code_kind,
-                      char *cp, location_t location);
-static void handle_syncline (char *args, location_t *location);
+static void handle_dollar (int token_type, char *cp, location loc);
+static void handle_at (int token_type, char *cp, location loc);
+static void handle_syncline (char *args);

 static int convert_ucn_to_byte (char const *hex_text);
 static int convert_ucn_to_byte (char const *hex_text);

-static void unexpected_end_of_file (location_t *, char const *);
+static void unexpected_eof (boundary, char const *);

 
 %}
 %x SC_COMMENT SC_LINE_COMMENT SC_YACC_COMMENT
 %x SC_STRING SC_CHARACTER
 
 %}
 %x SC_COMMENT SC_LINE_COMMENT SC_YACC_COMMENT
 %x SC_STRING SC_CHARACTER

+%x SC_AFTER_IDENTIFIER

 %x SC_ESCAPED_STRING SC_ESCAPED_CHARACTER
 %x SC_ESCAPED_STRING SC_ESCAPED_CHARACTER

-%x SC_BRACED_CODE SC_PROLOGUE SC_EPILOGUE
+%x SC_PRE_CODE SC_BRACED_CODE SC_PROLOGUE SC_EPILOGUE

 
 letter   [.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_]
 id       {letter}({letter}|[0-9])*
 directive %{letter}({letter}|[0-9]|-)*
 
 letter   [.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_]
 id       {letter}({letter}|[0-9])*
 directive %{letter}({letter}|[0-9]|-)*

-int      [0-9]+
+int      [0-9]+

 
 /* POSIX says that a tag must be both an id and a C union member, but
    historically almost any character is allowed in a tag.  We disallow
 
 /* POSIX says that a tag must be both an id and a C union member, but
    historically almost any character is allowed in a tag.  We disallow


@@ -122,15 +130,51 @@ splice     (\\[ \f\t\v]*\n)*
   /* Nesting level of the current code in braces.  */
   int braces_level IF_LINT (= 0);
 
   /* Nesting level of the current code in braces.  */
   int braces_level IF_LINT (= 0);
 

-  /* Scanner context when scanning C code.  */
-  int c_context IF_LINT (= 0);
+  /* Parent context state, when applicable.  */
+  int context_state IF_LINT (= 0);
+
+  /* Token type to return, when applicable.  */
+  int token_type IF_LINT (= 0);

 
 

-  /* At each yylex invocation, mark the current position as the
-     start of the next token.  */
-  STEP;
+  /* Location of most recent identifier, when applicable.  */
+  location id_loc IF_LINT (= empty_location);
+
+  /* Where containing code started, when applicable.  Its initial
+     value is relevant only when yylex is invoked in the SC_EPILOGUE
+     start condition.  */
+  boundary code_start = scanner_cursor;
+
+  /* Where containing comment or string or character literal started,
+     when applicable.  */
+  boundary token_start IF_LINT (= scanner_cursor);

 %}
 
 
 %}
 
 

+  /*-----------------------.
+  | Scanning white space.  |
+  `-----------------------*/
+
+<INITIAL,SC_AFTER_IDENTIFIER,SC_PRE_CODE>
+{
+  [ \f\n\t\v]  ;
+  ","         warn_at (*loc, _("stray `,' treated as white space"));
+
+  /* Comments. */
+  "//".*       ;
+  "/*" {
+    token_start = loc->start;
+    context_state = YY_START;
+    BEGIN SC_YACC_COMMENT;
+  }
+
+  /* #line directives are not documented, and may be withdrawn or
+     modified in future versions of Bison.  */
+  ^"#line "{int}" \"".*"\"\n" {
+    handle_syncline (yytext + sizeof "#line " - 1);
+  }
+}
+
+

   /*----------------------------.
   | Scanning Bison directives.  |
   `----------------------------*/
   /*----------------------------.
   | Scanning Bison directives.  |
   `----------------------------*/


@@ -140,7 +184,7 @@ splice       (\\[ \f\t\v]*\n)*
   "%debug"                return PERCENT_DEBUG;
   "%define"               return PERCENT_DEFINE;
   "%defines"              return PERCENT_DEFINES;
   "%debug"                return PERCENT_DEBUG;
   "%define"               return PERCENT_DEFINE;
   "%defines"              return PERCENT_DEFINES;

-  "%destructor"           return PERCENT_DESTRUCTOR;
+  "%destructor"                  token_type = PERCENT_DESTRUCTOR; BEGIN SC_PRE_CODE;

   "%dprec"               return PERCENT_DPREC;
   "%error"[-_]"verbose"   return PERCENT_ERROR_VERBOSE;
   "%expect"               return PERCENT_EXPECT;
   "%dprec"               return PERCENT_DPREC;
   "%error"[-_]"verbose"   return PERCENT_ERROR_VERBOSE;
   "%expect"               return PERCENT_EXPECT;


@@ -148,62 +192,50 @@ splice     (\\[ \f\t\v]*\n)*
   "%fixed"[-_]"output"[-_]"files"   return PERCENT_YACC;
   "%glr-parser"           return PERCENT_GLR_PARSER;
   "%left"                 return PERCENT_LEFT;
   "%fixed"[-_]"output"[-_]"files"   return PERCENT_YACC;
   "%glr-parser"           return PERCENT_GLR_PARSER;
   "%left"                 return PERCENT_LEFT;

+  "%lex-param"           token_type = PERCENT_LEX_PARAM; BEGIN SC_PRE_CODE;

   "%locations"            return PERCENT_LOCATIONS;
   "%merge"               return PERCENT_MERGE;
   "%name"[-_]"prefix"     return PERCENT_NAME_PREFIX;
   "%no"[-_]"lines"        return PERCENT_NO_LINES;
   "%nonassoc"             return PERCENT_NONASSOC;
   "%locations"            return PERCENT_LOCATIONS;
   "%merge"               return PERCENT_MERGE;
   "%name"[-_]"prefix"     return PERCENT_NAME_PREFIX;
   "%no"[-_]"lines"        return PERCENT_NO_LINES;
   "%nonassoc"             return PERCENT_NONASSOC;

+  "%nondeterministic-parser"   return PERCENT_NONDETERMINISTIC_PARSER;

   "%nterm"                return PERCENT_NTERM;
   "%output"               return PERCENT_OUTPUT;
   "%nterm"                return PERCENT_NTERM;
   "%output"               return PERCENT_OUTPUT;

-  "%parse-param"          return PERCENT_PARSE_PARAM;
+  "%parse-param"         token_type = PERCENT_PARSE_PARAM; BEGIN SC_PRE_CODE;

   "%prec"                 rule_length--; return PERCENT_PREC;
   "%prec"                 rule_length--; return PERCENT_PREC;

-  "%printer"              return PERCENT_PRINTER;
+  "%printer"              token_type = PERCENT_PRINTER; BEGIN SC_PRE_CODE;

   "%pure"[-_]"parser"     return PERCENT_PURE_PARSER;
   "%right"                return PERCENT_RIGHT;
   "%pure"[-_]"parser"     return PERCENT_PURE_PARSER;
   "%right"                return PERCENT_RIGHT;

-  "%lex-param"            return PERCENT_LEX_PARAM;

   "%skeleton"             return PERCENT_SKELETON;
   "%start"                return PERCENT_START;
   "%term"                 return PERCENT_TOKEN;
   "%token"                return PERCENT_TOKEN;
   "%token"[-_]"table"     return PERCENT_TOKEN_TABLE;
   "%type"                 return PERCENT_TYPE;
   "%skeleton"             return PERCENT_SKELETON;
   "%start"                return PERCENT_START;
   "%term"                 return PERCENT_TOKEN;
   "%token"                return PERCENT_TOKEN;
   "%token"[-_]"table"     return PERCENT_TOKEN_TABLE;
   "%type"                 return PERCENT_TYPE;

-  "%union"                return PERCENT_UNION;
+  "%union"               token_type = PERCENT_UNION; BEGIN SC_PRE_CODE;

   "%verbose"              return PERCENT_VERBOSE;
   "%yacc"                 return PERCENT_YACC;
 
   "%verbose"              return PERCENT_VERBOSE;
   "%yacc"                 return PERCENT_YACC;
 

-  {directive}             {
+  {directive} {

     complain_at (*loc, _("invalid directive: %s"), quote (yytext));
     complain_at (*loc, _("invalid directive: %s"), quote (yytext));

-    STEP;
-  }
-
-  ^"#line "{int}" \"".*"\"\n" {
-    handle_syncline (yytext + sizeof "#line " - 1, loc);
-    STEP;

   }
 
   "="                     return EQUAL;
   }
 
   "="                     return EQUAL;

-  ":"                     rule_length = 0; return COLON;

   "|"                     rule_length = 0; return PIPE;
   ";"                     return SEMICOLON;
 
   "|"                     rule_length = 0; return PIPE;
   ";"                     return SEMICOLON;
 

-  [ \f\n\t\v]  STEP;
-
-  "," {
-    warn_at (*loc, _("stray `,' treated as white space"));
-    STEP;
-  }
-
-  {id}        {
+  {id} {

     val->symbol = symbol_get (yytext, *loc);
     val->symbol = symbol_get (yytext, *loc);

+    id_loc = *loc;

     rule_length++;
     rule_length++;

-    return ID;
+    BEGIN SC_AFTER_IDENTIFIER;

   }
 
   {int} {
     unsigned long num;
   }
 
   {int} {
     unsigned long num;

-    errno = 0;
+    set_errno (0);

     num = strtoul (yytext, 0, 10);
     num = strtoul (yytext, 0, 10);

-    if (INT_MAX < num || errno)
+    if (INT_MAX < num || get_errno ())

       {
        complain_at (*loc, _("integer out of range: %s"), quote (yytext));
        num = INT_MAX;
       {
        complain_at (*loc, _("integer out of range: %s"), quote (yytext));
        num = INT_MAX;


@@ -213,26 +245,28 @@ splice     (\\[ \f\t\v]*\n)*
   }
 
   /* Characters.  We don't check there is only one.  */
   }
 
   /* Characters.  We don't check there is only one.  */

-  "'"         STRING_GROW; BEGIN SC_ESCAPED_CHARACTER;
+  "'"        STRING_GROW; token_start = loc->start; BEGIN SC_ESCAPED_CHARACTER;

 
   /* Strings. */
 
   /* Strings. */

-  "\""        STRING_GROW; BEGIN SC_ESCAPED_STRING;
-
-  /* Comments. */
-  "/*"        BEGIN SC_YACC_COMMENT;
-  "//".*      STEP;
+  "\""       STRING_GROW; token_start = loc->start; BEGIN SC_ESCAPED_STRING;

 
   /* Prologue. */
 
   /* Prologue. */

-  "%{"        BEGIN SC_PROLOGUE;
+  "%{"        code_start = loc->start; BEGIN SC_PROLOGUE;

 
   /* Code in between braces.  */
 
   /* Code in between braces.  */

-  "{"         STRING_GROW; braces_level = 0; BEGIN SC_BRACED_CODE;
+  "{" {
+    STRING_GROW;
+    token_type = BRACED_CODE;
+    braces_level = 0;
+    code_start = loc->start;
+    BEGIN SC_BRACED_CODE;
+  }

 
   /* A type. */
   "<"{tag}">" {
 
   /* A type. */
   "<"{tag}">" {

-    obstack_grow (&string_obstack, yytext + 1, yyleng - 2);
+    obstack_grow (&obstack_for_string, yytext + 1, yyleng - 2);

     STRING_FINISH;
     STRING_FINISH;

-    val->struniq = struniq_new (last_string);
+    val->uniqstr = uniqstr_new (last_string);

     STRING_FREE;
     return TYPE;
   }
     STRING_FREE;
     return TYPE;
   }


@@ -246,7 +280,38 @@ splice      (\\[ \f\t\v]*\n)*
 
   . {
     complain_at (*loc, _("invalid character: %s"), quote (yytext));
 
   . {
     complain_at (*loc, _("invalid character: %s"), quote (yytext));

-    STEP;
+  }
+
+  <<EOF>> {
+    loc->start = loc->end = scanner_cursor;
+    yyterminate ();
+  }
+}
+
+
+  /*-----------------------------------------------------------------.
+  | Scanning after an identifier, checking whether a colon is next.  |
+  `-----------------------------------------------------------------*/
+
+<SC_AFTER_IDENTIFIER>
+{
+  ":" {
+    rule_length = 0;
+    *loc = id_loc;
+    BEGIN INITIAL;
+    return ID_COLON;
+  }
+  . {
+    scanner_cursor.column -= mbsnwidth (yytext, yyleng, 0);
+    yyless (0);
+    *loc = id_loc;
+    BEGIN INITIAL;
+    return ID;
+  }
+  <<EOF>> {
+    *loc = id_loc;
+    BEGIN INITIAL;
+    return ID;

   }
 }
 
   }
 }
 


@@ -257,13 +322,9 @@ splice      (\\[ \f\t\v]*\n)*
 
 <SC_YACC_COMMENT>
 {
 
 <SC_YACC_COMMENT>
 {

-  "*/" {
-    STEP;
-    BEGIN INITIAL;
-  }
-
+  "*/"     BEGIN context_state;

   .|\n    ;
   .|\n    ;

-  <<EOF>>  unexpected_end_of_file (loc, "*/");
+  <<EOF>>  unexpected_eof (token_start, "*/"); BEGIN context_state;

 }
 
 
 }
 
 


@@ -273,8 +334,8 @@ splice       (\\[ \f\t\v]*\n)*
 
 <SC_COMMENT>
 {
 
 <SC_COMMENT>
 {

-  "*"{splice}"/"  STRING_GROW; BEGIN c_context;
-  <<EOF>>        unexpected_end_of_file (loc, "*/");
+  "*"{splice}"/"  STRING_GROW; BEGIN context_state;
+  <<EOF>>        unexpected_eof (token_start, "*/"); BEGIN context_state;

 }
 
 
 }
 
 


@@ -284,9 +345,9 @@ splice       (\\[ \f\t\v]*\n)*
 
 <SC_LINE_COMMENT>
 {
 
 <SC_LINE_COMMENT>
 {

-  "\n"          STRING_GROW; BEGIN c_context;
+  "\n"          STRING_GROW; BEGIN context_state;

   {splice}      STRING_GROW;
   {splice}      STRING_GROW;

-  <<EOF>>       BEGIN c_context;
+  <<EOF>>       BEGIN context_state;

 }
 
 
 }
 
 


@@ -300,14 +361,15 @@ splice     (\\[ \f\t\v]*\n)*
   "\"" {
     STRING_GROW;
     STRING_FINISH;
   "\"" {
     STRING_GROW;
     STRING_FINISH;

-    val->string = last_string;
+    loc->start = token_start;
+    val->chars = last_string;

     rule_length++;
     BEGIN INITIAL;
     return STRING;
   }
 
   .|\n     STRING_GROW;
     rule_length++;
     BEGIN INITIAL;
     return STRING;
   }
 
   .|\n     STRING_GROW;

-  <<EOF>>   unexpected_end_of_file (loc, "\"");
+  <<EOF>>   unexpected_eof (token_start, "\""); BEGIN INITIAL;

 }
 
   /*---------------------------------------------------------------.
 }
 
   /*---------------------------------------------------------------.


@@ -318,12 +380,14 @@ splice     (\\[ \f\t\v]*\n)*
 <SC_ESCAPED_CHARACTER>
 {
   "'" {
 <SC_ESCAPED_CHARACTER>
 {
   "'" {

+    unsigned char last_string_1;

     STRING_GROW;
     STRING_FINISH;
     STRING_GROW;
     STRING_FINISH;

+    loc->start = token_start;

     val->symbol = symbol_get (last_string, *loc);
     symbol_class_set (val->symbol, token_sym, *loc);
     val->symbol = symbol_get (last_string, *loc);
     symbol_class_set (val->symbol, token_sym, *loc);

-    symbol_user_token_number_set (val->symbol,
-                                 (unsigned char) last_string[1], *loc);
+    last_string_1 = last_string[1];
+    symbol_user_token_number_set (val->symbol, last_string_1, *loc);

     STRING_FREE;
     rule_length++;
     BEGIN INITIAL;
     STRING_FREE;
     rule_length++;
     BEGIN INITIAL;


@@ -331,7 +395,7 @@ splice       (\\[ \f\t\v]*\n)*
   }
 
   .|\n     STRING_GROW;
   }
 
   .|\n     STRING_GROW;

-  <<EOF>>   unexpected_end_of_file (loc, "'");
+  <<EOF>>   unexpected_eof (token_start, "'"); BEGIN INITIAL;

 }
 
 
 }
 
 


@@ -344,54 +408,41 @@ splice     (\\[ \f\t\v]*\n)*
   \\[0-7]{1,3} {
     unsigned long c = strtoul (yytext + 1, 0, 8);
     if (UCHAR_MAX < c)
   \\[0-7]{1,3} {
     unsigned long c = strtoul (yytext + 1, 0, 8);
     if (UCHAR_MAX < c)

-      {
-       complain_at (*loc, _("invalid escape sequence: %s"),
-                    quote (yytext));
-       STEP;
-      }
+      complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext));

     else
     else

-      obstack_1grow (&string_obstack, c);
+      obstack_1grow (&obstack_for_string, c);

   }
 
   \\x[0-9abcdefABCDEF]+ {
     unsigned long c;
   }
 
   \\x[0-9abcdefABCDEF]+ {
     unsigned long c;

-    errno = 0;
+    set_errno (0);

     c = strtoul (yytext + 2, 0, 16);
     c = strtoul (yytext + 2, 0, 16);

-    if (UCHAR_MAX < c || errno)
-      {
-       complain_at (*loc, _("invalid escape sequence: %s"),
-                    quote (yytext));
-       STEP;
-      }
+    if (UCHAR_MAX < c || get_errno ())
+      complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext));

     else
     else

-      obstack_1grow (&string_obstack, c);
+      obstack_1grow (&obstack_for_string, c);

   }
 
   }
 

-  \\a  obstack_1grow (&string_obstack, '\a');
-  \\b  obstack_1grow (&string_obstack, '\b');
-  \\f  obstack_1grow (&string_obstack, '\f');
-  \\n  obstack_1grow (&string_obstack, '\n');
-  \\r  obstack_1grow (&string_obstack, '\r');
-  \\t  obstack_1grow (&string_obstack, '\t');
-  \\v  obstack_1grow (&string_obstack, '\v');
+  \\a  obstack_1grow (&obstack_for_string, '\a');
+  \\b  obstack_1grow (&obstack_for_string, '\b');
+  \\f  obstack_1grow (&obstack_for_string, '\f');
+  \\n  obstack_1grow (&obstack_for_string, '\n');
+  \\r  obstack_1grow (&obstack_for_string, '\r');
+  \\t  obstack_1grow (&obstack_for_string, '\t');
+  \\v  obstack_1grow (&obstack_for_string, '\v');

 
   /* \\[\"\'?\\] would be shorter, but it confuses xgettext.  */
 
   /* \\[\"\'?\\] would be shorter, but it confuses xgettext.  */

-  \\("\""|"'"|"?"|"\\")  obstack_1grow (&string_obstack, yytext[1]);
+  \\("\""|"'"|"?"|"\\")  obstack_1grow (&obstack_for_string, yytext[1]);

 
   \\(u|U[0-9abcdefABCDEF]{4})[0-9abcdefABCDEF]{4} {
     int c = convert_ucn_to_byte (yytext);
     if (c < 0)
 
   \\(u|U[0-9abcdefABCDEF]{4})[0-9abcdefABCDEF]{4} {
     int c = convert_ucn_to_byte (yytext);
     if (c < 0)

-      {
-       complain_at (*loc, _("invalid escape sequence: %s"),
-                    quote (yytext));
-       STEP;
-      }
+      complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext));

     else
     else

-      obstack_1grow (&string_obstack, c);
+      obstack_1grow (&obstack_for_string, c);

   }
   \\(.|\n)     {
   }
   \\(.|\n)     {

-    complain_at (*loc, _("unrecognized escape sequence: %s"),
-                quote (yytext));
+    complain_at (*loc, _("unrecognized escape sequence: %s"), quote (yytext));

     STRING_GROW;
   }
 }
     STRING_GROW;
   }
 }


@@ -404,9 +455,9 @@ splice       (\\[ \f\t\v]*\n)*
 
 <SC_CHARACTER>
 {
 
 <SC_CHARACTER>
 {

-  "'"                  STRING_GROW; BEGIN c_context;
+  "'"                  STRING_GROW; BEGIN context_state;

   \\{splice}[^$@\[\]]  STRING_GROW;
   \\{splice}[^$@\[\]]  STRING_GROW;

-  <<EOF>>              unexpected_end_of_file (loc, "'");
+  <<EOF>>              unexpected_eof (token_start, "'"); BEGIN context_state;

 }
 
 
 }
 
 


@@ -417,9 +468,12 @@ splice      (\\[ \f\t\v]*\n)*
 
 <SC_STRING>
 {
 
 <SC_STRING>
 {

-  "\""                 STRING_GROW; BEGIN c_context;
+  "\""                 STRING_GROW; BEGIN context_state;

   \\{splice}[^$@\[\]]  STRING_GROW;
   \\{splice}[^$@\[\]]  STRING_GROW;

-  <<EOF>>              unexpected_end_of_file (loc, "\"");
+  <<EOF>> {
+    unexpected_eof (token_start, "\"");
+    BEGIN context_state;
+  }

 }
 
 
 }
 
 


@@ -429,10 +483,64 @@ splice     (\\[ \f\t\v]*\n)*
 
 <SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
 {
 
 <SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
 {

-  "'"            STRING_GROW; c_context = YY_START; BEGIN SC_CHARACTER;
-  "\""           STRING_GROW; c_context = YY_START; BEGIN SC_STRING;
-  "/"{splice}"*"  STRING_GROW; c_context = YY_START; BEGIN SC_COMMENT;
-  "/"{splice}"/"  STRING_GROW; c_context = YY_START; BEGIN SC_LINE_COMMENT;
+  "'" {
+    STRING_GROW;
+    context_state = YY_START;
+    token_start = loc->start;
+    BEGIN SC_CHARACTER;
+  }
+  "\"" {
+    STRING_GROW;
+    context_state = YY_START;
+    token_start = loc->start;
+    BEGIN SC_STRING;
+  }
+  "/"{splice}"*" {
+    STRING_GROW;
+    context_state = YY_START;
+    token_start = loc->start;
+    BEGIN SC_COMMENT;
+  }
+  "/"{splice}"/" {
+    STRING_GROW;
+    context_state = YY_START;
+    BEGIN SC_LINE_COMMENT;
+  }
+}
+
+
+  /*---------------------------------------------------------------.
+  | Scanning after %union etc., possibly followed by white space.  |
+  | For %union only, allow arbitrary C code to appear before the   |
+  | following brace, as an extension to POSIX.                    |
+  `---------------------------------------------------------------*/
+
+<SC_PRE_CODE>
+{
+  . {
+    bool valid = yytext[0] == '{' || token_type == PERCENT_UNION;
+    scanner_cursor.column -= mbsnwidth (yytext, yyleng, 0);
+    yyless (0);
+
+    if (valid)
+      {
+       braces_level = -1;
+       code_start = loc->start;
+       BEGIN SC_BRACED_CODE;
+      }
+    else
+      {
+       complain_at (*loc, _("missing `{' in `%s'"),
+                    token_name (token_type));
+       obstack_sgrow (&obstack_for_string, "{}");
+       STRING_FINISH;
+       val->chars = last_string;
+       BEGIN INITIAL;
+       return token_type;
+      }
+  }
+
+  <<EOF>>  unexpected_eof (scanner_cursor, "{}"); BEGIN INITIAL;

 }
 
 
 }
 
 


@@ -446,15 +554,35 @@ splice     (\\[ \f\t\v]*\n)*
   "{"|"<"{splice}"%"  STRING_GROW; braces_level++;
   "%"{splice}">"      STRING_GROW; braces_level--;
   "}" {
   "{"|"<"{splice}"%"  STRING_GROW; braces_level++;
   "%"{splice}">"      STRING_GROW; braces_level--;
   "}" {

-    STRING_GROW;
-    braces_level--;
-    if (braces_level < 0)
+    bool outer_brace = --braces_level < 0;
+
+    /* As an undocumented Bison extension, append `;' before the last
+       brace in braced code, so that the user code can omit trailing
+       `;'.  But do not append `;' if emulating Yacc, since Yacc does
+       not append one.
+
+       FIXME: Bison should warn if a semicolon seems to be necessary
+       here, and should omit the semicolon if it seems unnecessary
+       (e.g., after ';', '{', or '}', each followed by comments or
+       white space).  Such a warning shouldn't depend on --yacc; it
+       should depend on a new --pedantic option, which would cause
+       Bison to warn if it detects an extension to POSIX.  --pedantic
+       should also diagnose other Bison extensions like %yacc.
+       Perhaps there should also be a GCC-style --pedantic-errors
+       option, so that such warnings are diagnosed as errors.  */
+    if (outer_brace && token_type == BRACED_CODE && ! yacc_flag)
+      obstack_1grow (&obstack_for_string, ';');
+
+    obstack_1grow (&obstack_for_string, '}');
+
+    if (outer_brace)

       {
        STRING_FINISH;
       {
        STRING_FINISH;

-       val->string = last_string;

        rule_length++;
        rule_length++;

+       loc->start = code_start;
+       val->chars = last_string;

        BEGIN INITIAL;
        BEGIN INITIAL;

-       return BRACED_CODE;
+       return token_type;

       }
   }
 
       }
   }
 


@@ -462,12 +590,10 @@ splice     (\\[ \f\t\v]*\n)*
      (as `<' `<%').  */
   "<"{splice}"<"  STRING_GROW;
 
      (as `<' `<%').  */
   "<"{splice}"<"  STRING_GROW;
 

-  "$"("<"{tag}">")?(-?[0-9]+|"$") { handle_dollar (current_braced_code,
-                                                  yytext, *loc); }
-  "@"(-?[0-9]+|"$")               { handle_at (current_braced_code,
-                                              yytext, *loc); }
+  "$"("<"{tag}">")?(-?[0-9]+|"$")  handle_dollar (token_type, yytext, *loc);
+  "@"(-?[0-9]+|"$")               handle_at (token_type, yytext, *loc);

 
 

-  <<EOF>>  unexpected_end_of_file (loc, "}");
+  <<EOF>>  unexpected_eof (code_start, "}"); BEGIN INITIAL;

 }
 
 
 }
 
 


@@ -479,12 +605,13 @@ splice     (\\[ \f\t\v]*\n)*
 {
   "%}" {
     STRING_FINISH;
 {
   "%}" {
     STRING_FINISH;

-    val->string = last_string;
+    loc->start = code_start;
+    val->chars = last_string;

     BEGIN INITIAL;
     return PROLOGUE;
   }
 
     BEGIN INITIAL;
     return PROLOGUE;
   }
 

-  <<EOF>>  unexpected_end_of_file (loc, "%}");
+  <<EOF>>  unexpected_eof (code_start, "%}"); BEGIN INITIAL;

 }
 
 
 }
 
 


@@ -497,7 +624,8 @@ splice       (\\[ \f\t\v]*\n)*
 {
   <<EOF>> {
     STRING_FINISH;
 {
   <<EOF>> {
     STRING_FINISH;

-    val->string = last_string;
+    loc->start = code_start;
+    val->chars = last_string;

     BEGIN INITIAL;
     return EPILOGUE;
   }
     BEGIN INITIAL;
     return EPILOGUE;
   }


@@ -511,34 +639,38 @@ splice     (\\[ \f\t\v]*\n)*
 
 <SC_COMMENT,SC_LINE_COMMENT,SC_STRING,SC_CHARACTER,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
 {
 
 <SC_COMMENT,SC_LINE_COMMENT,SC_STRING,SC_CHARACTER,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
 {

-  \$   obstack_sgrow (&string_obstack, "$][");
-  \@   obstack_sgrow (&string_obstack, "@@");
-  \[   obstack_sgrow (&string_obstack, "@{");
-  \]   obstack_sgrow (&string_obstack, "@}");
+  \$   obstack_sgrow (&obstack_for_string, "$][");
+  \@   obstack_sgrow (&obstack_for_string, "@@");
+  \[   obstack_sgrow (&obstack_for_string, "@{");
+  \]   obstack_sgrow (&obstack_for_string, "@}");

   .|\n  STRING_GROW;
 }
 
 
 %%
 
   .|\n  STRING_GROW;
 }
 
 
 %%
 

-/* Extend *LOC to account for token TOKEN of size SIZE.  */
+/* Keeps track of the maximum number of semantic values to the left of 
+   a handle (those referenced by $0, $-1, etc.) are required by the 
+   semantic actions of this grammar. */
+int max_left_semantic_context = 0;
+
+/* Set *LOC and adjust scanner cursor to account for token TOKEN of
+   size SIZE.  */

 
 static void
 
 static void

-extend_location (location_t *loc, char const *token, int size)
+adjust_location (location *loc, char const *token, size_t size)

 {
 {

-  int line = loc->last_line;
-  int column = loc->last_column;
+  int line = scanner_cursor.line;
+  int column = scanner_cursor.column;

   char const *p0 = token;
   char const *p = token;
   char const *lim = token + size;
 
   char const *p0 = token;
   char const *p = token;
   char const *lim = token + size;
 

+  loc->start = scanner_cursor;
+

   for (p = token; p < lim; p++)
     switch (*p)
       {
   for (p = token; p < lim; p++)
     switch (*p)
       {

-      case '\r':
-       /* \r shouldn't survive no_cr_read.  */
-       abort ();
-

       case '\n':
        line++;
        column = 1;
       case '\n':
        line++;
        column = 1;


@@ -552,8 +684,10 @@ extend_location (location_t *loc, char const *token, int size)
        break;
       }
 
        break;
       }
 

-  loc->last_line = line;
-  loc->last_column = column + mbsnwidth (p0, p - p0, 0);
+  scanner_cursor.line = line;
+  scanner_cursor.column = column + mbsnwidth (p0, p - p0, 0);
+
+  loc->end = scanner_cursor;

 }
 
 
 }
 
 


@@ -564,14 +698,14 @@ extend_location (location_t *loc, char const *token, int size)
 static size_t
 no_cr_read (FILE *fp, char *buf, size_t size)
 {
 static size_t
 no_cr_read (FILE *fp, char *buf, size_t size)
 {

-  size_t s = fread (buf, 1, size, fp);
-  if (s)
+  size_t bytes_read = fread (buf, 1, size, fp);
+  if (bytes_read)

     {
     {

-      char *w = memchr (buf, '\r', s);
+      char *w = memchr (buf, '\r', bytes_read);

       if (w)
        {
          char const *r = ++w;
       if (w)
        {
          char const *r = ++w;

-         char const *lim = buf + s;
+         char const *lim = buf + bytes_read;

 
          for (;;)
            {
 
          for (;;)
            {


@@ -600,7 +734,7 @@ no_cr_read (FILE *fp, char *buf, size_t size)
        }
     }
 
        }
     }
 

-  return s;
+  return bytes_read;

 }
 
 
 }
 
 


@@ -609,15 +743,18 @@ no_cr_read (FILE *fp, char *buf, size_t size)
 |                                                                   |
 | Possible inputs: $[<TYPENAME>]($|integer)                         |
 |                                                                   |
 |                                                                   |
 | Possible inputs: $[<TYPENAME>]($|integer)                         |
 |                                                                   |

-| Output to the STRING_OBSTACK a reference to this semantic value.  |
+| Output to OBSTACK_FOR_STRING a reference to this semantic value.  |

 `------------------------------------------------------------------*/
 
 `------------------------------------------------------------------*/
 

-static inline void
-handle_action_dollar (char *text, location_t location)
+static inline bool
+handle_action_dollar (char *text, location loc)

 {
   const char *type_name = NULL;
   char *cp = text + 1;
 
 {
   const char *type_name = NULL;
   char *cp = text + 1;
 

+  if (! current_rule)
+    return false;
+

   /* Get the type name if explicit. */
   if (*cp == '<')
     {
   /* Get the type name if explicit. */
   if (*cp == '<')
     {


@@ -631,127 +768,110 @@ handle_action_dollar (char *text, location_t location)
   if (*cp == '$')
     {
       if (!type_name)
   if (*cp == '$')
     {
       if (!type_name)

-       type_name = symbol_list_n_type_name_get (current_rule, location, 0);
+       type_name = symbol_list_n_type_name_get (current_rule, loc, 0);

       if (!type_name && typed)
       if (!type_name && typed)

-       complain_at (location, _("$$ of `%s' has no declared type"),
+       complain_at (loc, _("$$ of `%s' has no declared type"),

                     current_rule->sym->tag);
       if (!type_name)
        type_name = "";
                     current_rule->sym->tag);
       if (!type_name)
        type_name = "";

-      obstack_fgrow1 (&string_obstack,
+      obstack_fgrow1 (&obstack_for_string,

                      "]b4_lhs_value([%s])[", type_name);
     }
   else
     {
       long num;
                      "]b4_lhs_value([%s])[", type_name);
     }
   else
     {
       long num;

-      errno = 0;
+      set_errno (0);

       num = strtol (cp, 0, 10);
 
       num = strtol (cp, 0, 10);
 

-      if (INT_MIN <= num && num <= rule_length && ! errno)
+      if (INT_MIN <= num && num <= rule_length && ! get_errno ())

        {
          int n = num;
        {
          int n = num;

+         if (1-n > max_left_semantic_context)
+           max_left_semantic_context = 1-n;

          if (!type_name && n > 0)
          if (!type_name && n > 0)

-           type_name = symbol_list_n_type_name_get (current_rule, location,
-                                                    n);
+           type_name = symbol_list_n_type_name_get (current_rule, loc, n);

          if (!type_name && typed)
          if (!type_name && typed)

-           complain_at (location, _("$%d of `%s' has no declared type"),
-                     n, current_rule->sym->tag);
+           complain_at (loc, _("$%d of `%s' has no declared type"),
+                        n, current_rule->sym->tag);

          if (!type_name)
            type_name = "";
          if (!type_name)
            type_name = "";

-         obstack_fgrow3 (&string_obstack,
+         obstack_fgrow3 (&obstack_for_string,

                          "]b4_rhs_value([%d], [%d], [%s])[",
                          rule_length, n, type_name);
        }
       else
                          "]b4_rhs_value([%d], [%d], [%s])[",
                          rule_length, n, type_name);
        }
       else

-       complain_at (location, _("integer out of range: %s"), quote (text));
+       complain_at (loc, _("integer out of range: %s"), quote (text));

     }
     }

-}
-
-
-/*---------------------------------------------------------------.
-| TEXT is expected to be $$ in some code associated to a symbol: |
-| destructor or printer.                                         |
-`---------------------------------------------------------------*/

 
 

-static inline void
-handle_symbol_code_dollar (char *text, location_t location)
-{
-  char *cp = text + 1;
-  if (*cp == '$')
-    obstack_sgrow (&string_obstack, "]b4_dollar_dollar[");
-  else
-    complain_at (location, _("invalid value: %s"), quote (text));
+  return true;

 }
 
 
 /*-----------------------------------------------------------------.
 | Dispatch onto handle_action_dollar, or handle_destructor_dollar, |
 }
 
 
 /*-----------------------------------------------------------------.
 | Dispatch onto handle_action_dollar, or handle_destructor_dollar, |

-| depending upon CODE_KIND.                                        |
+| depending upon TOKEN_TYPE.                                       |

 `-----------------------------------------------------------------*/
 
 static void
 `-----------------------------------------------------------------*/
 
 static void

-handle_dollar (braced_code_t braced_code_kind,
-              char *text, location_t location)
+handle_dollar (int token_type, char *text, location loc)

 {
 {

-  switch (braced_code_kind)
+  switch (token_type)

     {
     {

-    case action_braced_code:
-      handle_action_dollar (text, location);
+    case BRACED_CODE:
+      if (handle_action_dollar (text, loc))
+       return;

       break;
 
       break;
 

-    case destructor_braced_code:
-    case printer_braced_code:
-      handle_symbol_code_dollar (text, location);
+    case PERCENT_DESTRUCTOR:
+    case PERCENT_PRINTER:
+      if (text[1] == '$')
+       {
+         obstack_sgrow (&obstack_for_string, "]b4_dollar_dollar[");
+         return;
+       }
+      break;
+
+    default:

       break;
     }
       break;
     }

+
+  complain_at (loc, _("invalid value: %s"), quote (text));

 }
 
 
 /*------------------------------------------------------.
 | TEXT is a location token (i.e., a `@...').  Output to |
 }
 
 
 /*------------------------------------------------------.
 | TEXT is a location token (i.e., a `@...').  Output to |

-| STRING_OBSTACK a reference to this location.          |
+| OBSTACK_FOR_STRING a reference to this location.      |

 `------------------------------------------------------*/
 
 `------------------------------------------------------*/
 

-static inline void
-handle_action_at (char *text, location_t location)
+static inline bool
+handle_action_at (char *text, location loc)

 {
   char *cp = text + 1;
 {
   char *cp = text + 1;

-  locations_flag = 1;
+  locations_flag = true;
+
+  if (! current_rule)
+    return false;

 
   if (*cp == '$')
 
   if (*cp == '$')

-    {
-      obstack_sgrow (&string_obstack, "]b4_lhs_location[");
-    }
+    obstack_sgrow (&obstack_for_string, "]b4_lhs_location[");

   else
     {
       long num;
   else
     {
       long num;

-      errno = 0;
+      set_errno (0);

       num = strtol (cp, 0, 10);
 
       num = strtol (cp, 0, 10);
 

-      if (INT_MIN <= num && num <= rule_length && ! errno)
+      if (INT_MIN <= num && num <= rule_length && ! get_errno ())

        {
          int n = num;
        {
          int n = num;

-         obstack_fgrow2 (&string_obstack, "]b4_rhs_location([%d], [%d])[",
+         obstack_fgrow2 (&obstack_for_string, "]b4_rhs_location([%d], [%d])[",

                          rule_length, n);
        }
       else
                          rule_length, n);
        }
       else

-       complain_at (location, _("integer out of range: %s"), quote (text));
+       complain_at (loc, _("integer out of range: %s"), quote (text));

     }
     }

-}
-
-
-/*---------------------------------------------------------------.
-| TEXT is expected to be @$ in some code associated to a symbol: |
-| destructor or printer.                                         |
-`---------------------------------------------------------------*/

 
 

-static inline void
-handle_symbol_code_at (char *text, location_t location)
-{
-  char *cp = text + 1;
-  if (*cp == '$')
-    obstack_sgrow (&string_obstack, "]b4_at_dollar[");
-  else
-    complain_at (location, _("invalid value: %s"), quote (text));
+  return true;

 }
 
 
 }
 
 


@@ -761,20 +881,28 @@ handle_symbol_code_at (char *text, location_t location)
 `-------------------------------------------------------------------*/
 
 static void
 `-------------------------------------------------------------------*/
 
 static void

-handle_at (braced_code_t braced_code_kind,
-          char *text, location_t location)
+handle_at (int token_type, char *text, location loc)

 {
 {

-  switch (braced_code_kind)
+  switch (token_type)

     {
     {

-    case action_braced_code:
-      handle_action_at (text, location);
+    case BRACED_CODE:
+      handle_action_at (text, loc);
+      return;
+
+    case PERCENT_DESTRUCTOR:
+    case PERCENT_PRINTER:
+      if (text[1] == '$')
+       {
+         obstack_sgrow (&obstack_for_string, "]b4_at_dollar[");
+         return;
+       }

       break;
 
       break;
 

-    case destructor_braced_code:
-    case printer_braced_code:
-      handle_symbol_code_at (text, location);
+    default:

       break;
     }
       break;
     }

+
+  complain_at (loc, _("invalid value: %s"), quote (text));

 }
 
 
 }
 
 


@@ -839,37 +967,30 @@ convert_ucn_to_byte (char const *ucn)
 `----------------------------------------------------------------*/
 
 static void
 `----------------------------------------------------------------*/
 
 static void

-handle_syncline (char *args, location_t *location)
+handle_syncline (char *args)

 {
   int lineno = strtol (args, &args, 10);
   const char *file = NULL;
   file = strchr (args, '"') + 1;
   *strchr (file, '"') = 0;
 {
   int lineno = strtol (args, &args, 10);
   const char *file = NULL;
   file = strchr (args, '"') + 1;
   *strchr (file, '"') = 0;

-  current_file = xstrdup (file);
-  location->file = current_file;
-  location->last_line = lineno;
+  scanner_cursor.file = current_file = xstrdup (file);
+  scanner_cursor.line = lineno;
+  scanner_cursor.column = 1;

 }
 
 
 }
 
 

-/*-------------------------------------------------------------.
-| Report an unexpected end of file at LOC.  An end of file was |
-| encountered and the expected TOKEN_END was missing.  After   |
-| reporting the problem, pretend that TOKEN_END was found.     |
-`-------------------------------------------------------------*/
+/*------------------------------------------------------------------------.
+| Report an unexpected EOF in a token or comment starting at START.       |
+| An end of file was encountered and the expected TOKEN_END was missing.  |
+`------------------------------------------------------------------------*/

 
 static void
 
 static void

-unexpected_end_of_file (location_t *loc, char const *token_end)
+unexpected_eof (boundary start, char const *token_end)

 {
 {

-  size_t i = strlen (token_end);
-
-  complain_at (*loc, _("missing `%s' at end of file"), token_end);
-
-  /* Adjust location's last column so that any later message does not
-     mention the characters just inserted.  */
-  loc->last_column -= i;
-
-  while (i != 0)
-    unput (token_end[--i]);
+  location loc;
+  loc.start = start;
+  loc.end = scanner_cursor;
+  complain_at (loc, _("missing `%s' at end of file"), token_end);

 }
 
 
 }
 
 


@@ -880,7 +1001,7 @@ unexpected_end_of_file (location_t *loc, char const *token_end)
 void
 scanner_initialize (void)
 {
 void
 scanner_initialize (void)
 {

-  obstack_init (&string_obstack);
+  obstack_init (&obstack_for_string);

 }
 
 
 }
 
 


@@ -891,7 +1012,7 @@ scanner_initialize (void)
 void
 scanner_free (void)
 {
 void
 scanner_free (void)
 {

-  obstack_free (&string_obstack, 0);
+  obstack_free (&obstack_for_string, 0);

   /* Reclaim Flex's buffers.  */
   yy_delete_buffer (YY_CURRENT_BUFFER);
 }
   /* Reclaim Flex's buffers.  */
   yy_delete_buffer (YY_CURRENT_BUFFER);
 }