uuscan.c

来自「UUDeview是一个编码解码器」· C语言 代码 · 共 2,492 行 · 第 1/5 页

	result->begin = 1;
      }

      /*
       * Don't want auto-detection
       */

      result->uudet = YENC_ENCODED;
      continue;
    }

    if (strncmp (line, "=yend ", 6) == 0 &&
	result->uudet == YENC_ENCODED) {
      if (yepartends == 0 || yepartends >= yefilesize) {
	result->end = 1;
      }
#if 0
      if (!uu_more_mime)
	break;
#endif
      continue;
    }

    /*
     * if we haven't yet found anything encoded, try to find something
     */

    if (!(result->uudet)) {
      /*
       * Netscape-Repair code is the same as in uunconc.c
       */

      if ((vflag = UUValidData (line, 0, &bhflag)) == 0 && !ismime)
	vflag = UURepairData (datei, line, 0, &bhflag);

      /*
       * Check data against all possible encodings
       */

      islen[iscnt%10] = strlen(line);
      isb64[iscnt%10] = (UUValidData (line, B64ENCODED, &bhflag)==B64ENCODED);
      isuue[iscnt%10] = (UUValidData (line, UU_ENCODED, &bhflag)==UU_ENCODED);
      isxxe[iscnt%10] = (UUValidData (line, XX_ENCODED, &bhflag)==XX_ENCODED);
      isbhx[iscnt%10] = (UUValidData (line, BH_ENCODED, &bhflag)==BH_ENCODED);

      /*
       * If we've got a first valid encoded line, we get suspicious if
       * it's shorter than, say, 40 characters.
       */

      if (vflag == B64ENCODED &&
	  (dflag == 0 || encoding != B64ENCODED) &&
	  strlen (line) < 40 && !result->begin && !uu_tinyb64) {
	isb64[iscnt%10] = 0;
	vflag = 0;
      }

      if ((vflag == UU_ENCODED || vflag == XX_ENCODED) &&
	      (dflag == 0 || encoding != vflag) &&
	      strlen (line) < 40 && !result->begin) {
	isuue[iscnt%10] = isxxe[iscnt%10] = 0;
	vflag = 0;
      }

      iscnt++;

      /*
       * Ah, so we got an encoded line? How interesting!
       */

      if (vflag) {
	/*
	 * For BinHex data, we can use the initial colon ':' as begin
	 * and the terminating colon as ':'.
	 * If (vflag && !bhflag), this is the last line,
	 */
	if (vflag == BH_ENCODED) {
	  if (line[0] == ':' && result->end) {
	    fseek (datei, oldposition, SEEK_SET);
	    break;
	  }
	  if (line[0] == ':')
	    result->begin = 1;
	  if (bhflag == 0) {
	    result->uudet = BH_ENCODED;
	    result->end   = 1;
	  }
	}
	/*
	 * For BinHex files, the file name is encoded in the first encoded
	 * data bytes. We try to extract it here
	 */
	if (vflag == BH_ENCODED && bhnf == 0 && result->filename == NULL) {
	  if (bhdsp == bhds2 ||
	      ((bhdsp-bhds2) <= (int) bhds2[0] &&
	       (bhdsp-bhds2) <  256)) { 
	    dcc = UUDecodeLine (line, bhds1, BH_ENCODED);
	    UUbhdecomp (bhds1, bhdsp, &bhl, &bhrpc,
			dcc, 256-(bhdsp-bhds2), &bhopc);
	    bhdsp += bhopc;
	  }
	  if ((bhdsp-bhds2) > (int) bhds2[0] && bhds2[0]>0 &&
	      result->filename==NULL) {
	    memcpy (bhds1, bhds2+1, (int) bhds2[0]);
	    bhds1[(int)bhds2[0]]='\0';
	    result->filename = _FP_strdup (bhds1);
	    bhnf             = 1;
	  }
	  else if (bhdsp-bhds2 >= 256 && bhds2[0]>0) {
	    memcpy (bhds1, bhds2+1, 255);
	    bhds1[255]       = '\0';
	    result->filename = _FP_strdup (bhds1);
	    bhnf             = 1;
	  }
	  else if (bhds2[0] <= 0)
	    bhnf = 1;
	}

	/*
	 * We accept an encoding if it has been true for four consecutive
	 * lines. Check the is<enc> arrays to avoid mistaking one encoding
	 * for the other. Uuencoded data is rather easily mistaken for
	 * Base 64. If the data matches more than one encoding, we need to
	 * scan further.
	 *
	 * Since text can also rather easily be mistaken for UUencoded
	 * data if it just happens to have 4 lines in a row that have the
	 * correct first character for the length of the line, we also add
	 * a check that the first 3 lines must be the same length, and the
	 * 4th line must be less than or equal to that length. (since
	 * uuencoders use the same length for all lines except the last,
	 * this shouldn't increase the minimum size of UUdata we can
	 * detect, as it would if we tested all 4 lines for being the same
	 * length.)  - Matthew Mueller, 20030109
	 */

	if (iscnt > 3) {
	  cbb64 = (isb64[(iscnt-1)%10] && isb64[(iscnt-2)%10] &&
		   isb64[(iscnt-3)%10] && isb64[(iscnt-4)%10]);
	  cbuue = (isuue[(iscnt-1)%10] && isuue[(iscnt-2)%10] &&
		   isuue[(iscnt-3)%10] && isuue[(iscnt-4)%10] &&
		   islen[(iscnt-1)%10] <= islen[(iscnt-2)%10] &&
		   islen[(iscnt-2)%10] == islen[(iscnt-3)%10] &&
		   islen[(iscnt-3)%10] == islen[(iscnt-4)%10]);
	  cbxxe = (isxxe[(iscnt-1)%10] && isxxe[(iscnt-2)%10] &&
		   isxxe[(iscnt-3)%10] && isxxe[(iscnt-4)%10] &&
		   islen[(iscnt-1)%10] <= islen[(iscnt-2)%10] &&
		   islen[(iscnt-2)%10] == islen[(iscnt-3)%10] &&
		   islen[(iscnt-3)%10] == islen[(iscnt-4)%10]);
	  cbbhx = (isbhx[(iscnt-1)%10] && isbhx[(iscnt-2)%10] &&
		   isbhx[(iscnt-3)%10] && isbhx[(iscnt-4)%10]);
	}
	else {
	  cbb64 = cbuue = cbxxe = cbbhx = 0;
	}

	if (cbb64 && !cbuue && !cbxxe && !cbbhx) {
	  result->uudet = B64ENCODED;
	}
	else if (!cbb64 && cbuue && !cbxxe && !cbbhx) {
	  result->uudet = UU_ENCODED;
	}
	else if (!cbb64 && !cbuue && cbxxe && !cbbhx) {
	  result->uudet = XX_ENCODED;
	}
	else if (!cbb64 && !cbuue && !cbxxe && cbbhx) {
	  result->uudet = BH_ENCODED;
	}

	if (result->uudet) {
          encoding = dflag = 0;

	  /*
	   * If we're in fast mode, we're told we don't have to look
	   * for anything below, so we can as well break out of every-
	   * thing
	   * We cannot fast-scan if we have a boundary to look after.
	   */

	  if (uu_fast_scanning && boundary == NULL)
	    break;

	  /*
	   * Skip the encoded data. We simply wait for a boundary, for
	   * a header or for an empty line. But we also try to pick up
	   * an "end"
	   */

	  hcount = lcount = 0;

	  while (!feof (datei)) {
	    /*
	     * Make Busy Polls
	     */
	    if (UUBUSYPOLL(ftell(datei),progress.fsize)) {
	      UUMessage (uuscan_id, __LINE__, UUMSG_NOTE,
			 uustring (S_SCAN_CANCEL));
	      *errcode = UURET_CANCEL;
	      break;
	    }

	    oldposition = ftell (datei);
	    if (_FP_fgets (line, 255, datei) == NULL)
	      break;
	    if (ferror (datei))
	      break;

	    line[255] = '\0';

	    /*
	     * Stop scanning at an empty line or a MIME-boundary.
	     */
	    if (IsLineEmpty (line))
	      break;
	    if (boundary && line[0] == '-' && line[1] == '-' &&
		strncmp (line+2, boundary, blen) == 0) {
	      fseek (datei, oldposition, SEEK_SET);
	      break;
	    }
	    else if (line[0] == 'e' && (result->uudet == UU_ENCODED ||
					result->uudet == XX_ENCODED)) {
	      if (strncmp (line, "end", 3) == 0) {
		result->end = 1;
		break;
	      }
	    }
	    else if (line[0] == 'b') {
	      if (strncmp (line, "begin ", 6) == 0) {
		fseek (datei, oldposition, SEEK_SET);
		break;
	      }
	    }

	    if (checkheaders) {
	      if (IsKnownHeader (line)) {
		(void) ScanHeaderLine (datei, line);
		if (hcount == 0)
		  preheaders = oldposition;
		hcount++;
		lcount++;
		if ((hcount >= hlcount.restart) ||
		    (hcount >= hlcount.afterdata && result->uudet)) {
		  /*
		   * Hey, a new header starts here
		   */
		  fseek (datei, preheaders, SEEK_SET);
		  break;
		}
	      }
	      else if (lcount > WAITHEADER) {
		hcount = 0;
		lcount = 0;
	      }
	      else if (hcount) {
		lcount++;
	      }
	    }
	    if (result->uudet == BH_ENCODED) {
	      /* pick up ``EOF'' for BinHex files. Slow :-< */
	      if (line[0] && strchr (line+1, ':') != NULL) {
		result->end = 1;
		bhflag      = 0;
		break;
	      }
	    }
	  }

	  if (ferror (datei) || *errcode == UURET_CANCEL)
	    break;

	  if (line[0] == '-' && line[1] == '-')
	    haddh = 1;

	  /*
	   * Okay, got everything we need. If we had headers or a
	   * boundary, we break out of the outer loop immediately.
	   */

	  if (IsKnownHeader (line) ||
	      (boundary && line[0] == '-' && line[1] == '-' &&
	       strncmp (line+2, boundary, blen) == 0)) {
	    break;
	  }

	  /*
	   * Otherwise, we wait until finding something more interesting
	   * in the outer loop
	   */

	  continue;
	}
	
	/*
	 * Select the encoding with the best "history"
	 */

	cbb64 = isb64[(iscnt-1)%10];
	cbuue = isuue[(iscnt-1)%10];
	cbxxe = isxxe[(iscnt-1)%10];
	cbbhx = isbhx[(iscnt-1)%10];
	dflag = 0;

	if (cbb64 || cbuue || cbxxe || cbbhx) {
	  for (dflag=2; dflag<iscnt && dflag<4; dflag++) {
	    if ((!cbb64 || !isb64[(iscnt-dflag)%10]) &&
		(!cbuue || !isuue[(iscnt-dflag)%10]) &&
		(!cbxxe || !isxxe[(iscnt-dflag)%10]) &&
		(!cbbhx || !isbhx[(iscnt-dflag)%10])) {
	      dflag--;
	      break;
	    }
	    cbb64 &= isb64[(iscnt-dflag)%10];
	    cbuue &= isuue[(iscnt-dflag)%10];
	    cbxxe &= isxxe[(iscnt-dflag)%10];
	    cbbhx &= isbhx[(iscnt-dflag)%10];
	  }
	}

	/*
	 * clear-cut cases
	 */

	if (cbb64 && !cbuue && !cbxxe && !cbbhx) {
	  encoding = B64ENCODED;
	}
	else if (!cbb64 && cbuue && !cbxxe && !cbbhx) {
	  encoding = UU_ENCODED;
	}
	else if (!cbb64 && !cbuue && cbxxe && !cbbhx) {
	  encoding = XX_ENCODED;
	}
	else if (!cbb64 && !cbuue && !cbxxe && cbbhx) {
	  encoding = BH_ENCODED;
	}
	else {
	  encoding = 0;
	}

	/*
	 * Check a few common non-clear-cut cases
	 */

	if (!encoding && cbuue && result->begin) {
	  encoding = UU_ENCODED;
	}
	else if (!encoding && cbxxe && result->begin) {
	  encoding = XX_ENCODED;
	}
	else if (!encoding && cbb64) {
	  encoding = B64ENCODED;
	}
	else if (!encoding && cbuue) {
	  encoding = UU_ENCODED;
	}
	else if (!encoding && cbxxe) {
	  encoding = XX_ENCODED;
	}
	else if (!encoding && cbbhx) {
	  encoding = BH_ENCODED;
	}
      }
      else if (!dontcare) {
	encoding = 0;
        dflag = 0;
	haddh = 0;
      }
    } /* if (!uudet) */
    /*
     * End of scanning loop
     */
  } /* while (!feof (datei)) */

  if (feof (datei))
    oldposition = ftell (datei);

  if (dflag && encoding == B64ENCODED && haddh)
    result->uudet = B64ENCODED;
  else if (dflag && encoding == BH_ENCODED)
    result->uudet = BH_ENCODED;

  /* Base64 doesn't have begin or end, so it was probably XX */
  if (result->uudet == B64ENCODED && result->begin && result->end)
    result->uudet = XX_ENCODED;

  /* Base64 doesn't have begin or end */
  if (result->uudet == B64ENCODED)
    result->begin = result->end = 0;

  /* Base64 and BinHex don't have a file mode */
  if (result->uudet == B64ENCODED || result->uudet == BH_ENCODED ||
      result->uudet == YENC_ENCODED)
    result->mode  = 6*64+4*8+4;

  /*
   * When strict MIME adherance is set, throw out suspicious attachments
   */

  if (uu_more_mime) {
    /*
     * In a MIME message, Base64 should be appropriately tagged
     */

    if (result->uudet == B64ENCODED) {
      result->uudet = 0;
    }

    /*
     * Do not accept incomplete UU or XX encoded messages
     */

    if ((result->uudet != 0 && result->uudet != B64ENCODED) &&
	(!result->begin || !result->end)) {
      result->uudet = 0;
    }
  }

  /*
   * In fast mode, this length will yield a false value. We don't care.
   * This must be checked for in uunconc(), where we usually stop decoding
   * after reaching startpos+length
   */

  if (uu_fast_scanning)
    result->length = progress.fsize-result->startpos;
  else
    result->length = ftell(datei)-result->startpos;

  if (ferror (datei)) {
    *errcode = UURET_IOERR;
    uu_errno = errno;
    return -1;
  }
  if (*errcode != UURET_OK) {
    return -1;
  }

  if (boundary && line[0] == '-' && line[1] == '-' &&
      strncmp (line+2, boundary, blen) == 0)
    return 2;
  else if (boundary && p3 &&
	   line[0] == 'C' && line[1] == 'o' &&
	   _FP_strnicmp (line, "Content-Type:", 13) == 0 &&
	   strcmp (p3, boundary) == 0)
    return 2;
  else if (IsKnownHeader (line))
    return 1;

  return 0;
}

/*
 * This is the main scanning function.
 */

fileread *
ScanPart (FILE *datei, char *fname, int *errcode)
{
  int ecount, hcount, lcount;
  int bhflag, begflag, vflag, blen=0, res;
  long preheaders, prevpos=0, preenc, before;
  char *line=uuscan_spline;
  fileread *result;
  char *ptr1, *ptr2;

  (void) UUDecodeLine (NULL, NULL, 0);          /* init */
  if (datei == NULL || feof (datei)) {
    *errcode = UURET_OK;
    return NULL;
  }

  *errcode = UURET_OK;

  if ((result = (fileread *) malloc (sizeof (fileread))) == NULL) {
    *errcode = UURET_NOMEM;
    return NULL;
  }
  memset (result, 0, sizeof (fileread));
  result->startpos = ftell (datei);
  preheaders       = result->startpos;
  before           = result->startpos;

  /* if this is a new file, reset our scanning state */
  if (sstate.source == NULL || strcmp (fname, sstate.source) != 0) {
    sstate.isfolder  = 1;		/* assume yes            */
    sstate.ismime    = 0;		/* wait for MIME-Version */
    sstate.mimestate = MS_HEADERS;	/* assume headers follow */
    /* mimseqno      = 1; */

    while (mssdepth) {
      mssdepth--;
      UUkillheaders (&(multistack[mssdepth].envelope));
      _FP_free (multistack[mssdepth].source);
    }

    UUkillheaders (&sstate.envelope);
    memset (&sstate.envelope, 0, sizeof (headers));

    _FP_free (sstate.source);
    if ((sstate.source = _FP_strdup (fname)) == NULL) {