nkf.c

NKF是网络上写新闻或邮件的时候,汉字的编码转换程序

            return '\n';/* add one new line before this character */
        }
        return 1;  /* default no fold in ASCII */
    } else {
        if(c2=='!') {
            if(c1=='"')  return 1; /* $B!"(B */
            if(c1=='#')  return 1; /* $B!#(B */
            if(c1=='$')  return 1; /* $B!$(B */
            if(c1=='%')  return 1; /* $B!%(B */
            if(c1=='\'') return 1; /* $B!\(B */
            if(c1=='(')  return 1; /* $B!((B */
            if(c1==')')  return 1; /* $B!)(B */
            if(c1=='*')  return 1; /* $B!*(B */
            if(c1=='+')  return 1; /* $B!+(B */
            if(c1==',')  return 1; /* $B!,(B */
        }
        line = 2;
        return '\n'; /* add one new line before this character */
    }
}

int
pre_convert(c1,c2)
int c1,c2;
{
        if(c2) c1 &= 0x7f;
        c1_return = c1;
        if(c2==EOF) return c2;
        c2 &= 0x7f;
        if(rot_f) {
            if(c2) {
                c1 = rot47(c1);
                c2 = rot47(c2);
            } else {
                if(!(c1 & 0x80)) 
                    c1 = rot13(c1);
            }
            c1_return = c1;
        }
        /* JISX0208 Alphabet */
        if(alpha_f && c2 == 0x23 ) return 0; 
        /* JISX0208 Kigou */
        if(alpha_f && c2 == 0x21 ) { 
           if(0x21==c1) {
               if(alpha_f&0x2) {
                   c1_return = ' ';
                   return 0;
               } else if(alpha_f&0x4) {
                   c1_return = ' ';
                   return DOUBLE_SPACE;
               } else {
                   return c2;
               }
           } else if(0x20<c1 && c1<0x7f && fv[c1-0x20]) {
               c1_return = fv[c1-0x20];
               return 0;
           } 
        }
        return c2;
}


#ifdef STRICT_MIME
/* This converts  =?ISO-2022-JP?B?HOGE HOGE */

unsigned char *mime_pattern[] = {
   (unsigned char *)"\075?ISO-8859-1?Q?",
   (unsigned char *)"\075?ISO-2022-JP?B?",
   (unsigned char *)"\075?ISO-2022-JP?Q?",
   (unsigned char *)"\075?JAPANESE_EUC?B?",
   (unsigned char *)"\075?SHIFT_JIS?B?",
   NULL
};

int      mime_encode[] = {
    'Q', 'B', 'Q',
    0
};
#endif

#define MAXRECOVER 20
int iso8859_f_save;

#ifdef STRICT_MIME

#define nkf_toupper(c)  (('a'<=c && c<='z')?(c-('a'-'A')):c)
/* I don't trust portablity of toupper */

int
mime_begin(f)
FILE *f;
{
    int c1;
    int i,j,k;
    unsigned char *p,*q;
    int r[MAXRECOVER];    /* recovery buffer, max mime pattern lenght */

    mime_mode = FALSE;
    /* =? has been checked   */
    j = 0;
    p = mime_pattern[j];
    r[0]='='; r[1]='?';

    for(i=2;p[i]>' ';i++) {                   /* start at =? */
        if( ((r[i] = c1 = getc(f))==EOF) || nkf_toupper(c1) != p[i] ) {
            /* pattern fails, try next one */
            q = p;
            while (p = mime_pattern[++j]) {
                for(k=2;k<i;k++)              /* assume length(p) > i */
                    if(p[k]!=q[k]) break;
                if(k==i && nkf_toupper(c1)==p[k]) break;
            }
            if(p) continue;  /* found next one, continue */
            /* all fails, output from recovery buffer */
            ungetc(c1,f);
            for(j=0;j<i;j++) {
                (*oconv)(0,r[j]);
            }
            return c1;
        }
    }
    mime_mode = mime_encode[j];
    iso8859_f_save = iso8859_f;
    if(j==0) {
        iso8859_f = TRUE;
    }
    if(mime_mode=='B') {
        mimebuf_f = unbuf_f;
        if(!unbuf_f) {
            /* do MIME integrity check */
            return mime_integrity(f,mime_pattern[j]);
        } 
    }
    mimebuf_f = TRUE;
    return c1;
}

#define mime_getc0(f)   (mimebuf_f?getc(f):Fifo(mime_input++))
#define mime_ungetc0(c,f) (mimebuf_f?ungetc(c,f):mime_input--)

#else
int
mime_begin(f)
FILE *f;
{
    int c1;
    int i,j,k;
    int r[MAXRECOVER];    /* recovery buffer, max mime pattern lenght */

    mime_mode = FALSE;
    /* =? has been checked   */
    j = 0;
    r[0]='='; r[1]='?';
    for(i=2;i<MAXRECOVER;i++) {                   /* start at =? */
	/* We accept any charcter type even if it is breaked by new lines */
        if( (r[i] = c1 = getc(f))==EOF) break;
	if(c1=='=') break;
	if(c1<' '&& c1!='\r' && c1!='\n') break;
	if(!(('A'<=c1&&c1<='Z')||('a'<=c1&&c1<='z')||('0'<=c1&&c1<='9')
	    ||c1=='_'||c1=='-'||c1=='?')) break;
	if(c1=='?') {
	    i++;
	    if(!(i<MAXRECOVER) || (r[i] = c1 = getc(f))==EOF) break;
	    if(c1=='b'||c1=='B') {
		mime_mode = 'B';
	    } else if(c1=='q'||c1=='Q') {
		mime_mode = 'Q';
	    } else {
		break;
	    }
	    i++;
	    if(!(i<MAXRECOVER) || (r[i] = c1 = getc(f))==EOF) break;
	    if(c1=='?') {
		break;
	    } else {
		mime_mode = FALSE;
	    }
	    break;
	}
    }
    if(!mime_mode || c1==EOF || i==MAXRECOVER) {
	ungetc(c1,f);
	for(j=0;j<i;j++) {
	    (*oconv)(0,r[j]);
	}
	return c1;
    }
    iso8859_f_save = iso8859_f;
    /* do no MIME integrity check */
    return c1;   /* used only for checking EOF */
}

#define mime_getc0(f)   getc(f)
#define mime_ungetc0(c,f) ungetc(c,f)

#endif

int 
mime_getc(f)
FILE *f;
{
    int c1, c2, c3, c4, cc;
    int t1, t2, t3, t4, mode, exit_mode;

    if(mime_top != mime_last) {  /* Something is in FIFO */
        return  Fifo(mime_top++);
    }

    if(mimebuf_f == FIXED_MIME)
        exit_mode = mime_mode;
    else
        exit_mode = FALSE;
    if(mime_mode == 'Q') {
        if((c1 = mime_getc0(f)) == EOF) return (EOF);
        if(c1=='_') return ' ';
        if(c1!='=' && c1!='?') 
            return c1;
        mime_mode = exit_mode; /* prepare for quit */
        if(c1<=' ') return c1;
        if((c2 = mime_getc0(f)) == EOF) return (EOF);
        if(c2<=' ') return c2;
        if(c1=='?'&&c2=='=') {
            /* end Q encoding */
            input_mode = exit_mode;
            iso8859_f = iso8859_f_save;
            return getc(f);
        }
        if(c1=='?') {
            mime_mode = 'Q'; /* still in MIME */
            mime_ungetc0(c2,f);
            return c1;
        }
        if((c3 = mime_getc0(f)) == EOF) return (EOF);
        if(c2<=' ') return c2;
        mime_mode = 'Q'; /* still in MIME */
#define hex(c)   (('0'<=c&&c<='9')?(c-'0'):\
     ('A'<=c&&c<='F')?(c-'A'+10):('a'<=c&&c<='f')?(c-'a'+10):0)
        return ((hex(c2)<<4) + hex(c3));
    }

    if(mime_mode != 'B') {
        mime_mode = FALSE;
        return getc(f);
    }


    /* Base64 encoding */
    /* 
        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 
        less or editor's control sequence (such as ESC-[-K in unbuffered
        mode. ignore incomplete MIME.
    */
    mode = mime_mode;
    mime_mode = exit_mode;  /* prepare for quit */

    while ((c1 = mime_getc0(f))<=' ') {
        if(c1==EOF)
            return (EOF);
    }
    if((c2 = mime_getc0(f))<=' ') {
        if(c2==EOF)
            return (EOF);
        if(mimebuf_f!=FIXED_MIME) input_mode = ASCII;  
        return c2;
    }
    if((c1 == '?') && (c2 == '=')) {
        input_mode = ASCII;
        while((c1 =  getc(f))==' ' /* || c1=='\n' || c1=='\r' */);
        return c1;
    }
    if((c3 = mime_getc0(f))<=' ') {
        if(c3==EOF)
            return (EOF);
        if(mimebuf_f!=FIXED_MIME) input_mode = ASCII;  
        return c3;
    }
    if((c4 = mime_getc0(f))<=' ') {
        if(c4==EOF)
            return (EOF);
        if(mimebuf_f!=FIXED_MIME) input_mode = ASCII;  
        return c4;
    }

    mime_mode = mode; /* still in MIME sigh... */

    /* BASE 64 decoding */

    t1 = 0x3f & base64decode(c1);
    t2 = 0x3f & base64decode(c2);
    t3 = 0x3f & base64decode(c3);
    t4 = 0x3f & base64decode(c4);
    cc = ((t1 << 2) & 0x0fc) | ((t2 >> 4) & 0x03);
    if(c2 != '=') {
        Fifo(mime_last++) = cc;
        cc = ((t2 << 4) & 0x0f0) | ((t3 >> 2) & 0x0f);
        if(c3 != '=') {
            Fifo(mime_last++) = cc;
            cc = ((t3 << 6) & 0x0c0) | (t4 & 0x3f);
            if(c4 != '=') 
                Fifo(mime_last++) = cc;
        }
    } else {
        return c1;
    }
    return  Fifo(mime_top++);
}

int
mime_ungetc(c) 
unsigned int   c;
{
    Fifo(mime_last++) = c;
    return c;
}

#ifdef STRICT_MIME
int
mime_integrity(f,p)
FILE *f;
unsigned char *p;
{
    int c,d;
    unsigned int q;
    /* In buffered mode, read until =? or NL or buffer fffull
     */
    mime_input = mime_top;
    mime_last = mime_top;
    while(*p) Fifo(mime_input++) = *p++;
    d = 0;
    q = mime_input;
    while((c=getc(f))!=EOF) {
        if(((mime_input-mime_top)&MIME_BUF_MASK)==0) break;
        if(c=='=' && d=='?') {
            /* checked. skip header, start decode */
            Fifo(mime_input++) = c;
            mime_input = q; 
            return 1;
        }
        if(!( (c=='+'||c=='/'|| c=='=' || c=='?' ||
            ('a'<=c && c<='z')||('A'<= c && c<='Z')||('0'<=c && c<='9'))))
            break;
        /* Should we check length mod 4? */
        Fifo(mime_input++) = c;
        d=c;
    }
    /* In case of Incomplete MIME, no MIME decode  */
    Fifo(mime_input++) = c;
    mime_last = mime_input;     /* point undecoded buffer */
    mime_mode = 1;              /* no decode on Fifo last in mime_getc */
    return 1;
}
#endif

int
base64decode(c)
    int            c;
{
    int             i;
    if(c > '@')
        if(c < '[')
            i = c - 'A';                        /* A..Z 0-25 */
        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
        i = '?'             /* 63 */ ;          /* / 63 */
    return (i);
}

void 
reinit()
{
    unbuf_f = FALSE;
    estab_f = FALSE;
    nop_f = FALSE;
    binmode_f = TRUE;       
    rot_f = FALSE;         
    input_f = FALSE;      
    alpha_f = FALSE;     
    mime_f = TRUE;      
    mimebuf_f = FALSE; 
    broken_f = FALSE;  
    iso8859_f = FALSE; 
    x0201_f = TRUE;    
    x0201_f = NO_X0201; 
    fold_f  = FALSE;
    kanji_intro = DEFAULT_J;
    ascii_intro = DEFAULT_R;
    oconv = DEFAULT_CONV; 
    output_mode = ASCII;
    input_mode =  ASCII;
    shift_mode =  FALSE;
    mime_mode =   FALSE;
    file_out = FALSE;
    add_cr = FALSE;
    del_cr = FALSE;
}

#ifndef PERL_XS
int 
usage()   
{
    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 bufferred (DEFAULT),Output is unbufferred\n");
#ifdef DEFAULT_CODE_SJIS
    fprintf(stderr,"j,s,e    Outout code is JIS 7 bit, Shift JIS (DEFAULT), AT&T JIS (EUC)\n");
#endif
#ifdef DEFAULT_CODE_JIS
    fprintf(stderr,"j,s,e    Outout code is JIS 7 bit (DEFAULT), Shift JIS, AT&T JIS (EUC)\n");
#endif
#ifdef DEFAULT_CODE_EUC
    fprintf(stderr,"j,s,e    Outout code is JIS 7 bit, Shift JIS, AT&T JIS (EUC) (DEFAULT)\n");
#endif
    fprintf(stderr,"J,S,E    Input assumption is JIS 7 bit , Shift JIS, AT&T JIS (EUC)\n");
    fprintf(stderr,"t        no conversion\n");
    fprintf(stderr,"i_       Output sequence to designate JIS-kanji (DEFAULT B)\n");
    fprintf(stderr,"o_       Output sequence to designate ASCII (DEFAULT B)\n");
    fprintf(stderr,"r        {de/en}crypt ROT13/47\n");
    fprintf(stderr,"v        Show this usage\n");
    fprintf(stderr,"m[BQ0]   MIME decode [B:base64,Q:quoted,0:no decode]\n");
    fprintf(stderr,"l        ISO8859-1 (Latin-1) support\n");
    fprintf(stderr,"f        Folding: -f60 or -f\n");
    fprintf(stderr,"Z[0-2]   Convert X0208 alphabet to ASCII  1: Kankaku to space,2: 2 spaces\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");
#ifdef MSDOS
    fprintf(stderr,"T        Text mode output\n");
#endif
    fprintf(stderr,"O        Output to File (DEFAULT 'nkf.out')\n");
    fprintf(stderr,"d,c      Delete \\r in line feed, Add \\r in line feed\n");
    fprintf(stderr,"Network Kanji Filter Version %s (%s) "
#if defined(MSDOS) && !defined(_Windows)
                  "for DOS"
#endif
#if !defined(__WIN32__) && defined(_Windows)
                  "for Win16"
#endif
#if defined(__WIN32__) && defined(_Windows)
                  "for Win32"
#endif
#ifdef __OS2__
                  "for OS/2"
#endif
                  ,Version,Patchlevel);
    fprintf(stderr,"\n%s\n",CopyRight);
    return 0;
}
#endif

/**
 ** $B%Q%C%A@):n<T(B
 **  [e-mail] (Kusakabe Youichi)
 **  NIDE Naoyuki <[e-mail]>
 **  [e-mail] (Junn Ohta)
 **  [e-mail] (Tomoyuki Inoue)
 **  [e-mail] (Tetsuaki Kiriyama)
 **  Kimihiko Sato <[e-mail]>
 **  [e-mail] (Akihiko Kuroe)
 **  [e-mail] (Shinji Kono)
 **  [e-mail] (COW)
 **
 ** $B:G=*99?7F|(B
 **  1998.11.16
 **/

/* end */