Compression Programs

The `compress` program

The compress program performs a reversible transformation on a file such that the result is (almost always) a smaller file.
If bigfile.txt is the name of a large file, we can execute:
```
		compress bigfile.txt 
```
and this will create a new file named bigfile.txt.Z. The original file, bigfile.txt, is deleted.
Text files (English language text, program source code, etc.) are typically reduced to 40% of their original size. The figures vary for other kinds of files -- executable files often yield very little compression.

Occasionally, compress cannot make the file smaller at all. In this case, compress refuses to create the output file and leaves the original file unchanged.
To recover the original file again, use the uncompress command. For example:
```
		uncompress bigfile.txt.Z 
```
will create the file bigfile.txt (with the same contents as before) and will delete bigfile.txt.Z.
If you wish to use the original contents of a compressed file without destroying the compressed file, you can use uncompress as a Unix filter program. E.g., to browse the contents of our compressed file, we can execute:
```
		uncompress < bigfile.txt.Z | more 
```
This is such a common operation that an additional Unix command named zcat has been created. For example:
```
		zcat bigfile.txt.Z | more 
```
is equivalent to the previous command.
You cannot compress a file twice and make it even smaller. The compress program refuses to even attempt compression on a file whose name ends with the .Z suffix. And if you fool compress by renaming the file to lose the suffix, the program simply won't be able to achieve any noticeable compression.

The gzip Program

There are other compression programs available for Unix. The commonest is gzip. It is public-domain software.

Here are some examples of its usage:

gzip bigfile.txt creates a compressed file named bigfile.txt.gz.
gunzip bigfile.txt.gz recovers the original file.
gunzip < bigfile.txt.gz | more allows you to browse the file without
destroying the compressed file.

Gzip usually produces better compression than compress, and it executes significantly faster.

Using `compress` or `gzip` with `tar`

Compress and gzip are often used to reduce the sizes of tar archive files before they are shipped over the internet.

A compressed tar archive can be created in a single Unix command pipeline as in the following example:

	  tar cvf - README.txt MyProject | compress > MyProject.tar.Z

The hyphen indicates that the output from the tar command is to go to standard output, not into a file.

After the compressed tar file has been received at the other end, we can recover all the files using this command:

	  zcat MyProject.tar.Z | tar xvf -

[Of course, gzip and gunzip can be used instead.]

`gzip bigfile.txt`	creates a compressed file named `bigfile.txt.gz`.
`gunzip bigfile.txt.gz`	recovers the original file.
`gunzip < bigfile.txt.gz \| more`	allows you to browse the file without destroying the compressed file.

Compression Programs

The compress program

The gzip Program

Using compress or gzip with tar

The `compress` program

Using `compress` or `gzip` with `tar`