Datenbanktypen
Aus AOWiki
Inhaltsverzeichnis |
[bearbeiten] The message
The message consists of the following parts:
[Bytes 1-2] - This tells the client that its the start of a message that needs to be converted. Its always '~&'.
[Bytes 3-7] - This is the first ID
[Bytes 8-12] - This is the second ID
[Bytes 13-n] - If the string needs some arguments they are placed here. The format of these arguments looks like this:
[datatype][length][data]
datatype (1 byte) -
's' = string,
'i' = integer
.. there might be more here (like maybe 'd' for double) but im not sure.
length (1 byte) - This is the length of the data including the length char (so its length of data + 1).
data ('length-1' bytes) - This is the actual argument data.
The last byte of the message is '~', this is the stop-char.
[bearbeiten] The ID's
The client use these ID's to retrieve the real message from the database. Before we can start looking in the db for these id's we need to convert them to a 4
byte value which is done like this:
Each ID-byte can hold values 33 to 117 so the first thing we need to do is subtract 33 from each byte so we get values 0 to 84. Now we just calculate the decimal value of these 5 bytes, this can be done with the following formula:
Decimal value = (byte1 * (85^4)) + (byte2 * (85^3)) + (byte3 * (85^2)) + (byte4 * (85^1)) + (byte5 * (85^0))
TNow lets get the 4 bytes from this value.. i wrote a little function to do it for me:
$Byte1 = intval($decimal_value / pow(256,3)); $rest = $id_decimal - $Byte1*pow(256,3);
$Byte2 = intval($rest / pow(256,2)); $rest = $rest - $Byte2*pow(256,2);
$Byte3 = intval($rest / pow(256,1)); $Byte4 = $rest - $Byte3*pow(256,1);
The ID's in the database is stored in reversed byte order so $byte4 here is the most significant byte.
[bearbeiten] The databases
text.mdb and ctext.ldb - Both these files can be used to retrieve the real message, not exactly sure why there is 2 different files.
[bearbeiten] Ctext.ldb
This one was quite easy to understand. It's structure looks like this:
[Byte 1-4] - This is the header ('MLDB').
[byte 5-n] - This is the messages stored like this:
First 4 bytes = First ID
Next 4 bytes = Second ID
Next 4 bytes = String length
Next 'length' bytes = Message.
Note. The length is also stored in reversed byte order so make sure you reverse it
[bearbeiten] text.mdb
This one was a bit harder to understand, i don't think i fully understand it yet but i know how to pull messages from it.
Structure: [Byte 1-8] - Header ('MMDB' + 4 bytes that might not really be a part of the header) After the header comes a list of the first ID's and a position in the file where the second ID can be found. The first ID and fileposition are stored like this: [First ID (4 bytes)][Fileposition (4 bytes)]...
Now lets seek down to the position of the second ID. Same thing here 2 chunks of 4 byte values but the second chunk here contains the fileposition of the message. [Second ID (4 bytes)][Fileposition (4 bytes)]...
Seeking down to the message...
Now all we gotta do is read the message and be happy . I don't think the length of the string is stored anywhere but all strings ends with chr(0) so its no biggie.
The string we get here is a formatting string so it might contain '%s', '%d' and so on... So just use the arguments from the original message and use sprintf($formatting_string, $arg1,$arg2...) to get the real string
