WEBVTT

0
00:02.160 --> 00:03.620
Welcome back.

1
00:03.660 --> 00:06.420
A set contains only unique values.

2
00:06.420 --> 00:07.350
Just like a map.

3
00:07.620 --> 00:11.700
In this lecture you're going to learn how to use a map as a set.

4
00:11.730 --> 00:17.790
You're going to build a program that can answer whether a given word exists in a file or not.

5
00:17.790 --> 00:19.100
All right let's get started

6
00:23.210 --> 00:27.120
first of all let me show you the input file that you're going to use.

7
00:27.140 --> 00:31.300
This is a passage from Romeo and Juliet from Shakespeare.

8
00:31.310 --> 00:35.570
It contains 70 unique words and 99 words in total.

9
00:35.570 --> 00:41.600
So the map that you're going to create will store only 70 words because map keys are unique.

10
00:41.600 --> 00:47.110
OK now I'm going to create a new scanner that uses the standard input.

11
00:47.160 --> 00:53.250
How can you split the input by words? You can use strings dot split the function for each line.

12
00:53.280 --> 00:54.120
Right.

13
00:54.180 --> 01:00.990
However there is a more natural way you can configure the scanner to scan only for the words instead

14
01:00.990 --> 01:02.240
of lines.

15
01:02.400 --> 01:08.740
All you need to do is calling the split method and telling it to scan for the words like so.

16
01:08.760 --> 01:10.980
Let me show you how it works.

17
01:11.010 --> 01:16.470
First I need to scan for the words like so each time I call the scan method.

18
01:16.470 --> 01:19.710
It will scan for a word not a line.

19
01:19.710 --> 01:20.790
Let me show you.

20
01:21.030 --> 01:24.870
I'm going to print the scanned word by calling the text method like so

21
01:28.230 --> 01:33.520
as you can see it scans for the words instead of lines all right.

22
01:33.570 --> 01:40.530
Since you have a program that can scan for the words it is time to index the words using a map to do

23
01:40.530 --> 01:40.890
that.

24
01:40.890 --> 01:46.950
I'm going to create a new map like so the key type is a string and the element type is a bool.

25
01:47.250 --> 01:53.190
The keys will be the unique words the corresponding elements will tell you whether the word is in the

26
01:53.190 --> 01:54.900
map or not.

27
01:54.900 --> 02:01.170
So the user will be able to find out whether a word exists in the input or not.

28
02:01.170 --> 02:06.750
Now I'm going to remove the println and I'm going to get the word using the text method like so.

29
02:07.530 --> 02:14.740
Let me also change the word to lower case letters because a String map keys are case sensitive.

30
02:14.930 --> 02:17.660
I also don't want to index tiny words.

31
02:17.840 --> 02:23.020
So I'm going to index a word only if it has three letters at least.

32
02:23.080 --> 02:23.350
OK.

33
02:23.360 --> 02:30.680
Now it's time to index the words in the map like so but why do you need to set it to true.

34
02:30.920 --> 02:34.410
It's because this is what makes the map a set.

35
02:34.460 --> 02:35.950
I'll explain it in a minute.

36
02:35.960 --> 02:36.580
No worries.

37
02:37.550 --> 02:41.720
Let's say the user asks for the sun word.

38
02:41.900 --> 02:44.150
Let's check the word using the comma-ok.

39
02:44.390 --> 02:50.660
idiom like so if the word is in the map the ok variable will be true.

40
02:50.720 --> 02:56.660
So I'm going to say the input contains the word and I'm going to terminate the program using a return

41
02:56.660 --> 02:57.950
statement.

42
02:57.950 --> 03:01.880
Otherwise I'm gonna say sorry the input does not contain the word

43
03:05.600 --> 03:08.420
since there is a sun word in the input.

44
03:08.420 --> 03:11.220
It says the input contains sun.

45
03:11.360 --> 03:18.750
Let's see whether Shakespeare talks about gophers clearly it does not.

46
03:18.910 --> 03:21.250
I'm not surprised at this at all.

47
03:21.370 --> 03:23.800
However this code is not right.

48
03:23.980 --> 03:32.350
For example let me show you what the map returns when I ask for sun and Tesla as you can see sun returns

49
03:32.350 --> 03:35.350
true whereas Tesla returns false.

50
03:35.570 --> 03:40.850
It's because when the scanner returns the sun word I set it to true in the map.

51
03:40.900 --> 03:43.240
That's why it returns true here.

52
03:43.270 --> 03:50.410
However the map doesn't contain the Tesla word so it returns false because the zero value for bool is

53
03:50.530 --> 03:51.490
false.

54
03:51.490 --> 03:54.890
So I don't need to use the comma okay idiom here.

55
03:55.120 --> 04:02.680
I can directly check whether a word exists in the map or not like so. Let me also remove this code. As

56
04:02.680 --> 04:04.600
you can see it still works.

57
04:04.610 --> 04:09.480
Let's also try it with the sun word it works.

58
04:09.500 --> 04:12.580
Now it's time to get the word from the user.

59
04:12.590 --> 04:17.120
Now I'm going to fast forward it because you already know how it works.

60
04:17.180 --> 04:20.330
I'm going to remove the query variable from here.

61
04:20.430 --> 04:22.520
I'm going to get it from the user.

62
04:22.520 --> 04:32.260
All right let's try it as you can see it works although there are 99 words in the input file here.

63
04:32.270 --> 04:39.930
You only need to store 70 words because the map keys are unique and that's efficient by the way.

64
04:39.930 --> 04:41.760
You don't have to pipe a file.

65
04:41.880 --> 04:46.900
You can pipe anything but for now let's directly enter the input.

66
04:46.980 --> 04:49.290
I'm going to look for the beautiful word.

67
04:49.470 --> 04:57.030
Now I'm going to say today is a beautiful day as you can see it says that the input contains the beautiful

68
04:57.030 --> 04:59.200
word. Cool!

69
04:59.280 --> 05:04.130
You can even check whether a website contains a word that you are looking for.

70
05:04.260 --> 05:08.140
To do that you can pipe the curl command to your program.

71
05:08.300 --> 05:12.020
The Curl tool can retrieve the contents of a website.

72
05:12.180 --> 05:19.020
For example I can get a text document from the web like so "-s" argument here tells the curl that

73
05:19.020 --> 05:21.810
only to fetch the content.

74
05:21.810 --> 05:22.470
Normally

75
05:22.470 --> 05:30.570
It also shows us the download progress the s flag prevents that as you can see it retrieves the source

76
05:30.570 --> 05:32.580
code of the article.

77
05:32.580 --> 05:36.000
Now I can search for a word in it using our program.

78
05:36.210 --> 05:39.570
First I need to type the same command again.

79
05:39.570 --> 05:42.260
Then I need to pipe the content to our program.

80
05:42.300 --> 05:45.480
Like so I'm going to search for the outside word.

81
05:45.510 --> 05:46.030
Like so.

82
05:47.580 --> 05:49.710
As you can see the article contains the word.

83
05:50.430 --> 05:52.670
Let me show you the indexed words.

84
05:52.890 --> 06:01.460
I'm going to loop over the keys of the map then I'm going to print the word as you can see the program

85
06:01.500 --> 06:09.270
not only indexes the words it also indexes the numbers and punctuation characters or sometimes a word

86
06:09.420 --> 06:11.890
ends with a dot or a comma.

87
06:11.940 --> 06:18.870
So if you search for these words it won't be able to find them for example let me search for the inclusive

88
06:18.930 --> 06:23.290
word even though the document contains the inclusive word.

89
06:23.370 --> 06:25.000
The program couldn't find it.

90
06:25.010 --> 06:30.120
It's because the map is doing an exact match and the word answer with a dot.

91
06:30.270 --> 06:34.850
For example if I say inclusive that it can find the word.

92
06:34.860 --> 06:35.760
Now it works.

93
06:35.850 --> 06:37.290
But this is not good.

94
06:38.010 --> 06:42.140
So how can you remove the numbers and punctuation characters.

95
06:42.150 --> 06:48.210
Well there are a lot of ways but the simplest one for now is using a regular expression.

96
06:48.210 --> 06:51.360
This lecture is about maps and scanners as you know.

97
06:51.360 --> 06:55.050
So I'm not going to talk about the regular expressions in full detail.

98
06:55.090 --> 07:02.160
They're out of scope but put simply a regular expression allows us to find a pattern within a text

99
07:02.790 --> 07:04.770
to create a regular expression.

100
07:04.770 --> 07:08.470
You can use the regexp package.

101
07:08.700 --> 07:13.790
I'm going to create a pattern that matches any characters but letters.

102
07:14.330 --> 07:19.890
I've called the must compile function out of the loop because it's a costly operation.

103
07:19.890 --> 07:26.700
You only need to compile the regular expression pattern only once then you can use it as many times

104
07:26.820 --> 07:27.880
as you want.

105
07:27.990 --> 07:31.970
Compiling a pattern allows the matcher to work faster.

106
07:31.980 --> 07:32.240
Okay.

107
07:32.250 --> 07:34.250
Now let's see how it works.

108
07:34.380 --> 07:39.090
The must compile function takes a string and returns a matcher.

109
07:39.270 --> 07:46.070
So if we give it an incorrect pattern it will crash your program any function that starts with a must

110
07:46.080 --> 07:49.710
word does that it must work or it will crash.

111
07:49.710 --> 07:52.020
I know that my pattern is fine.

112
07:52.020 --> 07:55.450
So there is no harm in calling this function here.

113
07:55.470 --> 07:59.010
The square braces mean match to a character.

114
07:59.010 --> 08:05.760
The caret symbol means do not match. a to z means any character from a to z.

115
08:06.300 --> 08:10.740
And lastly the plus sign means match for the same pattern.

116
08:10.890 --> 08:12.420
One or more times.

117
08:12.900 --> 08:21.180
So in summary this pattern means find any character except letters one or more times the must compile

118
08:21.180 --> 08:27.860
function returns me a pattern matcher so I can use it to remove anything except letters.

119
08:27.900 --> 08:29.570
Let's use it down here.

120
08:29.580 --> 08:36.860
I'm going to call another function to replace anything but letters like so. You give it a string and

121
08:36.860 --> 08:40.020
you say search the pattern inside the word.

122
08:40.010 --> 08:43.300
And if you can find it replace it with an empty string.

123
08:43.400 --> 08:46.310
So this will remove anything but the letters.

124
08:46.310 --> 08:47.330
Let me show you.

125
08:47.450 --> 08:53.190
As you can see now there are only words without any numbers or punctuation characters.

126
08:53.260 --> 08:54.070
Awesome.

127
08:54.140 --> 09:00.380
For example the inclusive word doesn't contain dot anymore because I have removed it by using the regular

128
09:00.380 --> 09:01.050
expression.

129
09:01.730 --> 09:06.020
So now the program says that the document contains the inclusive word.

130
09:06.110 --> 09:06.800
Cool.

131
09:06.800 --> 09:13.250
These were a few examples to show you that using the buf io scanner you can get input from anywhere

132
09:13.640 --> 09:16.160
and you can use a map as a set.

133
09:16.200 --> 09:16.750
All right.

134
09:16.850 --> 09:19.740
That's how you can use a map as a set.

135
09:19.790 --> 09:20.790
That's all for now.

136
09:20.810 --> 09:21.830
See you in the next lecture.