[PATCH 1/5] posix: Sync regex code with gnulib

[Adhemerval Zanella]

It sync with gnulib commit 43ee1a6bf.  The main change is 9682f18e9.
(which does not have a meaniful description).

Checked on x86_64-linux-gnu.
---
 posix/regcomp.c        |  2 +-
 posix/regex.h          | 17 ++++++++++++-----
 posix/regex_internal.c | 19 ++++++++++---------
 posix/regex_internal.h | 16 ++++++++++++----
 4 files changed, 35 insertions(+), 19 deletions(-)

diff --git a/posix/regcomp.c b/posix/regcomp.c
index 93bb0a0538..692928b0db 100644
--- a/posix/regcomp.c
+++ b/posix/regcomp.c
@@ -558,7 +558,7 @@ weak_alias (__regerror, regerror)
 static const bitset_t utf8_sb_map =
 {
   /* Set the first 128 bits.  */
-# if defined __GNUC__ && !defined __STRICT_ANSI__
+# if (defined __GNUC__ || __clang_major__ >= 4) && !defined __STRICT_ANSI__
   [0 ... 0x80 / BITSET_WORD_BITS - 1] = BITSET_WORD_MAX
 # else
 #  if 4 * BITSET_WORD_BITS < ASCII_CHARS
diff --git a/posix/regex.h b/posix/regex.h
index 5fe41c8685..7418e6c76f 100644
--- a/posix/regex.h
+++ b/posix/regex.h
@@ -612,7 +612,9 @@ extern int re_exec (const char *);
    'configure' might #define 'restrict' to those words, so pick a
    different name.  */
 #ifndef _Restrict_
-# if defined __restrict || 2 < __GNUC__ + (95 <= __GNUC_MINOR__)
+# if defined __restrict \
+     || 2 < __GNUC__ + (95 <= __GNUC_MINOR__) \
+     || __clang_major__ >= 3
 #  define _Restrict_ __restrict
 # elif 199901L <= __STDC_VERSION__ || defined restrict
 #  define _Restrict_ restrict
@@ -620,13 +622,18 @@ extern int re_exec (const char *);
 #  define _Restrict_
 # endif
 #endif
-/* For [restrict], use glibc's __restrict_arr if available.
-   Otherwise, GCC 3.1 (not in C++ mode) and C99 support [restrict].  */
+/* For the ISO C99 syntax
+     array_name[restrict]
+   use glibc's __restrict_arr if available.
+   Otherwise, GCC 3.1 and clang support this syntax (but not in C++ mode).
+   Other ISO C99 compilers support it as well.  */
 #ifndef _Restrict_arr_
 # ifdef __restrict_arr
 #  define _Restrict_arr_ __restrict_arr
-# elif ((199901L <= __STDC_VERSION__ || 3 < __GNUC__ + (1 <= __GNUC_MINOR__)) \
-        && !defined __GNUG__)
+# elif ((199901L <= __STDC_VERSION__ \
+         || 3 < __GNUC__ + (1 <= __GNUC_MINOR__) \
+         || __clang_major__ >= 3) \
+        && !defined __cplusplus)
 #  define _Restrict_arr_ _Restrict_
 # else
 #  define _Restrict_arr_
diff --git a/posix/regex_internal.c b/posix/regex_internal.c
index e1b6b4d5af..ed0a13461b 100644
--- a/posix/regex_internal.c
+++ b/posix/regex_internal.c
@@ -300,18 +300,20 @@ build_wcs_upper_buffer (re_string_t *pstr)
       while (byte_idx < end_idx)
 	{
 	  wchar_t wc;
+	  unsigned char ch = pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx];
 
-	  if (isascii (pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx])
-	      && mbsinit (&pstr->cur_state))
+	  if (isascii (ch) && mbsinit (&pstr->cur_state))
 	    {
-	      /* In case of a singlebyte character.  */
-	      pstr->mbs[byte_idx]
-		= toupper (pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]);
 	      /* The next step uses the assumption that wchar_t is encoded
 		 ASCII-safe: all ASCII values can be converted like this.  */
-	      pstr->wcs[byte_idx] = (wchar_t) pstr->mbs[byte_idx];
-	      ++byte_idx;
-	      continue;
+	      wchar_t wcu = __towupper (ch);
+	      if (isascii (wcu))
+		{
+		  pstr->mbs[byte_idx] = wcu;
+		  pstr->wcs[byte_idx] = wcu;
+		  byte_idx++;
+		  continue;
+		}
 	    }
 
 	  remain_len = end_idx - byte_idx;
@@ -348,7 +350,6 @@ build_wcs_upper_buffer (re_string_t *pstr)
 	    {
 	      /* It is an invalid character, an incomplete character
 		 at the end of the string, or '\0'.  Just use the byte.  */
-	      int ch = pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx];
 	      pstr->mbs[byte_idx] = ch;
 	      /* And also cast it to wide char.  */
 	      pstr->wcs[byte_idx++] = (wchar_t) ch;
diff --git a/posix/regex_internal.h b/posix/regex_internal.h
index 8c42586c42..4a3cf779bf 100644
--- a/posix/regex_internal.h
+++ b/posix/regex_internal.h
@@ -77,6 +77,14 @@
 # define isblank(ch) ((ch) == ' ' || (ch) == '\t')
 #endif
 
+/* regex code assumes isascii has its usual numeric meaning,
+   even if the portable character set uses EBCDIC encoding,
+   and even if wint_t is wider than int.  */
+#ifndef _LIBC
+# undef isascii
+# define isascii(c) (((c) & ~0x7f) == 0)
+#endif
+
 #ifdef _LIBC
 # ifndef _RE_DEFINE_LOCALE_FUNCTIONS
 #  define _RE_DEFINE_LOCALE_FUNCTIONS 1
@@ -335,7 +343,7 @@ typedef struct
     Idx idx;			/* for BACK_REF */
     re_context_type ctx_type;	/* for ANCHOR */
   } opr;
-#if __GNUC__ >= 2 && !defined __STRICT_ANSI__
+#if (__GNUC__ >= 2 || defined __clang__) && !defined __STRICT_ANSI__
   re_token_type_t type : 8;
 #else
   re_token_type_t type;
@@ -841,10 +849,10 @@ re_string_elem_size_at (const re_string_t *pstr, Idx idx)
 #endif /* RE_ENABLE_I18N */
 
 #ifndef FALLTHROUGH
-# if __GNUC__ < 7
-#  define FALLTHROUGH ((void) 0)
-# else
+# if (__GNUC__ >= 7) || (__clang_major__ >= 10)
 #  define FALLTHROUGH __attribute__ ((__fallthrough__))
+# else
+#  define FALLTHROUGH ((void) 0)
 # endif
 #endif
 
-- 
2.25.1