📝 This post is part of my “Shorts” series, where each post is concise and hyper-focused on a single concept.

.gitignore is an indispensable part of any Git project. It is used to maintain a list of files/directories that Git should not track. Those files can be anything: build files, tmp directories, log files, etc.

You may also have used this file when working on a software development project. In this post, I want to tell you about some of the patterns that can help you:

1. Avoid weird bugs
2. Make your .gitignore more efficient

For example, did you know that if your .gitignore file contains the following line, then all directories in your project named nestedDir will be ignored. It does not matter if the directory is in the root directory.

nestedDir/

One of the fundamental things to learn about .gitignore is to know that any pattern found in the .gitignore file will be searched for recursively in your repository and then, ignored.

Patterns supported

• Blank lines and lines starting with # are ignored. Lines starting with # can be considered as comments.
• Standard glob patterns are supported (more on this below).
• Patterns starting with / avoid recursivity. For example, a pattern like /rootFile will only ignore the rootFile present in the root of the repository.
• In order to specify a directory, end the pattern with a /. Patterns starting with ! will be negated. For example, if your .gitignore contains *.log (ignoring all log files) but you want to track production.log file specifically, you need to add !production.log to your .gitignore.

Examples

# Ignore all log files present anywhere in the repository
*.log

# But, track the production.log file
!production.log

# Ignore the config.yaml file present in the root directory ONLY
/config.yaml

# Ignore all directories named `posts` present anywhere in the repository
posts/

# Ignore all txt files which are present directly inside manual directory
manual/*.txt

# Ignore all exe files which are present inside the bin directory and its subdirectories
bin/**/*.exe

You can also look up the list of good .gitignore file examples here.

Understanding basic glob patterns

• Asterisk (*) matches zero or more characters. Example: *.log would match any file ending with .log.
• Question mark (?) matches one character. Example: ?.log would match files like 1.log, p.log but not match test.log.
• If you list characters inside square brackets ([]), it would match all files/directories containing any of those characters. Example: [abc].log will match a.log, b.log, c.log but not abc.log or d.log.
• Square brackets also supports ranges. So, [12345] is equivalent to [1-5].
• Double asterisks (**) will match nested directories. Example: a/**/z will match a/z, a/b/z, a/b/c/z and so on.

One more thing

It is not necessary to have only one .gitignore file per repository. You can create separate .gitignore files for subdirectories. The rules in a .gitignore will be applied on the files under the directory where each .gitignore is located.

What's your ONE takeaway?

When I learned about these patterns, one of the biggest takeaways for me was that the rules are applied recursively on all subdirectories. What's yours? Tell me about it on my email.

Further exploration

Run man gitignore in your terminal to learn more about using .gitignore.

Sources

• https://git-scm.com/book/en/v2/Git-Basics-Recording-Changes-to-the-Repository