gzip Command Cheat Sheet

gzip (GNU zip) reduces the size of named files using Lempel-Ziv coding (LZ77). Included in almost every Linux distribution, it is the standard for single-file compression.

Synopsis

gzip [options] [file...]
gunzip [file...]

Basic Usage

Compress a File

gzip bigfile.txt

- Compresses bigfile.txt -> bigfile.txt.gz. - Replaces the original file. - Preserves timestamp, mode, and ownership.

Decompress a File

gzip -d bigfile.txt.gz
# OR
gunzip bigfile.txt.gz

- Restores bigfile.txt.gz -> bigfile.txt. - Removes the .gz file.

Keep Original File (-k)

If you don't want to delete the source file.

gzip -k bigfile.txt

Result: bigfile.txt AND bigfile.txt.gz both exist.

Compression Levels

Trade off between speed and size.

Flag	Meaning	Speed	Ratio
`-1` (`--fast`)	Fastest compression	Very Fast	Lower
`-6`	Default	Balanced	Good
`-9` (`--best`)	Best compression	Slow	Best

gzip -9 backup.iso

Advanced Usage

Recursive Compression (-r)

⚠️ Gotcha Alert: gzip -r folder does NOT create a single folder.gz archive (like zip or tar). Instead, it recursively descends into the directory and compresses every single file individually.

gzip -r my_logs/

Result: my_logs/access.log.gz, my_logs/error.log.gz, etc.

To create a single archive, use tar first:

tar -czf my_logs.tar.gz my_logs/

Verbose Output (-v)

See how much space you saved.

gzip -v file.txt
# Output: file.txt: 65.4% -- replaced with file.txt.gz

Test Integrity (-t)

Check if a compressed file is corrupted.

gzip -t archive.gz

(Silent if OK, prints error if corrupted).

View Details (-l)

List compression info without decompressing.

gzip -l archive.gz

Output:

compressed  uncompressed  ratio uncompressed_name
      1024          4096  75.0% archive

Piping and Redirection

gzip works great with standard streams (stdin / stdout).

Compress Stream to File

cat database.sql | gzip > database.sql.gz

Decompress Stream to Pipe

gunzip -c log.gz | grep "Error"

Force Output to Stdout (-c)

Useful for keeping the original file without using -k (on older systems) or redirecting elsewhere.

gzip -c file.txt > /backup/file.txt.gz

The "z" Tools using zlib

You don't always need to decompress files to read them.

Command	Usage
`zcat`	Print contents of gz (like `cat`)
`zless` / `zmore`	Page through gz (like `less`)
`zgrep`	Search inside gz (like `grep`)
`zdiff`	Compare gz files

# Search logs without extracting
zgrep "404 Not Found" access.log.gz

Concatenation

Gzip files can be concatenated.

gzip -c part1 > full.gz
gzip -c part2 >> full.gz

When you run gunzip full.gz, it decompresses part1 followed immediately by part2.

Comparison

Tool	Algorithm	Speed	Ratio	Use Case
`gzip`	DEFLATE	Fast	Good	General, Logs, Web
`bzip2`	Burrows-Wheeler	Slow	Better	Archived code
`xz`	LZMA2	Very Slow	Best	Distribution packages
`pigz`	Parallel Gzip	Very Fast	Good	Multi-core systems

Tip: If you have a multi-core CPU and compress large files, install pigz and use it exactly like gzip.

Exit Status

Code	Meaning
0	Success
1	Error
2	Warning

Notes

Suffix: Default is .gz. gzip refuses to compress files that already have this suffix.
Force: Use -f to overwrite existing output files or compress symlinks.