Quasijarus2Gzip: How A Program Starts By Accident
It turns out that recently I got an old *.tar.Z file with source code of various very old Unix programs. I was curious about some of them like a very old version of sendmail, but when I tried to decompress one of the files I found out that I couldn't.
Previously I had already decompressed *.tar.Z files with gzip so the
next thing I did was to see exactly what that *.tar.Z file was. To do
this I used the file command and found out that it was a Quasijarus
Strong Compression Data file. I didn't know what it was so I kept
investigating.
My next step was to find out exactly what the Quasijarus Strong Compression Data format was. First I found a brief description of the format1 which contained links to the Quasijarus project website. These links are no longer online, but thanks to archive.org2 I was able to see the Quasijarus project website3 and specifically the information on the compression format4.
After reading it all, the main points for me were:
- The file format is based on Gzip and both formats only differ in their headers.
- The header of the Quasijarus Strong Compression Data format is only composed by the 0x1f 0xa1 bytes (magic number).
So I decided to make a Quasijarus Strong Compression Data to Gzip converter as it couldn't be very complicated. It also provided to me an excuse to have some fun ;p.
The last step was to find out the format of the header of the Gzip file format. According to its specification5, it has a header of 10 bytes distributed as follows:
- 2 bytes for the magic number 0x1f 0x8b
- 1 byte for the compression algorithm (08 for DEFLATE).
- 1 byte for headers flags that could be set to 0.
- 4 bytes for the file timestamp that can also be set to 0.
- 1 byte for compression flags that can be set to 0 too.
- 1 byte for the operating system identifier (3 for Unix)
With all this information I made my Quasijarus Strong Compression Data to Gzip converter. You can download it here.
(This post is a translation of the original one in Spanish).