summary refs log tree commit diff
path: root/wcsmbs/mbrtowc.c
diff options
context:
space:
mode:
authorUlrich Drepper <drepper@redhat.com>1998-04-07 09:21:28 +0000
committerUlrich Drepper <drepper@redhat.com>1998-04-07 09:21:28 +0000
commit4bca4c174852bfc348f99e85684fc8f65631d125 (patch)
treea5779cb83844d32d94557e4b8613278d7d877c9b /wcsmbs/mbrtowc.c
parent91641c65220646ae2ad8b4ec6972adc7e6720d36 (diff)
downloadglibc-4bca4c174852bfc348f99e85684fc8f65631d125.tar.gz
glibc-4bca4c174852bfc348f99e85684fc8f65631d125.tar.xz
glibc-4bca4c174852bfc348f99e85684fc8f65631d125.zip
Update.
1998-04-07 08:51  Ulrich Drepper  <drepper@cygnus.com>

	* iconv/gconv.c: Take care for NULL arguments.
	* iconv/gconv_dl.c: Make find_func function global.
	* iconv/gconv_int.h: Add prototype for __gconv_find_func.
	Add prototypes for ASCII conversion functions.
	* iconv/gconv_simple.c: Add ASCII conversion functions.
	* locale/C-ctype.c: Correct charset name.

	* wcsmbs/Makefile (distribute): Add wcsmbsload.h.
	(routines): Remove wmemrtowcs and wmemrtombs, add wcsnlen and
	wcsmbsload.
	* wcsmbs/btowc.c: Rewrite to use iconv functionality.
	* wcsmbs/mbrtowc.c: Likewise.
	* wcsmbs/mbsnrtowcs.c: Likewise.
	* wcsmbs/mbsrtowcs.c: Likewise.
	* wcsmbs/wcrtomb.c: Likewise.
	* wcsmbs/wcsnrtombs.c: Likewise.
	* wcsmbs/wcsrtombs.c: Likewise.
	* wcsmbs/wctob.c: Likewise.
	* wcsmbs/wchar.h: Add prototype for __wcslen, wcsnlen.  Remove
	prototypes for wmemr*.
	* wcsmbs/wcslen.c: Rename to __wcslen and make wcslen weak alias.
	* wcsmbs/wcsnlen.c: New file.
	* wcsmbs/wcsmbsload.c: New file.
	* wcsmbs/wcsmbsload.h: New file.

	* manual/filesys.texi: Mention risks of tmpnam and mktemp.

	* manual/install.texi: Describe some more critical points.

	* string/string.h: Add prototype for __strnlen.
	* string/strnlen.c: Rename to __strnlen and make strnlen weak alias.

	* sysdeps/posix/mktemp.c: Rewrite to allow many more files and
	much less predictable names.
	* sysdeps/posix/mkstemp.c: Likewise.

1998-04-05  Andreas Jaeger  <aj@arthur.rhein-neckar.de>

	* manual/libc.texinfo (Top): Change "file namespace" to "local
	namespace".

	* manual/socket.texi: Change file namespace to local namespace.
	(Out-of-Band Data): Remove unneeded variable link.
	(Host Address Functions): Use uint32_t consequently and add a
	number of clarifications for IPv4/IPv6, classless addresses.
	(Internet Namespace): Added some paragraphs about IPv6.
	Based on suggestions by Francesco Potorti` <F.Potorti@cnuce.cnr.it>.

1998-04-05  Philip Blundell  <Philip.Blundell@pobox.com>

	Update for draft-ietf-ipngwg-bsd-api-new-01.txt:

	* resolv/netdb.h (getnameinfo): Use `socklen_t' not `size_t'.
	(NI_NUMERICHOST, et al.): Tidy up and add comments.
	(AI_NUMERICHOST): Define.
	(getnodebyname): New prototype.
	(AI_V4MAPPED, et al.): New constants.

	* sysdeps/unix/bsd/bsd4.4/bits/sockaddr.h (SA_LEN): New macro.<
	* sysdeps/generic/bits/sockaddr.h (SA_LEN): Likewise.
	* sysdeps/unix/sysv/linux/Makefile (sysdep_routines): Add sa_len
	for socket.
	* sysdeps/unix/sysv/linux/sa_len.c: New file.
	* sysdeps/unix/sysv/linux/Dist: Add sa_len.c.

	* sysdeps/unix/sysv/linux/bits/socket.h: Make multiple #inclusion safe.

	* sysdeps/generic/bits/sockunion.h: New file, defining sockaddr_union.
	* sysdeps/unix/sysv/linux/bits/sockunion.h: Likewise.
	* socket/Makefile (headers): Add bits/sockunion.h.
	* socket/sys/socket.h: Include <bits/sockunion.h>
	* inet/netinet/in.h: Include <bits/sockaddr.h> rather than
	<sys/socket.h>, to avoid getting sockaddr_union defined.
	* sysdeps/generic/bits/socket.h: Allow inclusion from netinet.in.h.
	* sysdeps/unix/sysv/linux/bits/socket.h: Likewise.
	* sysdeps/unix/sysv/linux/mips/bits/socket.h: Likewise.

	* sysdeps/unix/sysv/linux/if_index.c: Remove use of SIOCGIFCOUNT
	(2.2.x kernels won't have it).

1998-04-06 21:21  Ulrich Drepper  <drepper@cygnus.com>

	* sysdeps/mach/hurd/bits/ioctls.h: Allow inclusion from hurd/ioctl.h.
	* sysdeps/mach/hurd/dl-sysdep.c: Use __ptr_t not caddr_t.
	* iconv/gconv_conf.c: Define MAXPATHLEN if not available before.
	Patches by UCHIYAMA Yasushi <uch@nop.or.jp>.

1998-04-05  Philip Blundell  <Philip.Blundell@pobox.com>

	* manual/socket.texi (Socket Addresses): Fix a typo.
	(Interface Naming): New section.

1998-04-05 23:29  Zack Weinberg  <zack@rabi.phys.columbia.edu>

	* configure.in: Check for CC using $ac_tool_prefix, to handle
	cross-compilation.  Use AC_CHECK_TOOL to find MiG.
	* config.make.in: Add MIG to be substituted.
	* mach/Machrules: Don't define MIG variable.

	Patch by Gordon Matzigkeit <gord@profitpress.com>.

1998-04-05  Mark Kettenis  <kettenis@phys.uva.nl>

	* elf/dl-load.c (decompose_rpath): Use local_strdup instead of
	strdupa to copy rpath.

1998-04-05  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* configure.in: Fix gcc version check.

	* aclocal.m4 (AC_PROG_CHECK_VER): Rewritten to make less confusing.

1998-04-05  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* configure.in: Fix gcc version check.

	* aclocal.m4 (AC_PROG_CHECK_VER): Rewritten to make less
	confusing.

1998-04-05  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* configure.in: Fix gcc version check.

	* aclocal.m4 (AC_PROG_CHECK_VER): Rewritten to make less
	confusing.

1998-03-31  Mark Kettenis  <kettenis@phys.uva.nl>

	* sysdeps/unix/bsd/unlockpt.c (unlockpt): Call __ptsname_r instead
	of ptsname_r.

	* stdlib/stdlib.h: Change prototype of ptsname_r to make it more
	like ttyname_r.
	* sysdeps/unix/sysv/linux/ptsname.c (__ptsname_r): Likewise.
	* sysdeps/generic/ptsname.c (__ptsname_r): Likewise.
	* sysdeps/unix/grantpt.c (grantpt): Change check of return value
	of __ptsname_r accordingly.
	* login/openpty.c (openpty): Likewise.

1998-06-04 18:47  H.J. Lu  <hjl@gnu.org>

	* libio/fileops.c (_IO_file_xsgetn): Adjust pointers.

1998-04-06 13:58  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* Makeconfig (libtypes): Redo H.J. Lu's change of 1998-03-27.

	* Makerules: Readd missing rule for $(objpfx)stamp.oS.

1998-04-06  Thorsten Kukuk  <kukuk@vt.uni-paderborn.de>

	* nis/ypclnt.c: Don't give YPERR_RPC back, use the do_ypcall
	return value for better error checking.
Diffstat (limited to 'wcsmbs/mbrtowc.c')
-rw-r--r--wcsmbs/mbrtowc.c144
1 files changed, 54 insertions, 90 deletions
diff --git a/wcsmbs/mbrtowc.c b/wcsmbs/mbrtowc.c
index cf0bbd6ce2..99bc5a4797 100644
--- a/wcsmbs/mbrtowc.c
+++ b/wcsmbs/mbrtowc.c
@@ -1,4 +1,4 @@
-/* Copyright (C) 1996, 1997 Free Software Foundation, Inc.
+/* Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
    Contributed by Ulrich Drepper <drepper@gnu.ai.mit.edu>, 1996.
 
@@ -18,113 +18,77 @@
    Boston, MA 02111-1307, USA.  */
 
 #include <errno.h>
+#include <gconv.h>
 #include <wchar.h>
+#include <wcsmbsload.h>
+
+#include <assert.h>
 
 #ifndef EILSEQ
-#define EILSEQ EINVAL
+# define EILSEQ EINVAL
 #endif
 
 
-static mbstate_t internal;
+/* This is the private state used if PS is NULL.  */
+static mbstate_t state;
 
 size_t
 __mbrtowc (wchar_t *pwc, const char *s, size_t n, mbstate_t *ps)
 {
-  size_t used = 0;
-
-  if (ps == NULL)
-    ps = &internal;
-
+  wchar_t buf[1];
+  struct gconv_step_data data;
+  size_t inbytes;
+  int status;
+  size_t result;
+
+  /* Tell where we want the result.  */
+  data.outbuf = (char *) (pwc ?: buf);
+  data.outbufavail = 0;
+  data.outbufsize = sizeof (wchar_t);
+  data.is_last = 1;
+  data.statep = ps ?: &state;
+
+  /* A first special case is if S is NULL.  This means put PS in the
+     initial state.  */
   if (s == NULL)
     {
-      /* See first paragraph of description in 7.16.6.3.2.  */
-      ps->count = 0;
-      return 0;
+      data.outbuf = (char *) buf;
+      s = "";
+      n = 1;
     }
 
-  if (n > 0)
-    {
-      if (ps->count == 0)
-	{
-	  unsigned char byte = (unsigned char) *s++;
-	  ++used;
-
-	  /* We must look for a possible first byte of a UTF8 sequence.  */
-	  if (byte < 0x80)
-	    {
-	      /* One byte sequence.  */
-	      if (pwc != NULL)
-		*pwc = (wchar_t) byte;
-	      return byte ? used : 0;
-	    }
-
-	  if ((byte & 0xc0) == 0x80 || (byte & 0xfe) == 0xfe)
-	    {
-	      /* Oh, oh.  An encoding error.  */
-	      __set_errno (EILSEQ);
-	      return (size_t) -1;
-	    }
-
-	  if ((byte & 0xe0) == 0xc0)
-	    {
-	      /* We expect two bytes.  */
-	      ps->count = 1;
-	      ps->value = byte & 0x1f;
-	    }
-	  else if ((byte & 0xf0) == 0xe0)
-	    {
-	      /* We expect three bytes.  */
-	      ps->count = 2;
-	      ps->value = byte & 0x0f;
-	    }
-	  else if ((byte & 0xf8) == 0xf0)
-	    {
-	      /* We expect four bytes.  */
-	      ps->count = 3;
-	      ps->value = byte & 0x07;
-	    }
-	  else if ((byte & 0xfc) == 0xf8)
-	    {
-	      /* We expect five bytes.  */
-	      ps->count = 4;
-	      ps->value = byte & 0x03;
-	    }
-	  else
-	    {
-	      /* We expect six bytes.  */
-	      ps->count = 5;
-	      ps->value = byte & 0x01;
-	    }
-	}
+  /* Make sure we use the correct function.  */
+  update_conversion_ptrs ();
 
-      /* We know we have to handle a multibyte character and there are
-	 some more bytes to read.  */
-      while (used < n)
+  /* Do a normal conversion.  */
+  inbytes = n;
+  status = (*__wcsmbs_gconv_fcts.towc->fct) (__wcsmbs_gconv_fcts.towc,
+					     &data, s, &inbytes, NULL, 0);
+
+  /* There must not be any problems with the conversion but illegal input
+     characters.  The output buffer must be large enough, otherwise the
+     definition of MB_CUR_MAX is not correct.  All the other possible
+     errors also must not happen.  */
+  assert (status == GCONV_OK || status == GCONV_ILLEGAL_INPUT
+	  || status == GCONV_INCOMPLETE_INPUT);
+
+  if (status == GCONV_OK)
+    {
+      if (*(wchar_t *)data.outbuf == L'\0')
 	{
-	  /* The second to sixths byte must be of the form 10xxxxxx.  */
-	  unsigned char byte = (unsigned char) *s++;
-	  ++used;
-
-	  if ((byte & 0xc0) != 0x80)
-	    {
-	      /* Oh, oh.  An encoding error.  */
-	      __set_errno (EILSEQ);
-	      return (size_t) -1;
-	    }
-
-	  ps->value <<= 6;
-	  ps->value |= byte & 0x3f;
-
-	  if (--ps->count == 0)
-	    {
-	      /* The character is finished.  */
-	      if (pwc != NULL)
-		*pwc = (wchar_t) ps->value;
-	      return ps->value ? used : 0;
-	    }
+	  /* The converted character is the NUL character.  */
+	  assert (mbsinit (data.statep));
+	  result = 0;
 	}
+      else
+	result = n - inbytes;
+    }
+  else
+    {
+      result = status == GCONV_INCOMPLETE_INPUT ? (size_t) -2 : (size_t) -1;
+      __set_errno (EILSEQ);
     }
 
-  return (size_t) -2;
+  return result;
 }
 weak_alias (__mbrtowc, mbrtowc)