by Alex Nadalin

Makefiles 101: how to use make as a task automation tool

TcD0rnzbcNYghR240XeuA429p4gdt93FhJ1o
Photo by Agto Nugroho on Unsplash

It seems like developers are afraid of using make as they associate it with the painful experience of compiling things from scratch — the dreaded ./configure && make && make install.

Part of this fear is due to the description of what make(1) does:

The purpose of the make utility is to determine automatically which pieces of a large program need to be recompiled, and issue the commands to recompile them.
Free Software Foundation The Linux Man Pages

Not everyone is aware that make can easily be used to manage tasks in your projects. In this article, I’d like to share a brief introduction to how Makefiles help me automate some tasks in my day to day activities. This brief guide will focus on using make as an automation tool for tasks rather than a tool for compiling code.

Executing tasks

Let’s start by simply creating a Makefile, and defining a task to run:

task:  date

If you run make task you will bump into the following error:

/tmp ᐅ make taskMakefile:2: *** missing separator.  Stop.

And that’s because Makefiles use tabs to indent code. Let’s update our example by using tabs rather than spaces and… Voilà.

/tmp ᐅ make taskdateFri Jun 15 08:34:15 +04 2018

What kind of sorcery is this? Well, make understood you wanted to run the section task of your makefile, and ran the code (date) within that section in a shell, outputting both the command and its output. If you want to skip outputting the command that’s being executed, you can simply prefix it with an @ sign:

task:  @date

Running the make command again:

/tmp ᐅ make taskFri Jun 15 08:34:15 +04 2018

The first task in a Makefile is the default one, meaning that we can run make without any arguments:

/tmp ᐅ make       Fri Jun 15 08:37:11 +04 2018

Running additional tasks

You can add additional tasks in your Makefile and call them with make $TASK:

task:  @datesome:  sleep 1  echo "Slept"thing:  cal

Running the make command again:

/tmp ᐅ make thingcal     June 2018        Su Mo Tu We Th Fr Sa                  1  2   3  4  5  6  7  8  9  10 11 12 13 14 15 16  17 18 19 20 21 22 23  24 25 26 27 28 29 30

Running tasks in a specific order

A lot of times you will want to execute a task before the current one. Think of it as before or after hooks in your automated tests. This can be done by specifying a list of tasks right after your task’s name:

task: thing some  @date...

Running the make command again:

/tmp ᐅ make taskcal     June 2018        Su Mo Tu We Th Fr Sa                  1  2   3  4  5  6  7  8  9  10 11 12 13 14 15 16  17 18 19 20 21 22 23  24 25 26 27 28 29 30
sleep 1echo "Slept"SleptFri Jun 15 08:40:23 +04 2018

Using variables with make

Defining and using variables is fairly straightforward:

VAR=123
print_var:        echo ${VAR}...

Running the make command again:

/tmp ᐅ make print_var    echo 123123

But watch out, as your shell variables won’t work out of the box:

print_user:        echo $USER

Running the make command again:

/tmp ᐅ make print_user   echo SERSER

You’ll will need to escape them with either ${VAR} or $$VAR.

Passing flags is also a bit different from what you might be used to. They’re positioned as flags but use the same syntax as environment variables:

print_foo:  echo $$FOO

Running the make command again:

/tmp ᐅ make print_fooecho $FOO
/tmp ᐅ make print_foo FOO=barecho $FOObar

Make and the shell

5.3.1 Choosing the Shell------------------------
The program used as the shell is taken from the variable `SHELL'.  If this variable is not set in your makefile, the program `/bin/sh' is used as the shell.

Make will use sh to execute code in a task. This means that some stuff might not work, as you’re probably using some syntax that’s specific to bash. In order to switch, you can simply specify the SHELL variable. In our case, we would want to use SHELL:=/bin/bash.

As seen before, sometimes you will need to use a quirky, custom syntax to get a regular shell command to work in make. Just like variables need to be escaped with a $$ or ${...}, you will need to use shell when using command substitution:

subshell:  echo $(shell echo ${USER})

Running the make command again:

/tmp ᐅ make subshellecho alexalex

Don’t believe me? Try removing the shell instruction. Here’s what you’re going to get:

/tmp ᐅ make subshellecho

Conclusion

There’s so much more make can do, and so many more quirky things you might need to find out to decrease the WPS (WTF per second) when working with it. ?

That doesn’t negate the fact that make is an extremely helpful tool that allows us to automate workflows with ease (without having to setup very complicated pipelines) by writing tab-separated lines with a bunch of shell commands instead.

Originally published at odino.org (15th June 2018).
You can follow me on Twitter - rants are welcome! ?