.

[gnulib.git] / regex.c

diff --git a/regex.c b/regex.c
index 2421f5a..ed4cfc9 100644 (file)
--- a/regex.c
+++ b/regex.c
@@ -2,7 +2,7 @@
    0.12.  (Implements POSIX draft P10003.2/D11.2, except for
    internationalization features.)
 
-   Copyright (C) 1993, 1994, 1995, 1996, 1997 Free Software Foundation, Inc.
+   Copyright (C) 1993, 1994, 1995, 1996, 1997, 1998 Free Software Foundation, Inc.
 
    This program is free software; you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
@@ -68,6 +68,7 @@
 #include "category.h"
 
 #define malloc xmalloc
+#define realloc xrealloc
 #define free xfree
 
 #else  /* not emacs */
@@ -1539,7 +1540,7 @@ static reg_errcode_t compile_range ();
 #define PATFETCH(c)                                                    \
   do {if (p == pend) return REG_EEND;                                  \
     c = (unsigned char) *p++;                                          \
-    if (translate) c = RE_TRANSLATE (translate, c);                    \
+    if (RE_TRANSLATE_P (translate)) c = RE_TRANSLATE (translate, c);   \
   } while (0)
 #endif
 
@@ -1560,7 +1561,8 @@ static reg_errcode_t compile_range ();
    when we use a character as a subscript we must make it unsigned.  */
 #ifndef TRANSLATE
 #define TRANSLATE(d) \
-  (translate ? (unsigned) RE_TRANSLATE (translate, (unsigned) (d)) : (d))
+  (RE_TRANSLATE_P (translate) \
+   ? (unsigned) RE_TRANSLATE (translate, (unsigned) (d)) : (d))
 #endif
 
 
@@ -2905,14 +2907,14 @@ regex_compile (pattern, size, syntax, bufp)
              || *pending_exact >= (1 << BYTEWIDTH) - (p - p1)
 
              /* If followed by a repetition operator.  */
-             || *p == '*' || *p == '^'
+             || (p != pend && (*p == '*' || *p == '^'))
              || ((syntax & RE_BK_PLUS_QM)
-                 ? *p == '\\' && (p[1] == '+' || p[1] == '?')
-                 : (*p == '+' || *p == '?'))
+                 ? p + 1 < pend && *p == '\\' && (p[1] == '+' || p[1] == '?')
+                 : p != pend && (*p == '+' || *p == '?'))
              || ((syntax & RE_INTERVALS)
                  && ((syntax & RE_NO_BK_BRACES)
-                     ? *p == '{'
-                     : (p[0] == '\\' && p[1] == '{'))))
+                     ? p != pend && *p == '{'
+                     : p + 1 < pend && p[0] == '\\' && p[1] == '{')))
            {
              /* Start building a new exactn.  */
 
@@ -3708,13 +3710,13 @@ re_search_2 (bufp, string1, size1, string2, size2, startpos, range, regs, stop)
     range = total_size - startpos;
 
   /* If the search isn't to be a backwards one, don't waste time in a
-     search for a pattern that must be anchored.  */
+     search for a pattern anchored at beginning of buffer.  */
   if (bufp->used > 0 && (re_opcode_t) bufp->buffer[0] == begbuf && range > 0)
     {
       if (startpos > 0)
        return -1;
       else
-       range = 1;
+       range = 0;
     }
 
 #ifdef emacs
@@ -3722,8 +3724,8 @@ re_search_2 (bufp, string1, size1, string2, size2, startpos, range, regs, stop)
      don't keep searching past point.  */
   if (bufp->used > 0 && (re_opcode_t) bufp->buffer[0] == at_dot && range > 0)
     {
-      range = PT - startpos;
-      if (range <= 0)
+      range = PT_BYTE - BEGV_BYTE - startpos;
+      if (range < 0)
        return -1;
     }
 #endif /* emacs */
@@ -3738,10 +3740,13 @@ re_search_2 (bufp, string1, size1, string2, size2, startpos, range, regs, stop)
     anchored_start = 1;
 
 #ifdef emacs
-  SETUP_SYNTAX_TABLE_FOR_OBJECT (re_match_object,
-                                POS_AS_IN_BUFFER (startpos > 0
-                                                  ? startpos - 1 : startpos),
-                                1);
+  gl_state.object = re_match_object;
+  {
+    int charpos
+      = SYNTAX_TABLE_BYTE_TO_CHAR (startpos > 0 ? startpos : startpos + 1);
+
+    SETUP_SYNTAX_TABLE_FOR_OBJECT (re_match_object, charpos, 1);
+  }
 #endif
 
   /* Loop through the string, looking for a place to start matching.  */
@@ -3781,7 +3786,7 @@ re_search_2 (bufp, string1, size1, string2, size2, startpos, range, regs, stop)
 
              /* Written out as an if-else to avoid testing `translate'
                 inside the loop.  */
-             if (translate)
+             if (RE_TRANSLATE_P (translate))
                {
                  if (multibyte)
                    while (range > lim)
@@ -3818,7 +3823,7 @@ re_search_2 (bufp, string1, size1, string2, size2, startpos, range, regs, stop)
                          : size1 - startpos);
 
              buf_ch = STRING_CHAR (d, room);
-             if (translate)
+             if (RE_TRANSLATE_P (translate))
                buf_ch = RE_TRANSLATE (translate, buf_ch);
 
              if (! (buf_ch >= 0400
@@ -4051,13 +4056,14 @@ re_match_2 (bufp, string1, size1, string2, size2, pos, regs, stop)
   int result;
 
 #ifdef emacs
-  SETUP_SYNTAX_TABLE_FOR_OBJECT (re_match_object,
-                                POS_AS_IN_BUFFER (pos > 0 ? pos - 1 : pos),
-                                1);
+  int charpos;
+  gl_state.object = re_match_object;
+  charpos = SYNTAX_TABLE_BYTE_TO_CHAR (POS_AS_IN_BUFFER (pos));
+  SETUP_SYNTAX_TABLE_FOR_OBJECT (re_match_object, charpos, 1);
 #endif
 
   result = re_match_2_internal (bufp, string1, size1, string2, size2,
-                                   pos, regs, stop);
+                               pos, regs, stop);
   alloca (0);
   return result;
 }
@@ -4493,7 +4499,7 @@ re_match_2_internal (bufp, string1, size1, string2, size2, pos, regs, stop)
 
          /* This is written out as an if-else so we don't waste time
             testing `translate' inside the loop.  */
-         if (translate)
+         if (RE_TRANSLATE_P (translate))
            {
 #ifdef emacs
              if (multibyte)
@@ -4555,7 +4561,7 @@ re_match_2_internal (bufp, string1, size1, string2, size2, pos, regs, stop)
            else
 #endif /* not emacs */
              {
-               buf_ch = *d;
+               buf_ch = (unsigned char) *d;
                buf_charlen = 1;
              }
 
@@ -4868,7 +4874,7 @@ re_match_2_internal (bufp, string1, size1, string2, size2, pos, regs, stop)
 
                /* Compare that many; failure if mismatch, else move
                   past them.  */
-               if (translate
+               if (RE_TRANSLATE_P (translate)
                    ? bcmp_translate (d, d2, mcnt, translate)
                    : bcmp (d, d2, mcnt))
                  goto fail;