comparison src/coding.c @ 23315:86a8b8566369

(decode_eol): While decoding DOS-like eol, handle a single CR code correctly. (shrink_decoding_region) <CODING_CATEGORY_IDX_ISO_7 or CODING_CATEGORY_IDX_ISO_7_TIGHT>: Skip also 8-bit codes.

23314:87acd2b6ce97

	    c = *src++;
	    if (c == '\r')
	      {
		ONE_MORE_BYTE (c);
		if (c != '\n')
		  {
		    if (coding->mode & CODING_MODE_INHIBIT_INCONSISTENT_EOL)
		      {
			result = CODING_FINISH_INCONSISTENT_EOL;
			goto label_end_of_loop_2;
		      }
		    *dst++ = '\r';
		    if (BASE_LEADING_CODE_P (c))
		      coding->fake_multibyte = 1;
		  }
		*dst++ = c;
	      }
	    else if (c == '\n'
		     && (coding->mode & CODING_MODE_INHIBIT_INCONSISTENT_EOL))
	      {
		result = CODING_FINISH_INCONSISTENT_EOL;

	    result = CODING_FINISH_INSUFFICIENT_SRC;
	  label_end_of_loop_2:
	    src = src_base;
	    break;
	  }
	if (result == CODING_FINISH_NORMAL
	    && src < src_end)
	  result = CODING_FINISH_INSUFFICIENT_DST;
      }
      break;

    case CODING_EOL_CR:
      if (coding->mode & CODING_MODE_INHIBIT_INCONSISTENT_EOL)

	case CODING_CATEGORY_IDX_ISO_7_TIGHT:
	  /* We can skip all charactes at the tail except for ESC and
             the following 2-byte at the tail.  */
	  if (eol_conversion)
	    while (begp < endp
		   && (c = endp[-1]) < 0x80 && c != ISO_CODE_ESC && c != '\r')
	      endp--;
	  else
	    while (begp < endp
		   && (c = endp[-1]) < 0x80 && c != ISO_CODE_ESC)
	      endp--;
	  /* Do not consider LF as ascii if preceded by CR, since that
             confuses eol decoding. */
	  if (begp < endp && endp < endp_orig && endp[-1] == '\r' && endp[0] == '\n')
	    endp++;

23315:86a8b8566369

	    c = *src++;
	    if (c == '\r')
	      {
		ONE_MORE_BYTE (c);
		if (c == '\n')
		  *dst++ = c;
		else
		  {
		    if (coding->mode & CODING_MODE_INHIBIT_INCONSISTENT_EOL)
		      {
			result = CODING_FINISH_INCONSISTENT_EOL;
			goto label_end_of_loop_2;
		      }
		    src--;
		    *dst++ = '\r';
		    if (BASE_LEADING_CODE_P (c))
		      coding->fake_multibyte = 1;
		  }
	      }
	    else if (c == '\n'
		     && (coding->mode & CODING_MODE_INHIBIT_INCONSISTENT_EOL))
	      {
		result = CODING_FINISH_INCONSISTENT_EOL;

	    result = CODING_FINISH_INSUFFICIENT_SRC;
	  label_end_of_loop_2:
	    src = src_base;
	    break;
	  }
	if (src < src_end)
	  {
	    if (result == CODING_FINISH_NORMAL)
	      result = CODING_FINISH_INSUFFICIENT_DST;
	    else if (result != CODING_FINISH_INCONSISTENT_EOL
		     && coding->mode & CODING_MODE_LAST_BLOCK)
	      {
		/* This is the last block of the text to be decoded.
		   We flush out all remaining codes.  */
		src_bytes = src_end - src;
		if (dst_bytes && (dst_end - dst < src_bytes))
		  src_bytes = dst_end - dst;
		bcopy (src, dst, src_bytes);
		dst += src_bytes;
		src += src_bytes;
	      }
	  }
      }
      break;

    case CODING_EOL_CR:
      if (coding->mode & CODING_MODE_INHIBIT_INCONSISTENT_EOL)

	case CODING_CATEGORY_IDX_ISO_7_TIGHT:
	  /* We can skip all charactes at the tail except for ESC and
             the following 2-byte at the tail.  */
	  if (eol_conversion)
	    while (begp < endp
		   && (c = endp[-1]) != ISO_CODE_ESC && c != '\r')
	      endp--;
	  else
	    while (begp < endp
		   && (c = endp[-1]) != ISO_CODE_ESC)
	      endp--;
	  /* Do not consider LF as ascii if preceded by CR, since that
             confuses eol decoding. */
	  if (begp < endp && endp < endp_orig && endp[-1] == '\r' && endp[0] == '\n')
	    endp++;