Do not allow identifiers that start with a dash.

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

diff --git a/src/scan-gram.l b/src/scan-gram.l
index 4ed30c6f3162e10b50289130132e7223d4ffdce5..83d76506a85a5b5092b57a32a9122de2bf3a9a55 100644 (file)
--- a/src/scan-gram.l
+++ b/src/scan-gram.l
@@ -1,7 +1,6 @@
 /* Bison Grammar Scanner                             -*- C -*-
 
-   Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009
-   Free Software Foundation, Inc.
+   Copyright (C) 2002-2011 Free Software Foundation, Inc.
 
    This file is part of Bison, the GNU Compiler Compiler.
 
@@ -56,10 +55,16 @@ static boundary scanner_cursor;
 static size_t no_cr_read (FILE *, char *, size_t);
 #define YY_INPUT(buf, result, size) ((result) = no_cr_read (yyin, buf, size))
 
-#define RETURN_PERCENT_FLAG(Value)              \
+#define RETURN_PERCENT_PARAM(Value)                     \
+  RETURN_VALUE(PERCENT_PARAM, param, param_ ## Value)
+
+#define RETURN_PERCENT_FLAG(Value)                              \
+  RETURN_VALUE(PERCENT_FLAG, uniqstr, uniqstr_new (Value))
+
+#define RETURN_VALUE(Token, Field, Value)       \
   do {                                          \
-    val->uniqstr = uniqstr_new (Value);         \
-    return PERCENT_FLAG;                        \
+    val->Field = Value;                         \
+    return Token;                               \
   } while (0)
 
 #define ROLLBACK_CURRENT_TOKEN                                  \
@@ -100,11 +105,12 @@ static void unexpected_newline (boundary, char const *);
  /* A complex tag, with nested angles brackets. */
 %x SC_TAG
 
- /* Three types of user code:
+ /* Four types of user code:
     - prologue (code between `%{' `%}' in the first section, before %%);
     - actions, printers, union, etc, (between braced in the middle section);
-    - epilogue (everything after the second %%). */
-%x SC_PROLOGUE SC_BRACED_CODE SC_EPILOGUE
+    - epilogue (everything after the second %%). 
+    - predicate (code between `%?{' and `{' in middle section); */
+%x SC_PROLOGUE SC_BRACED_CODE SC_EPILOGUE SC_PREDICATE
  /* C and C++ comments in code. */
 %x SC_COMMENT SC_LINE_COMMENT
  /* Strings and characters in code. */
@@ -112,8 +118,8 @@ static void unexpected_newline (boundary, char const *);
  /* Bracketed identifiers support. */
 %x SC_BRACKETED_ID SC_RETURN_BRACKETED_ID
 
-letter   [-.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_]
-id       {letter}({letter}|[0-9])*
+letter   [.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_]
+id       {letter}({letter}|[-0-9])*
 directive %{id}
 int      [0-9]+
 
@@ -204,7 +210,7 @@ splice       (\\[ \f\t\v]*\n)*
   "%glr-parser"                     return PERCENT_GLR_PARSER;
   "%language"                       return PERCENT_LANGUAGE;
   "%left"                           return PERCENT_LEFT;
-  "%lex-param"                      return PERCENT_LEX_PARAM;
+  "%lex-param"                      RETURN_PERCENT_PARAM(lex);
   "%locations"                      RETURN_PERCENT_FLAG("locations");
   "%merge"                          return PERCENT_MERGE;
   "%name"[-_]"prefix"               return PERCENT_NAME_PREFIX;
@@ -214,7 +220,8 @@ splice       (\\[ \f\t\v]*\n)*
   "%nondeterministic-parser"        return PERCENT_NONDETERMINISTIC_PARSER;
   "%nterm"                          return PERCENT_NTERM;
   "%output"                         return PERCENT_OUTPUT;
-  "%parse-param"                    return PERCENT_PARSE_PARAM;
+  "%param"                          RETURN_PERCENT_PARAM(both);
+  "%parse-param"                    RETURN_PERCENT_PARAM(parse);
   "%prec"                           return PERCENT_PREC;
   "%precedence"                     return PERCENT_PRECEDENCE;
   "%printer"                        return PERCENT_PRINTER;
@@ -278,6 +285,13 @@ splice      (\\[ \f\t\v]*\n)*
     BEGIN SC_BRACED_CODE;
   }
 
+  /* Semantic predicate. */
+  "%?"[ \f\n\t\v]*"{" {
+    nesting = 0;
+    code_start = loc->start;
+    BEGIN SC_PREDICATE;
+  }
+
   /* A type. */
   "<*>"       return TAG_ANY;
   "<>"        return TAG_NONE;
@@ -617,12 +631,9 @@ splice      (\\[ \f\t\v]*\n)*
   }
   \\(.|\n)     {
     char const *p = yytext + 1;
-    char quoted_ws[] = "` '";
-    if (isspace (*p) && isprint (*p))
-      {
-        quoted_ws[1] = *p;
-        p = quoted_ws;
-      }
+    /* Quote only if escaping won't make the character visible.  */
+    if (isspace ((unsigned char) *p) && isprint ((unsigned char) *p))
+      p = quote (p);
     else
       p = quotearg_style_mem (escape_quoting_style, p, 1);
     complain_at (*loc, _("invalid character after \\-escape: %s"), p);
@@ -657,7 +668,7 @@ splice       (\\[ \f\t\v]*\n)*
   | Strings, comments etc. can be found in user code.  |
   `---------------------------------------------------*/
 
-<SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
+<SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE,SC_PREDICATE>
 {
   "'" {
     STRING_GROW;
@@ -687,14 +698,32 @@ splice     (\\[ \f\t\v]*\n)*
 
 
   /*-----------------------------------------------------------.
-  | Scanning some code in braces (actions). The initial "{" is |
-  | already eaten.                                             |
+  | Scanning some code in braces (actions, predicates). The    |
+  | initial "{" is already eaten.                              |
   `-----------------------------------------------------------*/
 
-<SC_BRACED_CODE>
+<SC_BRACED_CODE,SC_PREDICATE>
 {
   "{"|"<"{splice}"%"  STRING_GROW; nesting++;
   "%"{splice}">"      STRING_GROW; nesting--;
+
+  /* Tokenize `<<%' correctly (as `<<' `%') rather than incorrrectly
+     (as `<' `<%').  */
+  "<"{splice}"<"  STRING_GROW;
+
+  <<EOF>> {
+    int token = (YY_START == SC_BRACED_CODE) ? BRACED_CODE : BRACED_PREDICATE;
+    unexpected_eof (code_start, "}");
+    STRING_FINISH;
+    loc->start = code_start;
+    val->code = last_string;
+    BEGIN INITIAL;
+    return token;
+  }
+}
+
+<SC_BRACED_CODE>
+{
   "}" {
     obstack_1grow (&obstack_for_string, '}');
 
@@ -708,22 +737,25 @@ splice     (\\[ \f\t\v]*\n)*
        return BRACED_CODE;
       }
   }
+}
 
-  /* Tokenize `<<%' correctly (as `<<' `%') rather than incorrrectly
-     (as `<' `<%').  */
-  "<"{splice}"<"  STRING_GROW;
-
-  <<EOF>> {
-    unexpected_eof (code_start, "}");
-    STRING_FINISH;
-    loc->start = code_start;
-    val->code = last_string;
-    BEGIN INITIAL;
-    return BRACED_CODE;
+<SC_PREDICATE>
+{
+  "}" {
+    --nesting;
+    if (nesting < 0)
+      {
+       STRING_FINISH;
+       loc->start = code_start;
+       val->code = last_string;
+       BEGIN INITIAL;
+       return BRACED_PREDICATE;
+      }
+    else
+      obstack_1grow (&obstack_for_string, '}');
   }
 }
 
-
   /*--------------------------------------------------------------.
   | Scanning some prologue: from "%{" (already scanned) to "%}".  |
   `--------------------------------------------------------------*/
@@ -770,8 +802,8 @@ splice       (\\[ \f\t\v]*\n)*
   | By default, grow the string obstack with the input.  |
   `-----------------------------------------------------*/
 
-<SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE,SC_STRING,SC_CHARACTER,SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER>.     |
-<SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>\n  STRING_GROW;
+<SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PREDICATE,SC_PROLOGUE,SC_EPILOGUE,SC_STRING,SC_CHARACTER,SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER>.        |
+  <SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PREDICATE,SC_PROLOGUE,SC_EPILOGUE>\n   STRING_GROW;
 
 %%