0

I saw a post on SO about using boolean types in C. One of the answers listed four ways to implement boolean types, the second and third way being:

typedef enum { false, true } bool;

and

typedef int bool;
enum { false, true };

The first one confused me at first because I had never seen an enum being used as a type. Since enum defines integer constants, I thought it was the same thing as the second one but slightly more confusing.

So, how does using enum as a type work? Is there a difference between using enum as a type and using an integer type, defining the enum elsewhere? Most importantly, why would someone use enum as a type?

  • 2
    Every `enum` is a type. Maybe you are confused by the fact that you can implicitly convert an integer to an `enum` value? – UnholySheep Jan 22 '22 at 21:50
  • 1
    Maybe the second answer from [here](https://stackoverflow.com/questions/63249307/difference-between-using-enum-and-int-variables-in-c) can help. – Shahriar Jan 22 '22 at 21:53
  • I'm not sure why you think the first example is confusing. It's more clear than the second one. It's saying there is a type, called `bool`, and it has two possible values: `true` and `false` – Botahamec Jan 22 '22 at 21:54
  • Does this answer your question? [Difference between using enum and int variables in C](https://stackoverflow.com/questions/63249307/difference-between-using-enum-and-int-variables-in-c) – user2864740 Jan 22 '22 at 21:57
  • The “why” appears answered in the linked question. If the focus was on comparing the differences (if any) between the two forms shown in the question, then consider making that the focus in question *and* title. – user2864740 Jan 22 '22 at 21:58
  • Enums as types in C are kind of half-assed. Because they're "compatible with an integer type" (even worse than being weakly convertible to/from integers), you can't really use them to force limit their value to one of the enumerated constants. For me, the only (nonportable) thing they've got going for them is that gcc and clang autosuggest case labels if you use an enum-typed variable as the controlling expression of a switch. – Petr Skocik Jan 22 '22 at 22:23
  • @Botahamec I think the first one is cleaner, I just want to understand it before I start using it. The second one merely conforms more to my previous understanding of enums. –  Jan 22 '22 at 23:36

1 Answers1

1

Generally in programming, enums are used for creating lists of defined final values. And because booleans can be either true or false, it is a good fit to implement it this way.

The first code bit you included is quite simple, it defines a type called bool as an enum that can be either true or false. Why it's defined here in the enum is because this way you won't get an error, when using true/false as values in your code. Also, there's a reason, why false is listed first - on the index 0 and true is on the index 1. Can you guess? ...You can convert enum element to int by its index (meaning: this way when you work with false value, you are basically working with 0).

The second code is very similar, it defines type bool, but as an integer instead. That's because that's how you normally represent bools in C = 1 (or any other int to get true) or 0 (false). The definition of enum here is again just to prevent from getting an error, when using true and false as values. Again the values are positioned this way for a reason - thanks to which you can just do if (true) or if (false) and it will work as intended (the reason is described above).

Nogard
  • 310
  • 2
  • 10