Update.

	* posix/Makefile (distribute): Add regcomp.c, regexec.c,
	regex_internal.c, and regex_internal.h.
	(CFLAGS-regex.c): Replace -DMBS_SUPPORT with -DRE_ENABLE_I18N.
	* posix/regex.c: Complete rewrite.
	* posix/regexec.c: New file.
	* posix/regcomp.c: New file.
	* posix/regex_internal.c: New file.
	* posix/regex_internal.h: New file.
	* posix/regex.h (RE_ICASE): New macro.
	Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.

8 files changed, 6832 insertions, 8 deletions

diff --git a/ChangeLog b/ChangeLog
index b132ab7926..41b32438c0 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,16 @@
 2002-02-26  Ulrich Drepper  <drepper@redhat.com>
 
+	* posix/Makefile (distribute): Add regcomp.c, regexec.c,
+	regex_internal.c, and regex_internal.h.
+	(CFLAGS-regex.c): Replace -DMBS_SUPPORT with -DRE_ENABLE_I18N.
+	* posix/regex.c: Complete rewrite.
+	* posix/regexec.c: New file.
+	* posix/regcomp.c: New file.
+	* posix/regex_internal.c: New file.
+	* posix/regex_internal.h: New file.
+	* posix/regex.h (RE_ICASE): New macro.
+	Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
+
 	* stdio-common/vfscanf.c (_IO_vfwscanf): Always use ungetc, never
 	ungetwc.  It's a macro.
 	* libio/tst-swscanf.c (do_test): Adjust for now fixed wscanf
diff --git a/NEWS b/NEWS
index 42ca9c7493..6bd0f4b563 100644
--- a/NEWS
+++ b/NEWS
@@ -1,4 +1,4 @@
-GNU C Library NEWS -- history of user-visible changes.  2002-1-28
+GNU C Library NEWS -- history of user-visible changes.  2002-2-26
 Copyright (C) 1992-2000, 2001, 2002 Free Software Foundation, Inc.
 See the end for copying conditions.
 
@@ -25,6 +25,9 @@ Version 2.3
 
 * The malloc functions were completely rewritten by Wolfram Gloger based
   on Doug Lea's malloc-2.7.0.c.
+
+* Isamu Hasegawa contributed a completely new and POSIX conforming
+  implementation of regex.
 
 Version 2.2.5
 
diff --git a/posix/Makefile b/posix/Makefile
index 785c227ebb..008154de9a 100644
--- a/posix/Makefile
+++ b/posix/Makefile
@@ -1,4 +1,4 @@
-# Copyright (C) 1991-1999, 2000, 2001 Free Software Foundation, Inc.
+# Copyright (C) 1991-1999, 2000, 2001, 2002 Free Software Foundation, Inc.
 # This file is part of the GNU C Library.
 
 # The GNU C Library is free software; you can redistribute it and/or
@@ -32,7 +32,8 @@ headers	:= sys/utsname.h sys/times.h sys/wait.h sys/types.h unistd.h	      \
 distribute := confstr.h TESTS TESTS2C.sed testcases.h \
 	      PTESTS PTESTS2C.sed ptestcases.h \
 	      globtest.c globtest.sh wordexp-tst.sh annexc.c fnmatch_loop.c   \
-	      spawn_int.h tst-getconf.sh
+	      spawn_int.h tst-getconf.sh regcomp.c regexec.c regex_internal.c \
+	      regex_internal.h
 
 routines :=								      \
 	uname								      \
@@ -109,7 +110,7 @@ $(objpfx)wordexp-tst.out: wordexp-tst.sh $(objpfx)wordexp-test
 endif
 endif
 
-CFLAGS-regex.c = -Wno-strict-prototypes -DMBS_SUPPORT
+CFLAGS-regex.c = -Wno-strict-prototypes -DRE_ENABLE_I18N
 CFLAGS-getaddrinfo.c = -DRESOLVER
 tstgetopt-ARGS = -a -b -cfoobar --required foobar --optional=bazbug \
 		--none random --col --color --colour
diff --git a/posix/regcomp.c b/posix/regcomp.c
new file mode 100644
index 0000000000..12da043062
--- /dev/null
+++ b/posix/regcomp.c
@@ -0,0 +1,3092 @@
+/* Extended regular expression matching and search library.
+   Copyright (C) 2002 Free Software Foundation, Inc.
+   This file is part of the GNU C Library.
+   Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
+
+   The GNU C Library is free software; you can redistribute it and/or
+   modify it under the terms of the GNU Lesser General Public
+   License as published by the Free Software Foundation; either
+   version 2.1 of the License, or (at your option) any later version.
+
+   The GNU C Library is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+   Lesser General Public License for more details.
+
+   You should have received a copy of the GNU Lesser General Public
+   License along with the GNU C Library; if not, write to the Free
+   Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+   02111-1307 USA.  */
+
+#include <assert.h>
+#include <ctype.h>
+#include <limits.h>
+#include <locale.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <wchar.h>
+#include <wctype.h>
+
+#ifdef _LIBC
+# ifndef _RE_DEFINE_LOCALE_FUNCTIONS
+#  define _RE_DEFINE_LOCALE_FUNCTIONS 1
+#   include <locale/localeinfo.h>
+#   include <locale/elem-hash.h>
+#   include <locale/coll-lookup.h>
+# endif
+#endif
+
+/* This is for other GNU distributions with internationalized messages.  */
+#if HAVE_LIBINTL_H || defined _LIBC
+# include <libintl.h>
+# ifdef _LIBC
+#  undef gettext
+#  define gettext(msgid) __dcgettext ("libc", msgid, LC_MESSAGES)
+# endif
+#else
+# define gettext(msgid) (msgid)
+#endif
+
+#ifndef gettext_noop
+/* This define is so xgettext can find the internationalizable
+   strings.  */
+# define gettext_noop(String) String
+#endif
+
+#include "regex.h"
+#include "regex_internal.h"
+
+static reg_errcode_t re_compile_internal (regex_t *preg, const char * pattern,
+                                          int length, reg_syntax_t syntax);
+static void re_compile_fastmap_iter (regex_t *bufp,
+                                     const re_dfastate_t *init_state,
+                                     char *fastmap);
+static reg_errcode_t init_dfa (re_dfa_t *dfa, int pat_len);
+static void init_word_char (re_dfa_t *dfa);
+static void free_charset (re_charset_t *cset);
+static void free_workarea_compile (regex_t *preg);
+static reg_errcode_t create_initial_state (re_dfa_t *dfa);
+static reg_errcode_t analyze (re_dfa_t *dfa);
+static reg_errcode_t analyze_tree (re_dfa_t *dfa, bin_tree_t *node);
+static void calc_first (re_dfa_t *dfa, bin_tree_t *node);
+static void calc_next (re_dfa_t *dfa, bin_tree_t *node);
+static void calc_epsdest (re_dfa_t *dfa, bin_tree_t *node);
+static int duplicate_node (re_dfa_t *dfa, int org_idx,
+                           unsigned int constraint);
+static reg_errcode_t calc_eclosure (re_dfa_t *dfa);
+static re_node_set calc_eclosure_iter (re_dfa_t *dfa, int node, int root);
+static void calc_inveclosure (re_dfa_t *dfa);
+static int fetch_number (re_string_t *input, re_token_t *token,
+                         reg_syntax_t syntax);
+static re_token_t fetch_token (re_string_t *input, reg_syntax_t syntax);
+static int peek_token (re_token_t *token, re_string_t *input,
+                        reg_syntax_t syntax);
+static int peek_token_bracket (re_token_t *token, re_string_t *input,
+                               reg_syntax_t syntax);
+static bin_tree_t *parse (re_string_t *regexp, regex_t *preg,
+                          reg_syntax_t syntax, reg_errcode_t *err);
+static bin_tree_t *parse_reg_exp (re_string_t *regexp, regex_t *preg,
+                                  re_token_t *token, reg_syntax_t syntax,
+                                  int nest, reg_errcode_t *err);
+static bin_tree_t *parse_branch (re_string_t *regexp, regex_t *preg,
+                                 re_token_t *token, reg_syntax_t syntax,
+                                 int nest, reg_errcode_t *err);
+static bin_tree_t *parse_expression (re_string_t *regexp, regex_t *preg,
+                                     re_token_t *token, reg_syntax_t syntax,
+                                     int nest, reg_errcode_t *err);
+static bin_tree_t *parse_sub_exp (re_string_t *regexp, regex_t *preg,
+                                  re_token_t *token, reg_syntax_t syntax,
+                                  int nest, reg_errcode_t *err);
+static bin_tree_t *parse_dup_op (bin_tree_t *dup_elem, re_string_t *regexp,
+                                 re_dfa_t *dfa, re_token_t *token,
+                                 reg_syntax_t syntax, reg_errcode_t *err);
+static bin_tree_t *parse_bracket_exp (re_string_t *regexp, re_dfa_t *dfa,
+                                      re_token_t *token, reg_syntax_t syntax,
+                                      reg_errcode_t *err);
+static reg_errcode_t parse_bracket_element (bracket_elem_t *elem,
+                                            re_string_t *regexp,
+                                            re_token_t *token, int token_len,
+                                            re_dfa_t *dfa,
+                                            reg_syntax_t syntax);
+static reg_errcode_t parse_bracket_symbol (bracket_elem_t *elem,
+                                          re_string_t *regexp,
+                                          re_token_t *token);
+static reg_errcode_t build_equiv_class (re_charset_t *mbcset,
+                                        re_bitset_ptr_t sbcset,
+                                        int *equiv_class_alloc,
+                                        const unsigned char *name);
+static reg_errcode_t build_charclass (re_charset_t *mbcset,
+                                      re_bitset_ptr_t sbcset,
+                                      int *char_class_alloc,
+                                      const unsigned char *name);
+static bin_tree_t *build_word_op (re_dfa_t *dfa, int not, reg_errcode_t *err);
+static void free_bin_tree (bin_tree_t *tree);
+static bin_tree_t *create_tree (bin_tree_t *left, bin_tree_t *right,
+                                re_token_type_t type, int index);
+static bin_tree_t *duplicate_tree (const bin_tree_t *src, re_dfa_t *dfa);
+
+/* This table gives an error message for each of the error codes listed
+   in regex.h.  Obviously the order here has to be same as there.
+   POSIX doesn't require that we do anything for REG_NOERROR,
+   but why not be nice?  */
+
+const char re_error_msgid[] =
+  {
+#define REG_NOERROR_IDX	0
+    gettext_noop ("Success")	/* REG_NOERROR */
+    "\0"
+#define REG_NOMATCH_IDX (REG_NOERROR_IDX + sizeof "Success")
+    gettext_noop ("No match")	/* REG_NOMATCH */
+    "\0"
+#define REG_BADPAT_IDX	(REG_NOMATCH_IDX + sizeof "No match")
+    gettext_noop ("Invalid regular expression") /* REG_BADPAT */
+    "\0"
+#define REG_ECOLLATE_IDX (REG_BADPAT_IDX + sizeof "Invalid regular expression")
+    gettext_noop ("Invalid collation character") /* REG_ECOLLATE */
+    "\0"
+#define REG_ECTYPE_IDX	(REG_ECOLLATE_IDX + sizeof "Invalid collation character")
+    gettext_noop ("Invalid character class name") /* REG_ECTYPE */
+    "\0"
+#define REG_EESCAPE_IDX	(REG_ECTYPE_IDX + sizeof "Invalid character class name")
+    gettext_noop ("Trailing backslash") /* REG_EESCAPE */
+    "\0"
+#define REG_ESUBREG_IDX	(REG_EESCAPE_IDX + sizeof "Trailing backslash")
+    gettext_noop ("Invalid back reference") /* REG_ESUBREG */
+    "\0"
+#define REG_EBRACK_IDX	(REG_ESUBREG_IDX + sizeof "Invalid back reference")
+    gettext_noop ("Unmatched [ or [^")	/* REG_EBRACK */
+    "\0"
+#define REG_EPAREN_IDX	(REG_EBRACK_IDX + sizeof "Unmatched [ or [^")
+    gettext_noop ("Unmatched ( or \\(") /* REG_EPAREN */
+    "\0"
+#define REG_EBRACE_IDX	(REG_EPAREN_IDX + sizeof "Unmatched ( or \\(")
+    gettext_noop ("Unmatched \\{") /* REG_EBRACE */
+    "\0"
+#define REG_BADBR_IDX	(REG_EBRACE_IDX + sizeof "Unmatched \\{")
+    gettext_noop ("Invalid content of \\{\\}") /* REG_BADBR */
+    "\0"
+#define REG_ERANGE_IDX	(REG_BADBR_IDX + sizeof "Invalid content of \\{\\}")
+    gettext_noop ("Invalid range end")	/* REG_ERANGE */
+    "\0"
+#define REG_ESPACE_IDX	(REG_ERANGE_IDX + sizeof "Invalid range end")
+    gettext_noop ("Memory exhausted") /* REG_ESPACE */
+    "\0"
+#define REG_BADRPT_IDX	(REG_ESPACE_IDX + sizeof "Memory exhausted")
+    gettext_noop ("Invalid preceding regular expression") /* REG_BADRPT */
+    "\0"
+#define REG_EEND_IDX	(REG_BADRPT_IDX + sizeof "Invalid preceding regular expression")
+    gettext_noop ("Premature end of regular expression") /* REG_EEND */
+    "\0"
+#define REG_ESIZE_IDX	(REG_EEND_IDX + sizeof "Premature end of regular expression")
+    gettext_noop ("Regular expression too big") /* REG_ESIZE */
+    "\0"
+#define REG_ERPAREN_IDX	(REG_ESIZE_IDX + sizeof "Regular expression too big")
+    gettext_noop ("Unmatched ) or \\)") /* REG_ERPAREN */
+  };
+
+const size_t re_error_msgid_idx[] =
+  {
+    REG_NOERROR_IDX,
+    REG_NOMATCH_IDX,
+    REG_BADPAT_IDX,
+    REG_ECOLLATE_IDX,
+    REG_ECTYPE_IDX,
+    REG_EESCAPE_IDX,
+    REG_ESUBREG_IDX,
+    REG_EBRACK_IDX,
+    REG_EPAREN_IDX,
+    REG_EBRACE_IDX,
+    REG_BADBR_IDX,
+    REG_ERANGE_IDX,
+    REG_ESPACE_IDX,
+    REG_BADRPT_IDX,
+    REG_EEND_IDX,
+    REG_ESIZE_IDX,
+    REG_ERPAREN_IDX
+  };
+
+/* Entry points for GNU code.  */
+
+/* re_compile_pattern is the GNU regular expression compiler: it
+   compiles PATTERN (of length SIZE) and puts the result in BUFP.
+   Returns 0 if the pattern was valid, otherwise an error string.
+
+   Assumes the `allocated' (and perhaps `buffer') and `translate' fields
+   are set in BUFP on entry.  */
+
+const char *
+re_compile_pattern (pattern, length, bufp)
+    const char *pattern;
+    size_t length;
+    struct re_pattern_buffer *bufp;
+{
+  reg_errcode_t ret;
+
+  /* GNU code is written to assume at least RE_NREGS registers will be set
+     (and at least one extra will be -1).  */
+  bufp->regs_allocated = REGS_UNALLOCATED;
+
+  /* And GNU code determines whether or not to get register information
+     by passing null for the REGS argument to re_match, etc., not by
+     setting no_sub.  */
+  bufp->no_sub = 0;
+
+  /* Match anchors at newline.  */
+  bufp->newline_anchor = 1;
+
+  ret = re_compile_internal (bufp, (const unsigned char *) pattern, length,
+                             re_syntax_options);
+
+  if (!ret)
+    return NULL;
+  return gettext (re_error_msgid + re_error_msgid_idx[(int) ret]);
+}
+#ifdef _LIBC
+weak_alias (__re_compile_pattern, re_compile_pattern)
+#endif
+
+/* Set by `re_set_syntax' to the current regexp syntax to recognize.  Can
+   also be assigned to arbitrarily: each pattern buffer stores its own
+   syntax, so it can be changed between regex compilations.  */
+/* This has no initializer because initialized variables in Emacs
+   become read-only after dumping.  */
+reg_syntax_t re_syntax_options;
+
+
+/* Specify the precise syntax of regexps for compilation.  This provides
+   for compatibility for various utilities which historically have
+   different, incompatible syntaxes.
+
+   The argument SYNTAX is a bit mask comprised of the various bits
+   defined in regex.h.  We return the old syntax.  */
+
+reg_syntax_t
+re_set_syntax (syntax)
+    reg_syntax_t syntax;
+{
+  reg_syntax_t ret = re_syntax_options;
+
+  re_syntax_options = syntax;
+  return ret;
+}
+#ifdef _LIBC
+weak_alias (__re_set_syntax, re_set_syntax)
+#endif
+
+int
+re_compile_fastmap (bufp)
+    struct re_pattern_buffer *bufp;
+{
+  re_dfa_t *dfa = (re_dfa_t *) bufp->buffer;
+  char *fastmap = bufp->fastmap;
+
+  memset (fastmap, '\0', sizeof (char) * SBC_MAX);
+  re_compile_fastmap_iter (bufp, dfa->init_state, fastmap);
+  if (dfa->init_state != dfa->init_state_word)
+    re_compile_fastmap_iter (bufp, dfa->init_state_word, fastmap);
+  if (dfa->init_state != dfa->init_state_nl)
+    re_compile_fastmap_iter (bufp, dfa->init_state_nl, fastmap);
+  if (dfa->init_state != dfa->init_state_begbuf)
+    re_compile_fastmap_iter (bufp, dfa->init_state_begbuf, fastmap);
+  bufp->fastmap_accurate = 1;
+  return 0;
+}
+#ifdef _LIBC
+weak_alias (__re_compile_fastmap, re_compile_fastmap)
+#endif
+
+/* Helper function for re_compile_fastmap.
+   Compile fastmap for the initial_state INIT_STATE.  */
+
+static void
+re_compile_fastmap_iter (bufp, init_state, fastmap)
+     regex_t *bufp;
+     const re_dfastate_t *init_state;
+     char *fastmap;
+{
+  re_dfa_t *dfa = (re_dfa_t *) bufp->buffer;
+  int node_cnt;
+  for (node_cnt = 0; node_cnt < init_state->nodes.nelem; ++node_cnt)
+    {
+      int node = init_state->nodes.elems[node_cnt];
+      re_token_type_t type = dfa->nodes[node].type;
+      if (type == OP_CONTEXT_NODE)
+        {
+          node = dfa->nodes[node].opr.ctx_info->entity;
+          type = dfa->nodes[node].type;
+        }
+
+      if (type == CHARACTER)
+        fastmap[dfa->nodes[node].opr.c] = 1;
+      else if (type == SIMPLE_BRACKET)
+        {
+          int i, j, ch;
+          for (i = 0, ch = 0; i < BITSET_UINTS; ++i)
+            for (j = 0; j < UINT_BITS; ++j, ++ch)
+              if (dfa->nodes[node].opr.sbcset[i] & (1 << j))
+                fastmap[ch] = 1;
+        }
+      else if (type == COMPLEX_BRACKET)
+        {
+          int i;
+          re_charset_t *cset = dfa->nodes[node].opr.mbcset;
+          if (cset->non_match || cset->ncoll_syms || cset->nequiv_classes
+              || cset->nranges || cset->nchar_classes)
+            {
+              if (_NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES) != 0)
+                {
+                  /* In this case we want to catch the bytes which are
+                     the first byte of any collation elements.
+                     e.g. In da_DK, we want to catch 'a' since "aa"
+                          is a valid collation element, and don't catch
+                          'b' since 'b' is the only collation element
+                          which starts from 'b'.  */
+                  int j, ch;
+                  const int32_t *table = (const int32_t *)
+                    _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB);
+                  for (i = 0, ch = 0; i < BITSET_UINTS; ++i)
+                    for (j = 0; j < UINT_BITS; ++j, ++ch)
+                      if (table[ch] < 0)
+                        fastmap[ch] = 1;
+                }
+            }
+          for (i = 0; i < cset->nmbchars; ++i)
+            {
+              unsigned char buf[256];
+              wctomb (buf, cset->mbchars[i]);
+              fastmap[buf[0]] = 1;
+            }
+        }
+      else if (type == END_OF_RE || type == COMPLEX_BRACKET
+               || type == OP_PERIOD)
+        {
+          memset (fastmap, '\1', sizeof (char) * SBC_MAX);
+          if (type == END_OF_RE)
+            bufp->can_be_null = 1;
+          return;
+        }
+    }
+}
+
+/* Entry point for POSIX code.  */
+/* regcomp takes a regular expression as a string and compiles it.
+
+   PREG is a regex_t *.  We do not expect any fields to be initialized,
+   since POSIX says we shouldn't.  Thus, we set
+
+     `buffer' to the compiled pattern;
+     `used' to the length of the compiled pattern;
+     `syntax' to RE_SYNTAX_POSIX_EXTENDED if the
+       REG_EXTENDED bit in CFLAGS is set; otherwise, to
+       RE_SYNTAX_POSIX_BASIC;
+     `newline_anchor' to REG_NEWLINE being set in CFLAGS;
+     `fastmap' to an allocated space for the fastmap;
+     `fastmap_accurate' to zero;
+     `re_nsub' to the number of subexpressions in PATTERN.
+
+   PATTERN is the address of the pattern string.
+
+   CFLAGS is a series of bits which affect compilation.
+
+     If REG_EXTENDED is set, we use POSIX extended syntax; otherwise, we
+     use POSIX basic syntax.
+
+     If REG_NEWLINE is set, then . and [^...] don't match newline.
+     Also, regexec will try a match beginning after every newline.
+
+     If REG_ICASE is set, then we considers upper- and lowercase
+     versions of letters to be equivalent when matching.
+
+     If REG_NOSUB is set, then when PREG is passed to regexec, that
+     routine will report only success or failure, and nothing about the
+     registers.
+
+   It returns 0 if it succeeds, nonzero if it doesn't.  (See regex.h for
+   the return codes and their meanings.)  */
+
+int
+regcomp (preg, pattern, cflags)
+    regex_t *preg;
+    const char *pattern;
+    int cflags;
+{
+  reg_errcode_t ret;
+  reg_syntax_t syntax = ((cflags & REG_EXTENDED) ? RE_SYNTAX_POSIX_EXTENDED
+                         : RE_SYNTAX_POSIX_BASIC);
+
+  preg->buffer = NULL;
+  preg->allocated = 0;
+  preg->used = 0;
+
+  /* Try to allocate space for the fastmap.  */
+  preg->fastmap = re_malloc (char, SBC_MAX);
+  if (preg->fastmap == NULL)
+    return REG_ESPACE;
+
+  syntax |= (cflags & REG_ICASE) ? RE_ICASE : 0;
+
+  /* If REG_NEWLINE is set, newlines are treated differently.  */
+  if (cflags & REG_NEWLINE)
+    { /* REG_NEWLINE implies neither . nor [^...] match newline.  */
+      syntax &= ~RE_DOT_NEWLINE;
+      syntax |= RE_HAT_LISTS_NOT_NEWLINE;
+      /* It also changes the matching behavior.  */
+      preg->newline_anchor = 1;
+    }
+  else
+    preg->newline_anchor = 0;
+  preg->no_sub = !!(cflags & REG_NOSUB);
+  preg->translate = NULL;
+
+  ret = re_compile_internal (preg, pattern, strlen (pattern), syntax);
+
+  /* POSIX doesn't distinguish between an unmatched open-group and an
+     unmatched close-group: both are REG_EPAREN.  */
+  if (ret == REG_ERPAREN)
+    ret = REG_EPAREN;
+
+  /* XXX Why the test for preg->fastmap != NULL?  */
+  if (ret == REG_NOERROR && preg->fastmap != NULL)
+    {
+      /* Compute the fastmap now, since regexec cannot modify the pattern
+	 buffer.  */
+      if (re_compile_fastmap (preg) == -2)
+	{
+	  /* Some error occurred while computing the fastmap, just forget
+	     about it.  */
+	  re_free (preg->fastmap);
+	  preg->fastmap = NULL;
+	}
+    }
+
+  return (int) ret;
+}
+#ifdef _LIBC
+weak_alias (__regcomp, regcomp)
+#endif
+
+/* Returns a message corresponding to an error code, ERRCODE, returned
+   from either regcomp or regexec.   We don't use PREG here.  */
+
+size_t
+regerror (errcode, preg, errbuf, errbuf_size)
+    int errcode;
+    const regex_t *preg;
+    char *errbuf;
+    size_t errbuf_size;
+{
+  const char *msg;
+  size_t msg_size;
+
+  if (errcode < 0
+      || errcode >= (int) (sizeof (re_error_msgid_idx)
+			   / sizeof (re_error_msgid_idx[0])))
+    /* Only error codes returned by the rest of the code should be passed
+       to this routine.  If we are given anything else, or if other regex
+       code generates an invalid error code, then the program has a bug.
+       Dump core so we can fix it.  */
+    abort ();
+
+  msg = gettext (re_error_msgid + re_error_msgid_idx[errcode]);
+
+  msg_size = strlen (msg) + 1; /* Includes the null.  */
+
+  if (errbuf_size != 0)
+    {
+      if (msg_size > errbuf_size)
+        {
+#if defined HAVE_MEMPCPY || defined _LIBC
+	  *((char *) __mempcpy (errbuf, msg, errbuf_size - 1)) = '\0';
+#else
+          memcpy (errbuf, msg, errbuf_size - 1);
+          errbuf[errbuf_size - 1] = 0;
+#endif
+        }
+      else
+        memcpy (errbuf, msg, msg_size);
+    }
+
+  return msg_size;
+}
+#ifdef _LIBC
+weak_alias (__regerror, regerror)
+#endif
+
+/* Free dynamically allocated space used by PREG.  */
+
+void
+regfree (preg)
+    regex_t *preg;
+{
+  int i, j;
+  re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+  if (dfa != NULL)
+    {
+      re_free (dfa->subexps);
+
+      for (i = 0; i < dfa->nodes_len; ++i)
+        {
+