Exploring the CFB File Format 8

  • [アーティクル]

Exploring the CFB File Format 8

Range Lock Sectors

A range lock sector is a part of a CFB file that is used for concurrency and transactions. The range sector will only appear in files greater than 2 GB. As we have seen

in previous posts, there are files with 512-byte sectors, and others with 4096-byte sectors. A 512-byte sector CFB file is limited to 2 GB and *MUST* never be larger

than this size. Conversely, a 4096-byte file can grow as large as 16 terabytes. Therefore, only a 4096-byte CFB file

contains a range lock sector and it will ALWAYS appear at file offsets 0x7FFFFF00-0x7FFFFFFF (inclusive).Therefore, is merely only a small feat to consume the

bytes in question:

# ff bytes are found in the range: 0x7FFFFF00-0x7FFFFFFF (inclusive)
for i in range(0, 255):
   cur_byte = disk[i]

It is merely a good idea just to skip this sector as it will not contain any user-defined data. No other data structure contained in a CFB file points to this sector.The

information found in this segment will be subject to the producer, and *not* the consumer. The producer will use this for various transactions when it comes to the

process of obtaining objects. The way this sector is *actually* used is considered an implementation detail and is constructed differently from app to app as each

one will use this space in a different manner.

 

Please note this sector is *not* important when parsing a cfb file, it is only important for the producer that created the

file. Thus, you could always jump to the offset past the sector in question:

 

file.seek(0x7FFFFFFF)

 

Subsequently, the producer can continue consuming sectors immediately following the range lock sector per the FAT.

Comments

  • Anonymous
    June 17, 2011
    The comment has been removed