cpplib: don't modify the IStable for '$' in identifiers

[Zack Weinberg <zack at wolery dot cumb dot org>]

This patch changes the definitions of is_idchar and is_idstart so they
check dollars_in_ident directly, rather than expecting '$' to appear
in the IStable.  That, in turn, means we can make the IStable
read-only data, and eliminate the last bit of cpplib internal state
maintained outside the cpp_reader structure.

There's still a dependence on the globals user_label_prefix and
std_prefix.  This is only an issue if you use extremely obscure
options.  '$' in identifiers changes with -ansi, which is not that
uncommon.  Also, those globals are not really under cpplib's control;
they belong to the back end and the driver.

Also, a couple of unrelated static data structures are now constant.

I can't bootstrap because of the strlensi problem, but I have compiled
cpplib and verified that $ is still handled properly in both modes.

zw

	* cpplib.h (_dollar_ok): New macro.
	(is_idchar, is_idstart): Use it.
	(IStable): Rename to _cpp_IStable.  Declare it const if
	gcc >=2.7 or C99.  Delete all references to FAKE_CONST.
	(is_idchar, is_idstart, is_numchar, is_numstart, is_hspace,
	is_space): Update for renamed IStable.

	* cppinit.c: Delete all references to FAKE_CONST and CAT
	macros. Define init_IStable as empty macro if gcc >=2.7 or
	C99. Change TABLE() to ISTABLE and hardcode name of table.
	(cpp_start_read): Don't change the IStable based on
	dollars_in_ident.

	* cpphash.c (unsafe_chars): Add pfile argument.  All callers
	changed.  Handle '$' for char1 correctly.
	* cpplib.c (cpp_get_token): Use is_numchar when parsing numbers.

	* cppexp.c (tokentab2): Make const.
	(cpp_lex): Make toktab const.
	* cppinit.c (include_defaults_array): Make const.
	(initialize_standard_includes): Make default_include const.
	
===================================================================
Index: cppexp.c
--- cppexp.c	2000/03/01 00:57:08	1.34
+++ cppexp.c	2000/03/04 01:35:15
@@ -403,7 +403,7 @@ struct token {
   int token;
 };
 
-static struct token tokentab2[] = {
+static const struct token tokentab2[] = {
   {"&&", ANDAND},
   {"||", OROR},
   {"<<", LSH},
@@ -424,7 +424,7 @@ cpp_lex (pfile, skip_evaluation)
      cpp_reader *pfile;
      int skip_evaluation;
 {
-  struct token *toktab;
+  const struct token *toktab;
   enum cpp_token token;
   struct operation op;
   U_CHAR *tok_start, *tok_end;
===================================================================
Index: cpphash.c
--- cpphash.c	2000/03/02 20:14:32	1.44
+++ cpphash.c	2000/03/04 01:35:16
@@ -35,7 +35,7 @@ static int comp_def_part	 PARAMS ((int, 
 					  int, int));
 static void push_macro_expansion PARAMS ((cpp_reader *,
 					  U_CHAR *, int, HASHNODE *));
-static int unsafe_chars		 PARAMS ((int, int));
+static int unsafe_chars		 PARAMS ((cpp_reader *, int, int));
 static int macro_cleanup	 PARAMS ((cpp_buffer *, cpp_reader *));
 static enum cpp_token macarg	 PARAMS ((cpp_reader *, int));
 static struct tm *timestamp	 PARAMS ((cpp_reader *));
@@ -1317,7 +1317,7 @@ macroexpand (pfile, hp)
 	      U_CHAR *expanded = ARG_BASE + arg->expanded;
 	      if (!ap->raw_before && totlen > 0 && arg->expand_length
 		  && !CPP_TRADITIONAL (pfile)
-		  && unsafe_chars (xbuf[totlen - 1], expanded[0]))
+		  && unsafe_chars (pfile, xbuf[totlen - 1], expanded[0]))
 		{
 		  xbuf[totlen++] = '\r';
 		  xbuf[totlen++] = ' ';
@@ -1328,7 +1328,7 @@ macroexpand (pfile, hp)
 
 	      if (!ap->raw_after && totlen > 0 && offset < defn->length
 		  && !CPP_TRADITIONAL (pfile)
-		  && unsafe_chars (xbuf[totlen - 1], exp[offset]))
+		  && unsafe_chars (pfile, xbuf[totlen - 1], exp[offset]))
 		{
 		  xbuf[totlen++] = '\r';
 		  xbuf[totlen++] = ' ';
@@ -1382,7 +1382,8 @@ macroexpand (pfile, hp)
    could cause mis-tokenization.  */
 
 static int
-unsafe_chars (c1, c2)
+unsafe_chars (pfile, c1, c2)
+     cpp_reader *pfile;
      int c1, c2;
 {
   switch (c1)
@@ -1397,6 +1398,11 @@ unsafe_chars (c1, c2)
 	return 1;		/* could extend a pre-processing number */
       goto letter;
 
+    case '$':
+      if (CPP_OPTIONS (pfile)->dollars_in_ident)
+	goto letter;
+      return 0;
+
     case 'L':
       if (c2 == '\'' || c2 == '\"')
 	return 1;		/* Could turn into L"xxx" or L'xxx'.  */
@@ -1468,7 +1474,7 @@ push_macro_expansion (pfile, xbuf, xbuf_
     {
       int c1 = mbuf->rlimit[-3];
       int c2 = CPP_BUF_PEEK (CPP_PREV_BUFFER (CPP_BUFFER (pfile)));
-      if (c2 == EOF || !unsafe_chars (c1, c2))
+      if (c2 == EOF || !unsafe_chars (pfile, c1, c2))
 	mbuf->rlimit -= 2;
     }
 }
===================================================================
Index: cppinit.c
--- cppinit.c	2000/03/04 00:11:40	1.48
+++ cppinit.c	2000/03/04 01:35:16
@@ -22,7 +22,6 @@ Foundation, 59 Temple Place - Suite 330,
 #include "config.h"
 #include "system.h"
 
-#define FAKE_CONST
 #include "cpplib.h"
 #include "cpphash.h"
 #include "output.h"
@@ -103,7 +102,7 @@ static const char * const known_suffixes
    All these directories are treated as `system' include directories
    (they are not subject to pedantic warnings in some cases).  */
 
-static struct default_include
+struct default_include
 {
   const char *fname;		/* The name of the directory.  */
   const char *component;	/* The component containing the directory
@@ -112,8 +111,9 @@ static struct default_include
   int cxx_aware;		/* Includes in this directory don't need to
 				   be wrapped in extern "C" when compiling
 				   C++.  */
-}
-include_defaults_array[]
+};
+
+static const struct default_include include_defaults_array[]
 #ifdef INCLUDE_DEFAULTS
 = INCLUDE_DEFAULTS;
 #else
@@ -203,25 +203,22 @@ static void new_pending_define		PARAMS (
 
 /* Fourth argument to append_include_chain: chain to use */
 enum { QUOTE = 0, BRACKET, SYSTEM, AFTER };
-
-/* If gcc is in use (stage2/stage3) we can make this table initialized data. */
-#ifdef __STDC__
-#define CAT(a, b) a##b
-#else
-#define CAT(a, b) a/**/b
-#endif
 
-#if (GCC_VERSION >= 2007)
-#define TABLE(id) static inline void CAT(init_, id) PARAMS ((void)) {} \
-unsigned char id[256] = {
-#define s(p, v) [p] = v,
+/* If we have designated initializers (GCC >2.7, or C99) this table
+   can be initialized, constant data.  Otherwise, it has to be filled
+   in at runtime.  */
+
+#if (GCC_VERSION >= 2007) || (__STDC_VERSION__ >= 199901L)
+#define init_IStable()  /* nothing */
+#define ISTABLE const unsigned char _cpp_IStable[256] = {
 #define END };
+#define s(p, v) [p] = v,
 #else
-#define TABLE(id) unsigned char id[256] = { 0 }; \
-static void CAT(init_,id) PARAMS ((void)) { \
-unsigned char *x = id;
-#define s(p, v) x[p] = v;
+#define ISTABLE unsigned char _cpp_IStable[256] = { 0 }; \
+ static void init_IStable PARAMS ((void)) { \
+ unsigned char *x = id;
 #define END } 
+#define s(p, v) x[p] = v;
 #endif
 
 #define A(x) s(x, ISidnum|ISidstart)
@@ -229,7 +226,7 @@ unsigned char *x = id;
 #define H(x) s(x, IShspace|ISspace)
 #define S(x) s(x, ISspace)
 
-TABLE (IStable)
+ISTABLE
   A('_')
 
   A('a') A('b') A('c') A('d') A('e') A('f') A('g') A('h') A('i')
@@ -251,10 +248,9 @@ END
 #undef N
 #undef H
 #undef S
-#undef TABLE
-#undef END
 #undef s
-#undef CAT
+#undef ISTABLE
+#undef END
 
 /* Given a colon-separated list of file names PATH,
    add all the names to the search path for include files.  */
@@ -668,7 +664,7 @@ initialize_standard_includes (pfile)
 {
   cpp_options *opts = CPP_OPTIONS (pfile);
   char *path;
-  struct default_include *p = include_defaults_array;
+  const struct default_include *p;
   char *specd_prefix = opts->include_prefix;
 
   /* Several environment variables may add to the include search path.
@@ -793,14 +789,10 @@ cpp_start_read (pfile, fname)
      preprocessing.  */
   if (opts->preprocessed)
     pfile->no_macro_expand++;
-  
-  /* Now that we know dollars_in_ident, we can initialize the syntax
-     tables. */
+
+  /* Set up the IStable.  This doesn't do anything if we were compiled
+     with a compiler that supports C99 designated initializers.  */
   init_IStable ();
-  /* XXX Get rid of code that depends on this, then IStable can
-     be truly const.  */
-  if (opts->dollars_in_ident)
-    IStable['$'] = ISidstart|ISidnum;
 
   /* Set up the include search path now.  */
   if (! opts->no_standard_includes)
===================================================================
Index: cpplib.c
--- cpplib.c	2000/03/03 00:09:21	1.123
+++ cpplib.c	2000/03/04 01:35:16
@@ -2615,7 +2615,7 @@ cpp_get_token (pfile)
 	      c = PEEKC ();
 	      if (c == EOF)
 		break;
-	      if (!is_idchar(c) && c != '.'
+	      if (!is_numchar(c) && c != '.'
 		  && ((c2 != 'e' && c2 != 'E'
 		       && ((c2 != 'p' && c2 != 'P') || CPP_C89 (pfile)))
 		      || (c != '+' && c != '-')))
@@ -2640,7 +2640,7 @@ cpp_get_token (pfile)
 		  c = GETC();
 		  if (c == EOF)
 		    goto chill_number_eof;
-		  if (!is_idchar(c))
+		  if (!is_numchar(c))
 		    break;
 		  CPP_PUTC (pfile, c);
 		}
===================================================================
Index: cpplib.h
--- cpplib.h	2000/03/04 00:11:40	1.62
+++ cpplib.h	2000/03/04 01:35:17
@@ -612,22 +612,23 @@ enum node_type {
 #define IShspace	0x08	/* ' ' \t \f \v */
 #define ISspace		0x10	/* ' ' \t \f \v \n */
 
-#define is_idchar(x)	(IStable[x] & ISidnum)
-#define is_numchar(x)	(IStable[x] & ISidnum)
-#define is_idstart(x)	(IStable[x] & ISidstart)
-#define is_numstart(x)	(IStable[x] & ISnumstart)
-#define is_hspace(x)	(IStable[x] & IShspace)
-#define is_space(x)	(IStable[x] & ISspace)
+#define _dollar_ok(x)	((x) == '$' && CPP_OPTIONS (pfile)->dollars_in_ident)
 
-/* This table is not really `const', but it is only modified at
-   initialization time, in a separate translation unit from the rest
-   of the library.  We let the rest of the library think it is `const'
-   to get better code and some additional compile-time checks.  */
-#ifndef FAKE_CONST
-#define FAKE_CONST const
+#define is_idchar(x)	((_cpp_IStable[x] & ISidnum) || _dollar_ok(x))
+#define is_idstart(x)	((_cpp_IStable[x] & ISidstart) || _dollar_ok(x))
+#define is_numchar(x)	(_cpp_IStable[x] & ISidnum)
+#define is_numstart(x)	(_cpp_IStable[x] & ISnumstart)
+#define is_hspace(x)	(_cpp_IStable[x] & IShspace)
+#define is_space(x)	(_cpp_IStable[x] & ISspace)
+
+/* This table is constant if it can be initialized at compile time,
+   which is the case if cpp was compiled with GCC >=2.7, or another
+   compiler that supports C99.  */
+#if (GCC_VERSION >= 2007) || (__STDC_VERSION__ >= 199901L)
+extern const unsigned char _cpp_IStable[256];
+#else
+extern unsigned char _cpp_IStable[256];
 #endif
-extern FAKE_CONST unsigned char IStable[256];
-#undef FAKE_CONST
 
 /* Stack of conditionals currently in progress
    (including both successful and failing conditionals).  */