Skip to main content

​
Functions

Functions in Bash are named blocks of shell commands. Bash executes the block in the current shell environment when we call the function. Two common ways to define them:
Both work in Bash. The second form is also portable to POSIX-compatible shells, while the function keyword is Bash-specific. We call a function just by its name: my_func.

​
Local Variables

Assignments inside a function affect global variables by default. If we want a variable local to the function call, we use the local keyword:
The local var temporarily hides the global var, so changing it does not change the global value. Bash uses dynamic scoping, which means functions called from my_func can also see this local variable while my_func is running.

​
Passing Parameters

We pass parameters to functions the same way as script arguments — they are accessed with $1, $2, etc. inside the function:

​
Return Values

Bash functions return an exit status from 0 to 255, not a general value like a C function. Some common statuses are: We set a function’s status with return N. Without an explicit return, the function uses the status of its last command. We access the most recent exit status with $?.

​
Debugging

This enables xtrace debugging. Can’t really write the whole debug output here lmao :D. Basically, Bash prints commands after expanding variables and other substitutions, just before it executes them. Really useful when a script is silently failing and we have no idea where. One warning: xtrace can expose passwords, tokens and other secrets contained in commands, so we should check the output before saving or sharing it.