WEBVTT 00:00.390 --> 00:07.950 Hello again! In this video, we are going to look at operators and overloading. C++ defines internal operators 00:07.950 --> 00:09.060 for built-in types. 00:10.230 --> 00:16.740 We have the plus operator for adding two ints, for example. So we can say c equals a plus b. 00:17.370 --> 00:21.750 This will actually call a function called operator plus, which takes a and b as arguments. 00:22.380 --> 00:26.370 This returns an int, and then the return value is stored in s. 00:26.970 --> 00:30.050 So there is an internal function inside C++, 00:30.600 --> 00:33.270 operator plus, which takes two ints and returns another int. 00:34.920 --> 00:39.400 As another example, C++ has the equality operator for comparing two ints. 00:40.200 --> 00:45.900 If we have a equals equals b. Built-in types are not really objects, but you can think of this 00:46.200 --> 00:48.060 as if it is a call to a member function. 00:48.840 --> 00:52.680 So the equality operator is called on a, with the argument b. 00:54.510 --> 00:59.400 And there are examples in the Standard Library as well. For example, with the string. The plus operator 00:59.400 --> 01:06.630 will add two strings and it is exactly the same except for the type, so it takes two strings and returns 01:06.640 --> 01:07.260 another string. 01:08.800 --> 01:15.100 The equality operator, that really is a member function this time, so it is called on the first string 01:15.520 --> 01:17.560 and it takes the second string as argument. 01:17.920 --> 01:24.670 So really, you can see that there are two types of operators; ones which take two arguments and are 01:25.030 --> 01:31.090 called as non-member functions, and ones which take a single argument and are called as member functions. 01:34.590 --> 01:40.110 And there are actually special names for this, or special terms. An operator which takes a single argument 01:40.110 --> 01:47.850 is unary and an operator which takes two arguments is binary. And there is actually one example of 01:48.060 --> 01:51.690 an operator which can be both, or it has two different versions. One for each. 01:52.770 --> 01:54.200 There is an operator minus. 01:54.210 --> 01:58.830 So that does the obvious thing of subtracting two numbers, which is a binary operation. 01:59.730 --> 02:01.320 It takes two arguments, a and b. 02:01.710 --> 02:06.240 But there is also a unary operator, which just changes the sign of the argument. 02:06.690 --> 02:11.190 So that will take b, and it will return b, but with the opposite sign. 02:13.130 --> 02:17.180 In fact, C++ has a ternary operator, which takes three arguments. 02:18.750 --> 02:27.840 This is a shorthand way of watching an if statement. So "a query b colon c" is equivalent to "if a b, else c" and 02:27.840 --> 02:30.570 that is implemented as an operator, which has three arguments. 02:33.980 --> 02:40.650 So, ignoring the ternary operator for the moment, opeartors take either one or two arguments. 02:40.670 --> 02:43.550 Operators can be member functions or non-member functions. 02:43.940 --> 02:50.420 Normally, if you have a unary operator, it is implemented as a mamber function on an object. And binary operators 02:50.420 --> 02:54.530 are usually non-member functions, so they take both objects as arguments. 02:57.090 --> 03:03.270 The classes which are defined in the Standard Library, usually have operators which will act on objects 03:03.270 --> 03:03.930 of that class. 03:04.470 --> 03:10.080 So as we saw with the string, with the plus operator and the equality operator, these have the same 03:10.080 --> 03:15.270 syntax as operators for the built-in types, and they have similar behavior. 03:15.660 --> 03:17.790 So the plus sign will add two strings. 03:18.330 --> 03:23.550 It is not going to reverse the string or remove alternate elements. That would be very confusing! 03:26.640 --> 03:33.840 And C++ allows us to define our own operators. If we write a class in C++, it does not have any operators, 03:34.500 --> 03:40.710 except for the assignment operator, if the compiler synthesizes that. But we can define other operators 03:40.710 --> 03:43.470 if we want to, which will act on objects of our class. 03:44.040 --> 03:48.670 And this is known as "overloading" the operators. "Overloading" here 03:48.680 --> 03:51.630 means giving a different meaning to an existing symbol. 03:52.020 --> 03:57.120 So if we have overloaded functions, then we have different meanings for different arguments. And we 03:57.120 --> 03:58.320 have overloaded operators. 03:58.320 --> 04:01.560 We have different meanings for different types that they operate upon. 04:03.660 --> 04:09.180 We can only use the symbols which are supplied for built-in operators, and we have to use the same 04:09.180 --> 04:13.910 syntax. So we are not allowed to invent our own operators or our own syntax. 04:13.920 --> 04:14.550 Thank goodness! 04:15.900 --> 04:19.150 Life is confusing enough in C++ as it is, without this! 04:20.460 --> 04:24.960 And when we write our operators, they should have similar semantics to the built-in ones. 04:25.290 --> 04:30.120 So that is just to avoid confusing people who are going to use your class and these operators. 04:30.510 --> 04:32.370 Okay, so that is it for this video. 04:32.760 --> 04:33.680 I will see you next time. 04:33.690 --> 04:35.730 But until then, keep coding!