Header Files (C)

Header files are where you place shared functions, function prototypes, #define's, and anything else you want shared. Definitions go in a .h file, whereas the code, the action, goes in the .c file. Everything could go in one C file, but organizing with multiple files is better for humans.[1]

Short answer is it separates the interface from the implementation.

Example

I'm pulling this wholesale from this SO answer[3] because it's great.

Take this example. foo does not know what bar is when it is invoked because bar is declared after foo. We could solve this by moving bar above foo, but then we would have the same problem within bar.

baz.c

void foo() {
  //bar? what is this function???
  bar();
}

void bar() {
  //I know foo() because it appears before. 
  foo();
}

foo can call bar only if it is defined before.
bar can only call foo if it is defined before

Since we must define one of them first, the solution to this problem are function prototypes.

baz.c

//prototype of foo()
void foo();
//prototype of bar()
void bar();

void foo() {
  // I can call bar() because I know it exists
  bar();
}

void bar() {
  // I can call foo() because I know it exists
  foo();
}

With a large project, you can end up amassing a large amount of prototypes, along with #define's, #ifdef's, etc. To clean things up, we can put all of these things into an .h file, and #include that instead.

baz.h

//prototype of foo()
void foo();
//prototype of bar()
void bar();

baz.c

#include 'baz.h'

void foo() {
  // I can call bar() because I know it exists
  bar();
}

void bar() {
  // I can call foo() because I know it exists
  foo();
}

It then became a good practice as it separates the interface from the implementation.

References

Last modified: 202401040446