Re: compressed save files -- MI400

Hi Dave ,

I could finish my program.
To konw if a descriptor is compressed, i compare values at offset xcc and
x48 in descriptors.
Uncompress the datas was not so difficult i had think(thank?thunk?)
I could find those informations about compressed datas.


Run-length encoding (RLE) eliminates strings of repeated bytes.  With the
RLE algorithm, the first byte after the
compression header is a control byte, known as a string control byte (SCB).
The SCB has the format shown below.

+------------------------------------------------------------------------+
� SCB Format                                                             �
+------------------------------------------------------------------------�
�       � Bit � Content                                                  �
+------------------------------------------------------------------------�
+------------------------------------------------------------------------�
�       � 0-1 � SCB type:                                                �
+-------+-----+----------------------------------------------------------�
�       �     � 00 � Raw data:  the following bytes are uncompressed.    �
�       �     �    � The Count field (bits 2-7) indicates the number of  �
�       �     �    � uncompressed bytes.  If the RU is not exhausted,    �
�       �     �    � another SCB follows n+1 bytes after this SCB.       �
+-------+-----+----+-----------------------------------------------------�
�       �     � 01 � Reserved                                            �
+-------+-----+----+-----------------------------------------------------�
�       �     � 10 � Master-character:  the Count field indicates the    �
�       �     �    � number of space (X'40') characters compressed.  If  �
�       �     �    � the RU is not exhausted, another SCB follows this   �
�       �     �    � master-character SCB.                               �
+-------+-----+----+-----------------------------------------------------�
�       �     � 11 � Duplicated-character:  the character (called the    �
�       �     �    � duplicated character, or DC) that follows this SCB  �
�       �     �    � appears in the raw data in an n-byte run; the       �
�       �     �    � n-byte run is compressed to this SCB-DC pair.  If   �
�       �     �    � the RU is not exhausted, another SCB follows this   �
�       �     �    � SCB-DC pair.                                        �
+------------------------------------------------------------------------�
+------------------------------------------------------------------------�
�       � 2-7 � Count:  indicates the number (n), in binary, of          �
�       �     � uncompressed bytes that follow (in the case of SCB type  �
�       �     � 00) or that should be generated upon decompression of    �
�       �     � this SCB sequence.                                       �
+------------------------------------------------------------------------+

--------------------------------------------------------------------------

SCBs cannot span RUs.  In short:

   If the last SCB in an RU is a raw-data SCB, then all of the raw data to
which it refers must be in that RU.

   The master-character SCB is allowed to be the final byte in an RU.

   If the duplicate-character SCB is the last SCB in the RU, then that SCB
is always the next-to-last byte in the RU,
    the last byte being the DC.

Thanks a lot.