Untangling Complex Declarations in C

4 min readDec 29, 2016

Here is a list of C declarations of increasing complexity. Until recently I could not have told you what all of them meant.

Now I can explain all of them fairly confidently after having learning about the precedence rules for C declarations, a topic that seems important and yet was strangely absent from the highly rated book that I purchased when I initially set out to learn C. It wasn’t until I read a second book on advanced C topics that I learned about these rules.

For me, learning how to interpret complex declarations was an important step towards using complex data structures and expanding my capabilities as a programmer. If I can’t read a complex declaration, then I can’t write one either, and I therefor might avoid using them altogether and opt for less ideal solutions.

I originally assumed that becoming fluent with complex declarations was, like many things in life, a process that slowly adds clarity with repeated exposure and use. But I was surprised to find that there are only a few rules to memorize that allow you to easily break down any declaration into small comprehensible components.

They are referred to as the precedence rules for C declarations. From high to low, they are:

Parentheses grouping together parts of a declaration.
The postfix operators: parentheses () indicating a function, and square brackets [] indicating an array.
The prefix operator: the asterisk denoting “pointer to”

To put it another way: Parentheses that are grouping together multiple parts of a declaration have the highest precedence. Next are the postfix operators () and []. Last is the prefix operator *.

The 4th declaration on my list above might be a good example to start with.

char* foo[5];

The precedence rules say that the square brackets are higher than the asterisk. So foo is an array of pointers to a char, not a pointer to an array of characters.

The step-by-step process for applying the precedence rules to this example is as follows.

Starting with the name foo, the interpretation starts out as: “foo is a…”

Next, the precedence rules state that the square brackets on the right outweigh the asterisk on the left.

char* foo[5];

Now the interpretation is: “foo is an array of 5”

The only option next is the asterisk on the left.

char* foo[5];

“foo is an array of 5 pointers to”

And the last part of the declaration is the data type char.

char* foo[5];

The final interpretation is: “foo is an array of 5 pointers to a char”

If I had wanted foo to be a pointer to an array, instead of an array of pointers, I would have needed to include a set of parentheses like in the following example.

char(* foo)[5];
// foo is achar(* foo)[5];
// foo is a pointerchar(* foo)[5];
// foo is a pointer to an array of fivechar(* foo)[5];
// foo is a pointer to an array of five chars

This time the parentheses gave the asterisk a higher precedence than the square brackets, which changed foo from an array of pointers to a pointer to an array.

The 8th declaration on the list includes two sets of parentheses.

char* (*foo)(char*);
// foo is achar* (*foo)(char*);
// foo is a pointerchar* (*foo)(char*);
// foo is a pointer to a function that accepts a pointer to a charchar* (*foo)(char*);
// foo is a pointer to a function that accepts a pointer to a char and returns a pointerchar* (*foo)(char*);
// foo is a pointer to a function that accepts a pointer to a char and returns a pointer to a char

I took the liberty of interpreting the function parentheses and it’s parameters as a single step, but it may be easier to interpret each parameter separately. The same precedence rules can be applied for each parameter as they are for the declaration as a whole.

As a final example, here is the most complex declaration on the list:

char* (*(*foo[5])(char*))[];
// foo is achar* (*(*foo[5])(char*))[];
// foo is an array of 5char* (*(*foo[5])(char*))[];
// foo is an array of 5 pointers tochar* (*(*foo[5])(char*))[];
// foo is an array of 5 pointers to a function that accepts a pointer to a charchar* (*(*foo[5])(char*))[];
// foo is an array of 5 pointers to a function that accepts a pointer to a char and returns a pointer tochar* (*(*foo[5])(char*))[];
// foo is an array of 5 pointers to a function that accepts a pointer to a char and returns a pointer to an array ofchar* (*(*foo[5])(char*))[];
// foo is an array of 5 pointers to a function that accepts a pointer to a char and returns a pointer to an array of pointers tochar* (*(*foo[5])(char*))[];
// foo is an array of 5 pointers to a function that accepts a pointer to a char and returns a pointer to an array of pointers to a char.

I should probably mention that this last example is used to show how to apply these precedence rules in an extreme case, but using this sort of declaration in a real world scenario may be considered unnecessarily complex and poor practice.

Also note that none of these examples include keywords like const or volatile, and they don’t show examples of declarations for structs, enums, or unions. If you want to read more on this topic you can google “declaration precedence rules for C” to find lots of good resources to help complete your understanding of this topic.

Untangling Complex Declarations in C

Written by Brian Barto