OSDN Git Service

* fix some errors.
[nkf/nkf.git] / nkf.c
diff --git a/nkf.c b/nkf.c
index a56a912..505d192 100644 (file)
--- a/nkf.c
+++ b/nkf.c
@@ -1,7 +1,7 @@
 /** Network Kanji Filter. (PDS Version)
 ************************************************************************
 ** Copyright (C) 1987, Fujitsu LTD. (Itaru ICHIKAWA)
-** \e$BO"Mm@h!'\e(B \e$B!J3t!KIY;NDL8&5f=j!!%=%U%H#38&!!;T@n!!;j\e(B 
+** \e$BO"Mm@h!'\e(B \e$B!J3t!KIY;NDL8&5f=j!!%=%U%H#38&!!;T@n!!;j\e(B
 ** \e$B!J\e(BE-Mail Address: ichikawa@flab.fujitsu.co.jp\e$B!K\e(B
 ** Copyright (C) 1996,1998
 ** Copyright (C) 2002
 **    \e$B%P%$%J%j$NG[I[$N:]$K$O\e(Bversion message\e$B$rJ]B8$9$k$3$H$r>r7o$H$7$^$9!#\e(B
 **    \e$B$3$N%W%m%0%i%`$K$D$$$F$OFC$K2?$NJ]>Z$b$7$J$$!"0-$7$+$i$:!#\e(B
 **
-**    Everyone is permitted to do anything on this program 
+**    Everyone is permitted to do anything on this program
 **    including copying, modifying, improving,
 **    as long as you don't try to pretend that you wrote it.
-**    i.e., the above copyright notice has to appear in all copies.  
+**    i.e., the above copyright notice has to appear in all copies.
 **    Binary distribution requires original version messages.
 **    You don't have to ask before copying, redistribution or publishing.
 **    THE AUTHOR DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE.
 ***********************************************************************/
 
 /***********************************************************************
-** UTF-8 \e$B%5%]!<%H$K$D$$$F\e(B
-**    \e$B=>Mh$N\e(B nkf \e$B$HF~$l$+$($F$=$N$^$^;H$($k$h$&$K$J$C$F$$$^$9\e(B
-**    nkf -e \e$B$J$I$H$7$F5/F0$9$k$H!"<+F0H=JL$G\e(B UTF-8 \e$B$HH=Dj$5$l$l$P!"\e(B
-**    \e$B$=$N$^$^\e(B euc-jp \e$B$KJQ49$5$l$^$9\e(B
-**
-**    \e$B$^$@%P%0$,$"$k2DG=@-$,9b$$$G$9!#\e(B
-**    (\e$BFC$K<+F0H=JL!"%3!<%I:.:_!"%(%i!<=hM}7O\e(B)
-**
-**    \e$B2?$+LdBj$r8+$D$1$?$i!"\e(B
-**        E-Mail: furukawa@tcp-ip.or.jp
-**    \e$B$^$G8fO"Mm$r$*4j$$$7$^$9!#\e(B
+ * \e$B8=:_!"\e(Bnkf \e$B$O\e(B SorceForge \e$B$K$F%a%s%F%J%s%9$,B3$1$i$l$F$$$^$9!#\e(B
+ * http://sourceforge.jp/projects/nkf/
 ***********************************************************************/
-/* $Id: nkf.c,v 1.112 2006/10/11 16:59:31 naruse Exp $ */
+/* $Id: nkf.c,v 1.159 2007/12/23 07:55:20 naruse Exp $ */
 #define NKF_VERSION "2.0.8"
-#define NKF_RELEASE_DATE "2006-10-12"
+#define NKF_RELEASE_DATE "2007-12-22"
+#define COPY_RIGHT \
+    "Copyright (C) 1987, FUJITSU LTD. (I.Ichikawa),2000 S. Kono, COW\n" \
+    "Copyright (C) 2002-2007 Kono, Furukawa, Naruse, mastodon"
+
 #include "config.h"
 #include "utf8tbl.h"
 
-#define COPY_RIGHT \
-    "Copyright (C) 1987, FUJITSU LTD. (I.Ichikawa),2000 S. Kono, COW\n" \
-    "Copyright (C) 2002-2006 Kono, Furukawa, Naruse, mastodon"
+#if defined(DEFAULT_CODE_JIS)
+#elif defined(DEFAULT_CODE_SJIS)
+#elif defined(DEFAULT_CODE_EUC)
+#elif defined(DEFAULT_CODE_UTF8)
+#else
+#define DEFAULT_CODE_JIS 1
+#endif
 
+#ifndef MIME_DECODE_DEFAULT
+#define MIME_DECODE_DEFAULT STRICT_MIME
+#endif
+#ifndef X0201_DEFAULT
+#define X0201_DEFAULT TRUE
+#endif
 
-/*
-**
-**
-**
-** USAGE:       nkf [flags] [file] 
-**
-** Flags:
-** b    Output is buffered             (DEFAULT)
-** u    Output is unbuffered
-**
-** t    no operation
-**
-** j    Output code is JIS 7 bit        (DEFAULT SELECT) 
-** s    Output code is MS Kanji         (DEFAULT SELECT) 
-** e    Output code is AT&T JIS         (DEFAULT SELECT) 
-** w    Output code is AT&T JIS         (DEFAULT SELECT) 
-** l    Output code is JIS 7bit and ISO8859-1 Latin-1
-**
-** m    MIME conversion for ISO-2022-JP
-** I    Convert non ISO-2022-JP charactor to GETA by Pekoe <pekoe@lair.net>
-** i_ Output sequence to designate JIS-kanji (DEFAULT_J)
-** o_ Output sequence to designate single-byte roman characters (DEFAULT_R)
-** M    MIME output conversion 
-**
-** r  {de/en}crypt ROT13/47
-**
-** v  display Version
-**
-** T  Text mode output        (for MS-DOS)
-**
-** x    Do not convert X0201 kana into X0208
-** Z    Convert X0208 alphabet to ASCII
-**
-** f60  fold option
-**
-** m    MIME decode
-** B    try to fix broken JIS, missing Escape
-** B[1-9]  broken level
-**
-** O   Output to 'nkf.out' file or last file name
-** d   Delete \r in line feed 
-** c   Add \r in line feed 
-** -- other long option
-** -- ignore following option (don't use with -O )
-**
-**/
+#if DEFAULT_NEWLINE == 0x0D0A
+#define PUT_NEWLINE(func) do {\
+    func(0x0D);\
+    func(0x0A);\
+} while (0)
+#define OCONV_NEWLINE(func) do {\
+    func(0, 0x0D);\
+    func(0, 0x0A);\
+} while (0)
+#elif DEFAULT_NEWLINE == 0x0D
+#define PUT_NEWLINE(func) func(0x0D)
+#define OCONV_NEWLINE(func) func(0, 0x0D)
+#else
+#define DEFAULT_NEWLINE 0x0A
+#define PUT_NEWLINE(func) func(0x0A)
+#define OCONV_NEWLINE(func) func(0, 0x0A)
+#endif
+#ifdef HELP_OUTPUT_STDERR
+#define HELP_OUTPUT stderr
+#else
+#define HELP_OUTPUT stdout
+#endif
 
 #if (defined(__TURBOC__) || defined(_MSC_VER) || defined(LSI_C) || defined(__MINGW32__) || defined(__EMX__) || defined(__MSDOS__) || defined(__WINDOWS__) || defined(__DOS__) || defined(__OS2__)) && !defined(MSDOS)
 #define MSDOS
@@ -196,45 +177,32 @@ void  djgpp_setbinmode(FILE *fp)
 #define         FALSE   0
 #define         TRUE    1
 
-/* state of output_mode and input_mode  
+/* state of output_mode and input_mode
 
    c2           0 means ASCII
-                X0201
-                ISO8859_1
-                X0208
+                JIS_X_0201
+                ISO_8859_1
+                JIS_X_0208
                 EOF      all termination
    c1           32bit data
 
  */
 
-#define         ASCII           0
-#define         X0208           1
-#define         X0201           2
-#define         ISO8859_1       8
-#define         NO_X0201        3
-#define         X0212      0x2844
-#define         X0213_1    0x284F
-#define         X0213_2    0x2850
-
 /* Input Assumption */
 
 #define         JIS_INPUT       4
 #define         EUC_INPUT      16
 #define         SJIS_INPUT      5
 #define         LATIN1_INPUT    6
+#define                UTF8_INPUT     13
+#define                UTF16_INPUT    1015
+#define                UTF32_INPUT    1017
+
 #define         FIXED_MIME      7
 #define         STRICT_MIME     8
 
 /* MIME ENCODE */
 
-#define                ISO2022JP       9
-#define        JAPANESE_EUC   10
-#define                SHIFT_JIS      11
-
-#define                UTF8           12
-#define                UTF8_INPUT     13
-#define                UTF16_INPUT    1015
-#define                UTF32_INPUT    1017
 
 /* byte order */
 
@@ -243,16 +211,14 @@ void  djgpp_setbinmode(FILE *fp)
 #define                ENDIAN_2143     2143
 #define                ENDIAN_3412     3412
 
-#define         WISH_TRUE      15
-
 /* ASCII CODE */
 
 #define         BS      0x08
 #define         TAB     0x09
-#define         NL      0x0a
+#define         LF      0x0a
 #define         CR      0x0d
 #define         ESC     0x1b
-#define         SPACE   0x20
+#define         SP      0x20
 #define         AT      0x40
 #define         SSP     0xa0
 #define         DEL     0x7f
@@ -260,6 +226,181 @@ void  djgpp_setbinmode(FILE *fp)
 #define         SO      0x0e
 #define         SSO     0x8e
 #define         SS3     0x8f
+#define         CRLF    0x0D0A
+
+
+/* encodings */
+
+enum nkf_encodings {
+    ASCII,
+    ISO_8859_1,
+    ISO_2022_JP,
+    CP50220,
+    CP50221,
+    CP50222,
+    ISO_2022_JP_1,
+    ISO_2022_JP_3,
+    SHIFT_JIS,
+    WINDOWS_31J,
+    CP10001,
+    EUC_JP,
+    CP51932,
+    EUCJP_MS,
+    EUCJP_ASCII,
+    SHIFT_JISX0213,
+    SHIFT_JIS_2004,
+    EUC_JISX0213,
+    EUC_JIS_2004,
+    UTF_8,
+    UTF_8N,
+    UTF_8_BOM,
+    UTF8_MAC,
+    UTF_16,
+    UTF_16BE,
+    UTF_16BE_BOM,
+    UTF_16LE,
+    UTF_16LE_BOM,
+    UTF_32,
+    UTF_32BE,
+    UTF_32BE_BOM,
+    UTF_32LE,
+    UTF_32LE_BOM,
+    JIS_X_0201=0x1000,
+    JIS_X_0208,
+    JIS_X_0212,
+    JIS_X_0213_1,
+    JIS_X_0213_2,
+    BINARY
+};
+
+nkf_char s_iconv(nkf_char c2, nkf_char c1, nkf_char c0);
+nkf_char e_iconv(nkf_char c2, nkf_char c1, nkf_char c0);
+nkf_char w_iconv(nkf_char c2, nkf_char c1, nkf_char c0);
+nkf_char w_iconv16(nkf_char c2, nkf_char c1, nkf_char c0);
+nkf_char w_iconv32(nkf_char c2, nkf_char c1, nkf_char c0);
+void j_oconv(nkf_char c2, nkf_char c1);
+void s_oconv(nkf_char c2, nkf_char c1);
+void e_oconv(nkf_char c2, nkf_char c1);
+void w_oconv(nkf_char c2, nkf_char c1);
+void w_oconv16(nkf_char c2, nkf_char c1);
+void w_oconv32(nkf_char c2, nkf_char c1);
+
+typedef struct {
+    char *name;
+    nkf_char (*iconv_func)(nkf_char c2, nkf_char c1, nkf_char c0);
+    void (*oconv_func)(nkf_char c2, nkf_char c1);
+} nkf_native_encoding;
+
+nkf_native_encoding NkfEncodingASCII =         { "US_ASCII", e_iconv, e_oconv };
+nkf_native_encoding NkfEncodingISO_2022_JP =   { "ISO-2022-JP", e_iconv, j_oconv };
+nkf_native_encoding NkfEncodingShift_JIS =     { "Shift_JIS", s_iconv, s_oconv };
+nkf_native_encoding NkfEncodingEUC_JP =                { "EUC-JP", e_iconv, e_oconv };
+nkf_native_encoding NkfEncodingUTF_8 =         { "UTF-8", w_iconv, w_oconv };
+nkf_native_encoding NkfEncodingUTF_16 =                { "UTF-16", w_iconv16, w_oconv16 };
+nkf_native_encoding NkfEncodingUTF_32 =                { "UTF-32", w_iconv32, w_oconv32 };
+
+typedef struct {
+    int id;
+    char *name;
+    nkf_native_encoding *based_encoding;
+} nkf_encoding;
+nkf_encoding nkf_encoding_table[] = {
+    {ASCII,            "ASCII",                &NkfEncodingASCII},
+    {ISO_8859_1,       "ISO-8859-1",           &NkfEncodingASCII},
+    {ISO_2022_JP,      "ISO-2022-JP",          &NkfEncodingASCII},
+    {CP50220,          "CP50220",              &NkfEncodingISO_2022_JP},
+    {CP50221,          "CP50221",              &NkfEncodingISO_2022_JP},
+    {CP50222,          "CP50222",              &NkfEncodingISO_2022_JP},
+    {ISO_2022_JP_1,    "ISO-2022-JP-1",        &NkfEncodingISO_2022_JP},
+    {ISO_2022_JP_3,    "ISO-2022-JP-3",        &NkfEncodingISO_2022_JP},
+    {SHIFT_JIS,                "Shift_JIS",            &NkfEncodingShift_JIS},
+    {WINDOWS_31J,      "WINDOWS-31J",          &NkfEncodingShift_JIS},
+    {CP10001,          "CP10001",              &NkfEncodingShift_JIS},
+    {EUC_JP,           "EUC-JP",               &NkfEncodingEUC_JP},
+    {CP51932,          "CP51932",              &NkfEncodingEUC_JP},
+    {EUCJP_MS,         "eucJP-MS",             &NkfEncodingEUC_JP},
+    {EUCJP_ASCII,      "eucJP-ASCII",          &NkfEncodingEUC_JP},
+    {SHIFT_JISX0213,   "Shift_JISX0213",       &NkfEncodingShift_JIS},
+    {SHIFT_JIS_2004,   "Shift_JIS-2004",       &NkfEncodingShift_JIS},
+    {EUC_JISX0213,     "EUC-JISX0213",         &NkfEncodingEUC_JP},
+    {EUC_JIS_2004,     "EUC-JIS-2004",         &NkfEncodingEUC_JP},
+    {UTF_8,            "UTF-8",                &NkfEncodingUTF_8},
+    {UTF_8N,           "UTF-8N",               &NkfEncodingUTF_8},
+    {UTF_8_BOM,                "UTF-8-BOM",            &NkfEncodingUTF_8},
+    {UTF8_MAC,         "UTF8-MAC",             &NkfEncodingUTF_8},
+    {UTF_16,           "UTF-16",               &NkfEncodingUTF_16},
+    {UTF_16BE,         "UTF-16BE",             &NkfEncodingUTF_16},
+    {UTF_16BE_BOM,     "UTF-16BE-BOM",         &NkfEncodingUTF_16},
+    {UTF_16LE,         "UTF-16LE",             &NkfEncodingUTF_16},
+    {UTF_16LE_BOM,     "UTF-16LE-BOM",         &NkfEncodingUTF_16},
+    {UTF_32,           "UTF-32",               &NkfEncodingUTF_32},
+    {UTF_32BE,         "UTF-32BE",             &NkfEncodingUTF_32},
+    {UTF_32BE_BOM,     "UTF-32BE-BOM",         &NkfEncodingUTF_32},
+    {UTF_32LE,         "UTF-32LE",             &NkfEncodingUTF_32},
+    {UTF_32LE_BOM,     "UTF-32LE-BOM",         &NkfEncodingUTF_32},
+    {BINARY,           "BINARY",               &NkfEncodingASCII},
+    {-1,               NULL,                   NULL}
+};
+#define NKF_ENCODING_TABLE_SIZE 34
+struct {
+    const char *name;
+    const int id;
+} encoding_name_to_id_table[] = {
+    {"ASCII",                  ASCII},
+    {"ISO-2022-JP",            ISO_2022_JP},
+    {"X-ISO2022JP-CP932",      CP50220},
+    {"CP50220",                        CP50220},
+    {"CP50221",                        CP50221},
+    {"CP50222",                        CP50222},
+    {"ISO-2022-JP-1",          ISO_2022_JP_1},
+    {"ISO-2022-JP-3",          ISO_2022_JP_3},
+    {"SHIFT_JIS",              SHIFT_JIS},
+    {"SJIS",                   SHIFT_JIS},
+    {"WINDOWS-31J",            WINDOWS_31J},
+    {"CSWINDOWS31J",           WINDOWS_31J},
+    {"CP932",                  WINDOWS_31J},
+    {"MS932",                  WINDOWS_31J},
+    {"CP10001",                        CP10001},
+    {"EUCJP",                  EUC_JP},
+    {"EUC-JP",                 EUC_JP},
+    {"CP51932",                        CP51932},
+    {"EUC-JP-MS",              EUCJP_MS},
+    {"EUCJP-MS",               EUCJP_MS},
+    {"EUCJPMS",                        EUCJP_MS},
+    {"EUC-JP-ASCII",           EUCJP_ASCII},
+    {"EUCJP-ASCII",            EUCJP_ASCII},
+    {"SHIFT_JISX0213",         SHIFT_JISX0213},
+    {"SHIFT_JIS-2004",         SHIFT_JIS_2004},
+    {"EUC-JISX0213",           EUC_JISX0213},
+    {"EUC-JIS-2004",           EUC_JIS_2004},
+    {"UTF-8",                  UTF_8},
+    {"UTF-8N",                 UTF_8N},
+    {"UTF-8-BOM",              UTF_8_BOM},
+    {"UTF8-MAC",               UTF8_MAC},
+    {"UTF-8-MAC",              UTF8_MAC},
+    {"UTF-16",                 UTF_16},
+    {"UTF-16BE",               UTF_16BE},
+    {"UTF-16BE-BOM",           UTF_16BE_BOM},
+    {"UTF-16LE",               UTF_16LE},
+    {"UTF-16LE-BOM",           UTF_16LE_BOM},
+    {"UTF-32",                 UTF_32},
+    {"UTF-32BE",               UTF_32BE},
+    {"UTF-32BE-BOM",           UTF_32BE_BOM},
+    {"UTF-32LE",               UTF_32LE},
+    {"UTF-32LE-BOM",           UTF_32LE_BOM},
+    {"BINARY",                 BINARY},
+    {NULL,                     -1}
+};
+#if defined(DEFAULT_CODE_JIS)
+#define            DEFAULT_ENCODING ISO_2022_JP
+#elif defined(DEFAULT_CODE_SJIS)
+#define            DEFAULT_ENCODING SHIFT_JIS
+#elif defined(DEFAULT_CODE_EUC)
+#define            DEFAULT_ENCODING EUC_JP
+#elif defined(DEFAULT_CODE_UTF8)
+#define            DEFAULT_ENCODING UTF_8
+#endif
+
 
 #define                is_alnum(c)  \
             (('a'<=c && c<='z')||('A'<= c && c<='Z')||('0'<=c && c<='9'))
@@ -269,16 +410,20 @@ void  djgpp_setbinmode(FILE *fp)
 #define nkf_isoctal(c)  ('0'<=c && c<='7')
 #define nkf_isdigit(c)  ('0'<=c && c<='9')
 #define nkf_isxdigit(c)  (nkf_isdigit(c) || ('a'<=c && c<='f') || ('A'<=c && c <= 'F'))
-#define nkf_isblank(c) (c == SPACE || c == TAB)
-#define nkf_isspace(c) (nkf_isblank(c) || c == CR || c == NL)
+#define nkf_isblank(c) (c == SP || c == TAB)
+#define nkf_isspace(c) (nkf_isblank(c) || c == CR || c == LF)
 #define nkf_isalpha(c) (('a' <= c && c <= 'z') || ('A' <= c && c <= 'Z'))
 #define nkf_isalnum(c) (nkf_isdigit(c) || nkf_isalpha(c))
-#define nkf_isprint(c) (' '<=c && c<='~')
+#define nkf_isprint(c) (SP<=c && c<='~')
 #define nkf_isgraph(c) ('!'<=c && c<='~')
 #define hex2bin(c) (('0'<=c&&c<='9') ? (c-'0') : \
                     ('A'<=c&&c<='F') ? (c-'A'+10) : \
-                    ('a'<=c&&c<='f') ? (c-'a'+10) : 0 )
+                    ('a'<=c&&c<='f') ? (c-'a'+10) : 0)
+#define bin2hex(c) ("0123456789ABCDEF"[c&15])
 #define is_eucg3(c2) (((unsigned short)c2 >> 8) == SS3)
+#define nkf_noescape_mime(c) ((c == CR) || (c == LF) || \
+    ((c > SP) && (c < DEL) && (c != '?') && (c != '=') && (c != '_') \
+     && (c != '(') && (c != ')') && (c != '.') && (c != 0x22)))
 
 #define CP932_TABLE_BEGIN 0xFA
 #define CP932_TABLE_END   0xFC
@@ -330,11 +475,9 @@ struct input_code{
     int _file_stat;
 };
 
-static char *input_codename = "";
+static char *input_codename = NULL; /* NULL: unestablished, "": BINARY */
+static nkf_encoding *output_encoding;
 
-#ifndef PERL_XS
-static const char *CopyRight = COPY_RIGHT;
-#endif
 #if !defined(PERL_XS) && !defined(WIN32DLL)
 static  nkf_char     noconvert(FILE *f);
 #endif
@@ -343,18 +486,18 @@ static  nkf_char     kanji_convert(FILE *f);
 static  nkf_char     h_conv(FILE *f,nkf_char c2,nkf_char c1);
 static  nkf_char     push_hold_buf(nkf_char c2);
 static  void    set_iconv(nkf_char f, nkf_char (*iconv_func)(nkf_char c2,nkf_char c1,nkf_char c0));
-static  nkf_char     s_iconv(nkf_char c2,nkf_char c1,nkf_char c0);
 static  nkf_char     s2e_conv(nkf_char c2, nkf_char c1, nkf_char *p2, nkf_char *p1);
-static  nkf_char     e_iconv(nkf_char c2,nkf_char c1,nkf_char c0);
 #if defined(UTF8_INPUT_ENABLE) || defined(UTF8_OUTPUT_ENABLE)
 /* UCS Mapping
  * 0: Shift_JIS, eucJP-ascii
  * 1: eucJP-ms
  * 2: CP932, CP51932
+ * 3: CP10001
  */
-#define UCS_MAP_ASCII 0
-#define UCS_MAP_MS    1
-#define UCS_MAP_CP932 2
+#define UCS_MAP_ASCII   0
+#define UCS_MAP_MS      1
+#define UCS_MAP_CP932   2
+#define UCS_MAP_CP10001 3
 static int ms_ucs_map_f = UCS_MAP_ASCII;
 #endif
 #ifdef UTF8_INPUT_ENABLE
@@ -372,9 +515,6 @@ static  void    encode_fallback_perl(nkf_char c);
 static  void    encode_fallback_subchar(nkf_char c);
 static  void    (*encode_fallback)(nkf_char c) = NULL;
 static  nkf_char     w2e_conv(nkf_char c2,nkf_char c1,nkf_char c0,nkf_char *p2,nkf_char *p1);
-static  nkf_char     w_iconv(nkf_char c2,nkf_char c1,nkf_char c0);
-static  nkf_char     w_iconv16(nkf_char c2,nkf_char c1,nkf_char c0);
-static  nkf_char     w_iconv32(nkf_char c2,nkf_char c1,nkf_char c0);
 static  nkf_char       unicode_to_jis_common(nkf_char c2,nkf_char c1,nkf_char c0,nkf_char *p2,nkf_char *p1);
 static  nkf_char       w_iconv_common(nkf_char c1,nkf_char c0,const unsigned short *const *pp,nkf_char psize,nkf_char *p2,nkf_char *p1);
 static  void    w16w_conv(nkf_char val, nkf_char *p2, nkf_char *p1, nkf_char *p0);
@@ -386,16 +526,10 @@ static  void    w_status(struct input_code *, nkf_char);
 static  int     output_bom_f = FALSE;
 static  int     output_endian = ENDIAN_BIG;
 static  nkf_char     e2w_conv(nkf_char c2,nkf_char c1);
-static  void    w_oconv(nkf_char c2,nkf_char c1);
-static  void    w_oconv16(nkf_char c2,nkf_char c1);
-static  void    w_oconv32(nkf_char c2,nkf_char c1);
 #endif
-static  void    e_oconv(nkf_char c2,nkf_char c1);
 static  nkf_char     e2s_conv(nkf_char c2, nkf_char c1, nkf_char *p2, nkf_char *p1);
-static  void    s_oconv(nkf_char c2,nkf_char c1);
-static  void    j_oconv(nkf_char c2,nkf_char c1);
 static  void    fold_conv(nkf_char c2,nkf_char c1);
-static  void    cr_conv(nkf_char c2,nkf_char c1);
+static  void    nl_conv(nkf_char c2,nkf_char c1);
 static  void    z_conv(nkf_char c2,nkf_char c1);
 static  void    rot_conv(nkf_char c2,nkf_char c1);
 static  void    hira_conv(nkf_char c2,nkf_char c1);
@@ -435,11 +569,10 @@ static  void    mimeout_addchar(nkf_char c);
 #ifndef PERL_XS
 static  void    usage(void);
 static  void    version(void);
+static  void    show_configuration(void);
 #endif
 static  void    options(unsigned char *c);
-#if defined(PERL_XS) || defined(WIN32DLL)
 static  void    reinit(void);
-#endif
 
 /* buffers */
 
@@ -453,7 +586,7 @@ static int             hold_count = 0;
 /* MIME preprocessor fifo */
 
 #define MIME_BUF_SIZE   (1024)    /* 2^n ring buffer */
-#define MIME_BUF_MASK   (MIME_BUF_SIZE-1)   
+#define MIME_BUF_MASK   (MIME_BUF_SIZE-1)
 #define Fifo(n)         mime_buf[(n)&MIME_BUF_MASK]
 static unsigned char           mime_buf[MIME_BUF_SIZE];
 static unsigned int            mime_top = 0;
@@ -470,18 +603,14 @@ static int             rot_f = FALSE;          /* rot14/43 mode */
 static int             hira_f = FALSE;          /* hira/kata henkan */
 static int             input_f = FALSE;        /* non fixed input code  */
 static int             alpha_f = FALSE;        /* convert JIx0208 alphbet to ASCII */
-static int             mime_f = STRICT_MIME;   /* convert MIME B base64 or Q */
+static int             mime_f = MIME_DECODE_DEFAULT;   /* convert MIME B base64 or Q */
 static int             mime_decode_f = FALSE;  /* mime decode is explicitly on */
 static int             mimebuf_f = FALSE;      /* MIME buffered input */
 static int             broken_f = FALSE;       /* convert ESC-less broken JIS */
 static int             iso8859_f = FALSE;      /* ISO8859 through */
 static int             mimeout_f = FALSE;       /* base64 mode */
-#if defined(MSDOS) || defined(__OS2__) 
-static int             x0201_f = TRUE;         /* Assume JISX0201 kana */
-#else
-static int             x0201_f = NO_X0201;     /* Assume NO JISX0201 */
-#endif
-static int             iso2022jp_f = FALSE;    /* convert ISO-2022-JP */
+static int             x0201_f = X0201_DEFAULT; /* convert JIS X 0201 */
+static int             iso2022jp_f = FALSE;    /* replace non ISO-2022-JP with GETA */
 
 #ifdef UNICODE_NORMALIZATION
 static int nfc_f = FALSE;
@@ -529,18 +658,16 @@ static nkf_char numchar_ungetc(nkf_char c,FILE *f);
 #ifdef CHECK_OPTION
 static int noout_f = FALSE;
 static void no_putc(nkf_char c);
-static nkf_char debug_f = FALSE;
+static int debug_f = FALSE;
 static void debug(const char *str);
 static nkf_char (*iconv_for_check)(nkf_char c2,nkf_char c1,nkf_char c0) = 0;
 #endif
 
-static int guess_f = FALSE;
+static int guess_f = 0; /* 0: OFF, 1: ON, 2: VERBOSE */
 #if !defined PERL_XS
 static  void    print_guessed_code(char *filename);
 #endif
 static  void    set_input_codename(char *codename);
-static int is_inputcode_mixed = FALSE;
-static int is_inputcode_set   = FALSE;
 
 #ifdef EXEC_IO
 static int exec_f = 0;
@@ -548,7 +675,7 @@ static int exec_f = 0;
 
 #ifdef SHIFTJIS_CP932
 /* invert IBM extended characters to others */
-static int cp51932_f = TRUE;
+static int cp51932_f = FALSE;
 
 /* invert NEC-selected IBM extended characters to IBM extended characters */
 static int cp932inv_f = TRUE;
@@ -587,7 +714,7 @@ struct input_code input_code_list[] = {
     {0}
 };
 
-static int              mimeout_mode = 0;
+static int              mimeout_mode = 0; /* 0, -1, 'Q', 'B', 1, 2 */
 static int              base64_count = 0;
 
 /* X0208 -> ASCII converter */
@@ -634,7 +761,7 @@ static nkf_char (*iconv)(nkf_char c2,nkf_char c1,nkf_char c0) = no_connection2;
 
 static void (*o_zconv)(nkf_char c2,nkf_char c1) = no_connection;
 static void (*o_fconv)(nkf_char c2,nkf_char c1) = no_connection;
-static void (*o_crconv)(nkf_char c2,nkf_char c1) = no_connection;
+static void (*o_nlconv)(nkf_char c2,nkf_char c1) = no_connection;
 static void (*o_rot_conv)(nkf_char c2,nkf_char c1) = no_connection;
 static void (*o_hira_conv)(nkf_char c2,nkf_char c1) = no_connection;
 static void (*o_base64conv)(nkf_char c2,nkf_char c1) = no_connection;
@@ -669,8 +796,7 @@ static int mime_decode_mode =   FALSE;    /* MIME mode B base64, Q hex */
 
 /* X0201 kana conversion table */
 /* 90-9F A0-DF */
-static const
-unsigned char cv[]= {
+static const unsigned char cv[]= {
     0x21,0x21,0x21,0x23,0x21,0x56,0x21,0x57,
     0x21,0x22,0x21,0x26,0x25,0x72,0x25,0x21,
     0x25,0x23,0x25,0x25,0x25,0x27,0x25,0x29,
@@ -692,8 +818,7 @@ unsigned char cv[]= {
 
 /* X0201 kana conversion table for daguten */
 /* 90-9F A0-DF */
-static const
-unsigned char dv[]= { 
+static const unsigned char dv[]= {
     0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
     0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
     0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
@@ -714,8 +839,7 @@ unsigned char dv[]= {
 
 /* X0201 kana conversion table for han-daguten */
 /* 90-9F A0-DF */
-static const
-unsigned char ev[]= { 
+static const unsigned char ev[]= {
     0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
     0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
     0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
@@ -737,8 +861,7 @@ unsigned char ev[]= {
 
 /* X0208 kigou conversion table */
 /* 0x8140 - 0x819e */
-static const
-unsigned char fv[] = {
+static const unsigned char fv[] = {
 
     0x00,0x00,0x00,0x00,0x2c,0x2e,0x00,0x3a,
     0x3b,0x3f,0x21,0x00,0x00,0x27,0x60,0x00,
@@ -755,7 +878,6 @@ unsigned char fv[] = {
 } ;
 
 
-#define    CRLF      1
 
 static int             file_out_f = FALSE;
 #ifdef OVERWRITE
@@ -766,7 +888,9 @@ static char            *backup_suffix = "";
 static char *get_backup_filename(const char *suffix, const char *filename);
 #endif
 
-static int             crmode_f = 0;   /* CR, NL, CRLF */
+static int nlmode_f = 0;   /* CR, LF, CRLF */
+static int input_newline = 0; /* 0: unestablished, EOF: MIXED */
+static nkf_char prev_cr = 0; /* CR or 0 */
 #ifdef EASYWIN /*Easy Win */
 static int             end_check;
 #endif /*Easy Win */
@@ -775,6 +899,57 @@ static int             end_check;
 nkf_char std_gc_buf[STD_GC_BUFSIZE];
 nkf_char std_gc_ndx;
 
+char* nkf_strcpy(const char *str)
+{
+    char* result = malloc(strlen(str) + 1);
+    if (!result){
+       perror(str);
+       return "";
+    }
+    strcpy(result, str);
+    return result;
+}
+
+static void nkf_str_upcase(const char *src, char *dest, size_t length)
+{
+    int i = 0;
+    for (; i < length && dest[i]; i++) {
+       dest[i] = nkf_toupper(src[i]);
+    }
+    dest[i] = 0;
+}
+
+static nkf_encoding *nkf_enc_from_index(int idx)
+{
+    if (idx < 0 || NKF_ENCODING_TABLE_SIZE <= idx) {
+       return 0;
+    }
+    return &nkf_encoding_table[idx];
+}
+
+static int nkf_enc_find_index(const char *name)
+{
+    int i, index = -1;
+    for (i = 0; encoding_name_to_id_table[i].id >= 0; i++) {
+       if (strcmp(name, encoding_name_to_id_table[i].name) == 0) {
+           return encoding_name_to_id_table[i].id;
+       }
+    }
+    return index;
+}
+
+static nkf_encoding *nkf_enc_find(const char *name)
+{
+    int idx = -1;
+    idx = nkf_enc_find_index(name);
+    if (idx < 0) return 0;
+    return nkf_enc_from_index(idx);
+}
+
+#define nkf_enc_name(enc) (enc)->name
+#define nkf_enc_to_index(enc) (enc)->id
+#define nkf_enc_to_base_encoding(enc) (enc)->based_encoding
+
 #ifdef WIN32DLL
 #include "nkf32dll.c"
 #elif defined(PERL_XS)
@@ -794,6 +969,32 @@ int main(int argc, char **argv)
     for (argc--,argv++; (argc > 0) && **argv == '-'; argc--, argv++) {
         cp = (unsigned char *)*argv;
         options(cp);
+        if (guess_f) {
+#ifdef CHECK_OPTION
+           int debug_f_back = debug_f;
+#endif
+#ifdef EXEC_IO
+           int exec_f_back = exec_f;
+#endif
+#ifdef X0212_ENABLE
+           int x0212_f_back = x0212_f;
+#endif
+           int x0213_f_back = x0213_f;
+           int guess_f_back = guess_f;
+           reinit();
+           guess_f = guess_f_back;
+           mime_f = FALSE;
+#ifdef CHECK_OPTION
+           debug_f = debug_f_back;
+#endif
+#ifdef EXEC_IO
+            exec_f = exec_f_back;
+#endif
+#ifdef X0212_ENABLE
+           x0212_f = x0212_f_back;
+#endif
+           x0213_f = x0213_f_back;
+       }
 #ifdef EXEC_IO
         if (exec_f){
             int fds[2], pid;
@@ -822,12 +1023,10 @@ int main(int argc, char **argv)
         }
 #endif
     }
-    if(x0201_f == WISH_TRUE)
-         x0201_f = ((!iso2022jp_f)? TRUE : NO_X0201);
 
     if (binmode_f == TRUE)
 #if defined(__OS2__) && (defined(__IBMC__) || defined(__IBMCPP__))
-    if (freopen("","wb",stdout) == NULL) 
+    if (freopen("","wb",stdout) == NULL)
         return (-1);
 #else
     setbinmode(stdout);
@@ -854,16 +1053,16 @@ int main(int argc, char **argv)
       }
     } else {
       int nfiles = argc;
+       int is_argument_error = FALSE;
       while (argc--) {
-           is_inputcode_mixed = FALSE;
-           is_inputcode_set   = FALSE;
-           input_codename = "";
+           input_codename = NULL;
+           input_newline = 0;
 #ifdef CHECK_OPTION
            iconv_for_check = 0;
 #endif
           if ((fin = fopen((origfname = *argv++), "r")) == NULL) {
-              perror(*--argv);
-               *argv++;
+               perror(*(argv-1));
+               is_argument_error = TRUE;
                continue;
           } else {
 #ifdef OVERWRITE
@@ -911,7 +1110,7 @@ int main(int argc, char **argv)
                       }
                   }else
 #endif
-                 if(argc == 1 ) {
+                 if(argc == 1) {
                      outfname = *argv++;
                      argc--;
                  } else {
@@ -924,7 +1123,7 @@ int main(int argc, char **argv)
                  }
                   if (binmode_f == TRUE) {
 #if defined(__OS2__) && (defined(__IBMC__) || defined(__IBMCPP__))
-                      if (freopen("","wb",stdout) == NULL) 
+                      if (freopen("","wb",stdout) == NULL)
                            return (-1);
 #else
                       setbinmode(stdout);
@@ -933,11 +1132,11 @@ int main(int argc, char **argv)
               }
               if (binmode_f == TRUE)
 #if defined(__OS2__) && (defined(__IBMC__) || defined(__IBMCPP__))
-                 if (freopen("","rb",fin) == NULL) 
+                 if (freopen("","rb",fin) == NULL)
                     return (-1);
 #else
                  setbinmode(fin);
-#endif 
+#endif
               setvbuffer(fin, (char *) stdibuf, IOBUF_SIZE);
               if (nop_f)
                   noconvert(fin);
@@ -1012,14 +1211,16 @@ int main(int argc, char **argv)
 #endif
           }
       }
+       if (is_argument_error)
+           return(-1);
     }
 #ifdef EASYWIN /*Easy Win */
-    if (file_out_f == FALSE) 
+    if (file_out_f == FALSE)
         scanf("%d",&end_check);
-    else 
+    else
         fclose(stdout);
 #else /* for Other OS */
-    if (file_out_f == TRUE) 
+    if (file_out_f == TRUE)
         fclose(stdout);
 #endif /*Easy Win */
     return (0);
@@ -1067,8 +1268,7 @@ char *get_backup_filename(const char *suffix, const char *filename)
 }
 #endif
 
-static const
-struct {
+static const struct {
     const char *name;
     const char *alias;
 } long_option[] = {
@@ -1093,7 +1293,8 @@ struct {
     {"hiragana","h1"},
     {"katakana","h2"},
     {"katakana-hiragana","h3"},
-    {"guess", "g"},
+    {"guess=", ""},
+    {"guess", "g1"},
     {"cp932", ""},
     {"no-cp932", ""},
 #ifdef X0212_ENABLE
@@ -1155,6 +1356,7 @@ void options(unsigned char *cp)
     unsigned char *p;
     unsigned char *cp_back = NULL;
     char codeset[32];
+    nkf_encoding *enc;
 
     if (option_mode==1)
        return;
@@ -1168,297 +1370,335 @@ void options(unsigned char *cp)
        p = 0;
         switch (*cp++) {
         case '-':  /* literal options */
-           if (!*cp || *cp == SPACE) {        /* ignore the rest of arguments */
+           if (!*cp || *cp == SP) {        /* ignore the rest of arguments */
                option_mode = 1;
                return;
            }
             for (i=0;i<sizeof(long_option)/sizeof(long_option[0]);i++) {
                 p = (unsigned char *)long_option[i].name;
                 for (j=0;*p && *p != '=' && *p == cp[j];p++, j++);
-               if (*p == cp[j] || cp[j] == ' '){
+               if (*p == cp[j] || cp[j] == SP){
                    p = &cp[j] + 1;
                    break;
                }
                p = 0;
             }
-           if (p == 0) return;
-           while(*cp && *cp != SPACE && cp++);
+           if (p == 0) {
+               fprintf(stderr, "unknown long option: --%s\n", cp);
+               return;
+           }
+           while(*cp && *cp != SP && cp++);
             if (long_option[i].alias[0]){
                cp_back = cp;
                cp = (unsigned char *)long_option[i].alias;
            }else{
                 if (strcmp(long_option[i].name, "ic=") == 0){
-                   for (i=0; i < 16 && SPACE < p[i] && p[i] < DEL; i++){
-                       codeset[i] = nkf_toupper(p[i]);
-                   }
-                   codeset[i] = 0;
-                   if(strcmp(codeset, "ISO-2022-JP") == 0 ||
-                     strcmp(codeset, "X-ISO2022JP-CP932") == 0 ||
-                     strcmp(codeset, "CP50220") == 0 ||
-                     strcmp(codeset, "CP50221") == 0 ||
-                     strcmp(codeset, "CP50222") == 0 ||
-                     strcmp(codeset, "ISO-2022-JP-MS") == 0){
+                   nkf_str_upcase((char *)p, codeset, 32);
+                   enc = nkf_enc_find(codeset);
+                   switch (nkf_enc_to_index(enc)) {
+                   case ISO_2022_JP:
+                       input_f = JIS_INPUT;
+                       break;
+                   case CP50220:
+                   case CP50221:
+                   case CP50222:
                        input_f = JIS_INPUT;
-                   }else if(strcmp(codeset, "ISO-2022-JP-1") == 0){
+#ifdef SHIFTJIS_CP932
+                       cp51932_f = TRUE;
+#endif
+#ifdef UTF8_OUTPUT_ENABLE
+                       ms_ucs_map_f = UCS_MAP_CP932;
+#endif
+                       break;
+                   case ISO_2022_JP_1:
                        input_f = JIS_INPUT;
 #ifdef X0212_ENABLE
                        x0212_f = TRUE;
 #endif
-                   }else if(strcmp(codeset, "ISO-2022-JP-3") == 0){
+                       break;
+                   case ISO_2022_JP_3:
                        input_f = JIS_INPUT;
 #ifdef X0212_ENABLE
                        x0212_f = TRUE;
 #endif
                        x0213_f = TRUE;
-                   }else if(strcmp(codeset, "SHIFT_JIS") == 0){
+                       break;
+                   case SHIFT_JIS:
                        input_f = SJIS_INPUT;
-                       if (x0201_f==NO_X0201) x0201_f=TRUE;
-                   }else if(strcmp(codeset, "WINDOWS-31J") == 0 ||
-                            strcmp(codeset, "CSWINDOWS31J") == 0 ||
-                            strcmp(codeset, "CP932") == 0 ||
-                            strcmp(codeset, "MS932") == 0){
+                       break;
+                   case WINDOWS_31J:
                        input_f = SJIS_INPUT;
-                       x0201_f = FALSE;
 #ifdef SHIFTJIS_CP932
                        cp51932_f = TRUE;
 #endif
 #ifdef UTF8_OUTPUT_ENABLE
                        ms_ucs_map_f = UCS_MAP_CP932;
 #endif
-                   }else if(strcmp(codeset, "EUCJP") == 0 ||
-                            strcmp(codeset, "EUC-JP") == 0){
+                       break;
+                   case CP10001:
+                       input_f = SJIS_INPUT;
+#ifdef SHIFTJIS_CP932
+                       cp51932_f = TRUE;
+#endif
+#ifdef UTF8_OUTPUT_ENABLE
+                       ms_ucs_map_f = UCS_MAP_CP10001;
+#endif
+                       break;
+                   case EUC_JP:
                        input_f = EUC_INPUT;
-                   }else if(strcmp(codeset, "CP51932") == 0){
+                       break;
+                   case CP51932:
                        input_f = EUC_INPUT;
-                       x0201_f = FALSE;
 #ifdef SHIFTJIS_CP932
                        cp51932_f = TRUE;
 #endif
 #ifdef UTF8_OUTPUT_ENABLE
                        ms_ucs_map_f = UCS_MAP_CP932;
 #endif
-                   }else if(strcmp(codeset, "EUC-JP-MS") == 0 ||
-                            strcmp(codeset, "EUCJP-MS") == 0 ||
-                            strcmp(codeset, "EUCJPMS") == 0){
+                       break;
+                   case EUCJP_MS:
                        input_f = EUC_INPUT;
-                       x0201_f = FALSE;
 #ifdef SHIFTJIS_CP932
                        cp51932_f = FALSE;
 #endif
 #ifdef UTF8_OUTPUT_ENABLE
                        ms_ucs_map_f = UCS_MAP_MS;
 #endif
-                   }else if(strcmp(codeset, "EUC-JP-ASCII") == 0 ||
-                            strcmp(codeset, "EUCJP-ASCII") == 0){
+                       break;
+                   case EUCJP_ASCII:
                        input_f = EUC_INPUT;
-                       x0201_f = FALSE;
 #ifdef SHIFTJIS_CP932
                        cp51932_f = FALSE;
 #endif
 #ifdef UTF8_OUTPUT_ENABLE
                        ms_ucs_map_f = UCS_MAP_ASCII;
 #endif
-                   }else if(strcmp(codeset, "SHIFT_JISX0213") == 0 ||
-                            strcmp(codeset, "SHIFT_JIS-2004") == 0){
+                       break;
+                   case SHIFT_JISX0213:
+                   case SHIFT_JIS_2004:
                        input_f = SJIS_INPUT;
                        x0213_f = TRUE;
 #ifdef SHIFTJIS_CP932
                        cp51932_f = FALSE;
-                       cp932inv_f = FALSE;
 #endif
-                       if (x0201_f==NO_X0201) x0201_f=TRUE;
-                   }else if(strcmp(codeset, "EUC-JISX0213") == 0 ||
-                            strcmp(codeset, "EUC-JIS-2004") == 0){
+                       break;
+                   case EUC_JISX0213:
+                   case EUC_JIS_2004:
                        input_f = EUC_INPUT;
-                       x0201_f = FALSE;
                        x0213_f = TRUE;
 #ifdef SHIFTJIS_CP932
                        cp51932_f = FALSE;
-                       cp932inv_f = FALSE;
 #endif
+                       break;
 #ifdef UTF8_INPUT_ENABLE
-                   }else if(strcmp(codeset, "UTF-8") == 0 ||
-                            strcmp(codeset, "UTF-8N") == 0 ||
-                            strcmp(codeset, "UTF-8-BOM") == 0){
+                   case UTF_8:
+                   case UTF_8N:
+                   case UTF_8_BOM:
                        input_f = UTF8_INPUT;
+                       break;
 #ifdef UNICODE_NORMALIZATION
-                   }else if(strcmp(codeset, "UTF8-MAC") == 0 ||
-                            strcmp(codeset, "UTF-8-MAC") == 0){
+                   case UTF8_MAC:
                        input_f = UTF8_INPUT;
                        nfc_f = TRUE;
+                       break;
 #endif
-                   }else if(strcmp(codeset, "UTF-16") == 0 ||
-                            strcmp(codeset, "UTF-16BE") == 0 ||
-                            strcmp(codeset, "UTF-16BE-BOM") == 0){
+                   case UTF_16:
+                   case UTF_16BE:
+                   case UTF_16BE_BOM:
                        input_f = UTF16_INPUT;
                        input_endian = ENDIAN_BIG;
-                   }else if(strcmp(codeset, "UTF-16LE") == 0 ||
-                            strcmp(codeset, "UTF-16LE-BOM") == 0){
+                       break;
+                   case UTF_16LE:
+                   case UTF_16LE_BOM:
                        input_f = UTF16_INPUT;
                        input_endian = ENDIAN_LITTLE;
-                   }else if(strcmp(codeset, "UTF-32") == 0 ||
-                            strcmp(codeset, "UTF-32BE") == 0 ||
-                            strcmp(codeset, "UTF-32BE-BOM") == 0){
+                       break;
+                   case UTF_32:
+                   case UTF_32BE:
+                   case UTF_32BE_BOM:
                        input_f = UTF32_INPUT;
                        input_endian = ENDIAN_BIG;
-                   }else if(strcmp(codeset, "UTF-32LE") == 0 ||
-                            strcmp(codeset, "UTF-32LE-BOM") == 0){
+                       break;
+                   case UTF_32LE:
+                   case UTF_32LE_BOM:
                        input_f = UTF32_INPUT;
                        input_endian = ENDIAN_LITTLE;
+                       break;
 #endif
+                   default:
+                       fprintf(stderr, "unknown input encoding: %s\n", codeset);
+                       break;
                    }
                     continue;
                }
                 if (strcmp(long_option[i].name, "oc=") == 0){
-                   for (i=0; i < 16 && SPACE < p[i] && p[i] < DEL; i++){
-                       codeset[i] = nkf_toupper(p[i]);
-                   }
-                   codeset[i] = 0;
-                   if(strcmp(codeset, "ISO-2022-JP") == 0 ||
-                      strcmp(codeset, "CP50220") == 0){
+                   x0201_f = FALSE;
+                   nkf_str_upcase((char *)p, codeset, 32);
+                   output_encoding = nkf_enc_find(codeset);
+                   switch (nkf_enc_to_index(output_encoding)) {
+                   case ISO_2022_JP:
                        output_conv = j_oconv;
-                   }else if(strcmp(codeset, "X-ISO2022JP-CP932") == 0){
-                       output_conv = j_oconv;
-                       no_cp932ext_f = TRUE;
-                   }else if(strcmp(codeset, "CP50221") == 0 ||
-                            strcmp(codeset, "ISO-2022-JP-MS") == 0){
-                       output_conv = j_oconv;
-                       x0201_f = FALSE;
-                   }else if(strcmp(codeset, "ISO-2022-JP-1") == 0){
-                       output_conv = j_oconv;
-#ifdef X0212_ENABLE
-                       x0212_f = TRUE;
+                       break;
+                   case CP50220:
+                           output_conv = j_oconv;
+                           x0201_f = TRUE;
+#ifdef SHIFTJIS_CP932
+                           cp932inv_f = FALSE;
 #endif
+#ifdef UTF8_OUTPUT_ENABLE
+                           ms_ucs_map_f = UCS_MAP_CP932;
+#endif
+                       break;
+                   case CP50221:
+                       output_conv = j_oconv;
 #ifdef SHIFTJIS_CP932
-                       cp51932_f = FALSE;
+                       cp932inv_f = FALSE;
 #endif
-                   }else if(strcmp(codeset, "ISO-2022-JP-3") == 0){
+#ifdef UTF8_OUTPUT_ENABLE
+                       ms_ucs_map_f = UCS_MAP_CP932;
+#endif
+                       break;
+                   case ISO_2022_JP_1:
                        output_conv = j_oconv;
 #ifdef X0212_ENABLE
                        x0212_f = TRUE;
 #endif
-                       x0213_f = TRUE;
 #ifdef SHIFTJIS_CP932
-                       cp51932_f = FALSE;
+                       cp932inv_f = FALSE;
 #endif
-                   }else if(strcmp(codeset, "ISO-2022-JP-MS") == 0){
+                       break;
+                   case ISO_2022_JP_3:
                        output_conv = j_oconv;
-                       x0201_f = FALSE;
 #ifdef X0212_ENABLE
                        x0212_f = TRUE;
 #endif
+                       x0213_f = TRUE;
 #ifdef SHIFTJIS_CP932
-                       cp51932_f = FALSE;
+                       cp932inv_f = FALSE;
 #endif
-                   }else if(strcmp(codeset, "SHIFT_JIS") == 0){
+                       break;
+                   case SHIFT_JIS:
                        output_conv = s_oconv;
-                   }else if(strcmp(codeset, "WINDOWS-31J") == 0 ||
-                            strcmp(codeset, "CSWINDOWS31J") == 0 ||
-                            strcmp(codeset, "CP932") == 0 ||
-                            strcmp(codeset, "MS932") == 0){
+                       break;
+                   case WINDOWS_31J:
                        output_conv = s_oconv;
-                       x0201_f = FALSE;
-#ifdef SHIFTJIS_CP932
-                       cp51932_f = TRUE;
-                       cp932inv_f = TRUE;
-#endif
 #ifdef UTF8_OUTPUT_ENABLE
                        ms_ucs_map_f = UCS_MAP_CP932;
 #endif
-                   }else if(strcmp(codeset, "EUCJP") == 0 ||
-                            strcmp(codeset, "EUC-JP") == 0){
+                       break;
+                   case CP10001:
+                       output_conv = s_oconv;
+#ifdef UTF8_OUTPUT_ENABLE
+                       ms_ucs_map_f = UCS_MAP_CP10001;
+#endif
+                       break;
+                   case EUC_JP:
                        output_conv = e_oconv;
-                   }else if(strcmp(codeset, "CP51932") == 0){
+                       break;
+                   case CP51932:
                        output_conv = e_oconv;
-                       x0201_f = FALSE;
 #ifdef SHIFTJIS_CP932
-                       cp51932_f = TRUE;
+                       cp932inv_f = FALSE;
 #endif
 #ifdef UTF8_OUTPUT_ENABLE
                        ms_ucs_map_f = UCS_MAP_CP932;
 #endif
-                   }else if(strcmp(codeset, "EUC-JP-MS") == 0 ||
-                            strcmp(codeset, "EUCJP-MS") == 0 ||
-                            strcmp(codeset, "EUCJPMS") == 0){
+                       break;
+                   case EUCJP_MS:
                        output_conv = e_oconv;
-                       x0201_f = FALSE;
 #ifdef X0212_ENABLE
                        x0212_f = TRUE;
 #endif
-#ifdef SHIFTJIS_CP932
-                       cp51932_f = FALSE;
-#endif
 #ifdef UTF8_OUTPUT_ENABLE
                        ms_ucs_map_f = UCS_MAP_MS;
 #endif
-                   }else if(strcmp(codeset, "EUC-JP-ASCII") == 0 ||
-                            strcmp(codeset, "EUCJP-ASCII") == 0){
+                       break;
+                   case EUCJP_ASCII:
                        output_conv = e_oconv;
-                       x0201_f = FALSE;
 #ifdef X0212_ENABLE
                        x0212_f = TRUE;
 #endif
-#ifdef SHIFTJIS_CP932
-                       cp51932_f = FALSE;
-#endif
 #ifdef UTF8_OUTPUT_ENABLE
                        ms_ucs_map_f = UCS_MAP_ASCII;
 #endif
-                   }else if(strcmp(codeset, "SHIFT_JISX0213") == 0 ||
-                            strcmp(codeset, "SHIFT_JIS-2004") == 0){
-                       output_conv = s_oconv;
-                       x0213_f = TRUE;
+                       break;
+                   case SHIFT_JISX0213:
+                   case SHIFT_JIS_2004:
+                           output_conv = s_oconv;
+                           x0213_f = TRUE;
 #ifdef SHIFTJIS_CP932
-                       cp932inv_f = FALSE;
+                           cp932inv_f = FALSE;
 #endif
-                   }else if(strcmp(codeset, "EUC-JISX0213") == 0 ||
-                            strcmp(codeset, "EUC-JIS-2004") == 0){
+                       break;
+                   case EUC_JISX0213:
+                   case EUC_JIS_2004:
                        output_conv = e_oconv;
 #ifdef X0212_ENABLE
                        x0212_f = TRUE;
 #endif
                        x0213_f = TRUE;
 #ifdef SHIFTJIS_CP932
-                       cp51932_f = FALSE;
+                       cp932inv_f = FALSE;
 #endif
+                       break;
 #ifdef UTF8_OUTPUT_ENABLE
-                   }else if(strcmp(codeset, "UTF-8") == 0){
+                   case UTF_8:
+                   case UTF_8N:
                        output_conv = w_oconv;
-                   }else if(strcmp(codeset, "UTF-8N") == 0){
-                       output_conv = w_oconv;
-                   }else if(strcmp(codeset, "UTF-8-BOM") == 0){
+                       break;
+                   case UTF_8_BOM:
                        output_conv = w_oconv;
                        output_bom_f = TRUE;
-                   }else if(strcmp(codeset, "UTF-16BE") == 0){
+                       break;
+                   case UTF_16BE:
                        output_conv = w_oconv16;
-                   }else if(strcmp(codeset, "UTF-16") == 0 ||
-                            strcmp(codeset, "UTF-16BE-BOM") == 0){
+                       break;
+                   case UTF_16:
+                   case UTF_16BE_BOM:
                        output_conv = w_oconv16;
                        output_bom_f = TRUE;
-                   }else if(strcmp(codeset, "UTF-16LE") == 0){
+                       break;
+                   case UTF_16LE:
                        output_conv = w_oconv16;
                        output_endian = ENDIAN_LITTLE;
-                   }else if(strcmp(codeset, "UTF-16LE-BOM") == 0){
+                       break;
+                   case UTF_16LE_BOM:
                        output_conv = w_oconv16;
                        output_endian = ENDIAN_LITTLE;
                        output_bom_f = TRUE;
-                   }else if(strcmp(codeset, "UTF-32") == 0 ||
-                            strcmp(codeset, "UTF-32BE") == 0){
+                       break;
+                   case UTF_32:
+                   case UTF_32BE:
                        output_conv = w_oconv32;
-                   }else if(strcmp(codeset, "UTF-32BE-BOM") == 0){
+                       break;
+                   case UTF_32BE_BOM:
                        output_conv = w_oconv32;
                        output_bom_f = TRUE;
-                   }else if(strcmp(codeset, "UTF-32LE") == 0){
+                       break;
+                   case UTF_32LE:
                        output_conv = w_oconv32;
                        output_endian = ENDIAN_LITTLE;
-                   }else if(strcmp(codeset, "UTF-32LE-BOM") == 0){
+                       break;
+                   case UTF_32LE_BOM:
                        output_conv = w_oconv32;
                        output_endian = ENDIAN_LITTLE;
                        output_bom_f = TRUE;
+                       break;
 #endif
+                   default:
+                       fprintf(stderr, "unknown output encoding: %s\n", codeset);
+                       break;
                    }
                     continue;
                }
+                if (strcmp(long_option[i].name, "guess=") == 0){
+                   if (p[0] == '1') {
+                       guess_f = 2;
+                   } else {
+                       guess_f = 1;
+                   }
+                    continue;
+                }
 #ifdef OVERWRITE
                 if (strcmp(long_option[i].name, "overwrite") == 0){
                     file_out_f = TRUE;
@@ -1578,7 +1818,7 @@ void options(unsigned char *cp)
                    encode_fallback = encode_fallback_html;
                     continue;
                 }
-                if (strcmp(long_option[i].name, "fb-xml" ) == 0){
+                if (strcmp(long_option[i].name, "fb-xml") == 0){
                    encode_fallback = encode_fallback_xml;
                     continue;
                 }
@@ -1654,7 +1894,7 @@ void options(unsigned char *cp)
             if (*cp=='1') {
                /* alias of -t */
                nop_f = TRUE;
-               *cp++;
+               *cp += 1;
            } else if (*cp=='2') {
                /*
                 * -t with put/get
@@ -1663,53 +1903,57 @@ void options(unsigned char *cp)
                 *
                 */
                nop_f = 2;
-               *cp++;
+               *cp += 1;
             } else
                nop_f = TRUE;
             continue;
         case 'j':           /* JIS output */
         case 'n':
             output_conv = j_oconv;
+            output_encoding = nkf_enc_from_index(ISO_2022_JP);
             continue;
         case 'e':           /* AT&T EUC output */
             output_conv = e_oconv;
+            cp932inv_f = FALSE;
+            output_encoding = nkf_enc_from_index(EUC_JP);
             continue;
         case 's':           /* SJIS output */
             output_conv = s_oconv;
+            output_encoding = nkf_enc_from_index(SHIFT_JIS);
             continue;
         case 'l':           /* ISO8859 Latin-1 support, no conversion */
             iso8859_f = TRUE;  /* Only compatible with ISO-2022-JP */
             input_f = LATIN1_INPUT;
             continue;
         case 'i':           /* Kanji IN ESC-$-@/B */
-            if (*cp=='@'||*cp=='B') 
+            if (*cp=='@'||*cp=='B')
                 kanji_intro = *cp++;
             continue;
         case 'o':           /* ASCII IN ESC-(-J/B */
-            if (*cp=='J'||*cp=='B'||*cp=='H') 
+            if (*cp=='J'||*cp=='B'||*cp=='H')
                 ascii_intro = *cp++;
             continue;
         case 'h':
-            /*  
+            /*
                 bit:1   katakana->hiragana
                 bit:2   hiragana->katakana
             */
-            if ('9'>= *cp && *cp>='0') 
+            if ('9'>= *cp && *cp>='0')
                 hira_f |= (*cp++ -'0');
-            else 
+            else
                 hira_f |= 1;
             continue;
         case 'r':
             rot_f = TRUE;
             continue;
-#if defined(MSDOS) || defined(__OS2__) 
+#if defined(MSDOS) || defined(__OS2__)
         case 'T':
             binmode_f = FALSE;
             continue;
 #endif
 #ifndef PERL_XS
         case 'V':
-            version();
+            show_configuration();
             exit(1);
             break;
         case 'v':
@@ -1723,16 +1967,22 @@ void options(unsigned char *cp)
                output_conv = w_oconv; cp++;
                if (cp[0] == '0'){
                    cp++;
+                   output_encoding = nkf_enc_from_index(UTF_8N);
                } else {
                    output_bom_f = TRUE;
+                   output_encoding = nkf_enc_from_index(UTF_8_BOM);
                }
            } else {
+               int enc_idx;
                if ('1'== cp[0] && '6'==cp[1]) {
                    output_conv = w_oconv16; cp+=2;
+                   enc_idx = UTF_16;
                } else if ('3'== cp[0] && '2'==cp[1]) {
                    output_conv = w_oconv32; cp+=2;
+                   enc_idx = UTF_32;
                } else {
                    output_conv = w_oconv;
+                   output_encoding = nkf_enc_from_index(UTF_8);
                    continue;
                }
                if (cp[0]=='L') {
@@ -1741,13 +1991,21 @@ void options(unsigned char *cp)
                } else if (cp[0] == 'B') {
                    cp++;
                 } else {
+                   output_encoding = nkf_enc_from_index(enc_idx);
                    continue;
                 }
                if (cp[0] == '0'){
                    cp++;
+                   enc_idx = enc_idx == UTF_16
+                       ? (output_endian == ENDIAN_LITTLE ? UTF_16LE : UTF_16BE)
+                       : (output_endian == ENDIAN_LITTLE ? UTF_32LE : UTF_32BE);
                } else {
                    output_bom_f = TRUE;
+                   enc_idx = enc_idx == UTF_16
+                       ? (output_endian == ENDIAN_LITTLE ? UTF_16LE_BOM : UTF_16BE_BOM)
+                       : (output_endian == ENDIAN_LITTLE ? UTF_32LE_BOM : UTF_32BE_BOM);
                }
+               output_encoding = nkf_enc_from_index(enc_idx);
            }
             continue;
 #endif
@@ -1787,18 +2045,19 @@ void options(unsigned char *cp)
             continue;
         case 'S':   /* MS Kanji input */
             input_f = SJIS_INPUT;
-            if (x0201_f==NO_X0201) x0201_f=TRUE;
             continue;
         case 'Z':   /* Convert X0208 alphabet to asii */
-            /*  bit:0   Convert X0208
-                bit:1   Convert Kankaku to one space
-                bit:2   Convert Kankaku to two spaces
-                bit:3   Convert HTML Entity
+            /* alpha_f
+              bit:0   Convert JIS X 0208 Alphabet to ASCII
+              bit:1   Convert Kankaku to one space
+              bit:2   Convert Kankaku to two spaces
+              bit:3   Convert HTML Entity
+              bit:4   Convert JIS X 0208 Katakana to JIS X 0201 Katakana
             */
-            if ('9'>= *cp && *cp>='0') 
-                alpha_f |= 1<<(*cp++ -'0');
-            else 
-                alpha_f |= TRUE;
+           while ('0'<= *cp && *cp <='9') {
+               alpha_f |= 1 << (*cp++ - '0');
+           }
+            if (!alpha_f) alpha_f = 1;
             continue;
         case 'x':   /* Convert X0201 kana to X0208 or X0201 Conversion */
             x0201_f = FALSE;    /* No X0201->X0208 conversion */
@@ -1806,15 +2065,14 @@ void options(unsigned char *cp)
                     ESC-(-I     in JIS, EUC, MS Kanji
                     SI/SO       in JIS, EUC, MS Kanji
                     SSO         in EUC, JIS, not in MS Kanji
-                    MS Kanji (0xa0-0xdf) 
+                    MS Kanji (0xa0-0xdf)
                output  X0201
                     ESC-(-I     in JIS (0x20-0x5f)
                     SSO         in EUC (0xa0-0xdf)
-                    0xa0-0xd    in MS Kanji (0xa0-0xdf) 
+                    0xa0-0xd    in MS Kanji (0xa0-0xdf)
             */
             continue;
-        case 'X':   /* Assume X0201 kana */
-            /* Default value is NO_X0201 for EUC/MS-Kanji mix */
+        case 'X':   /* Convert X0201 kana to X0208 */
             x0201_f = TRUE;
             continue;
         case 'F':   /* prserve new lines */
@@ -1826,7 +2084,7 @@ void options(unsigned char *cp)
                fold_len *= 10;
                fold_len += *cp++ - '0';
            }
-            if (!(0<fold_len && fold_len<BUFSIZ)) 
+            if (!(0<fold_len && fold_len<BUFSIZ))
                 fold_len = DEFAULT_FOLD;
            if (*cp=='-') {
                fold_margin = 0;
@@ -1867,9 +2125,9 @@ void options(unsigned char *cp)
                 bit:1   allow any x on ESC-(-x or ESC-$-x
                 bit:2   reset to ascii on NL
             */
-            if ('9'>= *cp && *cp>='0') 
+            if ('9'>= *cp && *cp>='0')
                 broken_f |= 1<<(*cp++ -'0');
-            else 
+            else
                 broken_f |= TRUE;
             continue;
 #ifndef PERL_XS
@@ -1878,35 +2136,44 @@ void options(unsigned char *cp)
             continue;
 #endif
         case 'c':/* add cr code */
-            crmode_f = CRLF;
+            nlmode_f = CRLF;
             continue;
         case 'd':/* delete cr code */
-            crmode_f = NL;
-            continue;
-        case 'I':   /* ISO-2022-JP output */
-            iso2022jp_f = TRUE;
+            nlmode_f = LF;
             continue;
+       case 'I':   /* ISO-2022-JP output */
+           iso2022jp_f = TRUE;
+           continue;
         case 'L':  /* line mode */
             if (*cp=='u') {         /* unix */
-                crmode_f = NL; cp++;
+                nlmode_f = LF; cp++;
             } else if (*cp=='m') { /* mac */
-                crmode_f = CR; cp++;
+                nlmode_f = CR; cp++;
             } else if (*cp=='w') { /* windows */
-                crmode_f = CRLF; cp++;
+                nlmode_f = CRLF; cp++;
             } else if (*cp=='0') { /* no conversion  */
-                crmode_f = 0; cp++;
+                nlmode_f = 0; cp++;
             }
             continue;
-        case 'g':
 #ifndef PERL_XS
-            guess_f = TRUE;
-#endif
+        case 'g':
+            if (*cp == '1') {
+                guess_f = 2;
+                cp++;
+            } else if (*cp == '0') {
+               guess_f = 1;
+                cp++;
+            } else {
+               guess_f = 1;
+            }
             continue;
-        case ' ':    
+#endif
+        case SP:
         /* module muliple options in a string are allowed for Perl moudle  */
            while(*cp && *cp++!='-');
             continue;
         default:
+           fprintf(stderr, "unknown option: -%c\n", *(cp-1));
             /* bogus option but ignored */
             continue;
         }
@@ -1948,7 +2215,7 @@ void set_iconv(nkf_char f, nkf_char (*iconv_func)(nkf_char c2,nkf_char c1,nkf_ch
         struct input_code *p = find_inputcode_byfunc(iconv);
         if (p){
             set_input_codename(p->name);
-            debug(input_codename);
+            debug(p->name);
         }
         iconv_for_check = iconv;
     }
@@ -1958,29 +2225,26 @@ void set_iconv(nkf_char f, nkf_char (*iconv_func)(nkf_char c2,nkf_char c1,nkf_ch
 #define SCORE_L2       (1)                   /* \e$BBh\e(B2\e$B?e=`4A;z\e(B */
 #define SCORE_KANA     (SCORE_L2 << 1)       /* \e$B$$$o$f$kH>3Q%+%J\e(B */
 #define SCORE_DEPEND   (SCORE_KANA << 1)     /* \e$B5!<o0MB8J8;z\e(B */
-#ifdef SHIFTJIS_CP932
-#define SCORE_CP932    (SCORE_DEPEND << 1)   /* CP932 \e$B$K$h$kFI$_49$(\e(B */
-#define SCORE_NO_EXIST (SCORE_CP932 << 1)    /* \e$BB8:_$7$J$$J8;z\e(B */
-#else
-#define SCORE_NO_EXIST (SCORE_DEPEND << 1)   /* \e$BB8:_$7$J$$J8;z\e(B */
-#endif
+#define SCORE_CP932    (SCORE_DEPEND << 1)   /* CP932 \e$B$K$h$kFI$_49$(\e(B (IBM extended characters) */
+#define SCORE_X0212    (SCORE_CP932 << 1)    /* JIS X 0212 */
+#define SCORE_NO_EXIST (SCORE_X0212 << 1)    /* \e$BB8:_$7$J$$J8;z\e(B */
 #define SCORE_iMIME    (SCORE_NO_EXIST << 1) /* MIME \e$B$K$h$k;XDj\e(B */
 #define SCORE_ERROR    (SCORE_iMIME << 1) /* \e$B%(%i!<\e(B */
 
 #define SCORE_INIT (SCORE_iMIME)
 
-const nkf_char score_table_A0[] = {
+static const char score_table_A0[] = {
     0, 0, 0, 0,
     0, 0, 0, 0,
     0, SCORE_DEPEND, SCORE_DEPEND, SCORE_DEPEND,
     SCORE_DEPEND, SCORE_DEPEND, SCORE_DEPEND, SCORE_NO_EXIST,
 };
 
-const nkf_char score_table_F0[] = {
+static const char score_table_F0[] = {
     SCORE_L2, SCORE_L2, SCORE_L2, SCORE_L2,
     SCORE_L2, SCORE_DEPEND, SCORE_NO_EXIST, SCORE_NO_EXIST,
-    SCORE_DEPEND, SCORE_DEPEND, SCORE_DEPEND, SCORE_DEPEND,
-    SCORE_DEPEND, SCORE_NO_EXIST, SCORE_NO_EXIST, SCORE_ERROR,
+    SCORE_DEPEND, SCORE_DEPEND, SCORE_CP932, SCORE_CP932,
+    SCORE_CP932, SCORE_NO_EXIST, SCORE_NO_EXIST, SCORE_ERROR,
 };
 
 void set_code_score(struct input_code *ptr, nkf_char score)
@@ -2007,6 +2271,8 @@ void code_score(struct input_code *ptr)
         set_code_score(ptr, SCORE_ERROR);
     }else if (c2 == SSO){
         set_code_score(ptr, SCORE_KANA);
+    }else if (c2 == 0x8f){
+        set_code_score(ptr, SCORE_X0212);
 #ifdef UTF8_OUTPUT_ENABLE
     }else if (!e2w_conv(c2, c1)){
         set_code_score(ptr, SCORE_NO_EXIST);
@@ -2076,17 +2342,19 @@ void s_status(struct input_code *ptr, nkf_char c)
               status_push_ch(ptr, c);
               code_score(ptr);
               status_clear(ptr);
-          }else if ((0x81 <= c && c < 0xa0) || (0xe0 <= c && c <= 0xef)){
+          }else if ((0x81 <= c && c < 0xa0) || (0xe0 <= c && c <= 0xea)){
               ptr->stat = 1;
               status_push_ch(ptr, c);
+          }else if (0xed <= c && c <= 0xee){
+              ptr->stat = 3;
+              status_push_ch(ptr, c);
 #ifdef SHIFTJIS_CP932
-          }else if (cp51932_f
-                    && is_ibmext_in_sjis(c)){
+          }else if (is_ibmext_in_sjis(c)){
               ptr->stat = 2;
               status_push_ch(ptr, c);
 #endif /* SHIFTJIS_CP932 */
 #ifdef X0212_ENABLE
-          }else if (x0212_f && 0xf0 <= c && c <= 0xfc){
+          }else if (0xf0 <= c && c <= 0xfc){
               ptr->stat = 1;
               status_push_ch(ptr, c);
 #endif /* X0212_ENABLE */
@@ -2106,18 +2374,26 @@ void s_status(struct input_code *ptr, nkf_char c)
           break;
       case 2:
 #ifdef SHIFTJIS_CP932
+       if ((0x40 <= c && c <= 0x7e) || (0x80 <= c && c <= 0xfc)) {
+           status_push_ch(ptr, c);
+           if (s2e_conv(ptr->buf[0], ptr->buf[1], &ptr->buf[0], &ptr->buf[1]) == 0) {
+               set_code_score(ptr, SCORE_CP932);
+               status_clear(ptr);
+               break;
+           }
+       }
+#endif /* SHIFTJIS_CP932 */
+       status_disable(ptr);
+          break;
+      case 3:
           if ((0x40 <= c && c <= 0x7e) || (0x80 <= c && c <= 0xfc)){
               status_push_ch(ptr, c);
-              if (s2e_conv(ptr->buf[0], ptr->buf[1], &ptr->buf[0], &ptr->buf[1]) == 0){
-                  set_code_score(ptr, SCORE_CP932);
-                  status_clear(ptr);
-                  break;
-              }
+              s2e_conv(ptr->buf[0], ptr->buf[1], &ptr->buf[0], &ptr->buf[1]);
+           set_code_score(ptr, SCORE_CP932);
+           status_clear(ptr);
+          }else{
+              status_disable(ptr);
           }
-#endif /* SHIFTJIS_CP932 */
-#ifndef X0212_ENABLE
-          status_disable(ptr);
-#endif
           break;
     }
 }
@@ -2309,7 +2585,7 @@ nkf_char noconvert(FILE *f)
 
 void module_connection(void)
 {
-    oconv = output_conv; 
+    oconv = output_conv;
     o_putc = std_putc;
 
     /* replace continucation module, from output side */
@@ -2329,8 +2605,8 @@ void module_connection(void)
        /* base64_count = 0; */
     }
 
-    if (crmode_f) {
-       o_crconv = oconv; oconv = cr_conv;
+    if (nlmode_f || guess_f) {
+       o_nlconv = oconv; oconv = nl_conv;
     }
     if (rot_f) {
        o_rot_conv = oconv; oconv = rot_conv;
@@ -2518,7 +2794,7 @@ void check_bom(FILE *f)
 }
 
 /*
-   Conversion main loop. Code detection only. 
+   Conversion main loop. Code detection only.
  */
 
 nkf_char kanji_convert(FILE *f)
@@ -2552,28 +2828,30 @@ nkf_char kanji_convert(FILE *f)
            code_status(c1);
         if (c2) {
             /* second byte */
-            if (c2 > DEL) {
+            if (c2 > ((input_f == JIS_INPUT && ms_ucs_map_f) ? 0x92 : DEL)) {
                 /* in case of 8th bit is on */
                 if (!estab_f&&!mime_decode_mode) {
                     /* in case of not established yet */
                     /* It is still ambiguious */
-                    if (h_conv(f, c2, c1)==EOF) 
+                    if (h_conv(f, c2, c1)==EOF)
                         LAST;
-                    else 
+                    else
                         c2 = 0;
                     NEXT;
-                } else
-                    /* in case of already established */
-                    if (c1 < AT) {
-                        /* ignore bogus code */
-                        c2 = 0;
-                        NEXT;
-                    } else
-                        SEND;
+                } else {
+                   /* in case of already established */
+                   if (c1 < AT) {
+                       /* ignore bogus code and not CP5022x UCD */
+                       c2 = 0;
+                       NEXT;
+                   } else {
+                       SEND;
+                   }
+               }
             } else
                 /* second byte, 7 bit code */
                 /* it might be kanji shitfted */
-                if ((c1 == DEL) || (c1 <= SPACE)) {
+                if ((c1 == DEL) || (c1 <= SP)) {
                     /* ignore bogus first code */
                     c2 = 0;
                     NEXT;
@@ -2639,7 +2917,7 @@ nkf_char kanji_convert(FILE *f)
                 SEND;
            } else
 #endif
-           if (c1 > DEL) {
+           if (c1 > ((input_f == JIS_INPUT && ms_ucs_map_f) ? 0x92 : DEL)) {
                 /* 8 bit code */
                 if (!estab_f && !iso8859_f) {
                     /* not established yet */
@@ -2647,16 +2925,16 @@ nkf_char kanji_convert(FILE *f)
                     NEXT;
                 } else { /* estab_f==TRUE */
                     if (iso8859_f) {
-                        c2 = ISO8859_1;
+                        c2 = ISO_8859_1;
                         c1 &= 0x7f;
                         SEND;
                     } else if (SSP<=c1 && c1<0xe0 && iconv == s_iconv) {
                         /* SJIS X0201 Case... */
-                        if(iso2022jp_f && x0201_f==NO_X0201) {
+                        if (iso2022jp_f && !x0201_f) {
                             (*oconv)(GETA1, GETA2);
                             NEXT;
                         } else {
-                           c2 = X0201;
+                           c2 = JIS_X_0201;
                            c1 &= 0x7f;
                            SEND;
                        }
@@ -2665,49 +2943,55 @@ nkf_char kanji_convert(FILE *f)
                         c1 = (*i_getc)(f);  /* skip SSO */
                         code_status(c1);
                         if (SSP<=c1 && c1<0xe0) {
-                           if(iso2022jp_f &&  x0201_f==NO_X0201) {
+                           if (iso2022jp_f && !x0201_f) {
                                (*oconv)(GETA1, GETA2);
                                NEXT;
                            } else {
-                               c2 = X0201;
+                               c2 = JIS_X_0201;
                                c1 &= 0x7f;
                                SEND;
                            }
                         } else  { /* bogus code, skip SSO and one byte */
                             NEXT;
                         }
+                   } else if (ms_ucs_map_f == UCS_MAP_CP10001 &&
+                              (c1 == 0xFD || c1 == 0xFE)) {
+                       /* CP10001 */
+                       c2 = JIS_X_0201;
+                       c1 &= 0x7f;
+                       SEND;
                     } else {
                        /* already established */
                        c2 = c1;
                        NEXT;
                     }
                 }
-            } else if ((c1 > SPACE) && (c1 != DEL)) {
+            } else if ((c1 > SP) && (c1 != DEL)) {
                 /* in case of Roman characters */
-                if (shift_mode) { 
+                if (shift_mode) {
                     /* output 1 shifted byte */
                     if (iso8859_f) {
-                        c2 = ISO8859_1;
+                        c2 = ISO_8859_1;
                         SEND;
-                    } else if (SPACE<=c1 && c1<(0xe0&0x7f) ){
+                    } else if (SP <= c1 && c1 < (0xe0&0x7f)){
                       /* output 1 shifted byte */
-                       if(iso2022jp_f && x0201_f==NO_X0201) {
+                       if (iso2022jp_f && !x0201_f) {
                            (*oconv)(GETA1, GETA2);
                            NEXT;
                        } else {
-                           c2 = X0201;
+                           c2 = JIS_X_0201;
                            SEND;
                        }
                     } else {
                         /* look like bogus code */
                         NEXT;
                     }
-                } else if (input_mode == X0208 || input_mode == X0212 ||
-                          input_mode == X0213_1 || input_mode == X0213_2) {
+                } else if (input_mode == JIS_X_0208 || input_mode == JIS_X_0212 ||
+                          input_mode == JIS_X_0213_1 || input_mode == JIS_X_0213_2) {
                     /* in case of Kanji shifted */
                     c2 = c1;
                     NEXT;
-                } else if (c1 == '=' && mime_f && !mime_decode_mode ) {
+                } else if (c1 == '=' && mime_f && !mime_decode_mode) {
                     /* Check MIME code */
                     if ((c1 = (*i_getc)(f)) == EOF) {
                         (*oconv)(0, '=');
@@ -2716,11 +3000,11 @@ nkf_char kanji_convert(FILE *f)
                         /* =? is mime conversion start sequence */
                        if(mime_f == STRICT_MIME) {
                            /* check in real detail */
-                           if (mime_begin_strict(f) == EOF) 
+                           if (mime_begin_strict(f) == EOF)
                                LAST;
                            else
                                NEXT;
-                       } else if (mime_begin(f) == EOF) 
+                       } else if (mime_begin(f) == EOF)
                             LAST;
                         else
                             NEXT;
@@ -2730,37 +3014,37 @@ nkf_char kanji_convert(FILE *f)
                         NEXT;
                     }
                 } else {
-                    /* normal ASCII code */ 
+                    /* normal ASCII code */
                     SEND;
                 }
-            } else if (!is_8bit && c1 == SI) {
-                shift_mode = FALSE; 
+            } else if (c1 == SI && (!is_8bit || mime_decode_mode)) {
+                shift_mode = FALSE;
                 NEXT;
-            } else if (!is_8bit && c1 == SO) {
-                shift_mode = TRUE; 
+            } else if (c1 == SO && (!is_8bit || mime_decode_mode)) {
+                shift_mode = TRUE;
                 NEXT;
-            } else if (!is_8bit && c1 == ESC ) {
+            } else if (c1 == ESC && (!is_8bit || mime_decode_mode)) {
                 if ((c1 = (*i_getc)(f)) == EOF) {
                     /*  (*oconv)(0, ESC); don't send bogus code */
                     LAST;
                 } else if (c1 == '$') {
                     if ((c1 = (*i_getc)(f)) == EOF) {
                         /*
-                        (*oconv)(0, ESC); don't send bogus code 
+                        (*oconv)(0, ESC); don't send bogus code
                         (*oconv)(0, '$'); */
                         LAST;
                     } else if (c1 == '@'|| c1 == 'B') {
                         /* This is kanji introduction */
-                        input_mode = X0208;
+                        input_mode = JIS_X_0208;
                         shift_mode = FALSE;
                         set_input_codename("ISO-2022-JP");
 #ifdef CHECK_OPTION
-                        debug(input_codename);
+                        debug("ISO-2022-JP");
 #endif
                         NEXT;
                     } else if (c1 == '(') {
                         if ((c1 = (*i_getc)(f)) == EOF) {
-                            /* don't send bogus code 
+                            /* don't send bogus code
                             (*oconv)(0, ESC);
                             (*oconv)(0, '$');
                             (*oconv)(0, '(');
@@ -2768,21 +3052,21 @@ nkf_char kanji_convert(FILE *f)
                             LAST;
                         } else if (c1 == '@'|| c1 == 'B') {
                             /* This is kanji introduction */
-                            input_mode = X0208;
+                            input_mode = JIS_X_0208;
                             shift_mode = FALSE;
                             NEXT;
 #ifdef X0212_ENABLE
                         } else if (c1 == 'D'){
-                            input_mode = X0212;
+                            input_mode = JIS_X_0212;
                             shift_mode = FALSE;
                             NEXT;
 #endif /* X0212_ENABLE */
-                        } else if (c1 == (X0213_1&0x7F)){
-                            input_mode = X0213_1;
+                        } else if (c1 == 0x4F){
+                            input_mode = JIS_X_0213_1;
                             shift_mode = FALSE;
                             NEXT;
-                        } else if (c1 == (X0213_2&0x7F)){
-                            input_mode = X0213_2;
+                        } else if (c1 == 0x50){
+                            input_mode = JIS_X_0213_2;
                             shift_mode = FALSE;
                             NEXT;
                         } else {
@@ -2795,7 +3079,7 @@ nkf_char kanji_convert(FILE *f)
                         }
                     } else if (broken_f&0x2) {
                         /* accept any ESC-(-x as broken code ... */
-                        input_mode = X0208;
+                        input_mode = JIS_X_0208;
                         shift_mode = FALSE;
                         NEXT;
                     } else {
@@ -2806,14 +3090,14 @@ nkf_char kanji_convert(FILE *f)
                     }
                 } else if (c1 == '(') {
                     if ((c1 = (*i_getc)(f)) == EOF) {
-                        /* don't send bogus code 
+                        /* don't send bogus code
                         (*oconv)(0, ESC);
                         (*oconv)(0, '('); */
                         LAST;
                     } else {
                         if (c1 == 'I') {
                             /* This is X0201 kana introduction */
-                            input_mode = X0201; shift_mode = X0201;
+                            input_mode = JIS_X_0201; shift_mode = JIS_X_0201;
                             NEXT;
                         } else if (c1 == 'B' || c1 == 'J' || c1 == 'H') {
                             /* This is X0208 kanji introduction */
@@ -2829,12 +3113,12 @@ nkf_char kanji_convert(FILE *f)
                             SEND;
                         }
                     }
-               } else if ( c1 == 'N' || c1 == 'n' ){
+               } else if ( c1 == 'N' || c1 == 'n'){
                    /* SS2 */
                    c3 = (*i_getc)(f);  /* skip SS2 */
-                   if ( (SPACE<=c3 && c3 < 0x60) || (0xa0<=c3 && c3 < 0xe0)){
+                   if ( (SP<=c3 && c3 < 0x60) || (0xa0<=c3 && c3 < 0xe0)){
                        c1 = c3;
-                       c2 = X0201;
+                       c2 = JIS_X_0201;
                        SEND;
                    }else{
                        (*i_ungetc)(c3, f);
@@ -2847,36 +3131,82 @@ nkf_char kanji_convert(FILE *f)
                     (*oconv)(0, ESC);
                     SEND;
                 }
-            } else if ((c1 == NL || c1 == CR) && broken_f&4) {
-                input_mode = ASCII; set_iconv(FALSE, 0);
-                SEND;
-           } else if (c1 == NL && mime_decode_f && !mime_decode_mode ) {
-               if ((c1=(*i_getc)(f))!=EOF && c1 == SPACE) {
-                   i_ungetc(SPACE,f);
-                   continue;
-               } else {
-                   i_ungetc(c1,f);
-               }
-               c1 = NL;
-               SEND;
-           } else if (c1 == CR && mime_decode_f && !mime_decode_mode ) {
-               if ((c1=(*i_getc)(f))!=EOF) {
-                   if (c1==SPACE) {
-                       i_ungetc(SPACE,f);
-                       continue;
-                   } else if (c1 == NL && (c1=(*i_getc)(f))!=EOF && c1 == SPACE) {
-                       i_ungetc(SPACE,f);
-                       continue;
+           } else if (c1 == ESC && iconv == s_iconv) {
+               /* ESC in Shift_JIS */
+               if ((c1 = (*i_getc)(f)) == EOF) {
+                   /*  (*oconv)(0, ESC); don't send bogus code */
+                   LAST;
+               } else if (c1 == '$') {
+                   /* J-PHONE emoji */
+                   if ((c1 = (*i_getc)(f)) == EOF) {
+                       /*
+                          (*oconv)(0, ESC); don't send bogus code
+                          (*oconv)(0, '$'); */
+                       LAST;
                    } else {
-                       i_ungetc(c1,f);
+                       if (('E' <= c1 && c1 <= 'G') ||
+                           ('O' <= c1 && c1 <= 'Q')) {
+                           /*
+                              NUM : 0 1 2 3 4 5
+                              BYTE: G E F O P Q
+                              C%7 : 1 6 0 2 3 4
+                              C%7 : 0 1 2 3 4 5 6
+                              NUM : 2 0 3 4 5 X 1
+                            */
+                           static const char jphone_emoji_first_table[7] = {2, 0, 3, 4, 5, 0, 1};
+                           c0 = (jphone_emoji_first_table[c1 % 7] << 8) - SP + 0xE000 + CLASS_UNICODE;
+                           while ((c1 = (*i_getc)(f)) != EOF) {
+                               if (SP <= c1 && c1 <= 'z') {
+                                   (*oconv)(0, c1 + c0);
+                               } else break; /* c1 == SO */
+                           }
+                       }
                    }
-                   i_ungetc(NL,f);
+                   if (c1 == EOF) LAST;
+                   NEXT;
                } else {
-                   i_ungetc(c1,f);
+                   /* lonely ESC  */
+                   (*oconv)(0, ESC);
+                   SEND;
                }
-               c1 = CR;
-               SEND;
-           } else 
+           } else if (c1 == LF || c1 == CR) {
+               if (broken_f&4) {
+                   input_mode = ASCII; set_iconv(FALSE, 0);
+                   SEND;
+               } else if (mime_decode_f && !mime_decode_mode){
+                   if (c1 == LF) {
+                       if ((c1=(*i_getc)(f))!=EOF && c1 == SP) {
+                           i_ungetc(SP,f);
+                           continue;
+                       } else {
+                           i_ungetc(c1,f);
+                       }
+                       c1 = LF;
+                       SEND;
+                   } else  { /* if (c1 == CR)*/
+                       if ((c1=(*i_getc)(f))!=EOF) {
+                           if (c1==SP) {
+                               i_ungetc(SP,f);
+                               continue;
+                           } else if (c1 == LF && (c1=(*i_getc)(f))!=EOF && c1 == SP) {
+                               i_ungetc(SP,f);
+                               continue;
+                           } else {
+                               i_ungetc(c1,f);
+                           }
+                           i_ungetc(LF,f);
+                       } else {
+                           i_ungetc(c1,f);
+                       }
+                       c1 = CR;
+                       SEND;
+                   }
+               }
+           } else if (c1 == DEL && input_mode == JIS_X_0208) {
+               /* CP5022x */
+               c2 = c1;
+               NEXT;
+           } else
                 SEND;
         }
         /* send: */
@@ -2903,16 +3233,24 @@ nkf_char kanji_convert(FILE *f)
                break;
            }
            break;
-       case X0208:
-       case X0213_1:
+       case JIS_X_0208:
+       case JIS_X_0213_1:
+           if (ms_ucs_map_f &&
+               0x7F <= c2 && c2 <= 0x92 &&
+               0x21 <= c1 && c1 <= 0x7E) {
+               /* CP932 UDC */
+               if(c1 == 0x7F) return 0;
+               c1 = (c2 - 0x7F) * 94 + c1 - 0x21 + 0xE000 + CLASS_UNICODE;
+               c2 = 0;
+           }
            (*oconv)(c2, c1); /* this is JIS, not SJIS/EUC case */
            break;
 #ifdef X0212_ENABLE
-       case X0212:
+       case JIS_X_0212:
            (*oconv)(PREFIX_EUCG3 | c2, c1);
            break;
 #endif /* X0212_ENABLE */
-       case X0213_2:
+       case JIS_X_0213_2:
            (*oconv)(PREFIX_EUCG3 | c2, c1);
            break;
        default:
@@ -2927,7 +3265,7 @@ nkf_char kanji_convert(FILE *f)
 
     /* epilogue */
     (*iconv)(EOF, 0, 0);
-    if (!is_inputcode_set)
+    if (!input_codename)
     {
        if (is_8bit) {
            struct input_code *p = input_code_list;
@@ -2937,6 +3275,9 @@ nkf_char kanji_convert(FILE *f)
                ++p;
            }
            set_input_codename(result->name);
+#ifdef CHECK_OPTION
+           debug(result->name);
+#endif
        }
     }
     return 1;
@@ -2975,12 +3316,12 @@ h_conv(FILE *f, nkf_char c2, nkf_char c1)
             code_status(c1);
         }
         while (p->name){
-            if (p->score < result->score){
+            if (p->status_func && p->score < result->score){
                 result = p;
             }
             ++p;
         }
-        set_iconv(FALSE, result->iconv_func);
+        set_iconv(TRUE, result->iconv_func);
     }
 
 
@@ -3005,7 +3346,7 @@ h_conv(FILE *f, nkf_char c2, nkf_char c1)
             (*iconv)(0, c2, 0);
             continue;
         }else if (iconv == s_iconv && 0xa1 <= c2 && c2 <= 0xdf){
-            (*iconv)(X0201, c2, 0);
+            (*iconv)(JIS_X_0201, c2, 0);
             continue;
         }
         if (hold_index < hold_count){
@@ -3072,28 +3413,30 @@ nkf_char s2e_conv(nkf_char c2, nkf_char c1, nkf_char *p2, nkf_char *p1)
 #if defined(SHIFTJIS_CP932) || defined(X0212_ENABLE)
     nkf_char val;
 #endif
-    static const nkf_char shift_jisx0213_s1a3_table[5][2] ={ { 1, 8}, { 3, 4}, { 5,12}, {13,14}, {15, 0} };
+    static const char shift_jisx0213_s1a3_table[5][2] ={ { 1, 8}, { 3, 4}, { 5,12}, {13,14}, {15, 0} };
 #ifdef SHIFTJIS_CP932
-    if (cp51932_f && is_ibmext_in_sjis(c2)){
-#if 0
-        extern const unsigned short shiftjis_cp932[3][189];
-#endif
+    if (!cp932inv_f && is_ibmext_in_sjis(c2)){
         val = shiftjis_cp932[c2 - CP932_TABLE_BEGIN][c1 - 0x40];
         if (val){
             c2 = val >> 8;
             c1 = val & 0xff;
         }
     }
+    if (cp932inv_f
+        && CP932INV_TABLE_BEGIN <= c2 && c2 <= CP932INV_TABLE_END){
+        nkf_char c = cp932inv[c2 - CP932INV_TABLE_BEGIN][c1 - 0x40];
+        if (c){
+            c2 = c >> 8;
+            c1 = c & 0xff;
+        }
+    }
 #endif /* SHIFTJIS_CP932 */
 #ifdef X0212_ENABLE
     if (!x0213_f && is_ibmext_in_sjis(c2)){
-#if 0
-        extern const unsigned short shiftjis_x0212[3][189];
-#endif
         val = shiftjis_x0212[c2 - 0xfa][c1 - 0x40];
         if (val){
             if (val > 0x7FFF){
-                c2 = PREFIX_EUCG3 | (val >> 8);
+                c2 = PREFIX_EUCG3 | ((val >> 8) & 0x7f);
                 c1 = val & 0xff;
             }else{
                 c2 = val >> 8;
@@ -3118,7 +3461,7 @@ nkf_char s2e_conv(nkf_char c2, nkf_char c1, nkf_char *p2, nkf_char *p1)
            if (0x9E < c1) c2++;
        }
        if (c1 < 0x9F)
-           c1 = c1 - ((c1 > DEL) ? SPACE : 0x1F);
+           c1 = c1 - ((c1 > DEL) ? SP : 0x1F);
        else {
            c1 = c1 - 0x7E;
        }
@@ -3134,10 +3477,15 @@ nkf_char s2e_conv(nkf_char c2, nkf_char c1, nkf_char *p2, nkf_char *p1)
 
 nkf_char s_iconv(nkf_char c2, nkf_char c1, nkf_char c0)
 {
-    if (c2 == X0201) {
+    if (c2 == JIS_X_0201) {
        c1 &= 0x7f;
-    } else if ((c2 == EOF) || (c2 == 0) || c2 < SPACE) {
+    } else if ((c2 == EOF) || (c2 == 0) || c2 < SP) {
         /* NOP */
+    } else if (!x0213_f && 0xF0 <= c2 && c2 <= 0xF9 && 0x40 <= c1 && c1 <= 0xFC) {
+       /* CP932 UDC */
+       if(c1 == 0x7F) return 0;
+       c1 = (c2 - 0xF0) * 188 + (c1 - 0x40 - (0x7E < c1)) + 0xE000 + CLASS_UNICODE;
+       c2 = 0;
     } else {
         nkf_char ret = s2e_conv(c2, c1, &c2, &c1);
         if (ret) return ret;
@@ -3148,36 +3496,60 @@ nkf_char s_iconv(nkf_char c2, nkf_char c1, nkf_char c0)
 
 nkf_char e_iconv(nkf_char c2, nkf_char c1, nkf_char c0)
 {
-    if (c2 == X0201) {
+    if (c2 == JIS_X_0201) {
        c1 &= 0x7f;
 #ifdef X0212_ENABLE
     }else if (c2 == 0x8f){
         if (c0 == 0){
             return -1;
         }
-        c2 = (c2 << 8) | (c1 & 0x7f);
-        c1 = c0 & 0x7f;
+       if (!cp51932_f && !x0213_f && 0xF5 <= c1 && c1 <= 0xFE && 0xA1 <= c0 && c0 <= 0xFE) {
+           /* encoding is eucJP-ms, so invert to Unicode Private User Area */
+           c1 = (c1 - 0xF5) * 94 + c0 - 0xA1 + 0xE3AC + CLASS_UNICODE;
+           c2 = 0;
+       } else {
+           c2 = (c2 << 8) | (c1 & 0x7f);
+           c1 = c0 & 0x7f;
 #ifdef SHIFTJIS_CP932
-        if (cp51932_f){
-            nkf_char s2, s1;
-            if (e2s_conv(c2, c1, &s2, &s1) == 0){
-                s2e_conv(s2, s1, &c2, &c1);
-                if (c2 < 0x100){
-                    c1 &= 0x7f;
-                    c2 &= 0x7f;
-                }
-            }
-        }
+           if (cp51932_f){
+               nkf_char s2, s1;
+               if (e2s_conv(c2, c1, &s2, &s1) == 0){
+                   s2e_conv(s2, s1, &c2, &c1);
+                   if (c2 < 0x100){
+                       c1 &= 0x7f;
+                       c2 &= 0x7f;
+                   }
+               }
+           }
 #endif /* SHIFTJIS_CP932 */
+        }
 #endif /* X0212_ENABLE */
     } else if (c2 == SSO){
-        c2 = X0201;
+        c2 = JIS_X_0201;
         c1 &= 0x7f;
-    } else if ((c2 == EOF) || (c2 == 0) || c2 < SPACE) {
+    } else if ((c2 == EOF) || (c2 == 0) || c2 < SP) {
         /* NOP */
     } else {
-        c1 &= 0x7f;
-        c2 &= 0x7f;
+       if (!cp51932_f && ms_ucs_map_f && 0xF5 <= c2 && c2 <= 0xFE && 0xA1 <= c1 && c1 <= 0xFE) {
+           /* encoding is eucJP-ms, so invert to Unicode Private User Area */
+           c1 = (c2 - 0xF5) * 94 + c1 - 0xA1 + 0xE000 + CLASS_UNICODE;
+           c2 = 0;
+       } else {
+           c1 &= 0x7f;
+           c2 &= 0x7f;
+#ifdef SHIFTJIS_CP932
+           if (cp51932_f && 0x79 <= c2 && c2 <= 0x7c){
+               nkf_char s2, s1;
+               if (e2s_conv(c2, c1, &s2, &s1) == 0){
+                   s2e_conv(s2, s1, &c2, &c1);
+                   if (c2 < 0x100){
+                       c1 &= 0x7f;
+                       c2 &= 0x7f;
+                   }
+               }
+           }
+#endif /* SHIFTJIS_CP932 */
+        }
     }
     (*oconv)(c2, c1);
     return 0;
@@ -3207,13 +3579,13 @@ nkf_char w2e_conv(nkf_char c2, nkf_char c1, nkf_char c0, nkf_char *p2, nkf_char
 nkf_char w_iconv(nkf_char c2, nkf_char c1, nkf_char c0)
 {
     nkf_char ret = 0;
-    static const int w_iconv_utf8_1st_byte[] =
+    static const char w_iconv_utf8_1st_byte[] =
     { /* 0xC0 - 0xFF */
        20, 20, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
        21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
        30, 31, 31, 31, 31, 31, 31, 31, 31, 31, 31, 31, 31, 32, 33, 33,
        40, 41, 41, 41, 42, 43, 43, 43, 50, 50, 50, 50, 60, 60, 70, 70};
-    
+
     if (c2 < 0 || 0xff < c2) {
     }else if (c2 == 0) { /* 0 : 1 byte*/
        c0 = 0;
@@ -3391,32 +3763,24 @@ nkf_char w_iconv32(nkf_char c2, nkf_char c1, nkf_char c0)
 
 nkf_char unicode_to_jis_common(nkf_char c2, nkf_char c1, nkf_char c0, nkf_char *p2, nkf_char *p1)
 {
-#if 0
-    extern const unsigned short *const utf8_to_euc_2bytes[];
-    extern const unsigned short *const utf8_to_euc_2bytes_ms[];
-    extern const unsigned short *const utf8_to_euc_2bytes_932[];
-    extern const unsigned short *const *const utf8_to_euc_3bytes[];
-    extern const unsigned short *const *const utf8_to_euc_3bytes_ms[];
-    extern const unsigned short *const *const utf8_to_euc_3bytes_932[];
-#endif
     const unsigned short *const *pp;
     const unsigned short *const *const *ppp;
-    static const int no_best_fit_chars_table_C2[] =
+    static const char no_best_fit_chars_table_C2[] =
     {1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
        1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
        1, 1, 0, 0, 1, 0, 0, 0, 0, 1, 1, 1, 2, 1, 1, 2,
        0, 0, 1, 1, 0, 1, 0, 1, 2, 1, 1, 1, 1, 1, 1, 1};
-    static const int no_best_fit_chars_table_C2_ms[] =
+    static const char no_best_fit_chars_table_C2_ms[] =
     {1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
        1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
        1, 0, 1, 1, 0, 1, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0,
        0, 0, 1, 1, 0, 1, 0, 1, 0, 1, 0, 1, 1, 1, 1, 0};
-    static const int no_best_fit_chars_table_932_C2[] =
+    static const char no_best_fit_chars_table_932_C2[] =
     {1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
        1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
        1, 1, 1, 1, 0, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1,
        0, 0, 1, 1, 0, 1, 0, 1, 1, 1, 1, 1, 0, 0, 0, 0};
-    static const int no_best_fit_chars_table_932_C3[] =
+    static const char no_best_fit_chars_table_932_C3[] =
     {1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
        1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1,
        1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
@@ -3437,7 +3801,7 @@ nkf_char unicode_to_jis_common(nkf_char c2, nkf_char c1, nkf_char c0, nkf_char *
                    if(no_best_fit_chars_table_932_C3[c1&0x3F]) return 1;
                    break;
                }
-           }else if(cp51932_f){
+           }else if(!cp932inv_f){
                switch(c2){
                case 0xC2:
                    if(no_best_fit_chars_table_C2[c1&0x3F]) return 1;
@@ -3448,11 +3812,27 @@ nkf_char unicode_to_jis_common(nkf_char c2, nkf_char c1, nkf_char c0, nkf_char *
                }
            }else if(ms_ucs_map_f == UCS_MAP_MS){
                if(c2 == 0xC2 && no_best_fit_chars_table_C2_ms[c1&0x3F]) return 1;
+           }else if(ms_ucs_map_f == UCS_MAP_CP10001){
+               switch(c2){
+               case 0xC2:
+                   switch(c1){
+                   case 0xA2:
+                   case 0xA3:
+                   case 0xA5:
+                   case 0xA6:
+                   case 0xAC:
+                   case 0xAF:
+                   case 0xB8:
+                       return 1;
+                   }
+                   break;
+               }
            }
        }
        pp =
            ms_ucs_map_f == UCS_MAP_CP932 ? utf8_to_euc_2bytes_932 :
            ms_ucs_map_f == UCS_MAP_MS ? utf8_to_euc_2bytes_ms :
+           ms_ucs_map_f == UCS_MAP_CP10001 ? utf8_to_euc_2bytes_mac :
            utf8_to_euc_2bytes;
        ret =  w_iconv_common(c2, c1, pp, sizeof_utf8_to_euc_2bytes, p2, p1);
     }else if(c0 < 0xF0){
@@ -3475,6 +3855,19 @@ nkf_char unicode_to_jis_common(nkf_char c2, nkf_char c1, nkf_char c0, nkf_char *
                    if(c1 == 0x80 || c0 == 0x9C) return 1;
                    break;
                }
+           }else if(ms_ucs_map_f == UCS_MAP_CP10001){
+               switch(c2){
+               case 0xE3:
+                   switch(c1){
+                   case 0x82:
+                           if(c0 == 0x94) return 1;
+                       break;
+                   case 0x83:
+                           if(c0 == 0xBB) return 1;
+                       break;
+                   }
+                   break;
+               }
            }else{
                switch(c2){
                case 0xE2:
@@ -3493,7 +3886,7 @@ nkf_char unicode_to_jis_common(nkf_char c2, nkf_char c1, nkf_char c0, nkf_char *
                        if(c0 == 0x8D) return 1;
                        break;
                    case 0xBD:
-                       if(c0 == 0x9E && cp51932_f) return 1;
+                       if(c0 == 0x9E && !cp932inv_f) return 1;
                        break;
                    case 0xBF:
                        if(0xA0 <= c0 && c0 <= 0xA5) return 1;
@@ -3506,11 +3899,12 @@ nkf_char unicode_to_jis_common(nkf_char c2, nkf_char c1, nkf_char c0, nkf_char *
        ppp =
            ms_ucs_map_f == UCS_MAP_CP932 ? utf8_to_euc_3bytes_932 :
            ms_ucs_map_f == UCS_MAP_MS ? utf8_to_euc_3bytes_ms :
+           ms_ucs_map_f == UCS_MAP_CP10001 ? utf8_to_euc_3bytes_mac :
            utf8_to_euc_3bytes;
        ret = w_iconv_common(c1, c0, ppp[c2 - 0xE0], sizeof_utf8_to_euc_C2, p2, p1);
     }else return -1;
 #ifdef SHIFTJIS_CP932
-    if (!ret && cp51932_f && is_eucg3(*p2)) {
+    if (!ret && !cp932inv_f && is_eucg3(*p2)) {
        nkf_char s2, s1;
        if (e2s_conv(*p2, *p1, &s2, &s1) == 0) {
            s2e_conv(s2, s1, p2, p1);
@@ -3549,7 +3943,7 @@ nkf_char w_iconv_common(nkf_char c1, nkf_char c0, const unsigned short *const *p
         c2 &= 0x7f;
         c2 |= PREFIX_EUCG3;
     }
-    if (c2 == SO) c2 = X0201;
+    if (c2 == SO) c2 = JIS_X_0201;
     c1 = val & 0x7f;
     if (p2) *p2 = c2;
     if (p1) *p1 = c1;
@@ -3558,13 +3952,12 @@ nkf_char w_iconv_common(nkf_char c1, nkf_char c0, const unsigned short *const *p
 
 void nkf_each_char_to_hex(void (*f)(nkf_char c2,nkf_char c1), nkf_char c)
 {
-    const char *hex = "0123456789ABCDEF";
     int shift = 20;
     c &= VALUE_MASK;
     while(shift >= 0){
        if(c >= 1<<shift){
            while(shift >= 0){
-               (*f)(0, hex[(c>>shift)&0xF]);
+               (*f)(0, bin2hex(c>>shift));
                shift -= 4;
            }
        }else{
@@ -3609,22 +4002,21 @@ void encode_fallback_xml(nkf_char c)
 
 void encode_fallback_java(nkf_char c)
 {
-    const char *hex = "0123456789ABCDEF";
     (*oconv)(0, '\\');
     c &= VALUE_MASK;
     if(!is_unicode_bmp(c)){
        (*oconv)(0, 'U');
        (*oconv)(0, '0');
        (*oconv)(0, '0');
-       (*oconv)(0, hex[(c>>20)&0xF]);
-       (*oconv)(0, hex[(c>>16)&0xF]);
+       (*oconv)(0, bin2hex(c>>20));
+       (*oconv)(0, bin2hex(c>>16));
     }else{
        (*oconv)(0, 'u');
     }
-    (*oconv)(0, hex[(c>>12)&0xF]);
-    (*oconv)(0, hex[(c>> 8)&0xF]);
-    (*oconv)(0, hex[(c>> 4)&0xF]);
-    (*oconv)(0, hex[ c     &0xF]);
+    (*oconv)(0, bin2hex(c>>12));
+    (*oconv)(0, bin2hex(c>> 8));
+    (*oconv)(0, bin2hex(c>> 4));
+    (*oconv)(0, bin2hex(c    ));
     return;
 }
 
@@ -3649,15 +4041,17 @@ void encode_fallback_subchar(nkf_char c)
 #ifdef UTF8_OUTPUT_ENABLE
 nkf_char e2w_conv(nkf_char c2, nkf_char c1)
 {
-#if 0
-    extern const unsigned short euc_to_utf8_1byte[];
-    extern const unsigned short *const euc_to_utf8_2bytes[];
-    extern const unsigned short *const euc_to_utf8_2bytes_ms[];
-    extern const unsigned short *const x0212_to_utf8_2bytes[];
-#endif
     const unsigned short *p;
 
-    if (c2 == X0201) {
+    if (c2 == JIS_X_0201) {
+       if (ms_ucs_map_f == UCS_MAP_CP10001) {
+           switch (c1) {
+           case 0x20:
+               return 0xA0;
+           case 0x7D:
+               return 0xA9;
+           }
+       }
         p = euc_to_utf8_1byte;
 #ifdef X0212_ENABLE
     } else if (is_eucg3(c2)){
@@ -3674,7 +4068,10 @@ nkf_char e2w_conv(nkf_char c2, nkf_char c1)
         c2 &= 0x7f;
         c2 = (c2&0x7f) - 0x21;
         if (0<=c2 && c2<sizeof_euc_to_utf8_2bytes)
-            p = ms_ucs_map_f != UCS_MAP_ASCII ? euc_to_utf8_2bytes_ms[c2] : euc_to_utf8_2bytes[c2];
+            p =
+               ms_ucs_map_f == UCS_MAP_ASCII ? euc_to_utf8_2bytes[c2] :
+               ms_ucs_map_f == UCS_MAP_CP10001 ? euc_to_utf8_2bytes_mac[c2] :
+               euc_to_utf8_2bytes_ms[c2];
        else
            return 0;
     }
@@ -3724,14 +4121,14 @@ void w_oconv(nkf_char c2, nkf_char c1)
     }
 #endif
 
-    if (c2 == 0) { 
+    if (c2 == 0) {
        output_mode = ASCII;
         (*o_putc)(c1);
-    } else if (c2 == ISO8859_1) {
-       output_mode = ISO8859_1;
+    } else if (c2 == ISO_8859_1) {
+       output_mode = UTF_8;
         (*o_putc)(c1 | 0x080);
     } else {
-        output_mode = UTF8;
+        output_mode = UTF_8;
        val = e2w_conv(c2, c1);
         if (val){
             w16w_conv(val, &c2, &c1, &c0);
@@ -3762,7 +4159,7 @@ void w_oconv16(nkf_char c2, nkf_char c1)
         return;
     }
 
-    if (c2 == ISO8859_1) {
+    if (c2 == ISO_8859_1) {
         c2 = 0;
         c1 |= 0x80;
 #ifdef NUMCHAR_OPTION
@@ -3794,6 +4191,7 @@ void w_oconv16(nkf_char c2, nkf_char c1)
         nkf_char val = e2w_conv(c2, c1);
         c2 = (val >> 8) & 0xff;
         c1 = val & 0xff;
+       if (!val) return;
     }
     if (output_endian == ENDIAN_LITTLE){
         (*o_putc)(c1);
@@ -3826,7 +4224,7 @@ void w_oconv32(nkf_char c2, nkf_char c1)
         return;
     }
 
-    if (c2 == ISO8859_1) {
+    if (c2 == ISO_8859_1) {
         c1 |= 0x80;
 #ifdef NUMCHAR_OPTION
     } else if (c2 == 0 && is_unicode_capsule(c1)) {
@@ -3834,6 +4232,7 @@ void w_oconv32(nkf_char c2, nkf_char c1)
 #endif
     } else if (c2) {
         c1 = e2w_conv(c2, c1);
+       if (!c1) return;
     }
     if (output_endian == ENDIAN_LITTLE){
         (*o_putc)( c1 & NKF_INT32_C(0x000000FF));
@@ -3855,28 +4254,46 @@ void e_oconv(nkf_char c2, nkf_char c1)
     if (c2 == 0 && is_unicode_capsule(c1)){
         w16e_conv(c1, &c2, &c1);
         if (c2 == 0 && is_unicode_capsule(c1)){
-           if(encode_fallback)(*encode_fallback)(c1);
-            return;
+           c2 = c1 & VALUE_MASK;
+           if (x0212_f && 0xE000 <= c2 && c2 <= 0xE757) {
+               /* eucJP-ms UDC */
+               c1 &= 0xFFF;
+               c2 = c1 / 94;
+               c2 += c2 < 10 ? 0x75 : 0x8FEB;
+               c1 = 0x21 + c1 % 94;
+               if (is_eucg3(c2)){
+                   (*o_putc)(0x8f);
+                   (*o_putc)((c2 & 0x7f) | 0x080);
+                   (*o_putc)(c1 | 0x080);
+               }else{
+                   (*o_putc)((c2 & 0x7f) | 0x080);
+                   (*o_putc)(c1 | 0x080);
+               }
+               return;
+           } else {
+               if (encode_fallback) (*encode_fallback)(c1);
+               return;
+           }
         }
     }
 #endif
     if (c2 == EOF) {
         (*o_putc)(EOF);
         return;
-    } else if (c2 == 0) { 
+    } else if (c2 == 0) {
        output_mode = ASCII;
         (*o_putc)(c1);
-    } else if (c2 == X0201) {
-       output_mode = JAPANESE_EUC;
+    } else if (c2 == JIS_X_0201) {
+       output_mode = EUC_JP;
         (*o_putc)(SSO); (*o_putc)(c1|0x80);
-    } else if (c2 == ISO8859_1) {
-       output_mode = ISO8859_1;
+    } else if (c2 == ISO_8859_1) {
+       output_mode = ISO_8859_1;
         (*o_putc)(c1 | 0x080);
 #ifdef X0212_ENABLE
     } else if (is_eucg3(c2)){
-       output_mode = JAPANESE_EUC;
+       output_mode = EUC_JP;
 #ifdef SHIFTJIS_CP932
-        if (cp51932_f){
+        if (!cp932inv_f){
             nkf_char s2, s1;
             if (e2s_conv(c2, c1, &s2, &s1) == 0){
                 s2e_conv(s2, s1, &c2, &c1);
@@ -3902,7 +4319,7 @@ void e_oconv(nkf_char c2, nkf_char c1)
             set_iconv(FALSE, 0);
             return; /* too late to rescue this char */
         }
-       output_mode = JAPANESE_EUC;
+       output_mode = EUC_JP;
         (*o_putc)(c2 | 0x080);
         (*o_putc)(c1 | 0x080);
     }
@@ -3942,7 +4359,7 @@ nkf_char e2s_conv(nkf_char c2, nkf_char c1, nkf_char *p2, nkf_char *p1)
 {
     nkf_char ndx;
     if (is_eucg3(c2)){
-       ndx = c2 & 0xff;
+       ndx = c2 & 0x7f;
        if (x0213_f){
            if((0x21 <= ndx && ndx <= 0x2F)){
                if (p2) *p2 = ((ndx - 1) >> 1) + 0xec - ndx / 8 * 3;
@@ -3959,9 +4376,6 @@ nkf_char e2s_conv(nkf_char c2, nkf_char c1, nkf_char *p2, nkf_char *p1)
        else if(nkf_isgraph(ndx)){
            nkf_char val = 0;
            const unsigned short *ptr;
-#if 0
-           extern const unsigned short *const x0212_shiftjis[];
-#endif
            ptr = x0212_shiftjis[ndx - 0x21];
            if (ptr){
                val = ptr[(c1 & 0x7f) - 0x21];
@@ -3989,9 +4403,21 @@ void s_oconv(nkf_char c2, nkf_char c1)
     if (c2 == 0 && is_unicode_capsule(c1)){
         w16e_conv(c1, &c2, &c1);
         if (c2 == 0 && is_unicode_capsule(c1)){
-           if(encode_fallback)(*encode_fallback)(c1);
-            return;
-        }
+           c2 = c1 & VALUE_MASK;
+           if (!x0213_f && 0xE000 <= c2 && c2 <= 0xE757) {
+               /* CP932 UDC */
+               c1 &= 0xFFF;
+               c2 = c1 / 188 + 0xF0;
+               c1 = c1 % 188;
+               c1 += 0x40 + (c1 > 0x3e);
+               (*o_putc)(c2);
+               (*o_putc)(c1);
+               return;
+           } else {
+               if(encode_fallback)(*encode_fallback)(c1);
+               return;
+           }
+       }
     }
 #endif
     if (c2 == EOF) {
@@ -4000,11 +4426,11 @@ void s_oconv(nkf_char c2, nkf_char c1)
     } else if (c2 == 0) {
        output_mode = ASCII;
         (*o_putc)(c1);
-    } else if (c2 == X0201) {
+    } else if (c2 == JIS_X_0201) {
        output_mode = SHIFT_JIS;
         (*o_putc)(c1|0x80);
-    } else if (c2 == ISO8859_1) {
-       output_mode = ISO8859_1;
+    } else if (c2 == ISO_8859_1) {
+       output_mode = ISO_8859_1;
         (*o_putc)(c1 | 0x080);
 #ifdef X0212_ENABLE
     } else if (is_eucg3(c2)){
@@ -4025,9 +4451,6 @@ void s_oconv(nkf_char c2, nkf_char c1)
 #ifdef SHIFTJIS_CP932
         if (cp932inv_f
             && CP932INV_TABLE_BEGIN <= c2 && c2 <= CP932INV_TABLE_END){
-#if 0
-            extern const unsigned short cp932inv[2][189];
-#endif
             nkf_char c = cp932inv[c2 - CP932INV_TABLE_BEGIN][c1 - 0x40];
             if (c){
                 c2 = c >> 8;
@@ -4050,13 +4473,21 @@ void j_oconv(nkf_char c2, nkf_char c1)
     if (c2 == 0 && is_unicode_capsule(c1)){
         w16e_conv(c1, &c2, &c1);
         if (c2 == 0 && is_unicode_capsule(c1)){
-           if(encode_fallback)(*encode_fallback)(c1);
-            return;
+           c2 = c1 & VALUE_MASK;
+           if (ms_ucs_map_f && 0xE000 <= c2 && c2 <= 0xE757) {
+               /* CP5022x UDC */
+               c1 &= 0xFFF;
+               c2 = 0x7F + c1 / 94;
+               c1 = 0x21 + c1 % 94;
+           } else {
+               if (encode_fallback) (*encode_fallback)(c1);
+               return;
+           }
         }
     }
 #endif
     if (c2 == EOF) {
-        if (output_mode !=ASCII && output_mode!=ISO8859_1) {
+        if (output_mode !=ASCII && output_mode!=ISO_8859_1) {
             (*o_putc)(ESC);
             (*o_putc)('(');
             (*o_putc)(ascii_intro);
@@ -4066,41 +4497,41 @@ void j_oconv(nkf_char c2, nkf_char c1)
 #ifdef X0212_ENABLE
     } else if (is_eucg3(c2)){
        if(x0213_f){
-           if(output_mode!=X0213_2){
-               output_mode = X0213_2;
+           if(output_mode!=JIS_X_0213_2){
+               output_mode = JIS_X_0213_2;
                (*o_putc)(ESC);
                (*o_putc)('$');
                (*o_putc)('(');
-               (*o_putc)(X0213_2&0x7F);
+               (*o_putc)(0x50);
            }
        }else{
-           if(output_mode!=X0212){
-               output_mode = X0212;
+           if(output_mode!=JIS_X_0212){
+               output_mode = JIS_X_0212;
                (*o_putc)(ESC);
                (*o_putc)('$');
                (*o_putc)('(');
-               (*o_putc)(X0212&0x7F);
+               (*o_putc)(0x44);
            }
         }
         (*o_putc)(c2 & 0x7f);
         (*o_putc)(c1);
 #endif
-    } else if (c2==X0201) {
-        if (output_mode!=X0201) {
-            output_mode = X0201;
+    } else if (c2==JIS_X_0201) {
+        if (output_mode!=JIS_X_0201) {
+            output_mode = JIS_X_0201;
             (*o_putc)(ESC);
             (*o_putc)('(');
             (*o_putc)('I');
         }
         (*o_putc)(c1);
-    } else if (c2==ISO8859_1) {
+    } else if (c2==ISO_8859_1) {
             /* iso8859 introduction, or 8th bit on */
-            /* Can we convert in 7bit form using ESC-'-'-A ? 
+            /* Can we convert in 7bit form using ESC-'-'-A ?
                Is this popular? */
-       output_mode = ISO8859_1;
+       output_mode = ISO_8859_1;
         (*o_putc)(c1|0x80);
     } else if (c2 == 0) {
-        if (output_mode !=ASCII && output_mode!=ISO8859_1) {
+        if (output_mode !=ASCII && output_mode!=ISO_8859_1) {
             (*o_putc)(ESC);
             (*o_putc)('(');
             (*o_putc)(ascii_intro);
@@ -4108,17 +4539,19 @@ void j_oconv(nkf_char c2, nkf_char c1)
         }
         (*o_putc)(c1);
     } else {
-       if(c2<0x20 || 0x7e<c2 || c1<0x20 || 0x7e<c1) return;
+       if(ms_ucs_map_f
+          ? c2<0x20 || 0x92<c2 || c1<0x20 || 0x7e<c1
+          : c2<0x20 || 0x7e<c2 || c1<0x20 || 0x7e<c1) return;
        if(x0213_f){
-           if (output_mode!=X0213_1) {
-               output_mode = X0213_1;
+           if (output_mode!=JIS_X_0213_1) {
+               output_mode = JIS_X_0213_1;
                (*o_putc)(ESC);
                (*o_putc)('$');
                (*o_putc)('(');
-               (*o_putc)(X0213_1&0x7F);
+               (*o_putc)(0x4F);
            }
-       }else if (output_mode != X0208) {
-            output_mode = X0208;
+       }else if (output_mode != JIS_X_0208) {
+            output_mode = JIS_X_0208;
             (*o_putc)(ESC);
             (*o_putc)('$');
             (*o_putc)(kanji_intro);
@@ -4146,20 +4579,20 @@ nkf_char broken_getc(FILE *f)
        return broken_buf[--broken_counter];
     }
     c= (*i_bgetc)(f);
-    if (c=='$' && broken_last != ESC 
-            && (input_mode==ASCII || input_mode==X0201)) {
+    if (c=='$' && broken_last != ESC
+            && (input_mode==ASCII || input_mode==JIS_X_0201)) {
        c1= (*i_bgetc)(f);
        broken_last = 0;
        if (c1=='@'|| c1=='B') {
-           broken_buf[0]=c1; broken_buf[1]=c; 
+           broken_buf[0]=c1; broken_buf[1]=c;
            broken_counter=2;
            return ESC;
        } else {
            (*i_bungetc)(c1,f);
            return c;
        }
-    } else if (c=='(' && broken_last != ESC 
-            && (input_mode==X0208 || input_mode==X0201)) { /* ) */
+    } else if (c=='(' && broken_last != ESC
+            && (input_mode==JIS_X_0208 || input_mode==JIS_X_0201)) { /* ) */
        c1= (*i_bgetc)(f);
        broken_last = 0;
        if (c1=='J'|| c1=='B') {
@@ -4183,48 +4616,41 @@ nkf_char broken_ungetc(nkf_char c, FILE *f)
     return c;
 }
 
-static nkf_char prev_cr = 0;
-
-void cr_conv(nkf_char c2, nkf_char c1)
+void nl_conv(nkf_char c2, nkf_char c1)
 {
-    if (prev_cr) {
+    if (guess_f && input_newline != EOF) {
+       if (c2 == 0 && c1 == LF) {
+           if (!input_newline) input_newline = prev_cr ? CRLF : LF;
+           else if (input_newline != (prev_cr ? CRLF : LF)) input_newline = EOF;
+       } else if (c2 == 0 && c1 == CR && input_newline == LF) input_newline = EOF;
+       else if (!prev_cr);
+       else if (!input_newline) input_newline = CR;
+       else if (input_newline != CR) input_newline = EOF;
+    }
+    if (prev_cr || (c2 == 0 && c1 == LF)) {
        prev_cr = 0;
-       if (! (c2==0&&c1==NL) ) {
-           cr_conv(0,'\n');
-       }
-    }
-    if (c2) {
-        (*o_crconv)(c2,c1);
-    } else if (c1=='\r') {
-       prev_cr = c1;
-    } else if (c1=='\n') {
-        if (crmode_f==CRLF) {
-            (*o_crconv)(0,'\r');
-       } else if (crmode_f==CR) {
-            (*o_crconv)(0,'\r');
-           return;
-       } 
-       (*o_crconv)(0,NL);
-    } else if (c1!='\032' || crmode_f!=NL){
-        (*o_crconv)(c2,c1);
+       if (nlmode_f != LF) (*o_nlconv)(0, CR);
+       if (nlmode_f != CR) (*o_nlconv)(0, LF);
     }
+    if (c2 == 0 && c1 == CR) prev_cr = CR;
+    else if (c2 != 0 || c1 != LF) (*o_nlconv)(c2, c1);
 }
 
-/* 
+/*
   Return value of fold_conv()
 
-       \n  add newline  and output char
-       \r  add newline  and output nothing
-       ' ' space
-       0   skip  
-       1   (or else) normal output 
+       LF  add newline  and output char
+       CR  add newline  and output nothing
+       SP  space
+       0   skip
+       1   (or else) normal output
 
   fold state in prev (previous character)
 
       >0x80 Japanese (X0208/X0201)
       <0x80 ASCII
-      \n    new line 
-      ' '   space
+      LF    new line
+      SP    space
 
   This fold algorthm does not preserve heading space in a line.
   This is the main difference from fmt.
@@ -4233,34 +4659,34 @@ void cr_conv(nkf_char c2, nkf_char c1)
 #define char_size(c2,c1) (c2?2:1)
 
 void fold_conv(nkf_char c2, nkf_char c1)
-{ 
+{
     nkf_char prev0;
     nkf_char fold_state;
 
-    if (c1== '\r' && !fold_preserve_f) {
+    if (c1== CR && !fold_preserve_f) {
        fold_state=0;  /* ignore cr */
-    }else if (c1== '\n'&&f_prev=='\r' && fold_preserve_f) {
-        f_prev = '\n';
+    }else if (c1== LF&&f_prev==CR && fold_preserve_f) {
+        f_prev = LF;
        fold_state=0;  /* ignore cr */
     } else if (c1== BS) {
         if (f_line>0) f_line--;
         fold_state =  1;
     } else if (c2==EOF && f_line != 0) {    /* close open last line */
-            fold_state = '\n';
-    } else if ((c1=='\n' && !fold_preserve_f)
-               || ((c1=='\r'||(c1=='\n'&&f_prev!='\r'))
+            fold_state = LF;
+    } else if ((c1==LF && !fold_preserve_f)
+               || ((c1==CR||(c1==LF&&f_prev!=CR))
                    && fold_preserve_f)) {
         /* new line */
-        if (fold_preserve_f) { 
+        if (fold_preserve_f) {
             f_prev = c1;
             f_line = 0;
-            fold_state =  '\r';
+            fold_state =  CR;
        } else if ((f_prev == c1 && !fold_preserve_f)
-                   || (f_prev == '\n' && fold_preserve_f)
+                   || (f_prev == LF && fold_preserve_f)
                    ) {        /* duplicate newline */
             if (f_line) {
                 f_line = 0;
-                fold_state =  '\n';    /* output two newline */
+                fold_state =  LF;    /* output two newline */
             } else {
                 f_line = 0;
                 fold_state =  1;
@@ -4269,41 +4695,41 @@ void fold_conv(nkf_char c2, nkf_char c1)
             if (f_prev&0x80) {     /* Japanese? */
                 f_prev = c1;
                 fold_state =  0;       /* ignore given single newline */
-            } else if (f_prev==' ') {
+            } else if (f_prev==SP) {
                 fold_state =  0;
             } else {
                 f_prev = c1;
-                if (++f_line<=fold_len) 
-                    fold_state =  ' ';
+                if (++f_line<=fold_len)
+                    fold_state =  SP;
                 else {
                     f_line = 0;
-                    fold_state =  '\r';        /* fold and output nothing */
+                    fold_state =  CR;        /* fold and output nothing */
                 }
             }
         }
     } else if (c1=='\f') {
-        f_prev = '\n';
+        f_prev = LF;
         f_line = 0;
-        fold_state =  '\n';            /* output newline and clear */
-    } else if ( (c2==0  && c1==' ')||
-               (c2==0  && c1=='\t')||
+        fold_state =  LF;            /* output newline and clear */
+    } else if ( (c2==0  && c1==SP)||
+               (c2==0  && c1==TAB)||
                (c2=='!'&& c1=='!')) {
         /* X0208 kankaku or ascii space */
-            if (f_prev == ' ') {
+            if (f_prev == SP) {
                 fold_state = 0;         /* remove duplicate spaces */
             } else {
-                f_prev = ' ';    
-                if (++f_line<=fold_len) 
-                    fold_state = ' ';         /* output ASCII space only */
+                f_prev = SP;
+                if (++f_line<=fold_len)
+                    fold_state = SP;         /* output ASCII space only */
                 else {
-                    f_prev = ' '; f_line = 0;
-                    fold_state = '\r';        /* fold and output nothing */
+                    f_prev = SP; f_line = 0;
+                    fold_state = CR;        /* fold and output nothing */
                 }
             }
     } else {
         prev0 = f_prev; /* we still need this one... , but almost done */
         f_prev = c1;
-        if (c2 || c2==X0201) 
+        if (c2 || c2==JIS_X_0201)
             f_prev |= 0x80;  /* this is Japanese */
         f_line += char_size(c2,c1);
         if (f_line<=fold_len) {   /* normal case */
@@ -4311,8 +4737,8 @@ void fold_conv(nkf_char c2, nkf_char c1)
         } else {
             if (f_line>fold_len+fold_margin) { /* too many kinsoku suspension */
                 f_line = char_size(c2,c1);
-                fold_state =  '\n';       /* We can't wait, do fold now */
-            } else if (c2==X0201) {
+                fold_state =  LF;       /* We can't wait, do fold now */
+            } else if (c2==JIS_X_0201) {
             /* simple kinsoku rules  return 1 means no folding  */
                 if (c1==(0xde&0x7f)) fold_state = 1; /* \e$B!+\e(B*/
                 else if (c1==(0xdf&0x7f)) fold_state = 1; /* \e$B!,\e(B*/
@@ -4320,15 +4746,15 @@ void fold_conv(nkf_char c2, nkf_char c1)
                 else if (c1==(0xa3&0x7f)) fold_state = 1; /* \e$B!$\e(B*/
                 else if (c1==(0xa1&0x7f)) fold_state = 1; /* \e$B!W\e(B*/
                 else if (c1==(0xb0&0x7f)) fold_state = 1; /* - */
-                else if (SPACE<=c1 && c1<=(0xdf&0x7f)) {      /* X0201 */
+                else if (SP<=c1 && c1<=(0xdf&0x7f)) {      /* X0201 */
                    f_line = 1;
-                   fold_state = '\n';/* add one new f_line before this character */
+                   fold_state = LF;/* add one new f_line before this character */
                } else {
                    f_line = 1;
-                   fold_state = '\n';/* add one new f_line before this character */
+                   fold_state = LF;/* add one new f_line before this character */
                }
             } else if (c2==0) {
-                /* kinsoku point in ASCII */ 
+                /* kinsoku point in ASCII */
                if (  c1==')'||    /* { [ ( */
                      c1==']'||
                      c1=='}'||
@@ -4338,17 +4764,17 @@ void fold_conv(nkf_char c2, nkf_char c1)
                      c1=='?'||
                      c1=='/'||
                      c1==':'||
-                     c1==';' ) {
+                     c1==';') {
                    fold_state = 1;
                /* just after special */
                } else if (!is_alnum(prev0)) {
                    f_line = char_size(c2,c1);
-                   fold_state = '\n';
-               } else if ((prev0==' ') ||   /* ignored new f_line */
-                      (prev0=='\n')||        /* ignored new f_line */
+                   fold_state = LF;
+               } else if ((prev0==SP) ||   /* ignored new f_line */
+                      (prev0==LF)||        /* ignored new f_line */
                       (prev0&0x80)) {        /* X0208 - ASCII */
                    f_line = char_size(c2,c1);
-                    fold_state = '\n';/* add one new f_line before this character */
+                    fold_state = LF;/* add one new f_line before this character */
                 } else {
                     fold_state = 1;  /* default no fold in ASCII */
                 }
@@ -4367,14 +4793,14 @@ void fold_conv(nkf_char c2, nkf_char c1)
                     else if (c1=='+')  fold_state = 1; /* \e$B!+\e(B */
                     else if (c1==',')  fold_state = 1; /* \e$B!,\e(B */
                          /* default no fold in kinsoku */
-                   else { 
-                       fold_state = '\n';
+                   else {
+                       fold_state = LF;
                        f_line = char_size(c2,c1);
                        /* add one new f_line before this character */
                    }
                 } else {
                    f_line = char_size(c2,c1);
-                    fold_state = '\n'; 
+                    fold_state = LF;
                     /* add one new f_line before this character */
                 }
             }
@@ -4382,18 +4808,18 @@ void fold_conv(nkf_char c2, nkf_char c1)
     }
     /* terminator process */
     switch(fold_state) {
-        case '\n': 
-            (*o_fconv)(0,'\n');
+        case LF:
+            OCONV_NEWLINE((*o_fconv));
             (*o_fconv)(c2,c1);
             break;
-        case 0:    
+        case 0:
             return;
-        case '\r': 
-            (*o_fconv)(0,'\n');
+        case CR:
+            OCONV_NEWLINE((*o_fconv));
             break;
-        case '\t': 
-        case ' ': 
-            (*o_fconv)(0,' ');
+        case TAB:
+        case SP:
+            (*o_fconv)(0,SP);
             break;
         default:
             (*o_fconv)(c2,c1);
@@ -4407,75 +4833,155 @@ void z_conv(nkf_char c2, nkf_char c1)
 
     /* if (c2) c1 &= 0x7f; assertion */
 
-    if (x0201_f && z_prev2==X0201) {  /* X0201 */
-        if (c1==(0xde&0x7f)) { /* \e$BByE@\e(B */
-            z_prev2=0;
-            (*o_zconv)(dv[(z_prev1-SPACE)*2],dv[(z_prev1-SPACE)*2+1]);
-            return;
-        } else if (c1==(0xdf&0x7f)&&ev[(z_prev1-SPACE)*2]) {  /* \e$BH>ByE@\e(B */
-            z_prev2=0;
-            (*o_zconv)(ev[(z_prev1-SPACE)*2],ev[(z_prev1-SPACE)*2+1]);
-            return;
-        } else {
-            z_prev2=0;
-            (*o_zconv)(cv[(z_prev1-SPACE)*2],cv[(z_prev1-SPACE)*2+1]);
-        }
+    if (c2 == JIS_X_0201 && (c1 == 0x20 || c1 == 0x7D || c1 == 0x7E)) {
+       (*o_zconv)(c2,c1);
+       return;
     }
 
-    if (c2==EOF) {
-        (*o_zconv)(c2,c1);
-        return;
+    if (x0201_f) {
+       if (z_prev2 == JIS_X_0201) {
+           if (c2 == JIS_X_0201) {
+               if (c1 == (0xde&0x7f)) { /* \e$BByE@\e(B */
+                   z_prev2 = 0;
+                   (*o_zconv)(dv[(z_prev1-SP)*2], dv[(z_prev1-SP)*2+1]);
+                   return;
+               } else if (c1 == (0xdf&0x7f) && ev[(z_prev1-SP)*2]) {  /* \e$BH>ByE@\e(B */
+                   z_prev2 = 0;
+                   (*o_zconv)(ev[(z_prev1-SP)*2], ev[(z_prev1-SP)*2+1]);
+                   return;
+               }
+           }
+           z_prev2 = 0;
+           (*o_zconv)(cv[(z_prev1-SP)*2], cv[(z_prev1-SP)*2+1]);
+       }
+       if (c2 == JIS_X_0201) {
+           if (dv[(c1-SP)*2] || ev[(c1-SP)*2]) {
+               /* wait for \e$BByE@\e(B or \e$BH>ByE@\e(B */
+               z_prev1 = c1;
+               z_prev2 = c2;
+               return;
+           } else {
+               (*o_zconv)(cv[(c1-SP)*2], cv[(c1-SP)*2+1]);
+               return;
+           }
+       }
     }
 
-    if (x0201_f && c2==X0201) {
-        if (dv[(c1-SPACE)*2]||ev[(c1-SPACE)*2]) {
-            /* wait for \e$BByE@\e(B or \e$BH>ByE@\e(B */
-            z_prev1 = c1; z_prev2 = c2;
-            return;
-        } else {
-            (*o_zconv)(cv[(c1-SPACE)*2],cv[(c1-SPACE)*2+1]);
-            return;
-        }
+    if (c2 == EOF) {
+        (*o_zconv)(c2, c1);
+        return;
     }
 
-    /* JISX0208 Alphabet */
-    if (alpha_f && c2 == 0x23 ) {
+    if (alpha_f&1 && c2 == 0x23) {
+       /* JISX0208 Alphabet */
         c2 = 0;
-    } else if (alpha_f && c2 == 0x21 ) { 
-    /* JISX0208 Kigou */
+    } else if (c2 == 0x21) {
+       /* JISX0208 Kigou */
        if (0x21==c1) {
-           if (alpha_f&0x2) {
-               c1 = ' ';
+           if (alpha_f&2) {
                c2 = 0;
-           } else if (alpha_f&0x4) {
-                (*o_zconv)(0,' ');
-                (*o_zconv)(0,' ');
+               c1 = SP;
+           } else if (alpha_f&4) {
+                (*o_zconv)(0, SP);
+                (*o_zconv)(0, SP);
                 return;
-           } 
-       } else if (0x20<c1 && c1<0x7f && fv[c1-0x20]) {
-           c1 = fv[c1-0x20];
-           c2 =  0;
-           if (alpha_f&0x8) {
-               char *entity = 0;
-               switch (c1){
-                 case '>': entity = "&gt;"; break;
-                 case '<': entity = "&lt;"; break;
-                 case '\"': entity = "&quot;"; break;
-                 case '&': entity = "&amp;"; break;
-               }
-               if (entity){
-                   while (*entity) (*o_zconv)(0, *entity++);
-                   return;
-               }
            }
-       } 
+       } else if (alpha_f&1 && 0x20<c1 && c1<0x7f && fv[c1-0x20]) {
+           c2 =  0;
+           c1 = fv[c1-0x20];
+       }
+    }
+
+    if (alpha_f&8 && c2 == 0) {
+       /* HTML Entity */
+       char *entity = 0;
+       switch (c1){
+       case '>': entity = "&gt;"; break;
+       case '<': entity = "&lt;"; break;
+       case '\"': entity = "&quot;"; break;
+       case '&': entity = "&amp;"; break;
+       }
+       if (entity){
+           while (*entity) (*o_zconv)(0, *entity++);
+           return;
+       }
+    }
+
+    if (alpha_f & 16) {
+       /* JIS X 0208 Katakana to JIS X 0201 Katakana */
+       if (c2 == 0x21) {
+           char c = 0;
+           switch (c1) {
+           case 0x23:
+               /* U+3002 (0x8142) Ideographic Full Stop -> U+FF61 (0xA1) Halfwidth Ideographic Full Stop */
+               c = 0xA1;
+               break;
+           case 0x56:
+               /* U+300C (0x8175) Left Corner Bracket -> U+FF62 (0xA2) Halfwidth Left Corner Bracket */
+               c = 0xA2;
+               break;
+           case 0x57:
+               /* U+300D (0x8176) Right Corner Bracket -> U+FF63 (0xA3) Halfwidth Right Corner Bracket */
+               c = 0xA3;
+               break;
+           case 0x22:
+               /* U+3001 (0x8141) Ideographic Comma -> U+FF64 (0xA4) Halfwidth Ideographic Comma */
+               c = 0xA4;
+               break;
+           case 0x26:
+               /* U+30FB (0x8145) Katakana Middle Dot -> U+FF65 (0xA5) Halfwidth Katakana Middle Dot */
+               c = 0xA5;
+               break;
+           case 0x3C:
+               /* U+30FC (0x815B) Katakana-Hiragana Prolonged Sound Mark -> U+FF70 (0xB0) Halfwidth Katakana-Hiragana Prolonged Sound Mark */
+               c = 0xB0;
+               break;
+           case 0x2B:
+               /* U+309B (0x814A) Katakana-Hiragana Voiced Sound Mark -> U+FF9E (0xDE) Halfwidth Katakana Voiced Sound Mark */
+               c = 0xDE;
+               break;
+           case 0x2C:
+               /* U+309C (0x814B) Katakana-Hiragana Semi-Voiced Sound Mark -> U+FF9F (0xDF) Halfwidth Katakana Semi-Voiced Sound Mark */
+               c = 0xDF;
+               break;
+           }
+           if (c) {
+               (*o_zconv)(JIS_X_0201, c);
+               return;
+           }
+       } else if (c2 == 0x25) {
+           /* JISX0208 Katakana */
+           static const int fullwidth_to_halfwidth[] =
+           {
+               0x0000, 0x2700, 0x3100, 0x2800, 0x3200, 0x2900, 0x3300, 0x2A00,
+               0x3400, 0x2B00, 0x3500, 0x3600, 0x365E, 0x3700, 0x375E, 0x3800,
+               0x385E, 0x3900, 0x395E, 0x3A00, 0x3A5E, 0x3B00, 0x3B5E, 0x3C00,
+               0x3C5E, 0x3D00, 0x3D5E, 0x3E00, 0x3E5E, 0x3F00, 0x3F5E, 0x4000,
+               0x405E, 0x4100, 0x415E, 0x2F00, 0x4200, 0x425E, 0x4300, 0x435E,
+               0x4400, 0x445E, 0x4500, 0x4600, 0x4700, 0x4800, 0x4900, 0x4A00,
+               0x4A5E, 0x4A5F, 0x4B00, 0x4B5E, 0x4B5F, 0x4C00, 0x4C5E, 0x4C5F,
+               0x4D00, 0x4D5E, 0x4D5F, 0x4E00, 0x4E5E, 0x4E5F, 0x4F00, 0x5000,
+               0x5100, 0x5200, 0x5300, 0x2C00, 0x5400, 0x2D00, 0x5500, 0x2E00,
+               0x5600, 0x5700, 0x5800, 0x5900, 0x5A00, 0x5B00, 0x0000, 0x5C00,
+               0x0000, 0x0000, 0x2600, 0x5D00, 0x335E, 0x0000, 0x0000, 0x0000,
+               0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000
+           };
+           if (fullwidth_to_halfwidth[c1-0x20]){
+               c2 = fullwidth_to_halfwidth[c1-0x20];
+               (*o_zconv)(JIS_X_0201, c2>>8);
+               if (c2 & 0xFF) {
+                   (*o_zconv)(JIS_X_0201, c2&0xFF);
+               }
+               return;
+           }
+       }
     }
     (*o_zconv)(c2,c1);
 }
 
 
 #define rot13(c)  ( \
-      ( c < 'A' ) ? c: \
+      ( c < 'A') ? c: \
       (c <= 'M')  ? (c + 13): \
       (c <= 'Z')  ? (c - 13): \
       (c < 'a')   ? (c): \
@@ -4485,15 +4991,15 @@ void z_conv(nkf_char c2, nkf_char c1)
 )
 
 #define  rot47(c) ( \
-      ( c < '!' ) ? c: \
-      ( c <= 'O' ) ? (c + 47) : \
-      ( c <= '~' ) ?  (c - 47) : \
+      ( c < '!') ? c: \
+      ( c <= 'O') ? (c + 47) : \
+      ( c <= '~') ?  (c - 47) : \
       c \
 )
 
 void rot_conv(nkf_char c2, nkf_char c1)
 {
-    if (c2==0 || c2==X0201 || c2==ISO8859_1) {
+    if (c2==0 || c2==JIS_X_0201 || c2==ISO_8859_1) {
        c1 = rot13(c1);
     } else if (c2) {
        c1 = rot47(c1);
@@ -4585,7 +5091,7 @@ void iso2022jp_check_conv(nkf_char c2, nkf_char c1)
 
 /* This converts  =?ISO-2022-JP?B?HOGE HOGE?= */
 
-const unsigned char *mime_pattern[] = {
+static const unsigned char *mime_pattern[] = {
     (const unsigned char *)"\075?EUC-JP?B?",
     (const unsigned char *)"\075?SHIFT_JIS?B?",
     (const unsigned char *)"\075?ISO-8859-1?Q?",
@@ -4610,16 +5116,16 @@ nkf_char (*mime_priority_func[])(nkf_char c2, nkf_char c1, nkf_char c0) = {
     0,
 };
 
-const nkf_char mime_encode[] = {
-    JAPANESE_EUC, SHIFT_JIS,ISO8859_1, ISO8859_1, X0208, X0201,
+static const nkf_char mime_encode[] = {
+    EUC_JP, SHIFT_JIS, ISO_8859_1, ISO_8859_1, JIS_X_0208, JIS_X_0201,
 #if defined(UTF8_INPUT_ENABLE)
-    UTF8, UTF8,
+    UTF_8, UTF_8,
 #endif
     ASCII,
     0
 };
 
-const nkf_char mime_encode_method[] = {
+static const nkf_char mime_encode_method[] = {
     'B', 'B','Q', 'B', 'B', 'Q',
 #if defined(UTF8_INPUT_ENABLE)
     'B', 'Q',
@@ -4668,8 +5174,8 @@ nkf_char mime_begin_strict(FILE *f)
     p = mime_pattern[j];
     r[0]='='; r[1]='?';
 
-    for(i=2;p[i]>' ';i++) {                   /* start at =? */
-        if ( ((r[i] = c1 = (*i_getc)(f))==EOF) || nkf_toupper(c1) != p[i] ) {
+    for(i=2;p[i]>SP;i++) {                   /* start at =? */
+        if (((r[i] = c1 = (*i_getc)(f))==EOF) || nkf_toupper(c1) != p[i]) {
             /* pattern fails, try next one */
             q = p;
             while (mime_pattern[++j]) {
@@ -4699,7 +5205,7 @@ nkf_char mime_begin_strict(FILE *f)
         if (!unbuf_f) {
             /* do MIME integrity check */
             return mime_integrity(f,mime_pattern[j]);
-        } 
+        }
     }
     switch_mime_getc();
     mimebuf_f = TRUE;
@@ -4718,7 +5224,7 @@ nkf_char mime_ungetc_buf(nkf_char c, FILE *f)
 {
     if (mimebuf_f)
        (*i_mungetc_buf)(c,f);
-    else 
+    else
        Fifo(--mime_input) = (unsigned char)c;
     return c;
 }
@@ -4737,8 +5243,8 @@ nkf_char mime_begin(FILE *f)
     for(i=2;i<MAXRECOVER;i++) {                   /* start at =? */
         /* We accept any character type even if it is breaked by new lines */
         c1 = (*i_getc)(f); Fifo(mime_last++) = (unsigned char)c1;
-        if (c1=='\n'||c1==' '||c1=='\r'||
-                c1=='-'||c1=='_'||is_alnum(c1) ) continue;
+        if (c1==LF||c1==SP||c1==CR||
+                c1=='-'||c1=='_'||is_alnum(c1)) continue;
         if (c1=='=') {
             /* Failed. But this could be another MIME preemble */
             (*i_ungetc)(c1,f);
@@ -4788,41 +5294,70 @@ void no_putc(nkf_char c)
 void debug(const char *str)
 {
     if (debug_f){
-        fprintf(stderr, "%s\n", str);
+        fprintf(stderr, "%s\n", str ? str : "NULL");
     }
 }
 #endif
 
 void set_input_codename(char *codename)
 {
-    if (guess_f && 
-        is_inputcode_set &&
-        strcmp(codename, "") != 0 && 
-        strcmp(codename, input_codename) != 0)
-    {
-        is_inputcode_mixed = TRUE;
+    if (!input_codename) {
+       input_codename = codename;
+    } else if (strcmp(codename, input_codename) != 0) {
+       input_codename = "";
+    }
+}
+
+static char* get_guessed_code(void)
+{
+    if (input_codename && !*input_codename) {
+       input_codename = "BINARY";
+    } else {
+       struct input_code *p = find_inputcode_byfunc(iconv);
+       if (!input_codename) {
+           input_codename = "ASCII";
+       } else if (strcmp(input_codename, "Shift_JIS") == 0) {
+           if (p->score & (SCORE_DEPEND|SCORE_CP932))
+               input_codename = "CP932";
+       } else if (strcmp(input_codename, "EUC-JP") == 0) {
+           if (p->score & (SCORE_X0212))
+               input_codename = "EUCJP-MS";
+           else if (p->score & (SCORE_DEPEND|SCORE_CP932))
+               input_codename = "CP51932";
+       } else if (strcmp(input_codename, "ISO-2022-JP") == 0) {
+           if (p->score & (SCORE_KANA))
+               input_codename = "CP50221";
+           else if (p->score & (SCORE_DEPEND|SCORE_CP932))
+               input_codename = "CP50220";
+       }
     }
-    input_codename = codename;
-    is_inputcode_set = TRUE;
+    return input_codename;
 }
 
 #if !defined(PERL_XS) && !defined(WIN32DLL)
 void print_guessed_code(char *filename)
 {
-    char *codename = "BINARY";
-    if (!is_inputcode_mixed) {
-        if (strcmp(input_codename, "") == 0) {
-            codename = "ASCII";
-        } else {
-            codename = input_codename;
-        }
+    if (filename != NULL) printf("%s: ", filename);
+    if (input_codename && !*input_codename) {
+       printf("BINARY\n");
+    } else {
+       input_codename = get_guessed_code();
+       if (guess_f == 1) {
+           printf("%s\n", input_codename);
+       } else {
+           printf("%s%s\n",
+                  input_codename,
+                  input_newline == CR   ? " (CR)" :
+                  input_newline == LF   ? " (LF)" :
+                  input_newline == CRLF ? " (CRLF)" :
+                  input_newline == EOF  ? " (MIXED NL)" :
+                  "");
+       }
     }
-    if (filename != NULL) printf("%s:", filename);
-    printf("%s\n", codename);
 }
 #endif /*WIN32DLL*/
 
-#ifdef INPUT_OPTION 
+#ifdef INPUT_OPTION
 
 nkf_char hex_getc(nkf_char ch, FILE *f, nkf_char (*g)(FILE *f), nkf_char (*u)(nkf_char c, FILE *f))
 {
@@ -4936,10 +5471,7 @@ nkf_char nfc_getc(FILE *f)
     int i=0, j, k=1, lower, upper;
     nkf_char buf[9];
     const nkf_nfchar *array;
-#if 0
-    extern const struct normalization_pair normalization_table[];
-#endif
-    
+
     buf[i] = (*g)(f);
     while (k > 0 && ((buf[i] & 0xc0) != 0x80)){
        lower=0, upper=NORMALIZATION_TABLE_LENGTH-1;
@@ -4975,7 +5507,7 @@ nkf_char nfc_ungetc(nkf_char c, FILE *f)
 #endif /* UNICODE_NORMALIZATION */
 
 
-nkf_char 
+nkf_char
 mime_getc(FILE *f)
 {
     nkf_char c1, c2, c3, c4, cc;
@@ -5001,15 +5533,15 @@ mime_getc(FILE *f)
     if (mime_decode_mode == 'Q') {
         if ((c1 = (*i_mgetc)(f)) == EOF) return (EOF);
 restart_mime_q:
-        if (c1=='_' && mimebuf_f != FIXED_MIME) return ' ';
-       if (c1<=' ' || DEL<=c1) {
+        if (c1=='_' && mimebuf_f != FIXED_MIME) return SP;
+       if (c1<=SP || DEL<=c1) {
            mime_decode_mode = exit_mode; /* prepare for quit */
            return c1;
        }
         if (c1!='=' && (c1!='?' || mimebuf_f == FIXED_MIME)) {
            return c1;
        }
-                
+
         mime_decode_mode = exit_mode; /* prepare for quit */
         if ((c2 = (*i_mgetc)(f)) == EOF) return (EOF);
         if (c1=='?'&&c2=='=' && mimebuf_f != FIXED_MIME) {
@@ -5023,32 +5555,32 @@ restart_mime_q:
            }
            while ((c1=(*i_getc)(f))!=EOF) {
                switch (c1) {
-               case NL:
+               case LF:
                case CR:
-                   if (c1==NL) {
-                       if ((c1=(*i_getc)(f))!=EOF && (c1==SPACE||c1==TAB)) {
-                           i_ungetc(SPACE,f);
+                   if (c1==LF) {
+                       if ((c1=(*i_getc)(f))!=EOF && (c1==SP||c1==TAB)) {
+                           i_ungetc(SP,f);
                            continue;
                        } else {
                            i_ungetc(c1,f);
                        }
-                       c1 = NL;
+                       c1 = LF;
                    } else {
-                       if ((c1=(*i_getc)(f))!=EOF && c1 == NL) {
-                           if ((c1=(*i_getc)(f))!=EOF && (c1==SPACE||c1==TAB)) {
-                               i_ungetc(SPACE,f);
+                       if ((c1=(*i_getc)(f))!=EOF && c1 == LF) {
+                           if ((c1=(*i_getc)(f))!=EOF && (c1==SP||c1==TAB)) {
+                               i_ungetc(SP,f);
                                continue;
                            } else {
                                i_ungetc(c1,f);
                            }
-                           i_ungetc(NL,f);
+                           i_ungetc(LF,f);
                        } else {
                            i_ungetc(c1,f);
                        }
                        c1 = CR;
                    }
                    break;
-               case SPACE:
+               case SP:
                case TAB:
                    lwsp_buf[lwsp_count] = (unsigned char)c1;
                    if (lwsp_count++>lwsp_size){
@@ -5065,7 +5597,7 @@ restart_mime_q:
                }
                break;
            }
-           if (lwsp_count > 0 && (c1 != '=' || (lwsp_buf[lwsp_count-1] != SPACE && lwsp_buf[lwsp_count-1] != TAB))) {
+           if (lwsp_count > 0 && (c1 != '=' || (lwsp_buf[lwsp_count-1] != SP && lwsp_buf[lwsp_count-1] != TAB))) {
                i_ungetc(c1,f);
                for(lwsp_count--;lwsp_count>0;lwsp_count--)
                    i_ungetc(lwsp_buf[lwsp_count],f);
@@ -5074,8 +5606,8 @@ restart_mime_q:
            free(lwsp_buf);
             return c1;
         }
-        if (c1=='='&&c2<' ') { /* this is soft wrap */
-            while((c1 =  (*i_mgetc)(f)) <=' ') {
+        if (c1=='='&&c2<SP) { /* this is soft wrap */
+            while((c1 =  (*i_mgetc)(f)) <=SP) {
                if ((c1 = (*i_mgetc)(f)) == EOF) return (EOF);
            }
             mime_decode_mode = 'Q'; /* still in MIME */
@@ -5087,7 +5619,7 @@ restart_mime_q:
             return c1;
         }
         if ((c3 = (*i_mgetc)(f)) == EOF) return (EOF);
-        if (c2<=' ') return c2;
+        if (c2<=SP) return c2;
         mime_decode_mode = 'Q'; /* still in MIME */
         return ((hex2bin(c2)<<4) + hex2bin(c3));
     }
@@ -5099,26 +5631,26 @@ restart_mime_q:
 
 
     /* Base64 encoding */
-    /* 
-        MIME allows line break in the middle of 
+    /*
+        MIME allows line break in the middle of
         Base64, but we are very pessimistic in decoding
-        in unbuf mode because MIME encoded code may broken by 
+        in unbuf mode because MIME encoded code may broken by
         less or editor's control sequence (such as ESC-[-K in unbuffered
         mode. ignore incomplete MIME.
     */
     mode = mime_decode_mode;
     mime_decode_mode = exit_mode;  /* prepare for quit */
 
-    while ((c1 = (*i_mgetc)(f))<=' ') {
+    while ((c1 = (*i_mgetc)(f))<=SP) {
         if (c1==EOF)
             return (EOF);
     }
 mime_c2_retry:
-    if ((c2 = (*i_mgetc)(f))<=' ') {
+    if ((c2 = (*i_mgetc)(f))<=SP) {
         if (c2==EOF)
             return (EOF);
        if (mime_f != STRICT_MIME) goto mime_c2_retry;
-        if (mimebuf_f!=FIXED_MIME) input_mode = ASCII;  
+        if (mimebuf_f!=FIXED_MIME) input_mode = ASCII;
         return c2;
     }
     if ((c1 == '?') && (c2 == '=')) {
@@ -5131,35 +5663,35 @@ mime_c2_retry:
        }
        while ((c1=(*i_getc)(f))!=EOF) {
            switch (c1) {
-           case NL:
+           case LF:
            case CR:
-               if (c1==NL) {
-                   if ((c1=(*i_getc)(f))!=EOF && (c1==SPACE||c1==TAB)) {
-                       i_ungetc(SPACE,f);
+               if (c1==LF) {
+                   if ((c1=(*i_getc)(f))!=EOF && (c1==SP||c1==TAB)) {
+                       i_ungetc(SP,f);
                        continue;
                    } else {
                        i_ungetc(c1,f);
                    }
-                   c1 = NL;
+                   c1 = LF;
                } else {
                    if ((c1=(*i_getc)(f))!=EOF) {
-                       if (c1==SPACE) {
-                           i_ungetc(SPACE,f);
+                       if (c1==SP) {
+                           i_ungetc(SP,f);
                            continue;
-                       } else if ((c1=(*i_getc)(f))!=EOF && (c1==SPACE||c1==TAB)) {
-                           i_ungetc(SPACE,f);
+                       } else if ((c1=(*i_getc)(f))!=EOF && (c1==SP||c1==TAB)) {
+                           i_ungetc(SP,f);
                            continue;
                        } else {
                            i_ungetc(c1,f);
                        }
-                       i_ungetc(NL,f);
+                       i_ungetc(LF,f);
                    } else {
                        i_ungetc(c1,f);
                    }
                    c1 = CR;
                }
                break;
-           case SPACE:
+           case SP:
            case TAB:
                lwsp_buf[lwsp_count] = (unsigned char)c1;
                if (lwsp_count++>lwsp_size){
@@ -5176,7 +5708,7 @@ mime_c2_retry:
            }
            break;
        }
-       if (lwsp_count > 0 && (c1 != '=' || (lwsp_buf[lwsp_count-1] != SPACE && lwsp_buf[lwsp_count-1] != TAB))) {
+       if (lwsp_count > 0 && (c1 != '=' || (lwsp_buf[lwsp_count-1] != SP && lwsp_buf[lwsp_count-1] != TAB))) {
            i_ungetc(c1,f);
            for(lwsp_count--;lwsp_count>0;lwsp_count--)
                i_ungetc(lwsp_buf[lwsp_count],f);
@@ -5186,19 +5718,19 @@ mime_c2_retry:
         return c1;
     }
 mime_c3_retry:
-    if ((c3 = (*i_mgetc)(f))<=' ') {
+    if ((c3 = (*i_mgetc)(f))<=SP) {
         if (c3==EOF)
             return (EOF);
        if (mime_f != STRICT_MIME) goto mime_c3_retry;
-        if (mimebuf_f!=FIXED_MIME) input_mode = ASCII;  
+        if (mimebuf_f!=FIXED_MIME) input_mode = ASCII;
         return c3;
     }
 mime_c4_retry:
-    if ((c4 = (*i_mgetc)(f))<=' ') {
+    if ((c4 = (*i_mgetc)(f))<=SP) {
         if (c4==EOF)
             return (EOF);
        if (mime_f != STRICT_MIME) goto mime_c4_retry;
-        if (mimebuf_f!=FIXED_MIME) input_mode = ASCII;  
+        if (mimebuf_f!=FIXED_MIME) input_mode = ASCII;
         return c4;
     }
 
@@ -5217,7 +5749,7 @@ mime_c4_retry:
         if (c3 != '=') {
             Fifo(mime_last++) = (unsigned char)cc;
             cc = ((t3 << 6) & 0x0c0) | (t4 & 0x3f);
-            if (c4 != '=') 
+            if (c4 != '=')
                 Fifo(mime_last++) = (unsigned char)cc;
         }
     } else {
@@ -5240,7 +5772,7 @@ nkf_char mime_integrity(FILE *f, const unsigned char *p)
      */
     mime_input = mime_top;
     mime_last = mime_top;
-    
+
     while(*p) Fifo(mime_input++) = *p++;
     d = 0;
     q = mime_input;
@@ -5252,7 +5784,7 @@ nkf_char mime_integrity(FILE *f, const unsigned char *p)
             /* checked. skip header, start decode */
             Fifo(mime_input++) = (unsigned char)c;
             /* mime_last_input = mime_input; */
-            mime_input = q; 
+            mime_input = q;
            switch_mime_getc();
             return 1;
         }
@@ -5276,13 +5808,15 @@ nkf_char base64decode(nkf_char c)
     if (c > '@') {
         if (c < '[') {
             i = c - 'A';                        /* A..Z 0-25 */
+       } else if (c == '_') {
+           i = '?'         /* 63 */ ;          /* _  63 */
         } else {
             i = c - 'G'     /* - 'a' + 26 */ ;  /* a..z 26-51 */
        }
     } else if (c > '/') {
         i = c - '0' + '4'   /* - '0' + 52 */ ;  /* 0..9 52-61 */
-    } else if (c == '+') {
-        i = '>'             /* 62 */ ;          /* +  62 */
+    } else if (c == '+' || c == '-') {
+        i = '>'             /* 62 */ ;          /* + and -  62 */
     } else {
         i = '?'             /* 63 */ ;          /* / 63 */
     }
@@ -5296,8 +5830,6 @@ static nkf_char b64c;
 #define MIMEOUT_BUF_LENGTH (60)
 char mimeout_buf[MIMEOUT_BUF_LENGTH+1];
 int mimeout_buf_count = 0;
-int mimeout_preserve_space = 0;
-#define itoh4(c)   (c>=10?c+'A'-10:c+'0')
 
 void open_mime(nkf_char mode)
 {
@@ -5305,42 +5837,37 @@ void open_mime(nkf_char mode)
     int i;
     int j;
     p  = mime_pattern[0];
-    for(i=0;mime_encode[i];i++) {
+    for(i=0;mime_pattern[i];i++) {
        if (mode == mime_encode[i]) {
            p = mime_pattern[i];
            break;
        }
     }
     mimeout_mode = mime_encode_method[i];
-    
     i = 0;
     if (base64_count>45) {
        if (mimeout_buf_count>0 && nkf_isblank(mimeout_buf[i])){
             (*o_mputc)(mimeout_buf[i]);
            i++;
        }
-       (*o_mputc)(NL);
-       (*o_mputc)(SPACE);
+       PUT_NEWLINE((*o_mputc));
+       (*o_mputc)(SP);
        base64_count = 1;
-       if (!mimeout_preserve_space && mimeout_buf_count>0
-           && (mimeout_buf[i]==SPACE || mimeout_buf[i]==TAB
-               || mimeout_buf[i]==CR || mimeout_buf[i]==NL )) {
+       if (mimeout_buf_count>0
+           && (mimeout_buf[i]==SP || mimeout_buf[i]==TAB
+               || mimeout_buf[i]==CR || mimeout_buf[i]==LF)) {
            i++;
        }
     }
-    if (!mimeout_preserve_space) {
-       for (;i<mimeout_buf_count;i++) {
-           if (mimeout_buf[i]==SPACE || mimeout_buf[i]==TAB
-               || mimeout_buf[i]==CR || mimeout_buf[i]==NL ) {
-               (*o_mputc)(mimeout_buf[i]);
-               base64_count ++;
-           } else {
-               break;
-           }
+    for (;i<mimeout_buf_count;i++) {
+       if (mimeout_buf[i]==SP || mimeout_buf[i]==TAB
+           || mimeout_buf[i]==CR || mimeout_buf[i]==LF) {
+           (*o_mputc)(mimeout_buf[i]);
+           base64_count ++;
+       } else {
+           break;
        }
     }
-    mimeout_preserve_space = FALSE;
-    
     while(*p) {
         (*o_mputc)(*p++);
         base64_count ++;
@@ -5378,9 +5905,9 @@ void eof_mime(void)
        base64_count += 2;
        break;
     }
-    if (mimeout_mode) {
+    if (mimeout_mode > 0) {
        if (mimeout_f!=FIXED_MIME) {
-           close_mime(); 
+           close_mime();
        } else if (mimeout_mode != 'Q')
            mimeout_mode = 'B';
     }
@@ -5390,13 +5917,13 @@ void mimeout_addchar(nkf_char c)
 {
     switch(mimeout_mode) {
     case 'Q':
-       if (c==CR||c==NL) {
+       if (c==CR||c==LF) {
            (*o_mputc)(c);
            base64_count = 0;
        } else if(!nkf_isalnum(c)) {
            (*o_mputc)('=');
-           (*o_mputc)(itoh4(((c>>4)&0xf)));
-           (*o_mputc)(itoh4((c&0xf)));
+           (*o_mputc)(bin2hex(((c>>4)&0xf)));
+           (*o_mputc)(bin2hex((c&0xf)));
            base64_count += 3;
        } else {
            (*o_mputc)(c);
@@ -5428,30 +5955,38 @@ void mimeout_addchar(nkf_char c)
     }
 }
 
-nkf_char mime_lastchar2, mime_lastchar1;
+/*nkf_char mime_lastchar2, mime_lastchar1;*/
 
 void mime_prechar(nkf_char c2, nkf_char c1)
 {
-    if (mimeout_mode){
-        if (c2){
-            if (base64_count + mimeout_buf_count/3*4> 66){
+    if (mimeout_mode > 0){
+        if (c2 == EOF){
+            if (base64_count + mimeout_buf_count/3*4> 73){
                 (*o_base64conv)(EOF,0);
-                (*o_base64conv)(0,NL);
-                (*o_base64conv)(0,SPACE);
+                OCONV_NEWLINE((*o_base64conv));
+                (*o_base64conv)(0,SP);
+                base64_count = 1;
             }
-        }/*else if (mime_lastchar2){
-            if (c1 <=DEL && !nkf_isspace(c1)){
-                (*o_base64conv)(0,SPACE);
+        } else {
+            if (base64_count + mimeout_buf_count/3*4> 66) {
+                (*o_base64conv)(EOF,0);
+                OCONV_NEWLINE((*o_base64conv));
+                (*o_base64conv)(0,SP);
+                base64_count = 1;
+                mimeout_mode = -1;
             }
-        }*/
-    }/*else{
-        if (c2 && mime_lastchar2 == 0
-            && mime_lastchar1 && !nkf_isspace(mime_lastchar1)){
-            (*o_base64conv)(0,SPACE);
         }
-    }*/
-    mime_lastchar2 = c2;
-    mime_lastchar1 = c1;
+    } else if (c2) {
+       if (c2 != EOF && base64_count + mimeout_buf_count/3*4> 60) {
+           mimeout_mode =  (output_mode==ASCII ||output_mode == ISO_8859_1) ? 'Q' : 'B';
+           open_mime(output_mode);
+           (*o_base64conv)(EOF,0);
+           OCONV_NEWLINE((*o_base64conv));
+           (*o_base64conv)(0,SP);
+           base64_count = 1;
+           mimeout_mode = -1;
+       }
+    }
 }
 
 void mime_putc(nkf_char c)
@@ -5462,16 +5997,16 @@ void mime_putc(nkf_char c)
     if (mimeout_f == FIXED_MIME){
         if (mimeout_mode == 'Q'){
             if (base64_count > 71){
-                if (c!=CR && c!=NL) {
+                if (c!=CR && c!=LF) {
                     (*o_mputc)('=');
-                    (*o_mputc)(NL);
+                    PUT_NEWLINE((*o_mputc));
                 }
                 base64_count = 0;
             }
         }else{
             if (base64_count > 71){
                 eof_mime();
-                (*o_mputc)(NL);
+                PUT_NEWLINE((*o_mputc));
                 base64_count = 0;
             }
             if (c == EOF) { /* c==EOF */
@@ -5483,23 +6018,31 @@ void mime_putc(nkf_char c)
         }
         return;
     }
-    
+
     /* mimeout_f != FIXED_MIME */
 
     if (c == EOF) { /* c==EOF */
+       if (mimeout_mode == -1 && mimeout_buf_count > 1) open_mime(output_mode);
        j = mimeout_buf_count;
        mimeout_buf_count = 0;
        i = 0;
-       if (mimeout_mode) {
-           for (;i<j;i++) {
-               if (nkf_isspace(mimeout_buf[i]) && base64_count < 71){
-                   break;
+       if (mimeout_mode > 0) {
+           if (!nkf_isblank(mimeout_buf[j-1])) {
+               for (;i<j;i++) {
+                   if (nkf_isspace(mimeout_buf[i]) && base64_count < 71){
+                       break;
+                   }
+                   mimeout_addchar(mimeout_buf[i]);
                }
-               mimeout_addchar(mimeout_buf[i]);
-           }
-           eof_mime();
-           for (;i<j;i++) {
-               mimeout_addchar(mimeout_buf[i]);
+               eof_mime();
+               for (;i<j;i++) {
+                   mimeout_addchar(mimeout_buf[i]);
+               }
+           } else {
+               for (;i<j;i++) {
+                   mimeout_addchar(mimeout_buf[i]);
+               }
+               eof_mime();
            }
        } else {
            for (;i<j;i++) {
@@ -5509,48 +6052,86 @@ void mime_putc(nkf_char c)
         return;
     }
 
-    if (mimeout_mode=='Q') {
-        if (c <= DEL && (output_mode==ASCII ||output_mode == ISO8859_1 ) ) {
-            if (c <= SPACE) {
-                close_mime();
-                (*o_mputc)(SPACE);
-                base64_count++;
-            }
-            (*o_mputc)(c);
-            base64_count++;
-        }
-        return;
-    }
-
     if (mimeout_buf_count > 0){
         lastchar = mimeout_buf[mimeout_buf_count - 1];
     }else{
         lastchar = -1;
     }
 
-    if (!mimeout_mode) {
-        if (c <= DEL && (output_mode==ASCII ||output_mode == ISO8859_1)) {
+    if (mimeout_mode=='Q') {
+        if (c <= DEL && (output_mode==ASCII ||output_mode == ISO_8859_1)) {
+           if (c == CR || c == LF) {
+               close_mime();
+               (*o_mputc)(c);
+               base64_count = 0;
+               return;
+            } else if (c <= SP) {
+                close_mime();
+               if (base64_count > 70) {
+                   PUT_NEWLINE((*o_mputc));
+                   base64_count = 0;
+               }
+               if (!nkf_isblank(c)) {
+                   (*o_mputc)(SP);
+                   base64_count++;
+               }
+            } else {
+               if (base64_count > 70) {
+                   close_mime();
+                   PUT_NEWLINE((*o_mputc));
+                   (*o_mputc)(SP);
+                   base64_count = 1;
+                   open_mime(output_mode);
+               }
+               if (!nkf_noescape_mime(c)) {
+                   mimeout_addchar(c);
+                   return;
+               }
+           }
+           (*o_mputc)(c);
+           base64_count++;
+        }
+        return;
+    }
+
+    if (mimeout_mode <= 0) {
+        if (c <= DEL && (output_mode==ASCII ||output_mode == ISO_8859_1)) {
             if (nkf_isspace(c)) {
-                if (c==CR || c==NL) {
-                    base64_count=0;
+               int flag = 0;
+               if (mimeout_mode == -1) {
+                   flag = 1;
+               }
+                if (c==CR || c==LF) {
+                   if (flag) {
+                       open_mime(output_mode);
+                       output_mode = 0;
+                   } else {
+                       base64_count = 0;
+                   }
                 }
                 for (i=0;i<mimeout_buf_count;i++) {
                     (*o_mputc)(mimeout_buf[i]);
-                    if (mimeout_buf[i] == CR || mimeout_buf[i] == NL){
+                    if (mimeout_buf[i] == CR || mimeout_buf[i] == LF){
                         base64_count = 0;
                     }else{
                         base64_count++;
                     }
                 }
-                mimeout_buf[0] = (char)c;
-                mimeout_buf_count = 1;
+               if (flag) {
+                   eof_mime();
+                   base64_count = 0;
+                   mimeout_mode = 0;
+                }
+               mimeout_buf[0] = (char)c;
+               mimeout_buf_count = 1;
             }else{
                 if (base64_count > 1
-                    && base64_count + mimeout_buf_count > 76){
-                    (*o_mputc)(NL);
+                    && base64_count + mimeout_buf_count > 76
+                   && mimeout_buf[0] != CR && mimeout_buf[0] != LF){
+                    PUT_NEWLINE((*o_mputc));
                     base64_count = 0;
                     if (!nkf_isspace(mimeout_buf[0])){
-                        (*o_mputc)(SPACE);
+                        (*o_mputc)(SP);
                         base64_count++;
                     }
                 }
@@ -5561,33 +6142,33 @@ void mime_putc(nkf_char c)
             }
             return;
         }else{
-            if (lastchar==CR || lastchar == NL){
+            if (lastchar==CR || lastchar == LF){
                 for (i=0;i<mimeout_buf_count;i++) {
                     (*o_mputc)(mimeout_buf[i]);
                 }
                 base64_count = 0;
                 mimeout_buf_count = 0;
             }
-            if (lastchar==SPACE) {
+            if (lastchar==SP) {
                 for (i=0;i<mimeout_buf_count-1;i++) {
                     (*o_mputc)(mimeout_buf[i]);
                     base64_count++;
                 }
-                mimeout_buf[0] = SPACE;
+                mimeout_buf[0] = SP;
                 mimeout_buf_count = 1;
             }
             open_mime(output_mode);
         }
     }else{
         /* mimeout_mode == 'B', 1, 2 */
-        if ( c<=DEL && (output_mode==ASCII ||output_mode == ISO8859_1 ) ) {
-            if (lastchar == CR || lastchar == NL){
+        if ( c<=DEL && (output_mode==ASCII ||output_mode == ISO_8859_1)) {
+            if (lastchar == CR || lastchar == LF){
                 if (nkf_isblank(c)) {
                     for (i=0;i<mimeout_buf_count;i++) {
                         mimeout_addchar(mimeout_buf[i]);
                     }
                     mimeout_buf_count = 0;
-                } else if (SPACE<c && c<DEL) {
+                } else if (SP<c && c<DEL) {
                     eof_mime();
                     for (i=0;i<mimeout_buf_count;i++) {
                         (*o_mputc)(mimeout_buf[i]);
@@ -5595,10 +6176,12 @@ void mime_putc(nkf_char c)
                     base64_count = 0;
                     mimeout_buf_count = 0;
                 }
+                mimeout_buf[mimeout_buf_count++] = (char)c;
+               return;
             }
-            if (c==SPACE || c==TAB || c==CR || c==NL) {
+            if (c==SP || c==TAB || c==CR || c==LF) {
                 for (i=0;i<mimeout_buf_count;i++) {
-                    if (SPACE<mimeout_buf[i] && mimeout_buf[i]<DEL) {
+                    if (SP<mimeout_buf[i] && mimeout_buf[i]<DEL) {
                         eof_mime();
                         for (i=0;i<mimeout_buf_count;i++) {
                             (*o_mputc)(mimeout_buf[i]);
@@ -5617,8 +6200,8 @@ void mime_putc(nkf_char c)
                     mimeout_buf_count = 0;
                 }
                 return;
-            }
-            if (mimeout_buf_count>0 && SPACE<c && c!='=') {
+           }
+            if (mimeout_buf_count>0 && SP<c && c!='=') {
                 mimeout_buf[mimeout_buf_count++] = (char)c;
                 if (mimeout_buf_count>MIMEOUT_BUF_LENGTH) {
                     j = mimeout_buf_count;
@@ -5635,7 +6218,7 @@ void mime_putc(nkf_char c)
        j = mimeout_buf_count;
        mimeout_buf_count = 0;
        for (i=0;i<j;i++) {
-           if (mimeout_buf[i]==CR || mimeout_buf[i]==NL)
+           if (mimeout_buf[i]==CR || mimeout_buf[i]==LF)
                break;
            mimeout_addchar(mimeout_buf[i]);
        }
@@ -5652,7 +6235,6 @@ void mime_putc(nkf_char c)
 }
 
 
-#if defined(PERL_XS) || defined(WIN32DLL)
 void reinit(void)
 {
     {
@@ -5669,17 +6251,13 @@ void reinit(void)
     hira_f = FALSE;
     input_f = FALSE;
     alpha_f = FALSE;
-    mime_f = STRICT_MIME;
+    mime_f = MIME_DECODE_DEFAULT;
     mime_decode_f = FALSE;
     mimebuf_f = FALSE;
     broken_f = FALSE;
     iso8859_f = FALSE;
     mimeout_f = FALSE;
-#if defined(MSDOS) || defined(__OS2__)
-     x0201_f = TRUE;
-#else
-     x0201_f = NO_X0201;
-#endif
+    x0201_f = X0201_DEFAULT;
     iso2022jp_f = FALSE;
 #if defined(UTF8_INPUT_ENABLE) || defined(UTF8_OUTPUT_ENABLE)
     ms_ucs_map_f = UCS_MAP_ASCII;
@@ -5707,9 +6285,7 @@ void reinit(void)
     noout_f = FALSE;
     debug_f = FALSE;
 #endif
-    guess_f = FALSE;
-    is_inputcode_mixed = FALSE;
-    is_inputcode_set   = FALSE;
+    guess_f = 0;
 #ifdef EXEC_IO
     exec_f = 0;
 #endif
@@ -5743,7 +6319,7 @@ void reinit(void)
     oconv = DEFAULT_CONV;
     o_zconv = no_connection;
     o_fconv = no_connection;
-    o_crconv = no_connection;
+    o_nlconv = no_connection;
     o_rot_conv = no_connection;
     o_hira_conv = no_connection;
     o_base64conv = no_connection;
@@ -5763,7 +6339,9 @@ void reinit(void)
     shift_mode =  FALSE;
     mime_decode_mode = FALSE;
     file_out_f = FALSE;
-    crmode_f = 0;
+    nlmode_f = 0;
+    input_newline = 0;
+    prev_cr = 0;
     option_mode = 0;
     broken_counter = 0;
     broken_last = 0;
@@ -5771,12 +6349,12 @@ void reinit(void)
 #ifdef CHECK_OPTION
     iconv_for_check = 0;
 #endif
-    input_codename = "";
+    input_codename = NULL;
+    output_encoding = nkf_enc_from_index(DEFAULT_ENCODING);
 #ifdef WIN32DLL
     reinitdll();
 #endif /*WIN32DLL*/
 }
-#endif
 
 void no_connection(nkf_char c2, nkf_char c1)
 {
@@ -5796,113 +6374,131 @@ nkf_char no_connection2(nkf_char c2, nkf_char c1, nkf_char c0)
 #endif
 void usage(void)
 {
-    fprintf(stderr,"USAGE:  nkf(nkf32,wnkf,nkf2) -[flags] [in file] .. [out file for -O flag]\n");
-    fprintf(stderr,"Flags:\n");
-    fprintf(stderr,"b,u      Output is buffered (DEFAULT),Output is unbuffered\n");
+    fprintf(HELP_OUTPUT,"USAGE:  nkf(nkf32,wnkf,nkf2) -[flags] [in file] .. [out file for -O flag]\n");
+    fprintf(HELP_OUTPUT,"Flags:\n");
+    fprintf(HELP_OUTPUT,"b,u      Output is buffered (DEFAULT),Output is unbuffered\n");
 #ifdef DEFAULT_CODE_SJIS
-    fprintf(stderr,"j,s,e,w  Output code is JIS 7 bit, Shift_JIS (DEFAULT), EUC-JP, UTF-8N\n");
+    fprintf(HELP_OUTPUT,"j,s,e,w  Output code is JIS 7 bit, Shift_JIS (DEFAULT), EUC-JP, UTF-8N\n");
 #endif
 #ifdef DEFAULT_CODE_JIS
-    fprintf(stderr,"j,s,e,w  Output code is JIS 7 bit (DEFAULT), Shift JIS, EUC-JP, UTF-8N\n");
+    fprintf(HELP_OUTPUT,"j,s,e,w  Output code is JIS 7 bit (DEFAULT), Shift JIS, EUC-JP, UTF-8N\n");
 #endif
 #ifdef DEFAULT_CODE_EUC
-    fprintf(stderr,"j,s,e,w  Output code is JIS 7 bit, Shift JIS, EUC-JP (DEFAULT), UTF-8N\n");
+    fprintf(HELP_OUTPUT,"j,s,e,w  Output code is JIS 7 bit, Shift JIS, EUC-JP (DEFAULT), UTF-8N\n");
 #endif
 #ifdef DEFAULT_CODE_UTF8
-    fprintf(stderr,"j,s,e,w  Output code is JIS 7 bit, Shift JIS, EUC-JP, UTF-8N (DEFAULT)\n");
+    fprintf(HELP_OUTPUT,"j,s,e,w  Output code is JIS 7 bit, Shift JIS, EUC-JP, UTF-8N (DEFAULT)\n");
 #endif
 #ifdef UTF8_OUTPUT_ENABLE
-    fprintf(stderr,"         After 'w' you can add more options. -w[ 8 [0], 16 [[BL] [0]] ]\n");
+    fprintf(HELP_OUTPUT,"         After 'w' you can add more options. -w[ 8 [0], 16 [[BL] [0]] ]\n");
 #endif
-    fprintf(stderr,"J,S,E,W  Input assumption is JIS 7 bit , Shift JIS, EUC-JP, UTF-8\n");
+    fprintf(HELP_OUTPUT,"J,S,E,W  Input assumption is JIS 7 bit , Shift JIS, EUC-JP, UTF-8\n");
 #ifdef UTF8_INPUT_ENABLE
-    fprintf(stderr,"         After 'W' you can add more options. -W[ 8, 16 [BL] ] \n");
-#endif
-    fprintf(stderr,"t        no conversion\n");
-    fprintf(stderr,"i[@B]    Specify the Esc Seq for JIS X 0208-1978/83 (DEFAULT B)\n");
-    fprintf(stderr,"o[BJH]   Specify the Esc Seq for ASCII/Roman        (DEFAULT B)\n");
-    fprintf(stderr,"r        {de/en}crypt ROT13/47\n");
-    fprintf(stderr,"h        1 katakana->hiragana, 2 hiragana->katakana, 3 both\n");
-    fprintf(stderr,"v        Show this usage. V: show version\n");
-    fprintf(stderr,"m[BQN0]  MIME decode [B:base64,Q:quoted,N:non-strict,0:no decode]\n");
-    fprintf(stderr,"M[BQ]    MIME encode [B:base64 Q:quoted]\n");
-    fprintf(stderr,"l        ISO8859-1 (Latin-1) support\n");
-    fprintf(stderr,"f/F      Folding: -f60 or -f or -f60-10 (fold margin 10) F preserve nl\n");
-    fprintf(stderr,"Z[0-3]   Convert X0208 alphabet to ASCII\n");
-    fprintf(stderr,"         1: Kankaku to 1 space  2: to 2 spaces  3: Convert to HTML Entity\n");
-    fprintf(stderr,"X,x      Assume X0201 kana in MS-Kanji, -x preserves X0201\n");
-    fprintf(stderr,"B[0-2]   Broken input  0: missing ESC,1: any X on ESC-[($]-X,2: ASCII on NL\n");
+    fprintf(HELP_OUTPUT,"         After 'W' you can add more options. -W[ 8, 16 [BL] ] \n");
+#endif
+    fprintf(HELP_OUTPUT,"t        no conversion\n");
+    fprintf(HELP_OUTPUT,"i[@B]    Specify the Esc Seq for JIS X 0208-1978/83 (DEFAULT B)\n");
+    fprintf(HELP_OUTPUT,"o[BJH]   Specify the Esc Seq for ASCII/Roman        (DEFAULT B)\n");
+    fprintf(HELP_OUTPUT,"r        {de/en}crypt ROT13/47\n");
+    fprintf(HELP_OUTPUT,"h        1 katakana->hiragana, 2 hiragana->katakana, 3 both\n");
+    fprintf(HELP_OUTPUT,"m[BQN0]  MIME decode [B:base64,Q:quoted,N:non-strict,0:no decode]\n");
+    fprintf(HELP_OUTPUT,"M[BQ]    MIME encode [B:base64 Q:quoted]\n");
+    fprintf(HELP_OUTPUT,"l        ISO8859-1 (Latin-1) support\n");
+    fprintf(HELP_OUTPUT,"f/F      Folding: -f60 or -f or -f60-10 (fold margin 10) F preserve nl\n");
+    fprintf(HELP_OUTPUT,"Z[0-4]   Default/0: Convert JISX0208 Alphabet to ASCII\n");
+    fprintf(HELP_OUTPUT,"         1: Kankaku to one space  2: to two spaces  3: HTML Entity\n");
+    fprintf(HELP_OUTPUT,"         4: JISX0208 Katakana to JISX0201 Katakana\n");
+    fprintf(HELP_OUTPUT,"X,x      Assume X0201 kana in MS-Kanji, -x preserves X0201\n");
+    fprintf(HELP_OUTPUT,"B[0-2]   Broken input  0: missing ESC,1: any X on ESC-[($]-X,2: ASCII on NL\n");
 #ifdef MSDOS
-    fprintf(stderr,"T        Text mode output\n");
-#endif
-    fprintf(stderr,"O        Output to File (DEFAULT 'nkf.out')\n");
-    fprintf(stderr,"I        Convert non ISO-2022-JP charactor to GETA\n");
-    fprintf(stderr,"d,c      Convert line breaks  -d: LF  -c: CRLF\n");
-    fprintf(stderr,"-L[uwm]  line mode u:LF w:CRLF m:CR (DEFAULT noconversion)\n");
-    fprintf(stderr,"\n");
-    fprintf(stderr,"Long name options\n");
-    fprintf(stderr," --ic=<input codeset>  --oc=<output codeset>\n");
-    fprintf(stderr,"                   Specify the input or output codeset\n");
-    fprintf(stderr," --fj  --unix --mac  --windows\n");
-    fprintf(stderr," --jis  --euc  --sjis  --utf8  --utf16  --mime  --base64\n");
-    fprintf(stderr,"                   Convert for the system or code\n");
-    fprintf(stderr," --hiragana  --katakana  --katakana-hiragana\n");
-    fprintf(stderr,"                   To Hiragana/Katakana Conversion\n");
-    fprintf(stderr," --prefix=         Insert escape before troublesome characters of Shift_JIS\n");
+    fprintf(HELP_OUTPUT,"T        Text mode output\n");
+#endif
+    fprintf(HELP_OUTPUT,"O        Output to File (DEFAULT 'nkf.out')\n");
+    fprintf(HELP_OUTPUT,"I        Convert non ISO-2022-JP charactor to GETA\n");
+    fprintf(HELP_OUTPUT,"d,c      Convert line breaks  -d: LF  -c: CRLF\n");
+    fprintf(HELP_OUTPUT,"-L[uwm]  line mode u:LF w:CRLF m:CR (DEFAULT noconversion)\n");
+    fprintf(HELP_OUTPUT,"v, V     Show this usage. V: show configuration\n");
+    fprintf(HELP_OUTPUT,"\n");
+    fprintf(HELP_OUTPUT,"Long name options\n");
+    fprintf(HELP_OUTPUT," --ic=<input codeset>  --oc=<output codeset>\n");
+    fprintf(HELP_OUTPUT,"                   Specify the input or output codeset\n");
+    fprintf(HELP_OUTPUT," --fj  --unix --mac  --windows\n");
+    fprintf(HELP_OUTPUT," --jis  --euc  --sjis  --utf8  --utf16  --mime  --base64\n");
+    fprintf(HELP_OUTPUT,"                   Convert for the system or code\n");
+    fprintf(HELP_OUTPUT," --hiragana  --katakana  --katakana-hiragana\n");
+    fprintf(HELP_OUTPUT,"                   To Hiragana/Katakana Conversion\n");
+    fprintf(HELP_OUTPUT," --prefix=         Insert escape before troublesome characters of Shift_JIS\n");
 #ifdef INPUT_OPTION
-    fprintf(stderr," --cap-input, --url-input  Convert hex after ':' or '%%'\n");
+    fprintf(HELP_OUTPUT," --cap-input, --url-input  Convert hex after ':' or '%%'\n");
 #endif
 #ifdef NUMCHAR_OPTION
-    fprintf(stderr," --numchar-input   Convert Unicode Character Reference\n");
+    fprintf(HELP_OUTPUT," --numchar-input   Convert Unicode Character Reference\n");
 #endif
 #ifdef UTF8_INPUT_ENABLE
-    fprintf(stderr," --fb-{skip, html, xml, perl, java, subchar}\n");
-    fprintf(stderr,"                   Specify how nkf handles unassigned characters\n");
+    fprintf(HELP_OUTPUT," --fb-{skip, html, xml, perl, java, subchar}\n");
+    fprintf(HELP_OUTPUT,"                   Specify how nkf handles unassigned characters\n");
 #endif
 #ifdef OVERWRITE
-    fprintf(stderr," --in-place[=SUFFIX]  --overwrite[=SUFFIX]\n");
-    fprintf(stderr,"                   Overwrite original listed files by filtered result\n");
-    fprintf(stderr,"                   --overwrite preserves timestamp of original files\n");
-#endif
-    fprintf(stderr," -g  --guess       Guess the input code\n");
-    fprintf(stderr," --help  --version Show this help/the version\n");
-    fprintf(stderr,"                   For more information, see also man nkf\n");
-    fprintf(stderr,"\n");
+    fprintf(HELP_OUTPUT," --in-place[=SUFFIX]  --overwrite[=SUFFIX]\n");
+    fprintf(HELP_OUTPUT,"                   Overwrite original listed files by filtered result\n");
+    fprintf(HELP_OUTPUT,"                   --overwrite preserves timestamp of original files\n");
+#endif
+    fprintf(HELP_OUTPUT," -g  --guess       Guess the input code\n");
+    fprintf(HELP_OUTPUT," --help  --version Show this help/the version\n");
+    fprintf(HELP_OUTPUT,"                   For more information, see also man nkf\n");
+    fprintf(HELP_OUTPUT,"\n");
     version();
 }
 
-void version(void)
+void show_configuration(void)
 {
-    fprintf(stderr,"Network Kanji Filter Version %s (%s) "
-#if defined(MSDOS) && !defined(__WIN32__) && !defined(__WIN16__) && !defined(__OS2__)
-                  "for DOS"
+    fprintf(HELP_OUTPUT, "Summary of my nkf " NKF_VERSION " (" NKF_RELEASE_DATE ") configuration:\n");
+    fprintf(HELP_OUTPUT, "  Compile-time options:\n");
+    fprintf(HELP_OUTPUT, "    Default output encoding:     "
+#if defined(DEFAULT_CODE_JIS)
+           "ISO-2022-JP"
+#elif defined(DEFAULT_CODE_SJIS)
+           "Shift_JIS"
+#elif defined(DEFAULT_CODE_EUC)
+           "EUC-JP"
+#elif defined(DEFAULT_CODE_UTF8)
+           "UTF-8"
+#endif
+           "\n");
+    fprintf(HELP_OUTPUT, "    Default output newline:      "
+#if DEFAULT_NEWLINE == CR
+           "CR"
+#elif DEFAULT_NEWLINE == CRLF
+           "CRLF"
+#else
+           "LF"
 #endif
-#if defined(MSDOS) && defined(__WIN16__)
-                  "for Win16"
+           "\n");
+    fprintf(HELP_OUTPUT, "    Decode MIME encoded string:  "
+#if MIME_DECODE_DEFAULT
+           "ON"
+#else
+           "OFF"
 #endif
-#if defined(MSDOS) && defined(__WIN32__)
-                  "for Win32"
+           "\n");
+    fprintf(HELP_OUTPUT, "    Convert JIS X 0201 Katakana: "
+#if X0201_DEFAULT
+           "ON"
+#else
+           "OFF"
 #endif
-#ifdef __OS2__
-                  "for OS/2"
+           "\n");
+fprintf(HELP_OUTPUT, " --help, --version output: "
+#if HELP_OUTPUT_HELP_OUTPUT
+"HELP_OUTPUT"
+#else
+"STDOUT"
 #endif
-                  ,NKF_VERSION,NKF_RELEASE_DATE);
-    fprintf(stderr,"\n%s\n",CopyRight);
+"\n");
 }
-#endif /*PERL_XS*/
 
-/**
- ** \e$B%Q%C%A@):n<T\e(B
- **  void@merope.pleiades.or.jp (Kusakabe Youichi)
- **  NIDE Naoyuki <nide@ics.nara-wu.ac.jp>
- **  ohta@src.ricoh.co.jp (Junn Ohta)
- **  inouet@strl.nhk.or.jp (Tomoyuki Inoue)
- **  kiri@pulser.win.or.jp (Tetsuaki Kiriyama)
- **  Kimihiko Sato <sato@sail.t.u-tokyo.ac.jp>
- **  a_kuroe@kuroe.aoba.yokohama.jp (Akihiko Kuroe)
- **  kono@ie.u-ryukyu.ac.jp (Shinji Kono)
- **  GHG00637@nifty-serve.or.jp (COW)
- **
- **/
-
-/* end */
+void version(void)
+{
+    fprintf(HELP_OUTPUT,"Network Kanji Filter Version " NKF_VERSION " (" NKF_RELEASE_DATE ") \n" COPY_RIGHT "\n");
+}
+#endif /*PERL_XS*/