1 /** Network Kanji Filter. (PDS Version)
2 ************************************************************************
3 ** Copyright (C) 1987, Fujitsu LTD. (Itaru ICHIKAWA)
4 **
\e$BO"Mm@h!'
\e(B
\e$B!J3t!KIY;NDL8&5f=j!!%=%U%H#38&!!;T@n!!;j
\e(B
5 **
\e$B!J
\e(BE-Mail Address: ichikawa@flab.fujitsu.co.jp
\e$B!K
\e(B
6 ** Copyright (C) 1996,1998
8 **
\e$BO"Mm@h!'
\e(B
\e$BN05eBg3X>pJs9)3X2J
\e(B
\e$B2OLn
\e(B
\e$B??<#
\e(B mime/X0208 support
9 **
\e$B!J
\e(BE-Mail Address: kono@ie.u-ryukyu.ac.jp
\e$B!K
\e(B
10 **
\e$BO"Mm@h!'
\e(B COW for DOS & Win16 & Win32 & OS/2
11 **
\e$B!J
\e(BE-Mail Address: GHG00637@niftyserve.or.p
\e$B!K
\e(B
13 **
\e$B$3$N%=!<%9$N$$$+$J$kJ#<L!$2~JQ!$=$@5$b5vBz$7$^$9!#$?$@$7!"
\e(B
14 **
\e$B$=$N:]$K$O!"C/$,9W8%$7$?$r<($9$3$NItJ,$r;D$9$3$H!#
\e(B
15 **
\e$B:FG[I[$d;(;o$NIUO?$J$I$NLd$$9g$o$;$bI,MW$"$j$^$;$s!#
\e(B
16 **
\e$B1DMxMxMQ$b>e5-$KH?$7$J$$HO0O$G5v2D$7$^$9!#
\e(B
17 **
\e$B%P%$%J%j$NG[I[$N:]$K$O
\e(Bversion message
\e$B$rJ]B8$9$k$3$H$r>r7o$H$7$^$9!#
\e(B
18 **
\e$B$3$N%W%m%0%i%`$K$D$$$F$OFC$K2?$NJ]>Z$b$7$J$$!"0-$7$+$i$:!#
\e(B
20 ** Everyone is permitted to do anything on this program
21 ** including copying, modifying, improving,
22 ** as long as you don't try to pretend that you wrote it.
23 ** i.e., the above copyright notice has to appear in all copies.
24 ** Binary distribution requires original version messages.
25 ** You don't have to ask before copying, redistribution or publishing.
26 ** THE AUTHOR DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE.
27 ***********************************************************************/
29 /***********************************************************************
30 ** UTF-8
\e$B%5%]!<%H$K$D$$$F
\e(B
31 **
\e$B=>Mh$N
\e(B nkf
\e$B$HF~$l$+$($F$=$N$^$^;H$($k$h$&$K$J$C$F$$$^$9
\e(B
32 ** nkf -e
\e$B$J$I$H$7$F5/F0$9$k$H!"<+F0H=JL$G
\e(B UTF-8
\e$B$HH=Dj$5$l$l$P!"
\e(B
33 **
\e$B$=$N$^$^
\e(B euc-jp
\e$B$KJQ49$5$l$^$9
\e(B
35 **
\e$B$^$@%P%0$,$"$k2DG=@-$,9b$$$G$9!#
\e(B
36 ** (
\e$BFC$K<+F0H=JL!"%3!<%I:.:_!"%(%i!<=hM}7O
\e(B)
38 **
\e$B2?$+LdBj$r8+$D$1$?$i!"
\e(B
39 ** E-Mail: furukawa@tcp-ip.or.jp
40 **
\e$B$^$G8fO"Mm$r$*4j$$$7$^$9!#
\e(B
41 ***********************************************************************/
44 static char *CopyRight =
45 "Copyright (C) 1987, FUJITSU LTD. (I.Ichikawa),2000 S. Kono, COW, 2002-2003 Kono, Furukawa";
46 static char *Version =
48 static char *Patchlevel =
55 ** USAGE: nkf [flags] [file]
58 ** b Output is bufferred (DEFAULT)
59 ** u Output is unbufferred
63 ** j Outout code is JIS 7 bit (DEFAULT SELECT)
64 ** s Output code is MS Kanji (DEFAULT SELECT)
65 ** e Output code is AT&T JIS (DEFAULT SELECT)
66 ** w Output code is AT&T JIS (DEFAULT SELECT)
67 ** l Output code is JIS 7bit and ISO8859-1 Latin-1
69 ** m MIME conversion for ISO-2022-JP
70 ** I Convert non ISO-2022-JP charactor to GETA by Pekoe <pekoe@lair.net>
71 ** i_ Output sequence to designate JIS-kanji (DEFAULT_J)
72 ** o_ Output sequence to designate single-byte roman characters (DEFAULT_R)
73 ** M MIME output conversion
75 ** r {de/en}crypt ROT13/47
79 ** T Text mode output (for MS-DOS)
81 ** x Do not convert X0201 kana into X0208
82 ** Z Convert X0208 alphabet to ASCII
87 ** B try to fix broken JIS, missing Escape
88 ** B[1-9] broken level
90 ** O Output to 'nkf.out' file or last file name
91 ** d Delete \r in line feed
92 ** c Add \r in line feed
93 ** -- other long option
94 ** -- ignore following option (don't use with -O )
98 #if (defined(__TURBOC__) || defined(_MSC_VER) || defined(LSI_C) || defined(__MINGW32__)) && !defined(MSDOS)
100 #if (defined(__Win32__) || defined(_WIN32)) && !defined(__WIN32__)
116 #if defined(MSDOS) || defined(__OS2__)
123 #define setbinmode(fp) fsetbin(fp)
124 #else /* Microsoft C, Turbo C */
125 #define setbinmode(fp) setmode(fileno(fp), O_BINARY)
127 #else /* UNIX,OS/2 */
128 #define setbinmode(fp)
131 #ifdef _IOFBF /* SysV and MSDOS, Windows */
132 #define setvbuffer(fp, buf, size) setvbuf(fp, buf, _IOFBF, size)
134 #define setvbuffer(fp, buf, size) setbuffer(fp, buf, size)
137 /*Borland C++ 4.5 EasyWin*/
138 #if defined(__TURBOC__) && defined(_Windows) && !defined(__WIN32__) /*Easy Win */
147 /* added by satoru@isoternet.org */
148 #include <sys/stat.h>
149 #ifndef MSDOS /* UNIX, OS/2 */
153 #if defined(_MSC_VER) || defined(__MINGW32__) /* VC++, MinGW */
154 #include <sys/utime.h>
155 #elif defined(__TURBOC__) /* BCC */
157 #elif defined(LSI_C) /* LSI C */
169 /* state of output_mode and input_mode
186 /* Input Assumption */
190 #define LATIN1_INPUT 6
192 #define STRICT_MIME 8
197 #define JAPANESE_EUC 10
201 #define UTF8_INPUT 13
202 #define UTF16_INPUT 14
203 #define UTF16BE_INPUT 15
221 #define is_alnum(c) \
222 (('a'<=c && c<='z')||('A'<= c && c<='Z')||('0'<=c && c<='9'))
224 #define HOLD_SIZE 1024
225 #define IOBUF_SIZE 16384
227 #define DEFAULT_J 'B'
228 #define DEFAULT_R 'B'
230 #define SJ0162 0x00e1 /* 01 - 62 ku offset */
231 #define SJ6394 0x0161 /* 63 - 94 ku offset */
233 #define RANGE_NUM_MAX 18
238 #if defined( UTF8_OUTPUT_ENABLE ) || defined( UTF8_INPUT_ENABLE )
239 #define sizeof_euc_utf8 94
240 #define sizeof_euc_to_utf8_1byte 94
241 #define sizeof_euc_to_utf8_2bytes 94
242 #define sizeof_utf8_to_euc_C2 64
243 #define sizeof_utf8_to_euc_E5B8 64
244 #define sizeof_utf8_to_euc_2bytes 112
245 #define sizeof_utf8_to_euc_3bytes 112
248 /* MIME preprocessor */
251 #ifdef EASYWIN /*Easy Win */
252 extern POINT _BufferSize;
255 /* function prototype */
257 #ifdef ANSI_C_PROTOTYPE
259 #define STATIC static
271 void (*status_func)PROTO((struct input_code *, int));
272 int (*iconv_func)PROTO((int c2, int c1, int c0));
276 STATIC char *input_codename = "";
278 STATIC int noconvert PROTO((FILE *f));
279 STATIC int kanji_convert PROTO((FILE *f));
280 STATIC int h_conv PROTO((FILE *f,int c2,int c1));
281 STATIC int push_hold_buf PROTO((int c2));
282 STATIC void set_iconv PROTO((int f, int (*iconv_func)()));
283 STATIC int s_iconv PROTO((int c2,int c1,int c0));
284 STATIC int s2e_conv PROTO((int c2, int c1, int *p2, int *p1));
285 STATIC int e_iconv PROTO((int c2,int c1,int c0));
286 #ifdef UTF8_INPUT_ENABLE
287 STATIC int w2e_conv PROTO((int c2,int c1,int c0,int *p2,int *p1));
288 STATIC int w_iconv PROTO((int c2,int c1,int c0));
289 STATIC int w_iconv16 PROTO((int c2,int c1,int c0));
290 STATIC int w_iconv_common PROTO((int c1,int c0,unsigned short **pp,int psize,int *p2,int *p1));
291 STATIC int ww16_conv PROTO((int c2, int c1, int c0));
293 #ifdef UTF8_OUTPUT_ENABLE
294 STATIC int e2w_conv PROTO((int c2,int c1));
295 STATIC void w_oconv PROTO((int c2,int c1));
296 STATIC void w_oconv16 PROTO((int c2,int c1));
298 STATIC void e_oconv PROTO((int c2,int c1));
299 STATIC void e2s_conv PROTO((int c2, int c1, int *p2, int *p1));
300 STATIC void s_oconv PROTO((int c2,int c1));
301 STATIC void j_oconv PROTO((int c2,int c1));
302 STATIC void fold_conv PROTO((int c2,int c1));
303 STATIC void cr_conv PROTO((int c2,int c1));
304 STATIC void z_conv PROTO((int c2,int c1));
305 STATIC void rot_conv PROTO((int c2,int c1));
306 STATIC void hira_conv PROTO((int c2,int c1));
307 STATIC void base64_conv PROTO((int c2,int c1));
308 STATIC void iso2022jp_check_conv PROTO((int c2,int c1));
309 STATIC void no_connection PROTO((int c2,int c1));
310 STATIC int no_connection2 PROTO((int c2,int c1,int c0));
312 STATIC void code_score PROTO((struct input_code *ptr));
313 STATIC void code_status PROTO((int c));
315 STATIC void std_putc PROTO((int c));
316 STATIC int std_getc PROTO((FILE *f));
317 STATIC int std_ungetc PROTO((int c,FILE *f));
319 STATIC int broken_getc PROTO((FILE *f));
320 STATIC int broken_ungetc PROTO((int c,FILE *f));
322 STATIC int mime_begin PROTO((FILE *f));
323 STATIC int mime_getc PROTO((FILE *f));
324 STATIC int mime_ungetc PROTO((int c,FILE *f));
326 STATIC int mime_begin_strict PROTO((FILE *f));
327 STATIC int mime_getc_buf PROTO((FILE *f));
328 STATIC int mime_ungetc_buf PROTO((int c,FILE *f));
329 STATIC int mime_integrity PROTO((FILE *f,unsigned char *p));
331 STATIC int base64decode PROTO((int c));
332 STATIC void mime_putc PROTO((int c));
333 STATIC void open_mime PROTO((int c));
334 STATIC void close_mime PROTO(());
335 STATIC void usage PROTO(());
336 STATIC void version PROTO(());
337 STATIC void options PROTO((unsigned char *c));
339 STATIC void reinit PROTO(());
344 static unsigned char stdibuf[IOBUF_SIZE];
345 static unsigned char stdobuf[IOBUF_SIZE];
346 static unsigned char hold_buf[HOLD_SIZE*2];
347 static int hold_count;
349 /* MIME preprocessor fifo */
351 #define MIME_BUF_SIZE (1024) /* 2^n ring buffer */
352 #define MIME_BUF_MASK (MIME_BUF_SIZE-1)
353 #define Fifo(n) mime_buf[(n)&MIME_BUF_MASK]
354 static unsigned char mime_buf[MIME_BUF_SIZE];
355 static unsigned int mime_top = 0;
356 static unsigned int mime_last = 0; /* decoded */
357 static unsigned int mime_input = 0; /* undecoded */
360 static int unbuf_f = FALSE;
361 static int estab_f = FALSE;
362 static int nop_f = FALSE;
363 static int binmode_f = TRUE; /* binary mode */
364 static int rot_f = FALSE; /* rot14/43 mode */
365 static int hira_f = FALSE; /* hira/kata henkan */
366 static int input_f = FALSE; /* non fixed input code */
367 static int alpha_f = FALSE; /* convert JIx0208 alphbet to ASCII */
368 static int mime_f = STRICT_MIME; /* convert MIME B base64 or Q */
369 static int mimebuf_f = FALSE; /* MIME buffered input */
370 static int broken_f = FALSE; /* convert ESC-less broken JIS */
371 static int iso8859_f = FALSE; /* ISO8859 through */
372 static int mimeout_f = FALSE; /* base64 mode */
373 #if defined(MSDOS) || defined(__OS2__)
374 static int x0201_f = TRUE; /* Assume JISX0201 kana */
376 static int x0201_f = NO_X0201; /* Assume NO JISX0201 */
378 static int iso2022jp_f = FALSE; /* convert ISO-2022-JP */
379 #ifdef UTF8_OUTPUT_ENABLE
380 static int w_oconv16_begin_f= 0; /* utf-16 header */
381 static int w_oconv16_LE = 0; /* utf-16 little endian */
385 #ifdef NUMCHAR_OPTION
387 #define CLASS_MASK 0x0f000000
388 #define CLASS_UTF16 0x01000000
392 static int cap_f = FALSE;
393 static int (*i_cgetc)PROTO((FILE *)) = std_getc; /* input of cgetc */
394 static int (*i_cungetc)PROTO((int c ,FILE *f)) = std_ungetc;
395 STATIC int cap_getc PROTO((FILE *f));
396 STATIC int cap_ungetc PROTO((int c,FILE *f));
398 static int url_f = FALSE;
399 static int (*i_ugetc)PROTO((FILE *)) = std_getc; /* input of ugetc */
400 static int (*i_uungetc)PROTO((int c ,FILE *f)) = std_ungetc;
401 STATIC int url_getc PROTO((FILE *f));
402 STATIC int url_ungetc PROTO((int c,FILE *f));
404 static int numchar_f = FALSE;
405 static int (*i_ngetc)PROTO((FILE *)) = std_getc; /* input of ugetc */
406 static int (*i_nungetc)PROTO((int c ,FILE *f)) = std_ungetc;
407 STATIC int numchar_getc PROTO((FILE *f));
408 STATIC int numchar_ungetc PROTO((int c,FILE *f));
412 static int noout_f = FALSE;
413 STATIC void no_putc PROTO((int c));
414 static int debug_f = FALSE;
415 STATIC void debug PROTO((char *str));
419 static int exec_f = 0;
422 #ifdef SHIFTJIS_CP932
423 STATIC int cp932_f = TRUE;
424 #define CP932_TABLE_BEGIN (0xfa)
425 #define CP932_TABLE_END (0xfc)
427 #endif /* SHIFTJIS_CP932 */
429 STATIC void e_status PROTO((struct input_code *, int));
430 STATIC void s_status PROTO((struct input_code *, int));
432 #ifdef UTF8_INPUT_ENABLE
433 STATIC void w_status PROTO((struct input_code *, int));
434 STATIC void w16_status PROTO((struct input_code *, int));
435 static int utf16_mode = UTF16_INPUT;
438 struct input_code input_code_list[] = {
439 {"EUC-JP", 0, 0, 0, {0, 0, 0}, e_status, e_iconv, 0},
440 {"Shift_JIS", 0, 0, 0, {0, 0, 0}, s_status, s_iconv, 0},
441 {"UTF-8", 0, 0, 0, {0, 0, 0}, w_status, w_iconv, 0},
442 {"UTF-16", 0, 0, 0, {0, 0, 0}, w16_status, w_iconv16, 0},
446 static int mimeout_mode = 0;
447 static int base64_count = 0;
449 /* X0208 -> ASCII converter */
452 static int f_line = 0; /* chars in line */
453 static int f_prev = 0;
454 static int fold_preserve_f = FALSE; /* preserve new lines */
455 static int fold_f = FALSE;
456 static int fold_len = 0;
459 static unsigned char kanji_intro = DEFAULT_J,
460 ascii_intro = DEFAULT_R;
464 #define FOLD_MARGIN 10
465 #define DEFAULT_FOLD 60
467 static int fold_margin = FOLD_MARGIN;
471 #ifdef DEFAULT_CODE_JIS
472 # define DEFAULT_CONV j_oconv
474 #ifdef DEFAULT_CODE_SJIS
475 # define DEFAULT_CONV s_oconv
477 #ifdef DEFAULT_CODE_EUC
478 # define DEFAULT_CONV e_oconv
480 #ifdef DEFAULT_CODE_UTF8
481 # define DEFAULT_CONV w_oconv
484 /* process default */
485 static void (*output_conv)PROTO((int c2,int c1)) = DEFAULT_CONV;
487 static void (*oconv)PROTO((int c2,int c1)) = no_connection;
488 /* s_iconv or oconv */
489 static int (*iconv)PROTO((int c2,int c1,int c0)) = no_connection2;
491 static void (*o_zconv)PROTO((int c2,int c1)) = no_connection;
492 static void (*o_fconv)PROTO((int c2,int c1)) = no_connection;
493 static void (*o_crconv)PROTO((int c2,int c1)) = no_connection;
494 static void (*o_rot_conv)PROTO((int c2,int c1)) = no_connection;
495 static void (*o_hira_conv)PROTO((int c2,int c1)) = no_connection;
496 static void (*o_base64conv)PROTO((int c2,int c1)) = no_connection;
497 static void (*o_iso2022jp_check_conv)PROTO((int c2,int c1)) = no_connection;
499 /* static redirections */
501 static void (*o_putc)PROTO((int c)) = std_putc;
503 static int (*i_getc)PROTO((FILE *f)) = std_getc; /* general input */
504 static int (*i_ungetc)PROTO((int c,FILE *f)) =std_ungetc;
506 static int (*i_bgetc)PROTO((FILE *)) = std_getc; /* input of mgetc */
507 static int (*i_bungetc)PROTO((int c ,FILE *f)) = std_ungetc;
509 static void (*o_mputc)PROTO((int c)) = std_putc ; /* output of mputc */
511 static int (*i_mgetc)PROTO((FILE *)) = std_getc; /* input of mgetc */
512 static int (*i_mungetc)PROTO((int c ,FILE *f)) = std_ungetc;
514 /* for strict mime */
515 static int (*i_mgetc_buf)PROTO((FILE *)) = std_getc; /* input of mgetc_buf */
516 static int (*i_mungetc_buf)PROTO((int c,FILE *f)) = std_ungetc;
519 static int output_mode = ASCII, /* output kanji mode */
520 input_mode = ASCII, /* input kanji mode */
521 shift_mode = FALSE; /* TRUE shift out, or X0201 */
522 static int mime_decode_mode = FALSE; /* MIME mode B base64, Q hex */
524 /* X0201 / X0208 conversion tables */
526 /* X0201 kana conversion table */
529 unsigned char cv[]= {
530 0x21,0x21,0x21,0x23,0x21,0x56,0x21,0x57,
531 0x21,0x22,0x21,0x26,0x25,0x72,0x25,0x21,
532 0x25,0x23,0x25,0x25,0x25,0x27,0x25,0x29,
533 0x25,0x63,0x25,0x65,0x25,0x67,0x25,0x43,
534 0x21,0x3c,0x25,0x22,0x25,0x24,0x25,0x26,
535 0x25,0x28,0x25,0x2a,0x25,0x2b,0x25,0x2d,
536 0x25,0x2f,0x25,0x31,0x25,0x33,0x25,0x35,
537 0x25,0x37,0x25,0x39,0x25,0x3b,0x25,0x3d,
538 0x25,0x3f,0x25,0x41,0x25,0x44,0x25,0x46,
539 0x25,0x48,0x25,0x4a,0x25,0x4b,0x25,0x4c,
540 0x25,0x4d,0x25,0x4e,0x25,0x4f,0x25,0x52,
541 0x25,0x55,0x25,0x58,0x25,0x5b,0x25,0x5e,
542 0x25,0x5f,0x25,0x60,0x25,0x61,0x25,0x62,
543 0x25,0x64,0x25,0x66,0x25,0x68,0x25,0x69,
544 0x25,0x6a,0x25,0x6b,0x25,0x6c,0x25,0x6d,
545 0x25,0x6f,0x25,0x73,0x21,0x2b,0x21,0x2c,
549 /* X0201 kana conversion table for daguten */
552 unsigned char dv[]= {
553 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
554 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
555 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
556 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
557 0x00,0x00,0x00,0x00,0x00,0x00,0x25,0x74,
558 0x00,0x00,0x00,0x00,0x25,0x2c,0x25,0x2e,
559 0x25,0x30,0x25,0x32,0x25,0x34,0x25,0x36,
560 0x25,0x38,0x25,0x3a,0x25,0x3c,0x25,0x3e,
561 0x25,0x40,0x25,0x42,0x25,0x45,0x25,0x47,
562 0x25,0x49,0x00,0x00,0x00,0x00,0x00,0x00,
563 0x00,0x00,0x00,0x00,0x25,0x50,0x25,0x53,
564 0x25,0x56,0x25,0x59,0x25,0x5c,0x00,0x00,
565 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
566 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
567 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
568 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
571 /* X0201 kana conversion table for han-daguten */
574 unsigned char ev[]= {
575 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
576 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
577 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
578 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
579 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
580 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
581 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
582 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
583 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
584 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
585 0x00,0x00,0x00,0x00,0x25,0x51,0x25,0x54,
586 0x25,0x57,0x25,0x5a,0x25,0x5d,0x00,0x00,
587 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
588 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
589 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
590 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
594 /* X0208 kigou conversion table */
595 /* 0x8140 - 0x819e */
597 unsigned char fv[] = {
599 0x00,0x00,0x00,0x00,0x2c,0x2e,0x00,0x3a,
600 0x3b,0x3f,0x21,0x00,0x00,0x27,0x60,0x00,
601 0x5e,0x00,0x5f,0x00,0x00,0x00,0x00,0x00,
602 0x00,0x00,0x00,0x00,0x00,0x2d,0x00,0x2f,
603 0x5c,0x00,0x00,0x7c,0x00,0x00,0x60,0x27,
604 0x22,0x22,0x28,0x29,0x00,0x00,0x5b,0x5d,
605 0x7b,0x7d,0x3c,0x3e,0x00,0x00,0x00,0x00,
606 0x00,0x00,0x00,0x00,0x2b,0x2d,0x00,0x00,
607 0x00,0x3d,0x00,0x3c,0x3e,0x00,0x00,0x00,
608 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
609 0x24,0x00,0x00,0x25,0x23,0x26,0x2a,0x40,
610 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00
616 static int file_out = FALSE;
618 static int overwrite = FALSE;
621 static int crmode_f = 0; /* CR, NL, CRLF */
622 #ifdef EASYWIN /*Easy Win */
623 static int end_check;
635 #ifdef EASYWIN /*Easy Win */
636 _BufferSize.y = 400;/*Set Scroll Buffer Size*/
639 for (argc--,argv++; (argc > 0) && **argv == '-'; argc--, argv++) {
640 cp = (unsigned char *)*argv;
645 if (pipe(fds) < 0 || (pid = fork()) < 0){
656 execvp(argv[1], &argv[1]);
670 if(x0201_f == WISH_TRUE)
671 x0201_f = ((!iso2022jp_f)? TRUE : NO_X0201);
673 if (binmode_f == TRUE)
675 if (freopen("","wb",stdout) == NULL)
682 setbuf(stdout, (char *) NULL);
684 setvbuffer(stdout, stdobuf, IOBUF_SIZE);
687 if (binmode_f == TRUE)
689 if (freopen("","rb",stdin) == NULL) return (-1);
693 setvbuffer(stdin, stdibuf, IOBUF_SIZE);
697 kanji_convert(stdin);
703 if ((fin = fopen((origfname = *argv++), "r")) == NULL) {
712 /* reopen file for stdout */
713 if (file_out == TRUE) {
716 outfname = malloc(strlen(origfname)
717 + strlen(".nkftmpXXXXXX")
723 strcpy(outfname, origfname);
727 for (i = strlen(outfname); i; --i){
728 if (outfname[i - 1] == '/'
729 || outfname[i - 1] == '\\'){
735 strcat(outfname, "ntXXXXXX");
737 fd = open(outfname, O_WRONLY | O_CREAT | O_TRUNC,
740 strcat(outfname, ".nkftmpXXXXXX");
741 fd = mkstemp(outfname);
744 || (fd_backup = dup(fileno(stdout))) < 0
745 || dup2(fd, fileno(stdout)) < 0
756 outfname = "nkf.out";
759 if(freopen(outfname, "w", stdout) == NULL) {
763 if (binmode_f == TRUE) {
765 if (freopen("","wb",stdout) == NULL)
772 if (binmode_f == TRUE)
774 if (freopen("","rb",fin) == NULL)
779 setvbuffer(fin, stdibuf, IOBUF_SIZE);
788 #if defined(MSDOS) && !defined(__MINGW32__)
796 if (dup2(fd_backup, fileno(stdout)) < 0){
799 if (stat(origfname, &sb)) {
800 fprintf(stderr, "Can't stat %s\n", origfname);
802 /*
\e$B%Q!<%_%C%7%g%s$rI|85
\e(B */
803 if (chmod(outfname, sb.st_mode)) {
804 fprintf(stderr, "Can't set permission %s\n", outfname);
807 /*
\e$B%?%$%`%9%?%s%W$rI|85
\e(B */
808 #if defined(MSDOS) && !defined(__MINGW32__)
809 tb[0] = tb[1] = sb.st_mtime;
810 if (utime(outfname, tb)) {
811 fprintf(stderr, "Can't set timestamp %s\n", outfname);
814 tb.actime = sb.st_atime;
815 tb.modtime = sb.st_mtime;
816 if (utime(outfname, &tb)) {
817 fprintf(stderr, "Can't set timestamp %s\n", outfname);
821 if (unlink(origfname)){
825 if (rename(outfname, origfname)) {
827 fprintf(stderr, "Can't rename %s to %s\n",
828 outfname, origfname);
836 #ifdef EASYWIN /*Easy Win */
837 if (file_out == FALSE)
838 scanf("%d",&end_check);
841 #else /* for Other OS */
842 if (file_out == TRUE)
872 {"katakana-hiragana","h3"},
873 #ifdef UTF8_OUTPUT_ENABLE
877 #ifdef UTF8_INPUT_ENABLE
879 {"utf16-input", "W16"},
888 #ifdef NUMCHAR_OPTION
889 {"numchar-input", ""},
895 #ifdef SHIFTJIS_CP932
904 static int option_mode;
919 case '-': /* literal options */
920 if (!*cp) { /* ignore the rest of arguments */
924 for (i=0;i<sizeof(long_option)/sizeof(long_option[0]);i++) {
926 p = (unsigned char *)long_option[i].name;
927 for (j=0;*p && *p++ == cp[j];j++);
928 if (! *p && !cp[j]) break;
931 cp = (unsigned char *)long_option[i].alias;
934 if (strcmp(long_option[i].name, "overwrite") == 0){
941 if (strcmp(long_option[i].name, "cap-input") == 0){
945 if (strcmp(long_option[i].name, "url-input") == 0){
950 #ifdef NUMCHAR_OPTION
951 if (strcmp(long_option[i].name, "numchar-input") == 0){
957 if (strcmp(long_option[i].name, "no-output") == 0){
961 if (strcmp(long_option[i].name, "debug") == 0){
966 #ifdef SHIFTJIS_CP932
967 if (strcmp(long_option[i].name, "no-cp932") == 0){
973 if (strcmp(long_option[i].name, "exec-in") == 0){
977 if (strcmp(long_option[i].name, "exec-out") == 0){
984 case 'b': /* buffered mode */
987 case 'u': /* non bufferd mode */
990 case 't': /* transparent mode */
993 case 'j': /* JIS output */
995 output_conv = j_oconv;
997 case 'e': /* AT&T EUC output */
998 output_conv = e_oconv;
1000 case 's': /* SJIS output */
1001 output_conv = s_oconv;
1003 case 'l': /* ISO8859 Latin-1 support, no conversion */
1004 iso8859_f = TRUE; /* Only compatible with ISO-2022-JP */
1005 input_f = LATIN1_INPUT;
1007 case 'i': /* Kanji IN ESC-$-@/B */
1008 if (*cp=='@'||*cp=='B')
1009 kanji_intro = *cp++;
1011 case 'o': /* ASCII IN ESC-(-J/B */
1012 if (*cp=='J'||*cp=='B'||*cp=='H')
1013 ascii_intro = *cp++;
1020 if ('9'>= *cp && *cp>='0')
1021 hira_f |= (*cp++ -'0');
1028 #if defined(MSDOS) || defined(__OS2__)
1043 #ifdef UTF8_OUTPUT_ENABLE
1044 case 'w': /* UTF-8 output */
1045 if ('1'== cp[0] && '6'==cp[1]) {
1046 output_conv = w_oconv16; cp+=2;
1048 w_oconv16_begin_f=2; cp++;
1051 w_oconv16_begin_f=1; cp++;
1053 } else if (cp[0] == 'B') {
1054 w_oconv16_begin_f=2; cp++;
1056 w_oconv16_begin_f=1; cp++;
1060 output_conv = w_oconv;
1063 #ifdef UTF8_INPUT_ENABLE
1064 case 'W': /* UTF-8 input */
1065 if ('1'== cp[0] && '6'==cp[1]) {
1066 input_f = UTF16_INPUT;
1068 input_f = UTF8_INPUT;
1071 /* Input code assumption */
1072 case 'J': /* JIS input */
1073 case 'E': /* AT&T EUC input */
1074 input_f = JIS_INPUT;
1076 case 'S': /* MS Kanji input */
1077 input_f = SJIS_INPUT;
1078 if (x0201_f==NO_X0201) x0201_f=TRUE;
1080 case 'Z': /* Convert X0208 alphabet to asii */
1081 /* bit:0 Convert X0208
1082 bit:1 Convert Kankaku to one space
1083 bit:2 Convert Kankaku to two spaces
1084 bit:3 Convert HTML Entity
1086 if ('9'>= *cp && *cp>='0')
1087 alpha_f |= 1<<(*cp++ -'0');
1091 case 'x': /* Convert X0201 kana to X0208 or X0201 Conversion */
1092 x0201_f = FALSE; /* No X0201->X0208 conversion */
1094 ESC-(-I in JIS, EUC, MS Kanji
1095 SI/SO in JIS, EUC, MS Kanji
1096 SSO in EUC, JIS, not in MS Kanji
1097 MS Kanji (0xa0-0xdf)
1099 ESC-(-I in JIS (0x20-0x5f)
1100 SSO in EUC (0xa0-0xdf)
1101 0xa0-0xd in MS Kanji (0xa0-0xdf)
1104 case 'X': /* Assume X0201 kana */
1105 /* Default value is NO_X0201 for EUC/MS-Kanji mix */
1108 case 'F': /* prserve new lines */
1109 fold_preserve_f = TRUE;
1110 case 'f': /* folding -f60 or -f */
1113 while('0'<= *cp && *cp <='9') { /* we don't use atoi here */
1115 fold_len += *cp++ - '0';
1117 if (!(0<fold_len && fold_len<BUFSIZ))
1118 fold_len = DEFAULT_FOLD;
1122 while('0'<= *cp && *cp <='9') { /* we don't use atoi here */
1124 fold_margin += *cp++ - '0';
1128 case 'm': /* MIME support */
1129 if (*cp=='B'||*cp=='Q') {
1130 mime_decode_mode = *cp++;
1131 mimebuf_f = FIXED_MIME;
1132 } else if (*cp=='N') {
1133 mime_f = TRUE; cp++;
1134 } else if (*cp=='S') {
1135 mime_f = STRICT_MIME; cp++;
1136 } else if (*cp=='0') {
1137 mime_f = FALSE; cp++;
1140 case 'M': /* MIME output */
1143 mimeout_f = FIXED_MIME; cp++;
1144 } else if (*cp=='Q') {
1146 mimeout_f = FIXED_MIME; cp++;
1151 case 'B': /* Broken JIS support */
1153 bit:1 allow any x on ESC-(-x or ESC-$-x
1154 bit:2 reset to ascii on NL
1156 if ('9'>= *cp && *cp>='0')
1157 broken_f |= 1<<(*cp++ -'0');
1162 case 'O':/* for Output file */
1166 case 'c':/* add cr code */
1169 case 'd':/* delete cr code */
1172 case 'I': /* ISO-2022-JP output */
1175 case 'L': /* line mode */
1176 if (*cp=='u') { /* unix */
1177 crmode_f = NL; cp++;
1178 } else if (*cp=='m') { /* mac */
1179 crmode_f = CR; cp++;
1180 } else if (*cp=='w') { /* windows */
1181 crmode_f = CRLF; cp++;
1182 } else if (*cp=='0') { /* no conversion */
1187 /* module muliple options in a string are allowed for Perl moudle */
1188 while(*cp && *cp!='-') cp++;
1192 /* bogus option but ignored */
1198 #ifdef ANSI_C_PROTOTYPE
1199 void set_iconv(int f, int (*iconv_func)(int c2,int c1,int c0))
1201 void set_iconv(f, iconv_func)
1203 int (*iconv_func)();
1207 static int (*iconv_for_check)() = 0;
1209 #ifdef INPUT_CODE_FIX
1217 #ifdef INPUT_CODE_FIX
1218 && (f == -TRUE || !input_f) /* -TRUE means "FORCE" */
1224 if (estab_f && iconv_for_check != iconv){
1225 #ifdef UTF8_INPUT_ENABLE
1226 if (iconv == w_iconv) debug(input_codename = "UTF-8");
1227 if (iconv == w_iconv16) debug(input_codename = "UTF-16");
1229 if (iconv == s_iconv) debug(input_codename = "Shift_JIS");
1230 if (iconv == e_iconv) debug(input_codename = "EUC-JP");
1231 iconv_for_check = iconv;
1236 #define SCORE_KANA (1) /*
\e$B$$$o$f$kH>3Q%+%J
\e(B */
1237 #define SCORE_DEPEND (SCORE_KANA << 1) /*
\e$B5!<o0MB8J8;z
\e(B */
1238 #ifdef SHIFTJIS_CP932
1239 #define SCORE_CP932 (SCORE_DEPEND << 1) /* CP932
\e$B$K$h$kFI$_49$(
\e(B */
1240 #define SCORE_NO_EXIST (SCORE_CP932 << 1) /*
\e$BB8:_$7$J$$J8;z
\e(B */
1242 #define SCORE_NO_EXIST (SCORE_DEPEND << 1) /*
\e$BB8:_$7$J$$J8;z
\e(B */
1244 #define SCORE_ERROR (SCORE_NO_EXIST << 1) /*
\e$B%(%i!<
\e(B */
1245 int score_table_A0[] = {
1248 0, SCORE_DEPEND, SCORE_DEPEND, SCORE_DEPEND,
1249 SCORE_DEPEND, SCORE_DEPEND, SCORE_DEPEND, SCORE_NO_EXIST,
1252 int score_table_F0[] = {
1254 0, SCORE_DEPEND, SCORE_NO_EXIST, SCORE_NO_EXIST,
1255 SCORE_DEPEND, SCORE_DEPEND, SCORE_DEPEND, SCORE_DEPEND,
1256 SCORE_DEPEND, SCORE_NO_EXIST, SCORE_NO_EXIST, SCORE_ERROR,
1259 void set_code_score(ptr, score)
1260 struct input_code *ptr;
1263 ptr->score |= score;
1266 void code_score(ptr)
1267 struct input_code *ptr;
1269 int c2 = ptr->buf[0];
1270 int c1 = ptr->buf[1];
1272 set_code_score(ptr, SCORE_ERROR);
1273 }else if ((c2 & 0xf0) == 0xa0){
1274 set_code_score(ptr, score_table_A0[c2 & 0x0f]);
1275 }else if ((c2 & 0xf0) == 0xf0){
1276 set_code_score(ptr, score_table_F0[c2 & 0x0f]);
1277 }else if (c2 == SSO){
1278 set_code_score(ptr, SCORE_KANA);
1280 #ifdef UTF8_OUTPUT_ENABLE
1281 else if (!e2w_conv(c2, c1)){
1282 set_code_score(ptr, SCORE_NO_EXIST);
1287 void status_disable(ptr)
1288 struct input_code *ptr;
1293 if (iconv == ptr->iconv_func) set_iconv(FALSE, 0);
1296 void status_push_ch(ptr, c)
1297 struct input_code *ptr;
1300 ptr->buf[ptr->index++] = c;
1303 void status_reset(ptr)
1304 struct input_code *ptr;
1311 void status_reinit(ptr)
1312 struct input_code *ptr;
1315 ptr->_file_stat = 0;
1318 void status_check(ptr, c)
1319 struct input_code *ptr;
1322 if (c <= DEL && estab_f){
1327 void s_status(ptr, c)
1328 struct input_code *ptr;
1333 status_check(ptr, c);
1338 #ifdef NUMCHAR_OPTION
1339 }else if ((c & CLASS_MASK) == CLASS_UTF16){
1342 }else if (0xa1 <= c && c <= 0xdf){
1343 status_push_ch(ptr, SSO);
1344 status_push_ch(ptr, c);
1347 }else if ((0x81 <= c && c < 0xa0) || (0xe0 <= c && c <= 0xef)){
1349 status_push_ch(ptr, c);
1350 #ifdef SHIFTJIS_CP932
1352 && CP932_TABLE_BEGIN <= c && c <= CP932_TABLE_END){
1354 status_push_ch(ptr, c);
1355 #endif /* SHIFTJIS_CP932 */
1357 status_disable(ptr);
1361 if ((0x40 <= c && c <= 0x7e) || (0x80 <= c && c <= 0xfc)){
1362 status_push_ch(ptr, c);
1363 s2e_conv(ptr->buf[0], ptr->buf[1], &ptr->buf[0], &ptr->buf[1]);
1367 status_disable(ptr);
1370 #ifdef SHIFTJIS_CP932
1372 if ((0x40 <= c && c <= 0x7e) || (0x80 <= c && c <= 0xfc)){
1373 status_push_ch(ptr, c);
1374 if (s2e_conv(ptr->buf[0], ptr->buf[1], &ptr->buf[0], &ptr->buf[1]) == 0){
1375 set_code_score(ptr, SCORE_CP932);
1380 status_disable(ptr);
1382 #endif /* SHIFTJIS_CP932 */
1386 void e_status(ptr, c)
1387 struct input_code *ptr;
1392 status_check(ptr, c);
1397 #ifdef NUMCHAR_OPTION
1398 }else if ((c & CLASS_MASK) == CLASS_UTF16){
1401 }else if (SSO == c || (0xa1 <= c && c <= 0xfe)){
1403 status_push_ch(ptr, c);
1405 status_disable(ptr);
1409 if (0xa1 <= c && c <= 0xfe){
1410 status_push_ch(ptr, c);
1414 status_disable(ptr);
1420 #ifdef UTF8_INPUT_ENABLE
1421 void w16_status(ptr, c)
1422 struct input_code *ptr;
1429 if (ptr->_file_stat == 0){
1430 if (c == 0xfe || c == 0xff){
1432 status_push_ch(ptr, c);
1433 ptr->_file_stat = 1;
1435 status_disable(ptr);
1436 ptr->_file_stat = -1;
1438 }else if (ptr->_file_stat > 0){
1440 status_push_ch(ptr, c);
1441 }else if (ptr->_file_stat < 0){
1442 status_disable(ptr);
1448 status_disable(ptr);
1449 ptr->_file_stat = -1;
1451 status_push_ch(ptr, c);
1458 if (ptr->stat != c && (c == 0xfe || c == 0xff)){
1459 status_push_ch(ptr, c);
1462 status_disable(ptr);
1463 ptr->_file_stat = -1;
1469 void w_status(ptr, c)
1470 struct input_code *ptr;
1475 status_check(ptr, c);
1480 #ifdef NUMCHAR_OPTION
1481 }else if ((c & CLASS_MASK) == CLASS_UTF16){
1484 }else if (0xc0 <= c && c <= 0xdf){
1486 status_push_ch(ptr, c);
1487 }else if (0xe0 <= c && c <= 0xef){
1489 status_push_ch(ptr, c);
1491 status_disable(ptr);
1496 if (0x80 <= c && c <= 0xbf){
1497 status_push_ch(ptr, c);
1498 if (ptr->index > ptr->stat){
1499 w2e_conv(ptr->buf[0], ptr->buf[1], ptr->buf[2],
1500 &ptr->buf[0], &ptr->buf[1]);
1505 status_disable(ptr);
1516 int action_flag = 1;
1517 struct input_code *result = 0;
1518 struct input_code *p = input_code_list;
1520 (p->status_func)(p, c);
1523 }else if(p->stat == 0){
1534 if (result && !estab_f){
1535 set_iconv(TRUE, result->iconv_func);
1536 }else if (c <= DEL){
1537 struct input_code *ptr = input_code_list;
1547 #define STD_GC_BUFSIZE (256)
1548 int std_gc_buf[STD_GC_BUFSIZE];
1558 return std_gc_buf[--std_gc_ndx];
1570 if (std_gc_ndx == STD_GC_BUFSIZE){
1573 std_gc_buf[std_gc_ndx++] = c;
1593 while ((c = (*i_getc)(f)) != EOF)
1602 oconv = output_conv;
1605 /* replace continucation module, from output side */
1607 /* output redicrection */
1616 if (mimeout_f == TRUE) {
1617 o_base64conv = oconv; oconv = base64_conv;
1619 /* base64_count = 0; */
1623 o_crconv = oconv; oconv = cr_conv;
1626 o_rot_conv = oconv; oconv = rot_conv;
1629 o_iso2022jp_check_conv = oconv; oconv = iso2022jp_check_conv;
1632 o_hira_conv = oconv; oconv = hira_conv;
1635 o_fconv = oconv; oconv = fold_conv;
1638 if (alpha_f || x0201_f) {
1639 o_zconv = oconv; oconv = z_conv;
1643 /* input redicrection */
1646 i_cgetc = i_getc; i_getc = cap_getc;
1647 i_cungetc = i_ungetc; i_ungetc= cap_ungetc;
1650 i_ugetc = i_getc; i_getc = url_getc;
1651 i_uungetc = i_ungetc; i_ungetc= url_ungetc;
1654 #ifdef NUMCHAR_OPTION
1656 i_ngetc = i_getc; i_getc = numchar_getc;
1657 i_nungetc = i_ungetc; i_ungetc= numchar_ungetc;
1660 if (mime_f && mimebuf_f==FIXED_MIME) {
1661 i_mgetc = i_getc; i_getc = mime_getc;
1662 i_mungetc = i_ungetc; i_ungetc = mime_ungetc;
1665 i_bgetc = i_getc; i_getc = broken_getc;
1666 i_bungetc = i_ungetc; i_ungetc = broken_ungetc;
1668 if (input_f == JIS_INPUT || input_f == LATIN1_INPUT) {
1669 set_iconv(-TRUE, e_iconv);
1670 } else if (input_f == SJIS_INPUT) {
1671 set_iconv(-TRUE, s_iconv);
1672 #ifdef UTF8_INPUT_ENABLE
1673 } else if (input_f == UTF8_INPUT) {
1674 set_iconv(-TRUE, w_iconv);
1675 } else if (input_f == UTF16_INPUT) {
1676 set_iconv(-TRUE, w_iconv16);
1679 set_iconv(FALSE, e_iconv);
1683 struct input_code *p = input_code_list;
1691 Conversion main loop. Code detection only.
1701 module_connection();
1706 output_mode = ASCII;
1709 #define NEXT continue /* no output, get next */
1710 #define SEND ; /* output c1 and c2, get next */
1711 #define LAST break /* end of loop, go closing */
1713 while ((c1 = (*i_getc)(f)) != EOF) {
1718 /* in case of 8th bit is on */
1720 /* in case of not established yet */
1721 /* It is still ambiguious */
1722 if (h_conv(f, c2, c1)==EOF)
1728 /* in case of already established */
1730 /* ignore bogus code */
1736 /* second byte, 7 bit code */
1737 /* it might be kanji shitfted */
1738 if ((c1 == DEL) || (c1 <= SPACE)) {
1739 /* ignore bogus first code */
1747 #ifdef UTF8_INPUT_ENABLE
1756 #ifdef NUMCHAR_OPTION
1757 } else if ((c1 & CLASS_MASK) == CLASS_UTF16){
1760 } else if (c1 > DEL) {
1762 if (!estab_f && !iso8859_f) {
1763 /* not established yet */
1766 } else { /* estab_f==TRUE */
1771 } else if (SSP<=c1 && c1<0xe0 && iconv == s_iconv) {
1772 /* SJIS X0201 Case... */
1773 if(iso2022jp_f && x0201_f==NO_X0201) {
1774 (*oconv)(GETA1, GETA2);
1781 } else if (c1==SSO && iconv != s_iconv) {
1782 /* EUC X0201 Case */
1783 c1 = (*i_getc)(f); /* skip SSO */
1785 if (SSP<=c1 && c1<0xe0) {
1786 if(iso2022jp_f && x0201_f==NO_X0201) {
1787 (*oconv)(GETA1, GETA2);
1794 } else { /* bogus code, skip SSO and one byte */
1798 /* already established */
1803 } else if ((c1 > SPACE) && (c1 != DEL)) {
1804 /* in case of Roman characters */
1806 /* output 1 shifted byte */
1810 } else if (SPACE<=c1 && c1<(0xe0&0x7f) ){
1811 /* output 1 shifted byte */
1812 if(iso2022jp_f && x0201_f==NO_X0201) {
1813 (*oconv)(GETA1, GETA2);
1820 /* look like bogus code */
1823 } else if (input_mode == X0208) {
1824 /* in case of Kanji shifted */
1827 } else if (c1 == '=' && mime_f && !mime_decode_mode ) {
1828 /* Check MIME code */
1829 if ((c1 = (*i_getc)(f)) == EOF) {
1832 } else if (c1 == '?') {
1833 /* =? is mime conversion start sequence */
1834 if(mime_f == STRICT_MIME) {
1835 /* check in real detail */
1836 if (mime_begin_strict(f) == EOF)
1840 } else if (mime_begin(f) == EOF)
1850 /* normal ASCII code */
1853 } else if (c1 == SI) {
1856 } else if (c1 == SO) {
1859 } else if (c1 == ESC ) {
1860 if ((c1 = (*i_getc)(f)) == EOF) {
1861 /* (*oconv)(0, ESC); don't send bogus code */
1863 } else if (c1 == '$') {
1864 if ((c1 = (*i_getc)(f)) == EOF) {
1866 (*oconv)(0, ESC); don't send bogus code
1867 (*oconv)(0, '$'); */
1869 } else if (c1 == '@'|| c1 == 'B') {
1870 /* This is kanji introduction */
1873 debug(input_codename = "ISO-2022-JP");
1875 } else if (c1 == '(') {
1876 if ((c1 = (*i_getc)(f)) == EOF) {
1877 /* don't send bogus code
1883 } else if (c1 == '@'|| c1 == 'B') {
1884 /* This is kanji introduction */
1889 /* could be some special code */
1896 } else if (broken_f&0x2) {
1897 /* accept any ESC-(-x as broken code ... */
1907 } else if (c1 == '(') {
1908 if ((c1 = (*i_getc)(f)) == EOF) {
1909 /* don't send bogus code
1911 (*oconv)(0, '('); */
1915 /* This is X0201 kana introduction */
1916 input_mode = X0201; shift_mode = X0201;
1918 } else if (c1 == 'B' || c1 == 'J' || c1 == 'H') {
1919 /* This is X0208 kanji introduction */
1920 input_mode = ASCII; shift_mode = FALSE;
1922 } else if (broken_f&0x2) {
1923 input_mode = ASCII; shift_mode = FALSE;
1928 /* maintain various input_mode here */
1932 } else if ( c1 == 'N' || c1 == 'n' ){
1934 c1 = (*i_getc)(f); /* skip SS2 */
1935 if ( SPACE<=c1 && c1 < 0xe0 ) {
1944 } else if ((c1 == NL || c1 == CR) && broken_f&4) {
1945 input_mode = ASCII; set_iconv(FALSE, 0);
1951 if (input_mode == X0208)
1952 (*oconv)(c2, c1); /* this is JIS, not SJIS/EUC case */
1953 else if (input_mode)
1954 (*oconv)(input_mode, c1); /* other special case */
1955 else if ((*iconv)(c2, c1, 0) < 0){ /* can be EUC/SJIS */
1956 int c0 = (*i_getc)(f);
1959 (*iconv)(c2, c1, c0);
1965 /* goto next_word */
1969 (*iconv)(EOF, 0, 0);
1982 /** it must NOT be in the kanji shifte sequence */
1983 /** it must NOT be written in JIS7 */
1984 /** and it must be after 2 byte 8bit code */
1991 while ((c1 = (*i_getc)(f)) != EOF) {
1997 if (push_hold_buf(c1) == EOF || estab_f){
2003 struct input_code *p = input_code_list;
2004 struct input_code *result = p;
2009 if (p->score < result->score){
2014 set_iconv(FALSE, p->iconv_func);
2019 ** 1) EOF is detected, or
2020 ** 2) Code is established, or
2021 ** 3) Buffer is FULL (but last word is pushed)
2023 ** in 1) and 3) cases, we continue to use
2024 ** Kanji codes by oconv and leave estab_f unchanged.
2029 while (wc < hold_count){
2030 c2 = hold_buf[wc++];
2032 #ifdef NUMCHAR_OPTION
2033 || (c2 & CLASS_MASK) == CLASS_UTF16
2038 }else if (iconv == s_iconv && 0xa1 <= c2 && c2 <= 0xdf){
2039 (*iconv)(X0201, c2, 0);
2042 if (wc < hold_count){
2043 c1 = hold_buf[wc++];
2052 if ((*iconv)(c2, c1, 0) < 0){
2054 if (wc < hold_count){
2055 c0 = hold_buf[wc++];
2064 (*iconv)(c2, c1, c0);
2077 if (hold_count >= HOLD_SIZE*2)
2079 hold_buf[hold_count++] = c2;
2080 return ((hold_count >= HOLD_SIZE*2) ? EOF : hold_count);
2083 int s2e_conv(c2, c1, p2, p1)
2087 #ifdef SHIFTJIS_CP932
2088 if (CP932_TABLE_BEGIN <= c2 && c2 <= CP932_TABLE_END){
2089 extern unsigned short shiftjis_cp932[3][189];
2090 c1 = shiftjis_cp932[c2 - CP932_TABLE_BEGIN][c1 - 0x40];
2091 if (c1 == 0) return 1;
2095 #endif /* SHIFTJIS_CP932 */
2096 c2 = c2 + c2 - ((c2 <= 0x9f) ? SJ0162 : SJ6394);
2098 c1 = c1 - ((c1 > DEL) ? SPACE : 0x1f);
2115 } else if ((c2 == EOF) || (c2 == 0) || c2 < SPACE) {
2118 int ret = s2e_conv(c2, c1, &c2, &c1);
2119 if (ret) return ret;
2132 } else if (c2 == SSO){
2135 } else if ((c2 == EOF) || (c2 == 0) || c2 < SPACE) {
2145 #ifdef UTF8_INPUT_ENABLE
2147 w2e_conv(c2, c1, c0, p2, p1)
2151 extern unsigned short * utf8_to_euc_2bytes[];
2152 extern unsigned short ** utf8_to_euc_3bytes[];
2155 if (0xc0 <= c2 && c2 <= 0xef) {
2156 unsigned short **pp;
2159 if (c0 == 0) return -1;
2160 pp = utf8_to_euc_3bytes[c2 - 0x80];
2161 ret = w_iconv_common(c1, c0, pp, sizeof_utf8_to_euc_C2, p2, p1);
2163 ret = w_iconv_common(c2, c1, utf8_to_euc_2bytes, sizeof_utf8_to_euc_2bytes, p2, p1);
2165 #ifdef NUMCHAR_OPTION
2167 c1 = CLASS_UTF16 | ww16_conv(c2, c1, c0);
2173 } else if (c2 == X0201) {
2186 int ret = w2e_conv(c2, c1, c0, &c2, &c1);
2194 w16w_conv(val, p2, p1, p0)
2202 }else if (val < 0x800){
2203 *p2 = 0xc0 | (val >> 6);
2204 *p1 = 0x80 | (val & 0x3f);
2207 *p2 = 0xe0 | (val >> 12);
2208 *p1 = 0x80 | ((val >> 6) & 0x3f);
2209 *p0 = 0x80 | (val & 0x3f);
2214 ww16_conv(c2, c1, c0)
2219 val = (c2 & 0x0f) << 12;
2220 val |= (c1 & 0x3f) << 6;
2222 }else if (c2 >= 0xc0){
2223 val = (c2 & 0x1f) << 6;
2224 val |= (c1 & 0x3f) << 6;
2232 w16e_conv(val, p2, p1)
2236 extern unsigned short * utf8_to_euc_2bytes[];
2237 extern unsigned short ** utf8_to_euc_3bytes[];
2239 unsigned short **pp;
2243 w16w_conv(val, &c2, &c1, &c0);
2246 pp = utf8_to_euc_3bytes[c2 - 0x80];
2247 psize = sizeof_utf8_to_euc_C2;
2248 ret = w_iconv_common(c1, c0, pp, psize, p2, p1);
2250 pp = utf8_to_euc_2bytes;
2251 psize = sizeof_utf8_to_euc_2bytes;
2252 ret = w_iconv_common(c2, c1, pp, psize, p2, p1);
2254 #ifdef NUMCHAR_OPTION
2257 *p1 = CLASS_UTF16 | val;
2266 w_iconv16(c2, c1, c0)
2271 if (c2==0376 && c1==0377){
2272 utf16_mode = UTF16_INPUT;
2274 } else if (c2==0377 && c1==0376){
2275 utf16_mode = UTF16BE_INPUT;
2278 if (utf16_mode == UTF16BE_INPUT) {
2280 tmp=c1; c1=c2; c2=tmp;
2282 if ((c2==0 && c1 < 0x80) || c2==EOF) {
2286 ret = w16e_conv(((c2<<8)&0xff00) + c1, &c2, &c1);
2287 if (ret) return ret;
2293 w_iconv_common(c1, c0, pp, psize, p2, p1)
2295 unsigned short **pp;
2303 if (pp == 0) return 1;
2306 if (c1 < 0 || psize <= c1) return 1;
2308 if (p == 0) return 1;
2311 if (c0 < 0 || sizeof_utf8_to_euc_E5B8 <= c0) return 1;
2313 if (val == 0) return 1;
2316 if (c2 == SO) c2 = X0201;
2325 #ifdef UTF8_OUTPUT_ENABLE
2330 extern unsigned short euc_to_utf8_1byte[];
2331 extern unsigned short * euc_to_utf8_2bytes[];
2335 p = euc_to_utf8_1byte;
2338 c2 = (c2&0x7f) - 0x21;
2339 if (0<=c2 && c2<sizeof_euc_to_utf8_2bytes)
2340 p = euc_to_utf8_2bytes[c2];
2345 c1 = (c1 & 0x7f) - 0x21;
2346 if (0<=c1 && c1<sizeof_euc_to_utf8_1byte)
2357 #ifdef NUMCHAR_OPTION
2358 if (c2 == 0 && (c1 & CLASS_MASK) == CLASS_UTF16){
2359 w16w_conv(c1, &c2, &c1, &c0);
2363 if (c0) (*o_putc)(c0);
2370 } else if (c2 == 0) {
2371 output_mode = ASCII;
2373 } else if (c2 == ISO8859_1) {
2374 output_mode = ISO8859_1;
2375 (*o_putc)(c1 | 0x080);
2378 w16w_conv((unsigned short)e2w_conv(c2, c1), &c2, &c1, &c0);
2382 if (c0) (*o_putc)(c0);
2397 if (w_oconv16_begin_f==2) {
2399 (*o_putc)((unsigned char)'\377');
2403 (*o_putc)((unsigned char)'\377');
2405 w_oconv16_begin_f=1;
2408 if (c2 == ISO8859_1) {
2411 #ifdef NUMCHAR_OPTION
2412 } else if (c2 == 0 && (c1 & CLASS_MASK) == CLASS_UTF16) {
2413 c2 = (c1 >> 8) & 0xff;
2417 unsigned short val = (unsigned short)e2w_conv(c2, c1);
2418 c2 = (val >> 8) & 0xff;
2437 #ifdef NUMCHAR_OPTION
2438 if (c2 == 0 && (c1 & CLASS_MASK) == CLASS_UTF16){
2439 w16e_conv(c1, &c2, &c1);
2445 } else if (c2 == 0) {
2446 output_mode = ASCII;
2448 } else if (c2 == X0201) {
2449 output_mode = JAPANESE_EUC;
2450 (*o_putc)(SSO); (*o_putc)(c1|0x80);
2451 } else if (c2 == ISO8859_1) {
2452 output_mode = ISO8859_1;
2453 (*o_putc)(c1 | 0x080);
2455 if ((c1<0x20 || 0x7e<c1) ||
2456 (c2<0x20 || 0x7e<c2)) {
2457 set_iconv(FALSE, 0);
2458 return; /* too late to rescue this char */
2460 output_mode = JAPANESE_EUC;
2461 (*o_putc)(c2 | 0x080);
2462 (*o_putc)(c1 | 0x080);
2467 e2s_conv(c2, c1, p2, p1)
2468 int c2, c1, *p2, *p1;
2470 if (p2) *p2 = ((c2 - 1) >> 1) + ((c2 <= 0x5e) ? 0x71 : 0xb1);
2471 if (p1) *p1 = c1 + ((c2 & 1) ? ((c1 < 0x60) ? 0x1f : 0x20) : 0x7e);
2479 #ifdef NUMCHAR_OPTION
2480 if (c2 == 0 && (c1 & CLASS_MASK) == CLASS_UTF16){
2481 w16e_conv(c1, &c2, &c1);
2487 } else if (c2 == 0) {
2488 output_mode = ASCII;
2490 } else if (c2 == X0201) {
2491 output_mode = SHIFT_JIS;
2493 } else if (c2 == ISO8859_1) {
2494 output_mode = ISO8859_1;
2495 (*o_putc)(c1 | 0x080);
2497 if ((c1<0x20 || 0x7e<c1) ||
2498 (c2<0x20 || 0x7e<c2)) {
2499 set_iconv(FALSE, 0);
2500 return; /* too late to rescue this char */
2502 output_mode = SHIFT_JIS;
2503 e2s_conv(c2, c1, &c2, &c1);
2514 #ifdef NUMCHAR_OPTION
2515 if ((c1 & CLASS_MASK) == CLASS_UTF16){
2516 w16e_conv(c1, &c2, &c1);
2520 if (output_mode !=ASCII && output_mode!=ISO8859_1) {
2523 (*o_putc)(ascii_intro);
2524 output_mode = ASCII;
2527 } else if (c2==X0201) {
2528 if (output_mode!=X0201) {
2529 output_mode = X0201;
2535 } else if (c2==ISO8859_1) {
2536 /* iso8859 introduction, or 8th bit on */
2537 /* Can we convert in 7bit form using ESC-'-'-A ?
2539 output_mode = ISO8859_1;
2541 } else if (c2 == 0) {
2542 if (output_mode !=ASCII && output_mode!=ISO8859_1) {
2545 (*o_putc)(ascii_intro);
2546 output_mode = ASCII;
2550 if (output_mode != X0208) {
2551 output_mode = X0208;
2554 (*o_putc)(kanji_intro);
2556 if (c1<0x20 || 0x7e<c1)
2558 if (c2<0x20 || 0x7e<c2)
2570 if (base64_count>50 && !mimeout_mode && c2==0 && c1==SPACE) {
2572 } else if (base64_count>66 && mimeout_mode) {
2573 (*o_base64conv)(EOF,0);
2575 (*o_putc)('\t'); base64_count += 7;
2577 (*o_base64conv)(c2,c1);
2581 static int broken_buf[3];
2582 static int broken_counter = 0;
2583 static int broken_last = 0;
2590 if (broken_counter>0) {
2591 return broken_buf[--broken_counter];
2594 if (c=='$' && broken_last != ESC
2595 && (input_mode==ASCII || input_mode==X0201)) {
2598 if (c1=='@'|| c1=='B') {
2599 broken_buf[0]=c1; broken_buf[1]=c;
2606 } else if (c=='(' && broken_last != ESC
2607 && (input_mode==X0208 || input_mode==X0201)) { /* ) */
2610 if (c1=='J'|| c1=='B') {
2611 broken_buf[0]=c1; broken_buf[1]=c;
2629 if (broken_counter<2)
2630 broken_buf[broken_counter++]=c;
2634 static int prev_cr = 0;
2642 if (! (c2==0&&c1==NL) ) {
2648 } else if (c1=='\r') {
2650 } else if (c1=='\n') {
2651 if (crmode_f==CRLF) {
2652 (*o_crconv)(0,'\r');
2653 } else if (crmode_f==CR) {
2654 (*o_crconv)(0,'\r');
2658 } else if (c1!='\032' || crmode_f!=NL){
2664 Return value of fold_conv()
2666 \n add newline and output char
2667 \r add newline and output nothing
2670 1 (or else) normal output
2672 fold state in prev (previous character)
2674 >0x80 Japanese (X0208/X0201)
2679 This fold algorthm does not preserve heading space in a line.
2680 This is the main difference from fmt.
2683 #define char_size(c2,c1) (c2?2:1)
2692 if (c1== '\r' && !fold_preserve_f) {
2693 fold_state=0; /* ignore cr */
2694 }else if (c1== '\n'&&f_prev=='\r' && fold_preserve_f) {
2696 fold_state=0; /* ignore cr */
2697 } else if (c1== BS) {
2698 if (f_line>0) f_line--;
2700 } else if (c2==EOF && f_line != 0) { /* close open last line */
2702 } else if ((c1=='\n' && !fold_preserve_f)
2703 || ((c1=='\r'||(c1=='\n'&&f_prev!='\r'))
2704 && fold_preserve_f)) {
2706 if (fold_preserve_f) {
2710 } else if ((f_prev == c1 && !fold_preserve_f)
2711 || (f_prev == '\n' && fold_preserve_f)
2712 ) { /* duplicate newline */
2715 fold_state = '\n'; /* output two newline */
2721 if (f_prev&0x80) { /* Japanese? */
2723 fold_state = 0; /* ignore given single newline */
2724 } else if (f_prev==' ') {
2728 if (++f_line<=fold_len)
2732 fold_state = '\r'; /* fold and output nothing */
2736 } else if (c1=='\f') {
2741 fold_state = '\n'; /* output newline and clear */
2742 } else if ( (c2==0 && c1==' ')||
2743 (c2==0 && c1=='\t')||
2744 (c2=='!'&& c1=='!')) {
2745 /* X0208 kankaku or ascii space */
2746 if (f_prev == ' ') {
2747 fold_state = 0; /* remove duplicate spaces */
2750 if (++f_line<=fold_len)
2751 fold_state = ' '; /* output ASCII space only */
2753 f_prev = ' '; f_line = 0;
2754 fold_state = '\r'; /* fold and output nothing */
2758 prev0 = f_prev; /* we still need this one... , but almost done */
2760 if (c2 || c2==X0201)
2761 f_prev |= 0x80; /* this is Japanese */
2762 f_line += char_size(c2,c1);
2763 if (f_line<=fold_len) { /* normal case */
2766 if (f_line>=fold_len+fold_margin) { /* too many kinsou suspension */
2767 f_line = char_size(c2,c1);
2768 fold_state = '\n'; /* We can't wait, do fold now */
2769 } else if (c2==X0201) {
2770 /* simple kinsoku rules return 1 means no folding */
2771 if (c1==(0xde&0x7f)) fold_state = 1; /*
\e$B!+
\e(B*/
2772 else if (c1==(0xdf&0x7f)) fold_state = 1; /*
\e$B!,
\e(B*/
2773 else if (c1==(0xa4&0x7f)) fold_state = 1; /*
\e$B!#
\e(B*/
2774 else if (c1==(0xa3&0x7f)) fold_state = 1; /*
\e$B!$
\e(B*/
2775 else if (c1==(0xa1&0x7f)) fold_state = 1; /*
\e$B!W
\e(B*/
2776 else if (c1==(0xb0&0x7f)) fold_state = 1; /* - */
2777 else if (SPACE<=c1 && c1<=(0xdf&0x7f)) { /* X0201 */
2779 fold_state = '\n';/* add one new f_line before this character */
2782 fold_state = '\n';/* add one new f_line before this character */
2785 /* kinsoku point in ASCII */
2786 if ( c1==')'|| /* { [ ( */
2797 /* just after special */
2798 } else if (!is_alnum(prev0)) {
2799 f_line = char_size(c2,c1);
2801 } else if ((prev0==' ') || /* ignored new f_line */
2802 (prev0=='\n')|| /* ignored new f_line */
2803 (prev0&0x80)) { /* X0208 - ASCII */
2804 f_line = char_size(c2,c1);
2805 fold_state = '\n';/* add one new f_line before this character */
2807 fold_state = 1; /* default no fold in ASCII */
2811 if (c1=='"') fold_state = 1; /*
\e$B!"
\e(B */
2812 else if (c1=='#') fold_state = 1; /*
\e$B!#
\e(B */
2813 else if (c1=='W') fold_state = 1; /*
\e$B!W
\e(B */
2814 else if (c1=='K') fold_state = 1; /*
\e$B!K
\e(B */
2815 else if (c1=='$') fold_state = 1; /*
\e$B!$
\e(B */
2816 else if (c1=='%') fold_state = 1; /*
\e$B!%
\e(B */
2817 else if (c1=='\'') fold_state = 1; /*
\e$B!\
\e(B */
2818 else if (c1=='(') fold_state = 1; /*
\e$B!(
\e(B */
2819 else if (c1==')') fold_state = 1; /*
\e$B!)
\e(B */
2820 else if (c1=='*') fold_state = 1; /*
\e$B!*
\e(B */
2821 else if (c1=='+') fold_state = 1; /*
\e$B!+
\e(B */
2822 else if (c1==',') fold_state = 1; /*
\e$B!,
\e(B */
2823 /* default no fold in kinsoku */
2826 f_line = char_size(c2,c1);
2827 /* add one new f_line before this character */
2830 f_line = char_size(c2,c1);
2832 /* add one new f_line before this character */
2837 /* terminator process */
2838 switch(fold_state) {
2857 int z_prev2=0,z_prev1=0;
2864 /* if (c2) c1 &= 0x7f; assertion */
2866 if (x0201_f && z_prev2==X0201) { /* X0201 */
2867 if (c1==(0xde&0x7f)) { /*
\e$BByE@
\e(B */
2869 (*o_zconv)(dv[(z_prev1-SPACE)*2],dv[(z_prev1-SPACE)*2+1]);
2871 } else if (c1==(0xdf&0x7f)&&ev[(z_prev1-SPACE)*2]) { /*
\e$BH>ByE@
\e(B */
2873 (*o_zconv)(ev[(z_prev1-SPACE)*2],ev[(z_prev1-SPACE)*2+1]);
2877 (*o_zconv)(cv[(z_prev1-SPACE)*2],cv[(z_prev1-SPACE)*2+1]);
2886 if (x0201_f && c2==X0201) {
2887 if (dv[(c1-SPACE)*2]||ev[(c1-SPACE)*2]) {
2888 /* wait for
\e$BByE@
\e(B or
\e$BH>ByE@
\e(B */
2889 z_prev1 = c1; z_prev2 = c2;
2892 (*o_zconv)(cv[(c1-SPACE)*2],cv[(c1-SPACE)*2+1]);
2897 /* JISX0208 Alphabet */
2898 if (alpha_f && c2 == 0x23 ) {
2900 } else if (alpha_f && c2 == 0x21 ) {
2901 /* JISX0208 Kigou */
2906 } else if (alpha_f&0x4) {
2911 } else if (0x20<c1 && c1<0x7f && fv[c1-0x20]) {
2917 case '>': entity = ">"; break;
2918 case '<': entity = "<"; break;
2919 case '\"': entity = """; break;
2920 case '&': entity = "&"; break;
2923 while (*entity) (*o_zconv)(0, *entity++);
2933 #define rot13(c) ( \
2935 (c <= 'M') ? (c + 13): \
2936 (c <= 'Z') ? (c - 13): \
2938 (c <= 'm') ? (c + 13): \
2939 (c <= 'z') ? (c - 13): \
2943 #define rot47(c) ( \
2945 ( c <= 'O' ) ? (c + 47) : \
2946 ( c <= '~' ) ? (c - 47) : \
2954 if (c2==0 || c2==X0201 || c2==ISO8859_1) {
2960 (*o_rot_conv)(c2,c1);
2967 if ((hira_f & 1) && c2==0x25 && 0x20<c1 && c1<0x74) {
2969 } else if ((hira_f & 2) && c2==0x24 && 0x20<c1 && c1<0x74) {
2972 (*o_hira_conv)(c2,c1);
2977 iso2022jp_check_conv(c2,c1)
2980 static int range[RANGE_NUM_MAX][2] = {
3003 if(c2 >= 0x00 && c2 <= 0x20 && c1 >= 0x7f && c1 <= 0xff) {
3007 if((c2 >= 0x29 && c2 <= 0x2f) || (c2 >= 0x75 && c2 <= 0x7e)) {
3012 for (i = 0; i < RANGE_NUM_MAX; i++) {
3013 start = range[i][0];
3016 if (c >= start && c <= end) {
3021 (*o_iso2022jp_check_conv)(c2,c1);
3025 /* This converts =?ISO-2022-JP?B?HOGE HOGE?= */
3027 unsigned char *mime_pattern[] = {
3028 (unsigned char *)"\075?EUC-JP?B?",
3029 (unsigned char *)"\075?SHIFT_JIS?B?",
3030 (unsigned char *)"\075?ISO-8859-1?Q?",
3031 (unsigned char *)"\075?ISO-8859-1?B?",
3032 (unsigned char *)"\075?ISO-2022-JP?B?",
3033 (unsigned char *)"\075?ISO-2022-JP?Q?",
3034 #if defined(UTF8_INPUT_ENABLE) || defined(UTF8_OUTPUT_ENABLE)
3035 (unsigned char *)"\075?UTF-8?B?",
3040 int mime_encode[] = {
3041 JAPANESE_EUC, SHIFT_JIS,ISO8859_1, ISO8859_1, X0208, X0201,
3042 #if defined(UTF8_INPUT_ENABLE) || defined(UTF8_OUTPUT_ENABLE)
3048 int mime_encode_method[] = {
3049 'B', 'B','Q', 'B', 'B', 'Q',
3050 #if defined(UTF8_INPUT_ENABLE) || defined(UTF8_OUTPUT_ENABLE)
3057 #define MAXRECOVER 20
3059 /* I don't trust portablity of toupper */
3060 #define nkf_toupper(c) (('a'<=c && c<='z')?(c-('a'-'A')):c)
3061 #define nkf_isdigit(c) ('0'<=c && c<='9')
3062 #define nkf_isxdigit(c) (nkf_isdigit(c) || ('a'<=c && c<='f') || ('A'<=c && c <= 'F'))
3067 if (i_getc!=mime_getc) {
3068 i_mgetc = i_getc; i_getc = mime_getc;
3069 i_mungetc = i_ungetc; i_ungetc = mime_ungetc;
3070 if(mime_f==STRICT_MIME) {
3071 i_mgetc_buf = i_mgetc; i_mgetc = mime_getc_buf;
3072 i_mungetc_buf = i_mungetc; i_mungetc = mime_ungetc_buf;
3078 unswitch_mime_getc()
3080 if(mime_f==STRICT_MIME) {
3081 i_mgetc = i_mgetc_buf;
3082 i_mungetc = i_mungetc_buf;
3085 i_ungetc = i_mungetc;
3089 mime_begin_strict(f)
3094 unsigned char *p,*q;
3095 int r[MAXRECOVER]; /* recovery buffer, max mime pattern lenght */
3097 mime_decode_mode = FALSE;
3098 /* =? has been checked */
3100 p = mime_pattern[j];
3103 for(i=2;p[i]>' ';i++) { /* start at =? */
3104 if ( ((r[i] = c1 = (*i_getc)(f))==EOF) || nkf_toupper(c1) != p[i] ) {
3105 /* pattern fails, try next one */
3107 while ((p = mime_pattern[++j])) {
3108 for(k=2;k<i;k++) /* assume length(p) > i */
3109 if (p[k]!=q[k]) break;
3110 if (k==i && nkf_toupper(c1)==p[k]) break;
3112 if (p) continue; /* found next one, continue */
3113 /* all fails, output from recovery buffer */
3121 mime_decode_mode = p[i-2];
3122 if (mime_decode_mode=='B') {
3123 mimebuf_f = unbuf_f;
3125 /* do MIME integrity check */
3126 return mime_integrity(f,mime_pattern[j]);
3138 /* we don't keep eof of Fifo, becase it contains ?= as
3139 a terminator. It was checked in mime_integrity. */
3140 return ((mimebuf_f)?
3141 (*i_mgetc_buf)(f):Fifo(mime_input++));
3145 mime_ungetc_buf(c,f)
3150 (*i_mungetc_buf)(c,f);
3152 Fifo(--mime_input)=c;
3163 /* In NONSTRICT mode, only =? is checked. In case of failure, we */
3164 /* re-read and convert again from mime_buffer. */
3166 /* =? has been checked */
3168 Fifo(mime_last++)='='; Fifo(mime_last++)='?';
3169 for(i=2;i<MAXRECOVER;i++) { /* start at =? */
3170 /* We accept any character type even if it is breaked by new lines */
3171 c1 = (*i_getc)(f); Fifo(mime_last++)= c1 ;
3172 if (c1=='\n'||c1==' '||c1=='\r'||
3173 c1=='-'||c1=='_'||is_alnum(c1) ) continue;
3175 /* Failed. But this could be another MIME preemble */
3183 c1 = (*i_getc)(f); Fifo(mime_last++) = c1;
3184 if (!(++i<MAXRECOVER) || c1==EOF) break;
3185 if (c1=='b'||c1=='B') {
3186 mime_decode_mode = 'B';
3187 } else if (c1=='q'||c1=='Q') {
3188 mime_decode_mode = 'Q';
3192 c1 = (*i_getc)(f); Fifo(mime_last++) = c1;
3193 if (!(++i<MAXRECOVER) || c1==EOF) break;
3195 mime_decode_mode = FALSE;
3201 if (!mime_decode_mode) {
3202 /* false MIME premble, restart from mime_buffer */
3203 mime_decode_mode = 1; /* no decode, but read from the mime_buffer */
3204 /* Since we are in MIME mode until buffer becomes empty, */
3205 /* we never go into mime_begin again for a while. */
3208 /* discard mime preemble, and goto MIME mode */
3210 /* do no MIME integrity check */
3211 return c1; /* used only for checking EOF */
3226 fprintf(stderr, "%s\n", str);
3235 if (nkf_isdigit(x)) return x - '0';
3236 return nkf_toupper(x) - 'A' + 10;
3241 #ifdef ANSI_C_PROTOTYPE
3242 int hex_getc(int ch, FILE *f, int (*g)(FILE *f), int (*u)(int c, FILE *f))
3245 hex_getc(ch, f, g, u)
3258 if (!nkf_isxdigit(c2)){
3263 if (!nkf_isxdigit(c3)){
3268 return (hex2bin(c2) << 4) | hex2bin(c3);
3275 return hex_getc(':', f, i_cgetc, i_cungetc);
3283 return (*i_cungetc)(c, f);
3290 return hex_getc('%', f, i_ugetc, i_uungetc);
3298 return (*i_uungetc)(c, f);
3302 #ifdef NUMCHAR_OPTION
3307 int (*g)() = i_ngetc;
3308 int (*u)() = i_nungetc;
3319 if (buf[i] == 'x' || buf[i] == 'X'){
3320 for (j = 0; j < 5; j++){
3322 if (!nkf_isxdigit(buf[i])){
3329 c |= hex2bin(buf[i]);
3332 for (j = 0; j < 6; j++){
3336 if (!nkf_isdigit(buf[i])){
3343 c += hex2bin(buf[i]);
3349 return CLASS_UTF16 | c;
3359 numchar_ungetc(c, f)
3363 return (*i_nungetc)(c, f);
3372 int c1, c2, c3, c4, cc;
3373 int t1, t2, t3, t4, mode, exit_mode;
3375 if (mime_top != mime_last) { /* Something is in FIFO */
3376 return Fifo(mime_top++);
3378 if (mime_decode_mode==1 ||mime_decode_mode==FALSE) {
3379 mime_decode_mode=FALSE;
3380 unswitch_mime_getc();
3381 return (*i_getc)(f);
3384 if (mimebuf_f == FIXED_MIME)
3385 exit_mode = mime_decode_mode;
3388 if (mime_decode_mode == 'Q') {
3389 if ((c1 = (*i_mgetc)(f)) == EOF) return (EOF);
3391 if (c1=='_') return ' ';
3392 if (c1!='=' && c1!='?') {
3396 mime_decode_mode = exit_mode; /* prepare for quit */
3397 if (c1<=' ') return c1;
3398 if ((c2 = (*i_mgetc)(f)) == EOF) return (EOF);
3399 if (c1=='?'&&c2=='=' && mimebuf_f != FIXED_MIME) {
3400 /* end Q encoding */
3401 input_mode = exit_mode;
3402 while((c1=(*i_getc)(f))!=EOF && c1==SPACE
3403 /* && (c1==NL||c1==TAB||c1=='\r') */ ) ;
3406 if (c1=='='&&c2<' ') { /* this is soft wrap */
3407 while((c1 = (*i_mgetc)(f)) <=' ') {
3408 if ((c1 = (*i_mgetc)(f)) == EOF) return (EOF);
3410 mime_decode_mode = 'Q'; /* still in MIME */
3411 goto restart_mime_q;
3414 mime_decode_mode = 'Q'; /* still in MIME */
3418 if ((c3 = (*i_mgetc)(f)) == EOF) return (EOF);
3419 if (c2<=' ') return c2;
3420 mime_decode_mode = 'Q'; /* still in MIME */
3421 #define hex(c) (('0'<=c&&c<='9')?(c-'0'):\
3422 ('A'<=c&&c<='F')?(c-'A'+10):('a'<=c&&c<='f')?(c-'a'+10):0)
3423 return ((hex(c2)<<4) + hex(c3));
3426 if (mime_decode_mode != 'B') {
3427 mime_decode_mode = FALSE;
3428 return (*i_mgetc)(f);
3432 /* Base64 encoding */
3434 MIME allows line break in the middle of
3435 Base64, but we are very pessimistic in decoding
3436 in unbuf mode because MIME encoded code may broken by
3437 less or editor's control sequence (such as ESC-[-K in unbuffered
3438 mode. ignore incomplete MIME.
3440 mode = mime_decode_mode;
3441 mime_decode_mode = exit_mode; /* prepare for quit */
3443 while ((c1 = (*i_mgetc)(f))<=' ') {
3448 if ((c2 = (*i_mgetc)(f))<=' ') {
3451 if (mime_f != STRICT_MIME) goto mime_c2_retry;
3452 if (mimebuf_f!=FIXED_MIME) input_mode = ASCII;
3455 if ((c1 == '?') && (c2 == '=')) {
3457 while((c1=(*i_getc)(f))!=EOF && c1==SPACE
3458 /* && (c1==NL||c1==TAB||c1=='\r') */ ) ;
3462 if ((c3 = (*i_mgetc)(f))<=' ') {
3465 if (mime_f != STRICT_MIME) goto mime_c3_retry;
3466 if (mimebuf_f!=FIXED_MIME) input_mode = ASCII;
3470 if ((c4 = (*i_mgetc)(f))<=' ') {
3473 if (mime_f != STRICT_MIME) goto mime_c4_retry;
3474 if (mimebuf_f!=FIXED_MIME) input_mode = ASCII;
3478 mime_decode_mode = mode; /* still in MIME sigh... */
3480 /* BASE 64 decoding */
3482 t1 = 0x3f & base64decode(c1);
3483 t2 = 0x3f & base64decode(c2);
3484 t3 = 0x3f & base64decode(c3);
3485 t4 = 0x3f & base64decode(c4);
3486 cc = ((t1 << 2) & 0x0fc) | ((t2 >> 4) & 0x03);
3488 Fifo(mime_last++) = cc;
3489 cc = ((t2 << 4) & 0x0f0) | ((t3 >> 2) & 0x0f);
3491 Fifo(mime_last++) = cc;
3492 cc = ((t3 << 6) & 0x0c0) | (t4 & 0x3f);
3494 Fifo(mime_last++) = cc;
3499 return Fifo(mime_top++);
3507 Fifo(--mime_top) = c;
3518 /* In buffered mode, read until =? or NL or buffer full
3520 mime_input = mime_top;
3521 mime_last = mime_top;
3522 while(*p) Fifo(mime_input++) = *p++;
3525 while((c=(*i_getc)(f))!=EOF) {
3526 if (((mime_input-mime_top)&MIME_BUF_MASK)==0) {
3527 break; /* buffer full */
3529 if (c=='=' && d=='?') {
3530 /* checked. skip header, start decode */
3531 Fifo(mime_input++) = c;
3532 /* mime_last_input = mime_input; */
3537 if (!( (c=='+'||c=='/'|| c=='=' || c=='?' || is_alnum(c))))
3539 /* Should we check length mod 4? */
3540 Fifo(mime_input++) = c;
3543 /* In case of Incomplete MIME, no MIME decode */
3544 Fifo(mime_input++) = c;
3545 mime_last = mime_input; /* point undecoded buffer */
3546 mime_decode_mode = 1; /* no decode on Fifo last in mime_getc */
3547 switch_mime_getc(); /* anyway we need buffered getc */
3558 i = c - 'A'; /* A..Z 0-25 */
3560 i = c - 'G' /* - 'a' + 26 */ ; /* a..z 26-51 */
3562 } else if (c > '/') {
3563 i = c - '0' + '4' /* - '0' + 52 */ ; /* 0..9 52-61 */
3564 } else if (c == '+') {
3565 i = '>' /* 62 */ ; /* + 62 */
3567 i = '?' /* 63 */ ; /* / 63 */
3572 static char basis_64[] =
3573 "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
3583 p = mime_pattern[0];
3584 for(i=0;mime_encode[i];i++) {
3585 if (mode == mime_encode[i]) {
3586 p = mime_pattern[i];
3590 mimeout_mode = mime_encode_method[i];
3592 /* (*o_mputc)(' '); */
3609 #define itoh4(c) (c>=10?c+'A'-10:c+'0')
3615 if (mimeout_f==FIXED_MIME) {
3616 if (base64_count>71) {
3624 if ( c<=DEL &&(output_mode==ASCII ||output_mode == ISO8859_1 )
3625 && mimeout_f!=FIXED_MIME) {
3626 if (mimeout_mode=='Q') {
3633 if (mimeout_mode!='B' || c!=SPACE) {
3642 } else if (!mimeout_mode && mimeout_f!=FIXED_MIME) {
3643 open_mime(output_mode);
3645 } else { /* c==EOF */
3646 switch(mimeout_mode) {
3651 (*o_mputc)(basis_64[((b64c & 0x3)<< 4)]);
3657 (*o_mputc)(basis_64[((b64c & 0xF) << 2)]);
3663 if (mimeout_f!=FIXED_MIME) {
3665 } else if (mimeout_mode != 'Q')
3670 switch(mimeout_mode) {
3674 (*o_mputc)(itoh4(((c>>4)&0xf)));
3675 (*o_mputc)(itoh4((c&0xf)));
3682 (*o_mputc)(basis_64[c>>2]);
3687 (*o_mputc)(basis_64[((b64c & 0x3)<< 4) | ((c & 0xF0) >> 4)]);
3693 (*o_mputc)(basis_64[((b64c & 0xF) << 2) | ((c & 0xC0) >>6)]);
3694 (*o_mputc)(basis_64[c & 0x3F]);
3714 mime_f = STRICT_MIME;
3718 #if defined(MSDOS) || defined(__OS2__)
3723 iso2022jp_f = FALSE;
3725 kanji_intro = DEFAULT_J;
3726 ascii_intro = DEFAULT_R;
3728 output_conv = DEFAULT_CONV;
3729 oconv = DEFAULT_CONV;
3732 i_mungetc = std_ungetc;
3733 i_mgetc_buf = std_getc;
3734 i_mungetc_buf = std_ungetc;
3737 i_ungetc=std_ungetc;
3740 i_bungetc= std_ungetc;
3744 o_crconv = no_connection;
3745 o_rot_conv = no_connection;
3746 o_iso2022jp_check_conv = no_connection;
3747 o_hira_conv = no_connection;
3748 o_fconv = no_connection;
3749 o_zconv = no_connection;
3752 i_ungetc = std_ungetc;
3754 i_mungetc = std_ungetc;
3756 output_mode = ASCII;
3759 mime_decode_mode = FALSE;
3768 struct input_code *p = input_code_list;
3773 #ifdef UTF8_OUTPUT_ENABLE
3774 if (w_oconv16_begin_f) {
3775 w_oconv16_begin_f = 2;
3780 fold_preserve_f = FALSE;
3783 fold_margin = FOLD_MARGIN;
3786 z_prev2=0,z_prev1=0;
3788 input_codename = "";
3793 no_connection(c2,c1)
3796 no_connection2(c2,c1,0);
3800 no_connection2(c2,c1,c0)
3803 fprintf(stderr,"nkf internal module connection failure.\n");
3811 fprintf(stderr,"USAGE: nkf(nkf32,wnkf,nkf2) -[flags] [in file] .. [out file for -O flag]\n");
3812 fprintf(stderr,"Flags:\n");
3813 fprintf(stderr,"b,u Output is bufferred (DEFAULT),Output is unbufferred\n");
3814 #ifdef DEFAULT_CODE_SJIS
3815 fprintf(stderr,"j,s,e,w Outout code is JIS 7 bit, Shift JIS (DEFAULT), AT&T JIS (EUC), UTF-8\n");
3817 #ifdef DEFAULT_CODE_JIS
3818 fprintf(stderr,"j,s,e,w Outout code is JIS 7 bit (DEFAULT), Shift JIS, AT&T JIS (EUC), UTF-8\n");
3820 #ifdef DEFAULT_CODE_EUC
3821 fprintf(stderr,"j,s,e,w Outout code is JIS 7 bit, Shift JIS, AT&T JIS (EUC) (DEFAULT), UTF-8\n");
3823 #ifdef DEFAULT_CODE_UTF8
3824 fprintf(stderr,"j,s,e,w Outout code is JIS 7 bit, Shift JIS, AT&T JIS (EUC), UTF-8 (DEFAULT)\n");
3826 fprintf(stderr,"J,S,E,W Input assumption is JIS 7 bit , Shift JIS, AT&T JIS (EUC), UTF-8\n");
3827 fprintf(stderr,"t no conversion\n");
3828 fprintf(stderr,"i_/o_ Output sequence to designate JIS-kanji/ASCII (DEFAULT B)\n");
3829 fprintf(stderr,"r {de/en}crypt ROT13/47\n");
3830 fprintf(stderr,"h 1 hirakana->katakana, 2 katakana->hirakana,3 both\n");
3831 fprintf(stderr,"v Show this usage. V: show version\n");
3832 fprintf(stderr,"m[BQN0] MIME decode [B:base64,Q:quoted,N:non-strict,0:no decode]\n");
3833 fprintf(stderr,"M[BQ] MIME encode [B:base64 Q:quoted]\n");
3834 fprintf(stderr,"l ISO8859-1 (Latin-1) support\n");
3835 fprintf(stderr,"f/F Folding: -f60 or -f or -f60-10 (fold margin 10) F preserve nl\n");
3836 fprintf(stderr,"Z[0-3] Convert X0208 alphabet to ASCII 1: Kankaku to space,2: 2 spaces,\n");
3837 fprintf(stderr," 3: Convert HTML Entity\n");
3838 fprintf(stderr,"X,x Assume X0201 kana in MS-Kanji, -x preserves X0201\n");
3839 fprintf(stderr,"B[0-2] Broken input 0: missing ESC,1: any X on ESC-[($]-X,2: ASCII on NL\n");
3841 fprintf(stderr,"T Text mode output\n");
3843 fprintf(stderr,"O Output to File (DEFAULT 'nkf.out')\n");
3844 fprintf(stderr,"d,c Delete \\r in line feed and \\032, Add \\r in line feed\n");
3845 fprintf(stderr,"I Convert non ISO-2022-JP charactor to GETA\n");
3846 fprintf(stderr,"-L[uwm] line mode u:LF w:CRLF m:CR (DEFAULT noconversion)\n");
3847 fprintf(stderr,"long name options\n");
3848 fprintf(stderr," --fj,--unix,--mac,--windows convert for the system\n");
3849 fprintf(stderr," --jis,--euc,--sjis,--utf8,--utf16,--mime,--base64 convert for the code\n");
3851 fprintf(stderr," --overwrite Overwrite original listed files by filtered result\n");
3853 fprintf(stderr," --help,--version\n");
3860 fprintf(stderr,"Network Kanji Filter Version %s (%s) "
3861 #if defined(MSDOS) && !defined(__WIN32__) && !defined(__WIN16__)
3864 #if defined(MSDOS) && defined(__WIN16__)
3867 #if defined(MSDOS) && defined(__WIN32__)
3873 ,Version,Patchlevel);
3874 fprintf(stderr,"\n%s\n",CopyRight);
3879 **
\e$B%Q%C%A@):n<T
\e(B
3880 ** void@merope.pleiades.or.jp (Kusakabe Youichi)
3881 ** NIDE Naoyuki <nide@ics.nara-wu.ac.jp>
3882 ** ohta@src.ricoh.co.jp (Junn Ohta)
3883 ** inouet@strl.nhk.or.jp (Tomoyuki Inoue)
3884 ** kiri@pulser.win.or.jp (Tetsuaki Kiriyama)
3885 ** Kimihiko Sato <sato@sail.t.u-tokyo.ac.jp>
3886 ** a_kuroe@kuroe.aoba.yokohama.jp (Akihiko Kuroe)
3887 ** kono@ie.u-ryukyu.ac.jp (Shinji Kono)
3888 ** GHG00637@nifty-serve.or.jp (COW)