Expand beyond single-file programs with scope. Then make C++ functions more flexible using default parameters, overloading, and templates.

Functions: Scope & Flexibility

You've learned quite a bit! You now know a bit about how scope works for functions and files, as well as how to make functions more flexible for different use cases:
- Scope is the region of code that has access to an element.
  - Globally scoped variables are accessible everywhere.
  - A variable created inside a function has local scope and can't be accessed outside the function.
- C++ functions are usually split to make code more modular:
  - The declaration in a header file.
  - The definition in another **.cpp** file.
- Programs with multiple **.cpp** files need to be linked at compile time:
```
g++ main.cpp fns.cpp
```
- Header files must be `include`d in the file with `main()`:
```cpp
#include "fns.hpp"
```
- You can define a function inline using the `inline` keyword, which may or may not improve execution speed.
- Default arguments can be added to function declarations so that you can call the function without including those arguments.
- You can overload C++ functions so that they handle different types of input and return different types.
- A function template enables a function to behave the same with different types of parameters.

Review

If your program keeps growing, you may have to scroll through many declarations before you see `main()`. That doesn't seem like the best way to do things. Plus you don't want to keep declaring the same functions over and over for different files — making changes would be incredibly tiresome!

Well, you can take those function declarations and move them all over to a _header file_, another file — usually with the same name as the file with all of the function definitions — with the extension **.hpp** or **.h**. For example, if your function definitions are in **my_functions.cpp**, the corresponding header file would be **my_functions.hpp** or **my_functions.h**.

So how do you bring everything from a header file into scope for another file? Do you just link the header in the compilation statement like you did with the second **.cpp** file? 

As it turns out, with headers, you can just add `#include "my_functions.hpp"` to the very top of **main.cpp**:
```cpp
#include "my_functions.hpp"
```
Boom! This line pastes in everything from **my_functions.hpp**. Now you have access to all of the function declarations you stowed away in your header.


Getting a Header Yourself

Take a look at the program below. We have a `void` function named `favorite_animal()` and `main()` with a few statements inside.

```cpp
#include <iostream>

std::string sea_animal = "manatee";

void favorite_animal(std::string best_animal) {

  std::string animal = best_animal;
  std::cout << "Best animal: " << animal << "\n";

}

int main() {

  favorite_animal("jaguar");

  std::cout << sea_animal << "\n";
  std::cout << animal << "\n";

}
```

When this program is compiled and executed, `sea_animal` will print, but `animal` won’t. Why do you think that’s the case?

_Scope_ is the region of code that can access or view a given element.

 - Variables defined in _global scope_ are accessible throughout the program.
 - Variables defined in a function have _local scope_ and are only accessible inside the function.

`sea_animal` was defined in global scope at the top of the program, outside of `main()`. So `sea_animal` is defined everywhere in the program.

Because `animal` was only defined within `favorite_animal()` and not returned, it is not accessible to the rest of the program.

The Scope of Things

Once you set foot in the wild of C++ development, you may encounter the term "inline functions" with a couple different meanings. 
An _inline function_ is a function definition, usually in a header file, qualified by `inline` like this:
```cpp
inline 
void eat() {

  std::cout << "nom nom\n";

}
```
Using `inline` advises the compiler to insert the function's body where the function call is, which sometimes helps with execution speed (and sometimes hinders execution speed). If you do use it, we recommend testing how it affects the execution speed of your program. The bottom line is `inline` is something you'll probably encounter, but may never use.

However, you will sometimes also hear about "inline functions" that are just member functions (i.e. functions inside of classes — we'll explain classes later) which have been defined and declared in a single line in a header file because the function body is so short:
```cpp
// cookie_functions.hpp

// eat() belongs to the Cookie class:
void Cookie::eat() {std::cout << "nom nom\n";} 
```
Please note that you should ALWAYS add the `inline` keyword if you are inlining functions in a header (unless you are dealing with member functions, which are automatically inlined for you). 


How to Get Your Functions Inline

If you add a parameter to a function in C++, then an argument will be required when you call the function. What does "required" mean here? Well, you'll get an error. But what if 9 times out of 10, you know you'll use the same input value? It would be really annoying to have to enter the same value over and over again. 

To make your code more flexible for situations like this, you can add _default arguments_ to your function declarations. Default arguments are values assigned to parameters when the function is declared and defined: 
```cpp
// Declaration
void intro(std::string name, std::string lang = "C++");

// Definition
void intro(std::string name, std::string lang) {
  std::cout << "Hi, my name is "
            << name
            << " and I'm learning "
            << lang
            << ".\n";
}
```
Then, if you leave the argument blank in your function call, instead of an error, your function will run with the default value. Meanwhile, if you DO have an argument to add when you call the function, that argument will replace the default argument when your code executes.

Either of these will work for the function we defined:
```cpp
intro("Mariel")
// "Hi, my name is Mariel and I'm learning C++."
```
```cpp
intro("Mariel", "Python")
// "Hi, my name is Mariel and I'm learning Python."
```
**Important:** Your computer cannot read your mind; it determines which argument corresponds with which parameter based on order. 

Parameters without default arguments come first. This will work for `add_nums(3);` because the computer knows `3` corresponds to `num1`:
```cpp
int add_nums(int num1, int num2 = 0);
```
But the following does NOT work for `add_num(3);` the computer assumes that `3` still corresponds to `num1`:
```cpp
int add_nums(int num1 = 0, int num2);
```
Similarly, when a function has two default arguments, you still need to call with both arguments — if BOTH of the following are true:
- The first argument IS the default value.
- The second argument is NOT the default value.

Default Arguments

So how do we implement templates with actual code? Unlike regular functions, templates are entirely created in header files.

Templates let us choose the type implementation right when you call the function. The type we choose may apply to the return type, a parameter type, or both.

Here's how we could build a template for `print_cat_ears()` so that the parameter type is flexible:
```cpp
template <typename T>
void print_cat_ears(T item) {

  std::cout << " " << item << "   " << item << " " << "\n";
  std::cout << item << item << item << " " << item << item << item << "\n";

}
```
We can call the function for `int`, `char`, `std::string`, or `double`:
```cpp
print_cat_ears(2);

// the output:
//  2   2
// 222 222
```
Now we can use `print_cat_ears()` with a parameter of any type we want, as long as the type can be used with the methods expected. For example, we couldn't pass an `std::vector` into the current `print_cat_ears()` because `<<` won't work with `std::vector`.

**Note:** Using templates will slow down the program's compile time, but speed up the execution time.

How to Build Function Templates

Programs can grow quickly. With a few functions, you can declare the function above `main()` and then you can define the function below `main()` like this:
```cpp
#include <iostream>

// Declaration at the top:
void eat();

int main() {

  eat();

}

// Definition at the bottom:
void eat() {

  std::cout << "nom nom nom\n";

}
```
But this isn’t ideal when your code gets longer; it’s common to use the same function in more than one **.cpp** file.

To make your code cleaner and more modular, you can move the function definitions over to another specialized **.cpp** file (e.g., **my_functions.cpp**), leaving a list of declarations above `main()`.

But files, like functions, have scope. So, how does the `main()` program know about the function definitions?

Before your program even compiles, it links together any files you list in your compilation statement into a single executable:
```
g++ main.cpp my_functions.cpp
```
And voila! Your program knows the function definitions.

Multi-File Programs

What if you want a function to accept an argument that can be either an `int` OR a `double`? Or what if you want some function parameters to be optional? C++ has a trick up its sleeve just for such situations.

In a process known as _function overloading_, you can give multiple C++ functions the same name. Just make sure at least one of these conditions is true:
- Each has different type parameters.
- Each has a different number of parameters.

Overloading enables you to change the way a function behaves depending on what is passed in as an argument:
```cpp
void print_cat_ears(char let) {
  std::cout << " " << let << "   " << let << " " << "\n";
  std::cout << let << let << let << " " << let << let << let << "\n";
}

void print_cat_ears(int num) {
  std::cout << " " << num << "   " << num << " " << "\n";
  std::cout << num << num << num << " " << num << num << num << "\n";
}
```
Given the above functions, you could call the functions like so and C++ will know what to do:
```cpp
print_cat_ears('A');
print_cat_ears(4);
```
Output:
```
 A   A 
AAA AAA

 4   4
444 444
```


Function Overload!

Overloading can be really tedious. Imagine you want to create a new function that works with `int`, `float`, `double`, and other number types. Do you really need to redefine the SAME function body over and over again with different parameters? 

Thankfully not! When two functions have different types but the same body &mdash; as was the case with `print_cat_ears()` &mdash;, there is a cleaner solution you can use: templates.

A _template_ is a C++ tool that allows programmers to add data types as parameters. 

This feature comes in handy for classes as well as for functions. In fact, `std::string` and `std::vector` are both template-based types. The [C++ standard library](https://en.cppreference.com/w/cpp/header) is full of templates!
