Software:Quantum compression

From HandWiki

The Quantum compression format is a little-known data compression method created by David Stafford of Cinematronics, LLC. It was licensed to Borland, Microsoft and Novell.[1] Quantum is one of the possible compression methods in a Microsoft Windows CAB archive.[2] Quantum uses an extended LZ77 compression algorithm.[3] Quantum archive files are named with the filename extension .Q by convention.

Quantum archiver

PAQ.EXE is the Quantum compressor for MS-DOS and UNPAQ.EXE is the decompressor. These are 32-bit applications using a DOS extender from Borland.

QWIN.EXE is the Quantum decompressor for 16-bit Windows. It is much faster than the MS-DOS decompressor and has more features such as selective decompression.

A command-line tool COMPRESS.EXE supplied with Microsoft Windows or its resource kits up until Windows XP (up until version 5.00.2134.1 of COMPRESS.EXE) is able to compress files using the Quantum algorithm.

All the Quantum compressors and decompressors require at least a 386 CPU in order to run.

The last known version of Quantum released by Cinematronics is 0.97.

Quantum archive file structure

File header

Bytes Description
2 Quantum signature: 0x44 0x53
1 Quantum major version number
1 Quantum minor version number
2 number of files within this archive
1 table size required for decompression
1 compression flags

This is immediately followed by the list of files:

Bytes Description
variable length of file name, see (*) below
variable file name, variable length string, not zero-terminated
variable length of comment field, see (*) below
variable comment field, variable length string, not zero-terminated
4 fully expanded file size in bytes
2 file time (DOS format)
2 file date (DOS format)

Immediately following the list of files is the compressed data.

(*) Strings are prefixed with their length. If the length is less than 128 then it is stored directly in one byte. If it is greater than 127 then the high bit of the first byte is set to 1 and the remaining fifteen bits contain the actual length in big-endian format.

References

External links