uunconc.c

来自「UUDeview是一个编码解码器」· C语言 代码 · 共 1,686 行 · 第 1/3 页

   * the next line. Then we try this in our despair
   */
  if (vflag == 0) {
    if (UUNetscapeCollapse (line))
      vflag = UUValidData (line, encoding, bhflag);
  }

  /*
   * If this line looks uuencoded, but the line is one character short
   * of a valid line, it was probably broken by MS Exchange. According
   * to my test cases, there is at most one space character missing;
   * there are never two spaces together.
   * If adding a space character helps making this line uuencoded, do
   * it!
   */

  if (vflag == 0) {
    ptr    = line + strlen(line);
    while (ptr>line && (*(ptr-1)=='\012' || *(ptr-1)=='\015')) {
      ptr--;
    }
    *ptr++ = ' ';
    *ptr-- = '\0';
    if ((vflag = UUValidData (line, encoding, bhflag)) != UU_ENCODED) {
      *ptr  = '\0';
      vflag = 0;
    }
  }
  return vflag;
}

/*
 * Decode a single encoded line using method
 */

size_t
UUDecodeLine (char *s, char *d, int method)
{
  int i, j, c, cc, count=0, z1, z2, z3, z4;
  static int leftover=0;
  int *table;

  /*
   * for re-initialization
   */

  if (s == NULL || d == NULL) {
    leftover = 0;
    return 0;
  }

  /*
   * To shut up gcc -Wall
   */
  z1 = z2 = z3 = z4 = 0;

  if (method == UU_ENCODED || method == XX_ENCODED) {
    if (method == UU_ENCODED)
      table = UUxlat;
    else
      table = XXxlat;

    i = table [ACAST(*s++)];
    j = UUxlen[i] - 1;

    while(j > 0) {
      c  = table[ACAST(*s++)] << 2;
      cc = table[ACAST(*s++)];
      c |= (cc >> 4);

      if(i-- > 0)
	d[count++] = c;
      
      cc <<= 4;
      c    = table[ACAST(*s++)];
      cc  |= (c >> 2);
      
      if(i-- > 0)
	d[count++] = cc;
      
      c <<= 6;
      c |= table[ACAST(*s++)];
      
      if(i-- > 0)
	d[count++] = c;
      
      j -= 4;
    }
  }
  else if (method == B64ENCODED) {
    if (leftover) {
      strcpy (uuncdl_fulline+leftover, s);
      leftover = 0;
      s        = uuncdl_fulline;
    }

    while ((z1 = B64xlat[ACAST(*s)]) != -1) {
      if ((z2 = B64xlat[ACAST(*(s+1))]) == -1) break;
      if ((z3 = B64xlat[ACAST(*(s+2))]) == -1) break;
      if ((z4 = B64xlat[ACAST(*(s+3))]) == -1) break;

      d[count++] = (z1 << 2) | (z2 >> 4);
      d[count++] = (z2 << 4) | (z3 >> 2);
      d[count++] = (z3 << 6) | (z4);

      s += 4;
    }
    if (z1 != -1 && z2 != -1 && *(s+2) == '=') {
      d[count++] = (z1 << 2) | (z2 >> 4);
      s+=2;
    }
    else if (z1 != -1 && z2 != -1 && z3 != -1 && *(s+3) == '=') {
      d[count++] = (z1 << 2) | (z2 >> 4);
      d[count++] = (z2 << 4) | (z3 >> 2);
      s+=3;
    }
    while (B64xlat[ACAST(*s)] != -1)
      uuncdl_fulline[leftover++] = *s++;
  }
  else if (method == BH_ENCODED) {
    if (leftover) {
      strcpy (uuncdl_fulline+leftover, s);
      leftover = 0;
      s        = uuncdl_fulline;
    }
    else if (*s == ':')
      s++;

    while ((z1 = BHxlat[ACAST(*s)]) != -1) {
      if ((z2 = BHxlat[ACAST(*(s+1))]) == -1) break;
      if ((z3 = BHxlat[ACAST(*(s+2))]) == -1) break;
      if ((z4 = BHxlat[ACAST(*(s+3))]) == -1) break;

      d[count++] = (z1 << 2) | (z2 >> 4);
      d[count++] = (z2 << 4) | (z3 >> 2);
      d[count++] = (z3 << 6) | (z4);

      s += 4;
    }
    if (z1 != -1 && z2 != -1 && *(s+2) == ':') {
      d[count++] = (z1 << 2) | (z2 >> 4);
      s+=2;
    }
    else if (z1 != -1 && z2 != -1 && z3 != -1 && *(s+3) == ':') {
      d[count++] = (z1 << 2) | (z2 >> 4);
      d[count++] = (z2 << 4) | (z3 >> 2);
      s+=3;
    }
    while (BHxlat[ACAST(*s)] != -1)
      uuncdl_fulline[leftover++] = *s++;
  }
  else if (method == YENC_ENCODED) {
    while (*s) {
      if (*s == '=') {
	if (*++s != '\0') {
	  d[count++] = (char) ((int) *s - 64 - 42);
	  s++;
	}
      }
      else if (*s == '\n' || *s == '\r') {
	s++; /* ignore */
      }
      else {
	d[count++] = (char) ((int) *s++ - 42);
      }
    }
  }

  return count;
}

/*
 * ``Decode'' Quoted-Printable text
 */

int
UUDecodeQP (FILE *datain, FILE *dataout, int *state,
	    long maxpos, int method, int flags,
	    char *boundary)
{
  char *line=uugen_inbuffer, *p1, *p2;
  int val;

  uulboundary = -1;

  while (!feof (datain) && 
	 (ftell(datain)<maxpos || flags&FL_TOEND ||
	  (!(flags&FL_PROPER) && uu_fast_scanning))) {
    if (_FP_fgets (line, 255, datain) == NULL)
      break;
    if (ferror (datain)) {
      UUMessage (uunconc_id, __LINE__, UUMSG_ERROR,
		 uustring (S_SOURCE_READ_ERR),
		 strerror (uu_errno = errno));
      return UURET_IOERR;
    }
    line[255] = '\0';

    if (boundary && line[0]=='-' && line[1]=='-' &&
	strncmp (line+2, boundary, strlen (boundary)) == 0) {
      if (line[strlen(boundary)+2]=='-')
	uulboundary = 1;
      else
	uulboundary = 0;
      return UURET_OK;
    }

    if (UUBUSYPOLL(ftell(datain)-progress.foffset,progress.fsize)) {
      UUMessage (uunconc_id, __LINE__, UUMSG_NOTE,
		 uustring (S_DECODE_CANCEL));
      return UURET_CANCEL;
    }

    p1 = p2 = line;

    while (*p2) {
      while (*p2 && *p2 != '=')
	p2++;
      if (*p2 == '\0')
	break;
      *p2 = '\0';
      fprintf (dataout, "%s", p1);
      p1  = ++p2;

      if (isxdigit (*p2) && isxdigit (*(p2+1))) {
	val  = ((isdigit(*p2))    ?  (*p2-'0')   : (tolower(*p2)-'a'+10)) << 4;
	val |= ((isdigit(*(p2+1)))?(*(p2+1)-'0') : (tolower(*(p2+1))-'a'+10));

	fputc (val, dataout);
	p2 += 2;
	p1  = p2;
      }
      else if (*p2 == '\012' || *(p2+1) == '\015') {
	/* soft line break */
	*p2 = '\0';
	break;
      }
      else {
	/* huh? */
	fputc ('=', dataout);
      }
    }
    /*
     * p2 points to a nullbyte right after the CR/LF/CRLF
     */
    val = 0;
    while (p2>p1 && isspace (*(p2-1))) {
      if (*(p2-1) == '\012' || *(p2-1) == '\015')
	val = 1;
      p2--;
    }
    *p2 = '\0';

    /*
     * If the part ends directly after this line, the data does not end
     * with a linebreak. Or, as the docs put it, "the CRLF preceding the
     * encapsulation line is conceptually attached to the boundary.
     * So if the part ends here, don't print a line break"
     */
    if (val && (!feof (datain) && 
		(ftell(datain)<maxpos || flags&FL_TOEND ||
		 (!(flags&FL_PROPER) && uu_fast_scanning))))
      fprintf (dataout, "%s\n", p1);
    else
      fprintf (dataout, "%s", p1);
  }
  return UURET_OK;
}

/*
 * ``Decode'' plain text. Our job is to properly handle the EOL sequence
 */

int
UUDecodePT (FILE *datain, FILE *dataout, int *state,
	    long maxpos, int method, int flags,
	    char *boundary)
{
  char *line=uugen_inbuffer, *ptr;

  uulboundary = -1;

  while (!feof (datain) && 
	 (ftell(datain)<maxpos || flags&FL_TOEND ||
	  (!(flags&FL_PROPER) && uu_fast_scanning))) {
    if (_FP_fgets (line, 255, datain) == NULL)
      break;
    if (ferror (datain)) {
      UUMessage (uunconc_id, __LINE__, UUMSG_ERROR,
		 uustring (S_SOURCE_READ_ERR),
		 strerror (uu_errno = errno));
      return UURET_IOERR;
    }
    line[255] = '\0';

    if (boundary && line[0]=='-' && line[1]=='-' &&
	strncmp (line+2, boundary, strlen (boundary)) == 0) {
      if (line[strlen(boundary)+2]=='-')
	uulboundary = 1;
      else
	uulboundary = 0;
      return UURET_OK;
    }

    if (UUBUSYPOLL(ftell(datain)-progress.foffset,progress.fsize)) {
      UUMessage (uunconc_id, __LINE__, UUMSG_NOTE,
		 uustring (S_DECODE_CANCEL));
      return UURET_CANCEL;
    }

    ptr = line + strlen (line);

    while (ptr>line && (*(ptr-1) == '\012' || *(ptr-1) == '\015'))
      ptr--;


    /*
     * If the part ends directly after this line, the data does not end
     * with a linebreak. Or, as the docs put it, "the CRLF preceding the
     * encapsulation line is conceptually attached to the boundary.
     * So if the part ends here, don't print a line break"
     */
    if ((*ptr == '\012' || *ptr == '\015') &&
	(ftell(datain)<maxpos || flags&FL_TOEND || flags&FL_PARTIAL ||
	 !boundary || (!(flags&FL_PROPER) && uu_fast_scanning))) {
      *ptr = '\0';
      fprintf (dataout, "%s\n", line);
    }
    else {
      *ptr = '\0';
      fprintf (dataout, "%s", line);
    }
  }
  return UURET_OK;
}

/*
 * Decode a single field using method. For the moment, this supports
 * Base64 and Quoted Printable only, to support RFC 1522 header decoding.
 * Quit when seeing the RFC 1522 ?= end marker.
 */

int
UUDecodeField (char *s, char *d, int method)
{
  int z1, z2, z3, z4;
  int count=0;

  if (method == B64ENCODED) {
    while ((z1 = B64xlat[ACAST(*s)]) != -1) {
      if ((z2 = B64xlat[ACAST(*(s+1))]) == -1) break;
      if ((z3 = B64xlat[ACAST(*(s+2))]) == -1) break;
      if ((z4 = B64xlat[ACAST(*(s+3))]) == -1) break;

      d[count++] = (z1 << 2) | (z2 >> 4);
      d[count++] = (z2 << 4) | (z3 >> 2);
      d[count++] = (z3 << 6) | (z4);

      s+=4;
    }
    if (z1 != -1 && z2 != -1 && *(s+2) == '=') {
      d[count++] = (z1 << 2) | (z2 >> 4);
      s+=2;
    }
    else if (z1 != -1 && z2 != -1 && z3 != -1 && *(s+3) == '=') {
      d[count++] = (z1 << 2) | (z2 >> 4);
      d[count++] = (z2 << 4) | (z3 >> 2);
      s+=3;
    }
  }
  else if (method == QP_ENCODED) {
    while (*s && (*s != '?' || *(s+1) != '=')) {
      while (*s && *s != '=' && (*s != '?' || *(s+1) != '=')) {
	d[count++] = *s++;
      }
      if (*s == '=') {
	if (isxdigit (*(s+1)) && isxdigit (*(s+2))) {
	  d[count]  = (isdigit (*(s+1)) ? (*(s+1)-'0') : (tolower (*(s+1))-'a'+10)) << 4;
	  d[count] |= (isdigit (*(s+2)) ? (*(s+2)-'0') : (tolower (*(s+2))-'a'+10));
	  count++;
	  s+=3;
	}
	else if (*(s+1) == '\012' || *(s+1) == '\015') {
	  s+=2;
	}
	else {
	  d[count++] = *s++;
	}
      }
    }
  }
  else {
    return -1;
  }

  d[count] = '\0';
  return count;
}

int
UUDecodePart (FILE *datain, FILE *dataout, int *state,
	      long maxpos, int method, int flags,
	      char *boundary)
{
  char *line=uugen_fnbuffer, *oline=uuncdp_oline;
  int warning=0, vlc=0, lc[2], hadct=0;
  int tc=0, tf=0, vflag, haddata=0, haddh=0;
  long yefilesize=0, yepartends=0;
  crc32_t yepartcrc=crc32(0L, Z_NULL, 0);
  static crc32_t yefilecrc=0;
  static int bhflag=0;
  size_t count=0;
  size_t yepartsize=0;
  char *ptr;

  if (datain == NULL || dataout == NULL) {
    yefilecrc = crc32(0L, Z_NULL, 0);
    bhflag = 0;
    return UURET_OK;
  }

  /*
   * Use specialized functions for QP_ENCODED and PT_ENCODED plaintext
   */

  if (method == QP_ENCODED)
    return UUDecodeQP (datain, dataout, state, maxpos,
		       method, flags, boundary);
  else if (method == PT_ENCODED)
    return UUDecodePT (datain, dataout, state, maxpos,
		       method, flags, boundary);

  lc[0] = lc[1] = 0;
  vflag = 0;

  uulboundary = -1;

  if (method == YENC_ENCODED) {
    *state = BEGIN;
  }

  while (!feof (datain) && *state != DONE && 
	 (ftell(datain)<maxpos || flags&FL_TOEND || maxpos==-1 ||
	  (!(flags&FL_PROPER) && uu_fast_scanning))) {
    if (_FP_fgets (line, 255, datain) == NULL)
      break;

    if (ferror (datain)) {
      UUMessage (uunconc_id, __LINE__, UUMSG_ERROR,
		 uustring (S_SOURCE_READ_ERR),
		 strerror (uu_errno = errno));
      return UURET_IOERR;
    }

    if (line[0]=='\015' || line[0]=='\012') { /* Empty line? */
      if (*state == DATA &&
	  (method == UU_ENCODED || method == XX_ENCODED))
	*state = END;

      /*
       * if we had a whole block of valid lines before, we reset our
       * 'valid data' flag, tf. Without this 'if', we'd break decoding
       * files with interleaved blank lines. The value of 5 is chosen
       * quite arbitrarly.
       */

      if (vlc > 5)
	tf = tc = 0;
      vlc = 0;
      continue;
    }
    
    /*
     * Busy Polls
     */

    if (UUBUSYPOLL(ftell(datain)-progress.foffset,progress.fsize)) {
      UUMessage (uunconc_id, __LINE__, UUMSG_NOTE,
		 uustring (S_DECODE_CANCEL));
      return UURET_CANCEL;
    }

    /*
     * try to make sense of data
     */

    line[255] = '\0'; /* For Safety of string functions */
    count     =  0;

    if (boundary && line[0]=='-' && line[1]=='-' &&
	strncmp (line+2, boundary, strlen (boundary)) == 0) {
      if (line[strlen(boundary)+2]=='-')
	uulboundary = 1;
      else
	uulboundary = 0;
      return UURET_OK;
    }

    /*
     * Use this pseudo-handling only if !FL_PROPER
     */

    if ((flags&FL_PROPER) == 0) {
      if (strncmp (line, "BEGIN", 5) == 0 &&
	  _FP_strstr  (line, "CUT HERE")  && !tf) { /* I hate these lines */
	tc = tf = vlc = 0;
	continue;
      }
      /* MIME body boundary */
      if (line[0] == '-' && line[1] == '-' && method == B64ENCODED) {
	if ((haddata || tc) && (haddh || hadct)) {
	  *state = DONE;
	  vlc   = 0;
	  lc[0] = lc[1] = 0;
	  continue;
	}
	hadct = 0;
	haddh = 1;
	continue;
      }
      if (_FP_strnicmp (line, "Content-Type", 12) == 0)
	hadct = 1;
    }

    if (*state == BEGIN) {
      if ((method == UU_ENCODED || method == XX_ENCODED) &&
	  (strncmp      (line, "begin ",       6) == 0 ||
	   _FP_strnicmp (line, "<pre>begin ", 11) == 0)) { /* for LYNX */
	*state = DATA;
	continue;
      }
      else if (method == BH_ENCODED && line[0] == ':') {
	if (UUValidData (line, BH_ENCODED, &bhflag) == BH_ENCODED) {
	  bhflag = 0;
	  *state = DATA;
	}
	else
	  continue;
      }
      else if (method == YENC_ENCODED &&
	       strncmp (line, "=ybegin ", 8) == 0 &&
	       _FP_strstr (line, " name=") != NULL) {
	*state = DATA;

	if ((ptr = _FP_strstr (line, " size=")) != NULL) {
	  ptr += 6;
	  yefilesize = atoi (ptr);
	}
	else {
	  yefilesize = -1;
	}

	if (_FP_strstr (line, " part=") != NULL) {
	  if (_FP_fgets (line, 255, datain) == NULL) {
	    break;
	  }

	  if ((ptr = _FP_strstr (line, " end=")) == NULL) {
	    break;
	  }
       
	  yepartends = atoi (ptr + 5);