1
00:00:00,000 --> 00:00:03,800
In Python, decorators, let you quote, unquote 'decorate' a

2
00:00:03,800 --> 00:00:07,400
function. But let's actually explore what that means.

3
00:00:07,480 --> 00:00:11,480
Decorators are functions that wrap around your existing functions,

4
00:00:11,490 --> 00:00:14,180
and we saw in the last video that you can nest functions,

5
00:00:14,180 --> 00:00:16,700
and that's actually where this really comes into play

6
00:00:16,700 --> 00:00:18,300
here, is this thing called a decorator.

7
00:00:18,300 --> 00:00:19,700
So here's a scenario.

8
00:00:19,700 --> 00:00:23,380
You've written a function, but you want to add some functionality

9
00:00:23,390 --> 00:00:26,300
to it, so you can do one of two things really.

10
00:00:26,310 --> 00:00:30,020
You can add changes to your existing function and potentially

11
00:00:30,030 --> 00:00:32,840
break how it works down the road wherever else it's being

12
00:00:32,850 --> 00:00:36,250
called because you've modified your function, or you can

13
00:00:36,250 --> 00:00:40,500
write a new function that has the old code and modify it.

14
00:00:40,500 --> 00:00:43,700
But with a decorator, it's sort of like being able to turn

15
00:00:43,740 --> 00:00:45,420
extra features on and off.

16
00:00:45,420 --> 00:00:48,200
All you have to do is call the decorator, and it will give it

17
00:00:48,200 --> 00:00:49,200
extra features.

18
00:00:49,200 --> 00:00:52,500
And every now and then, you know, sometimes you just want to add

19
00:00:52,500 --> 00:00:54,800
those extra features to your base function,

20
00:00:54,860 --> 00:00:57,050
and that's exactly what a decorator is for.

21
00:00:57,060 --> 00:01:00,830
So a decorator typically you'll see it used with the '@' symbol,

22
00:01:00,840 --> 00:01:04,340
and then it's decorator, 'decorator',

23
00:01:04,599 --> 00:01:07,900
and then it goes around your 'function'.

24
00:01:07,900 --> 00:01:10,900
This is what a decorator typically looks like.

25
00:01:10,980 --> 00:01:14,550
And this can add more functionality to your function down

26
00:01:14,560 --> 00:01:17,430
here, and this itself would be another function.

27
00:01:17,430 --> 00:01:20,900
So all we're going to do is say, "Hey, there's a function called 'function()',

28
00:01:20,900 --> 00:01:24,200
and just inject it into the decorator that we called 'decorator'.

29
00:01:24,200 --> 00:01:28,100
And so we've essentially just nested a function inside of a function.

30
00:01:28,140 --> 00:01:31,170
Now, to create a decorator, we're going to need to know about

31
00:01:31,170 --> 00:01:32,400
function nesting,

32
00:01:32,400 --> 00:01:35,400
and that was from the previous video.

33
00:01:35,460 --> 00:01:39,120
So if you didn't watch that one, you're probably not going

34
00:01:39,130 --> 00:01:41,310
to totally understand what's going on in this one.

35
00:01:41,320 --> 00:01:45,390
So hop back a lesson if you didn't watch the video on nesting

36
00:01:45,400 --> 00:01:48,330
functions. Otherwise, if you're ready, let's go ahead and

37
00:01:48,340 --> 00:01:52,740
create our first decorator. So we can create a function in

38
00:01:52,740 --> 00:01:54,900
here, and let's call it 'my_decorator',

39
00:01:54,900 --> 00:01:58,400
and this decorator is going to take an 'original func'.

40
00:01:58,480 --> 00:02:01,750
Now, in Python, you don't have to pass a strict variable,

41
00:02:01,750 --> 00:02:05,600
you can also pass a function into another function.

42
00:02:06,060 --> 00:02:10,150
And in here we're going to create another function,

43
00:02:10,160 --> 00:02:11,770
it's a nested function,

44
00:02:12,100 --> 00:02:16,600
and this one's just going to be called 'func_wrapper',

45
00:02:16,630 --> 00:02:18,700
or it could even just be called 'wrapper'

46
00:02:18,710 --> 00:02:23,230
if you want. 'print("Do a thing here")'.

47
00:02:23,440 --> 00:02:28,780
And then it could run the 'original_func', and then it

48
00:02:28,780 --> 00:02:32,300
could "Do a thing after",

49
00:02:32,300 --> 00:02:34,200
And let's change 'here' to 'before'.

50
00:02:34,250 --> 00:02:37,210
So "Do a thing before", run the function, and then, "Do a thing

51
00:02:37,220 --> 00:02:42,630
after", and then in our main function called 'my_decorator',

52
00:02:42,640 --> 00:02:46,380
we are simply going to return 'wrapper' as the entire function,

53
00:02:46,500 --> 00:02:51,800
not as an executed function, but just as the function itself.

54
00:02:51,800 --> 00:02:53,400
So let's go ahead and save that.

55
00:02:53,400 --> 00:02:57,200
Let's create another function in here called 'greeting'.

56
00:02:57,200 --> 00:03:01,200
And all this is going to say is 'print("HELLO WORLD")'.

57
00:03:01,200 --> 00:03:02,400
So we're going to keep this very simple.

58
00:03:03,240 --> 00:03:08,700
And if we run 'greeting' now, just says, "HELLO WORLD", the decorator

59
00:03:08,700 --> 00:03:09,700
has not been applied.

60
00:03:09,700 --> 00:03:12,900
We can apply this decorator though, in two different ways.

61
00:03:12,920 --> 00:03:16,340
The first way, the most common way you're going to see, and

62
00:03:16,350 --> 00:03:19,910
probably the most elegant way, is using the '@' symbol, '@

63
00:03:20,000 --> 00:03:23,500
my_decorator', and that just matches the name here.

64
00:03:23,500 --> 00:03:28,100
So if we rerun this cell, and rerun this cell, we can now

65
00:03:28,100 --> 00:03:29,500
see "Do a thing before",

66
00:03:29,500 --> 00:03:31,800
"HELLO WORLD", that's our function here,

67
00:03:31,800 --> 00:03:35,800
"Do a thing after", and that's executing this code for us.

68
00:03:35,800 --> 00:03:38,560
Now, that's how you create a decorator in less than five

69
00:03:38,570 --> 00:03:40,840
minutes. There is, however, a second way to do this.

70
00:03:40,980 --> 00:03:45,930
You could also say 'new_greeting =', and then you

71
00:03:45,930 --> 00:03:50,000
could have 'my_decorator' as a function,

72
00:03:50,000 --> 00:03:54,320
and you could also pass in 'greeting' as the function's parameter,

73
00:03:55,210 --> 00:03:58,820
and now you can use 'new_greeting',

74
00:03:58,830 --> 00:04:01,750
and if you don't use it like a function, you'll see this

75
00:04:01,760 --> 00:04:02,740
typical thing here.

76
00:04:02,740 --> 00:04:08,200
But if you do use it as a function, it will also work for you as well.

77
00:04:08,240 --> 00:04:11,810
Actually, this is going to double work because this is still

78
00:04:11,820 --> 00:04:14,600
decorated. So let's rerun these cells without the decorator,

79
00:04:14,600 --> 00:04:16,700
and now we can see it works normally.

80
00:04:16,760 --> 00:04:20,149
So here just looks like a regular function, and we're just

81
00:04:20,149 --> 00:04:22,300
passing in a function instead of a regular variable.

82
00:04:22,370 --> 00:04:25,370
What I like to do though, and maybe this is just because

83
00:04:25,370 --> 00:04:26,899
I like writing a lot of Django,

84
00:04:26,899 --> 00:04:29,400
and so I see decorators quite a bit.

85
00:04:29,900 --> 00:04:32,300
I like 'my_decorator'.

86
00:04:32,900 --> 00:04:35,300
I like it being above the function.

87
00:04:35,300 --> 00:04:37,990
And at any point in time, you can just delete it, or you can

88
00:04:38,000 --> 00:04:41,600
comment it out, and you can toggle on and off other features.

89
00:04:41,650 --> 00:04:45,970
So let's go ahead and create a second dary decorator, something

90
00:04:45,980 --> 00:04:48,970
that's going to actually do something, because right now

91
00:04:48,980 --> 00:04:51,700
our decorator just prints stuff, and that's not actually useful,

92
00:04:51,710 --> 00:04:55,130
and it doesn't actually show you that this is modifying anything.

93
00:04:55,130 --> 00:04:58,700
It's just a 'print' statement, another 'print' statement, another 'print' statement.

94
00:04:58,700 --> 00:05:01,500
And frankly, that's not what you're going to see in real

95
00:05:01,500 --> 00:05:02,700
life Python code.

96
00:05:02,780 --> 00:05:06,740
So let's create a new decorator, and I'm going to call it

97
00:05:06,750 --> 00:05:13,240
'add_stuff', and it's going to take an 'original_func', and

98
00:05:13,250 --> 00:05:15,340
then there's going to be a function inside of this.

99
00:05:15,400 --> 00:05:18,700
We could call it 'wrapper', we could really call it anything.

100
00:05:18,710 --> 00:05:21,770
It doesn't have to be 'wrapper', so it could be like the 'function_

101
00:05:21,770 --> 00:05:26,300
wrapper', and due to scoping, we will have access to 'original_func' in here.

102
00:05:26,360 --> 00:05:28,520
So we can run this right away.

103
00:05:28,520 --> 00:05:31,400
We can run the 'original_func', and get that 'value'.

104
00:05:31,720 --> 00:05:34,750
The 'value' of the 'original_func' is going to be not 'greeting',

105
00:05:34,940 --> 00:05:36,470
'original_func',

106
00:05:36,480 --> 00:05:39,260
then if we wanted to, we could actually return something like

107
00:05:39,260 --> 00:05:46,000
an f-string where it says, "The original value was ",

108
00:05:46,620 --> 00:05:49,020
or even better, let's not return that,

109
00:05:49,020 --> 00:05:50,400
let's just print that,

110
00:05:50,400 --> 00:05:52,800
and then let's change that 'value'.

111
00:05:52,800 --> 00:05:54,800
So let's assume that this 'value' right now is going to be

112
00:05:54,800 --> 00:05:56,680
a number just because we're going to be adding stuff.

113
00:05:56,680 --> 00:06:03,000
So whatever that 'value' is, we're going to say 'value = value ** 2',

114
00:06:03,040 --> 00:06:06,090
and we're going to 'return value'.

115
00:06:06,100 --> 00:06:11,160
And let's go ahead, and just print out what the new 'value'

116
00:06:11,170 --> 00:06:12,870
is. "The original value was ".

117
00:06:12,880 --> 00:06:15,570
The new 'value' is going to be whatever 'value' to the power

118
00:06:15,570 --> 00:06:17,600
of 2 is, and it's going to return that for us.

119
00:06:17,600 --> 00:06:21,800
[no audio]

120
00:06:21,800 --> 00:06:26,530
Let's take that function and return it here.

121
00:06:26,540 --> 00:06:27,760
Okay, cool.

122
00:06:27,910 --> 00:06:29,620
So now we have a decorator.

123
00:06:29,620 --> 00:06:32,960
Let's go ahead and create a brand new function that simply

124
00:06:32,970 --> 00:06:33,950
returns a number.

125
00:06:33,960 --> 00:06:38,790
So we're gonna call this one 'def pie', and all it's going to

126
00:06:38,800 --> 00:06:41,100
do is 'return 3.14'.

127
00:06:41,190 --> 00:06:44,460
And yes, I know I spelt it 'pie', because I'm thinking

128
00:06:44,460 --> 00:06:46,200
about 'pi' as a food right now.

129
00:06:46,220 --> 00:06:49,040
And if we run 'pie', we see it's 3.14.

130
00:06:50,500 --> 00:06:56,100
But if we take this decorator and throw it up here, run that

131
00:06:56,100 --> 00:06:58,500
cell, and then rerun this cell,

132
00:06:58,560 --> 00:07:00,780
"The original value was 3.14".

133
00:07:00,790 --> 00:07:04,730
The new value is whatever 'pie' to the power of 2 is, which

134
00:07:04,730 --> 00:07:09,100
is 9.8596 something rather, and it gives us the actual 'value' here.

135
00:07:09,160 --> 00:07:12,190
Now, we can actually store this as well, because not only

136
00:07:12,200 --> 00:07:15,330
are there 'print' statements in here, but it's returning a

137
00:07:15,330 --> 00:07:17,900
'value' in which we can actually store into a variable.

138
00:07:17,900 --> 00:07:22,900
So we could say 'pi_times_pi =', whatever that is,

139
00:07:22,900 --> 00:07:28,600
and then 'pi_times_pi' as a variable is the actual output.

140
00:07:28,600 --> 00:07:31,000
That's what we got from this 'return' statement here.

141
00:07:31,000 --> 00:07:35,500
Now, in places like data science, you might not see this too often.

142
00:07:35,500 --> 00:07:38,710
You might not even see this too often in machine learning

143
00:07:38,710 --> 00:07:40,000
or artificial intelligence.

144
00:07:40,000 --> 00:07:44,950
But I can tell you for certain in web development, in Django

145
00:07:44,950 --> 00:07:47,800
and Flask, you're going to see these quite a bit.

146
00:07:49,000 --> 00:07:50,200
And I actually really like this.

147
00:07:50,290 --> 00:07:51,750
It just adds a little functionality.

148
00:07:51,760 --> 00:07:55,210
And, hey, if you ever wanted to turn it off, you can just

149
00:07:55,220 --> 00:07:56,650
comment it out or delete it.

150
00:07:56,660 --> 00:08:00,500
Let's rerun those cells, and it goes back to normal.

151
00:08:00,500 --> 00:08:01,800
Now, that's pretty cool.

152
00:08:01,840 --> 00:08:05,680
Now, what I would like you to do is go ahead and try to create

153
00:08:05,690 --> 00:08:06,790
your own decorator.

154
00:08:06,800 --> 00:08:08,650
It does not need to do anything fancy.

155
00:08:08,660 --> 00:08:11,890
And honestly, I'm going to suggest stick with something simple.

156
00:08:11,900 --> 00:08:14,880
Just do a 'print' statement before, a 'print' statement after,

157
00:08:14,880 --> 00:08:16,900
and run your function in here.

158
00:08:16,960 --> 00:08:18,160
Give that a shot.

159
00:08:18,160 --> 00:08:22,900
Don't forget, you can always download this Jupyter Notebook as well.