about summary refs log tree commit diff
path: root/iconvdata
diff options
context:
space:
mode:
authorUlrich Drepper <drepper@redhat.com>1998-04-26 14:03:20 +0000
committerUlrich Drepper <drepper@redhat.com>1998-04-26 14:03:20 +0000
commit8babd571bf5d07755947de8ad0bea1895df2df81 (patch)
treeb5786fbd48e24d4c28a849751a9ac4bb2682038a /iconvdata
parentbc900b1118e1fc7b065b37e4d28aa7c998ce5a66 (diff)
downloadglibc-8babd571bf5d07755947de8ad0bea1895df2df81.tar.gz
glibc-8babd571bf5d07755947de8ad0bea1895df2df81.tar.xz
glibc-8babd571bf5d07755947de8ad0bea1895df2df81.zip
Update.
	* iconvdata/Makefile: Add rules for ISO-2022-KR.
	* iconvdata/iso-2022-kr.c: New file.

	* iconvdata/gconv-modules: Add entry for ISO-2022-JP, ISO-2022-JP-2,
	and ISO-2022-KR.
Diffstat (limited to 'iconvdata')
-rw-r--r--iconvdata/Makefile9
-rw-r--r--iconvdata/gconv-modules4
-rw-r--r--iconvdata/iso-2022-jp.c8
-rw-r--r--iconvdata/iso-2022-kr.c273
4 files changed, 286 insertions, 8 deletions
diff --git a/iconvdata/Makefile b/iconvdata/Makefile
index d302250a8b..276ae886ed 100644
--- a/iconvdata/Makefile
+++ b/iconvdata/Makefile
@@ -31,7 +31,7 @@ modules	:= ISO8859-1 ISO8859-2 ISO8859-3 ISO8859-4 ISO8859-5		 \
 	   EBCDIC-FI-SE-A EBCDIC-FR EBCDIC-IS-FRISS EBCDIC-IT EBCDIC-PT	 \
 	   EBCDIC-UK EBCDIC-US IBM037 IBM038 IBM274 IBM275 IBM423 IBM424 \
 	   IBM500 IBM870 IBM871 IBM891 IBM903 IBM904 IBM905 IBM1047	 \
-	   CP874 CP737 CP775
+	   CP874 CP737 CP775 ISO-2022-KR
 ifneq ($(PERL),no)
 modules += KOI8-R LATIN-GREEK LATIN-GREEK-1 IBM256 IBM273 IBM277 IBM278	 \
 	   IBM280 IBM281 IBM284 IBM285 IBM290 IBM297 IBM420 IBM437	 \
@@ -143,6 +143,7 @@ EUC-JP-routines := eucjp
 EUC-CN-routines := euccn
 EUC-TW-routines := euctw
 ISO-2022-JP-routines := iso-2022-jp
+ISO-2022-KR-routines := iso-2022-kr
 libJIS-routines := jis0201 jis0208 jis0212
 libKSC-routines := ksc5601
 libGB-routines := gb2312
@@ -163,7 +164,9 @@ $(objpfx)EUC-TW.so: $(objpfx)libCNS.so
 
 LDFLAGS-ISO-2022-JP.so = -Wl,-rpath,$(gconvdir)
 $(objpfx)ISO-2022-JP.so: $(objpfx)libJIS.so $(objpfx)libGB.so \
-			 $(objpfx)libCNS.so
+			 $(objpfx)libCNS.so $(objpfx)libKSC.so
+LDFLAGS-ISO-2022-JP.so = -Wl,-rpath,$(gconvdir)
+$(objpfx)ISO-2022-JP.so: $(objpfx)libKSC.so
 
 LDFLAGS-libJIS.so = -Wl,-soname,$(@F)
 LDFLAGS-libKSC.so = -Wl,-soname,$(@F)
@@ -193,7 +196,7 @@ distribute := 8bit-generic.c 8bit-gap.c gap.pl gaptab.pl gconv-modules	    \
 	      ibm903.c ibm904.c ibm905.c ibm918.c ibm1004.c ibm1026.c	    \
 	      ibm1047.c cp1250.c cp1251.c cp1252.c cp1253.c cp1254.c	    \
 	      cp1255.c cp1256.c cp1257.c cp874.c cp874.h cp737.c cp737.h    \
-	      cp775.c cp775.h iso-2022-jp.c
+	      cp775.c cp775.h iso-2022-jp.c iso-2022-kr.c
 
 # We build the transformation modules only when we build shared libs.
 ifeq (yes,$(build-shared))
diff --git a/iconvdata/gconv-modules b/iconvdata/gconv-modules
index 169fbb2b52..fe47da1e9c 100644
--- a/iconvdata/gconv-modules
+++ b/iconvdata/gconv-modules
@@ -757,3 +757,7 @@ module	INTERNAL		ISO-2022-JP//		ISO-2022-JP	1
 
 module	ISO-2022-JP-2//		INTERNAL		ISO-2022-JP	1
 module	INTERNAL		ISO-2022-JP-2//		ISO-2022-JP	1
+
+#	from			  to			module		cost
+module	ISO-2022-KR//		INTERNAL		ISO-2022-KR	1
+module	INTERNAL		ISO-2022-KR//		ISO-2022-KR	1
diff --git a/iconvdata/iso-2022-jp.c b/iconvdata/iso-2022-jp.c
index afd6579d16..9ca648c3c2 100644
--- a/iconvdata/iso-2022-jp.c
+++ b/iconvdata/iso-2022-jp.c
@@ -54,6 +54,7 @@ struct gap
 #define PREPARE_LOOP \
   enum direction dir = ((struct iso2022jp_data *) step->data)->dir;	      \
   enum variant var = ((struct iso2022jp_data *) step->data)->var;	      \
+  int save_state;							      \
   int set = data->statep->count;
 #define END_LOOP \
   data->statep->count = set;
@@ -81,7 +82,6 @@ struct iso2022jp_data
 {
   enum direction dir;
   enum variant var;
-  mbstate_t save_state;
 };
 
 
@@ -211,11 +211,9 @@ gconv_end (struct gconv_step *data)
    and retore the state.  */
 #define SAVE_RESET_STATE(Save) \
   if (Save)								      \
-    ((struct iso2022jp_data *) step->data)->save_state.count		      \
-      = data->statep->count;						      \
+    save_set = set;							      \
   else									      \
-    data->statep->count							      \
-      = ((struct iso2022jp_data *) step->data)->save_state.count
+    set = save_set
 
 
 /* First define the conversion function from ISO-2022-JP to UCS4.  */
diff --git a/iconvdata/iso-2022-kr.c b/iconvdata/iso-2022-kr.c
new file mode 100644
index 0000000000..16e656a5d3
--- /dev/null
+++ b/iconvdata/iso-2022-kr.c
@@ -0,0 +1,273 @@
+/* Conversion module for ISO-2022-KR.
+   Copyright (C) 1998 Free Software Foundation, Inc.
+   This file is part of the GNU C Library.
+   Contributed by Ulrich Drepper <drepper@cygnus.com>, 1998.
+
+   The GNU C Library is free software; you can redistribute it and/or
+   modify it under the terms of the GNU Library General Public License as
+   published by the Free Software Foundation; either version 2 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
+   Library General Public License for more details.
+
+   You should have received a copy of the GNU Library General Public
+   License along with the GNU C Library; see the file COPYING.LIB.  If not,
+   write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+   Boston, MA 02111-1307, USA.  */
+
+#include <gconv.h>
+#include <stdint.h>
+#include <string.h>
+#include "ksc5601.h"
+
+/* This makes obvious what everybody knows: 0x1b is the Esc character.  */
+#define SI	0x0f
+#define SO	0x0e
+
+/* Definitions used in the body of the `gconv' function.  */
+#define DEFINE_INIT		1
+#define DEFINE_FINI		1
+#define FROM_LOOP		from_iso2022kr_loop
+#define TO_LOOP			to_iso2022kr_loop
+#define MIN_NEEDED_FROM		1
+#define MAX_NEEDED_FROM		3
+#define MIN_NEEDED_TO		4
+#define MAX_NEEDED_TO		4
+#define PREPARE_LOOP \
+  int save_set;
+  int set = data->statep->count;
+
+/* The COUNT element of the state keeps track of the currently selected
+   character set.  The possible values are:  */
+enum
+{
+  ASCII_set = 0,
+  KSC5601_set
+};
+
+
+/* Since this is a stateful encoding we have to provide code which resets
+   the output state to the initial state.  This has to be done during the
+   flushing.  */
+#define EMIT_SHIFT_TO_INIT \
+  if (data->statep->count != 0)						      \
+    {									      \
+      if (step->data == &from_object)					      \
+	/* It's easy, we don't have to emit anything, we just reset the	      \
+	   state for the input.  */					      \
+	set = 0;							      \
+      else								      \
+	{								      \
+	  char *outbuf = data->outbuf;					      \
+	  								      \
+	  /* We are not in the initial state.  To switch back we have	      \
+	     to emit `SO'.  */						      \
+	  if (outbuf == data->outbufend)				      \
+	    /* We don't have enough room in the output buffer.  */	      \
+	    status = GCONV_FULL_OUTPUT;					      \
+	  else								      \
+	    {								      \
+	      /* Write out the shift sequence.  */			      \
+	      *outbuf++ = SO;						      \
+	      data->outbuf = outbuf;					      \
+	      set = 0;							      \
+	    }								      \
+	}								      \
+    }
+
+
+/* Since we might have to reset input pointer we must be able to save
+   and retore the state.  */
+#define SAVE_RESET_STATE(Save) \
+  if (Save)								      \
+    save_set = set;							      \
+  else									      \
+    set = save_set
+
+
+/* First define the conversion function from ISO-2022-JP to UCS4.  */
+#define MIN_NEEDED_INPUT	MIN_NEEDED_FROM
+#define MAX_NEEDED_INPUT	MAX_NEEDED_FROM
+#define MIN_NEEDED_OUTPUT	MIN_NEEDED_TO
+#define LOOPFCT			FROM_LOOP
+#define BODY \
+  {									      \
+    uint32_t ch = *inptr;						      \
+									      \
+    /* This is a 7bit character set, disallow all 8bit characters.  */	      \
+    if (ch > 0x7f)							      \
+      {									      \
+	result = GCONV_ILLEGAL_INPUT;					      \
+	break;								      \
+      }									      \
+									      \
+    /* Recognize escape sequences.  */					      \
+    if (ch == ESC)							      \
+      {									      \
+	/* We don't really have to handle escape sequences since all the      \
+	   switching is done using the SI and SO bytes.  Butwe have to	      \
+	   recognize `Esc $ ) C' since this is a kind of flag for this	      \
+	   encoding.  We simply ignore it.  */				      \
+	if (inptr + 1 > inend						      \
+	    || (inptr[1] == '$'						      \
+		&& (inptr + 2 > inend					      \
+		    || (inptr[2] == ')' && inptr + 3 > inend))))	      \
+			    						      \
+	  {								      \
+	    result = GCONV_EMPTY_INPUT;					      \
+	    break;							      \
+	  }								      \
+	if (inptr[1] == '$' && inptr[2] == ')' && inptr[3] == 'C')	      \
+	  {								      \
+	    /* Yeah, yeah, we know this is ISO 2022-KR.  */		      \
+	    inptr += 4;							      \
+	    continue;							      \
+	  }								      \
+      }									      \
+    else if (ch == SI)							      \
+      {									      \
+	/* Switch to use KSC.  */					      \
+	++inptr;							      \
+	set = KSC5601_set;						      \
+	continue;							      \
+      }									      \
+    else if (ch == SO)							      \
+      {									      \
+	/* Switch to use ASCII.  */					      \
+	++inptr;							      \
+	set = ASCII_set;						      \
+	continue;							      \
+      }									      \
+									      \
+    if (set == ASCII_set || ch < 0x21 || ch == 0x7f)			      \
+      /* Almost done, just advance the input pointer.  */		      \
+      ++inptr;								      \
+    else								      \
+      {									      \
+	assert (set == KSC5601_set);					      \
+									      \
+	/* Use the KSC 5601 table.  */					      \
+	ch = ksc5601_to_ucs4 (&inptr,					      \
+			      NEED_LENGTH_TEST ? inend - inptr : 2, 0);	      \
+									      \
+	if (NEED_LENGTH_TEST && ch == 0)				      \
+	  {								      \
+	    result = GCONV_EMPTY_INPUT;					      \
+	    break;							      \
+	  }								      \
+	else if (ch == UNKNOWN_10646_CHAR)				      \
+	  {								      \
+	    result = GCONV_ILLEGAL_INPUT;				      \
+	    break;							      \
+	  }								      \
+      }									      \
+									      \
+    *((uint32_t *) outptr)++ = ch;					      \
+  }
+#define EXTRA_LOOP_DECLS	, int set
+#include <iconv/loop.c>
+
+
+/* Next, define the other direction.  */
+#define MIN_NEEDED_INPUT	MIN_NEEDED_TO
+#define MIN_NEEDED_OUTPUT	MIN_NEEDED_FROM
+#define MAX_NEEDED_OUTPUT	MAX_NEEDED_FROM
+#define LOOPFCT			TO_LOOP
+#define BODY \
+  {									      \
+    unsigned char ch;							      \
+    size_t written = 0;							      \
+									      \
+    ch = *((uint32_t *) inptr);						      \
+									      \
+    /* First see whether we can write the character using the currently	      \
+       selected character set.  */					      \
+    if (set == ASCII_set || (ch >= 0x01 && (ch < 0x21 || ch == 0x7f)))	      \
+      {									      \
+	/* Please note that the NUL byte is *not* matched if we are not	      \
+	   currently using the ASCII charset.  This is because we must	      \
+	   switch to the initial state whenever a NUL byte is written.  */    \
+	if (ch <= 0x7f)							      \
+	  {								      \
+	    *outptr++ = ch;						      \
+	    written = 1;						      \
+	  }								      \
+      }									      \
+    else								      \
+      {									      \
+	assert (set == KSC5601_set);					      \
+									      \
+	written = ucs4_to_ksc5601 (ch, outptr,				      \
+				   (NEED_LENGTH_TEST ? outend - outptr : 2)); \
+									      \
+	if (NEED_LENGTH_TEST && written == 0)				      \
+	  {								      \
+	    result = GCONV_FULL_OUTPUT;					      \
+	    break;							      \
+	  }								      \
+	if (written == UNKNOWN_10646_CHAR)				      \
+	  {								      \
+	    /* Either this is an unknown character or we have to switch	      \
+	       the currently selected character set.  The character sets      \
+	       do not code entirely separate parts of ISO 10646 and	      \
+	       therefore there is no single correct result.  If we choose     \
+	       the character set to use wrong we might be end up with	      \
+	       using yet another character set for the next character	      \
+	       though the current and the next could be encoded with one      \
+	       character set.  We leave this kind of optimization for	      \
+	       later and now simply use a fixed order in which we test for    \
+	       availability  */						      \
+									      \
+	if (ch <= 0x7f)							      \
+	  {								      \
+	    /* We must encode using ASCII.  First write out the		      \
+	       escape sequence.  */					      \
+	    *outptr++ = SO;						      \
+	    set = ASCII_set;						      \
+									      \
+	    if (NEED_LENGTH_TEST && outptr == outend)			      \
+	      {								      \
+		result = GCONV_FULL_OUTPUT;				      \
+		break;							      \
+	      }								      \
+									      \
+	    *outptr++ = ch;						      \
+	  }								      \
+	else								      \
+	  {								      \
+	    written = ucs4_to_ksc5601 (ch, buf, 2);			      \
+	    if (written != UNKNOWN_10646_CHAR)				      \
+	      {								      \
+		/* We use KSC 5601.  */					      \
+		*outptr++ = SI;						      \
+		set = KSC5601_set;					      \
+									      \
+		if (NEED_LENGTH_TEST && outptr + 2 > outend)		      \
+		  {							      \
+		    result = GCONV_FULL_OUTPUT;				      \
+		    break;						      \
+		  }							      \
+									      \
+		*outptr++ = buf[0];					      \
+		*outptr++ = buf[1];					      \
+	      }								      \
+	    else							      \
+	      {								      \
+		result = GCONV_ILLEGAL_INPUT;				      \
+		break;							      \
+	      }								      \
+	  }								      \
+      }									      \
+									      \
+    /* Now that we wrote the output increment the input pointer.  */	      \
+    inptr += 4;								      \
+  }
+#define EXTRA_LOOP_DECLS	, int set
+#include <iconv/loop.c>
+
+
+/* Now define the toplevel functions.  */
+#include <iconv/skeleton.c>