Compiler and Interpreter: Compiled Language vs Interpreted Programming Languages — Transcript

Full transcript

1. 0:00So, we need to get our source code
2. 0:02converted into machine code somehow
3. 0:05before it can run. And there are two
4. 0:07main ways of doing this. What's called
5. 0:09compiling the source code, and what's
6. 0:11called interpreting the source code.
7. 0:14Now, luckily, this is not a big decision
8. 0:16you have to worry about. Most languages
9. 0:18you'll deal with will naturally fall
10. 0:20into one or the other, but it is worth
11. 0:22knowing the difference. So, let's have a
12. 0:24simple scenario. Let's say it's just you
13. 0:27and me. You have your computer, and I
14. 0:30have my computer, and you're going to
15. 0:32write a program that you want me to run.
16. 0:35Now, with a compiled language, what
17. 0:37happens is you write your source code,
18. 0:39and then you have a program called a
19. 0:41compiler that will go through that
20. 0:43source code and create a separate file
21. 0:45that contains the machine code, and you
22. 0:47just give me that file. This end result
23. 0:50is sometimes referred to as an
24. 0:52executable or an executable file,
25. 0:54because I can directly execute it. I can
26. 0:56now just run your program. You keep your
27. 0:59source code, and I never see it. Now,
28. 1:02with an interpreted language on the
29. 1:04other hand, you don't compile your
30. 1:06source code beforehand. You just give me
31. 1:08a copy of it. So, I'll need my machine
32. 1:11to interpret it whenever I want to run
33. 1:14your program. Now, an interpreter is
34. 1:16different to a compiler. It does this on
35. 1:18the fly. We can think of it as going
36. 1:20through your source code line by line
37. 1:23and processing it on the spot. It does
38. 1:25not save it as a separate machine code
39. 1:27file.
40. 1:29Now, you've used interpreted languages
41. 1:30even if you don't know it. Whenever
42. 1:32you've looked at a web page with
43. 1:33JavaScript, which, if you've surfed the
44. 1:35web for more than 2 minutes in your
45. 1:37lifetime, you have, this is what's been
46. 1:40happening. The JavaScript has been sent
47. 1:42to you over the web, along with a bunch
48. 1:44of other files like web pages and
49. 1:46images, and it's been sent as source
50. 1:48code onto your machine, and your web
51. 1:50browser has just interpreted that
52. 1:52JavaScript so it can run that code. So,
53. 1:55which one's best? Well, they both have
54. 1:57their good and their bad points.
55. 1:59Benefits of compiled code. Once it's
56. 2:02compiled, it's immediately ready to run
57. 2:04and you could send it to 100 or 1,000 or
58. 2:06100,000 different people. It's ready to
59. 2:08go. It can be optimized for a CPU, so it
60. 2:12can actually be faster. And you don't
61. 2:14have to send your source code to
62. 2:16everybody, which might be a good thing.
63. 2:18However, the downsides are if I compile
64. 2:21it on a PC, that executable file won't
65. 2:24work on a Mac. In fact, it often needs
66. 2:26to be compiled separately for different
67. 2:28kinds of CPU even on the same platform.
68. 2:31And when you're writing code, to compile
69. 2:34is an extra step that you have to take
70. 2:36every time you want to test your
71. 2:37program. Now, with interpreted code, the
72. 2:40big benefits are I don't really care
73. 2:42what kind of machine is on the other end
74. 2:44because we don't provide machine code.
75. 2:46We just send the source code and we let
76. 2:48the other side take care of it. So, it
77. 2:50can be more portable and more flexible
78. 2:52across platforms.
79. 2:54It's also a little easier when testing
80. 2:56because you just write your source code
81. 2:58and then run it, letting the interpreter
82. 3:00take care of converting it. There is no
83. 3:02in-between compile step. And it can be
84. 3:05easier to debug when things go wrong
85. 3:07because you always have access to all
86. 3:09the source code. However, it has its
87. 3:11downsides, too. Because everyone who
88. 3:13needs to run that program on their
89. 3:15machine has to have an interpreter for
90. 3:17that language on their machine.
91. 3:19It also can be slower because you have
92. 3:21to interpret it every time the program
93. 3:23is run. And the source code is
94. 3:25effectively public because you're
95. 3:27sending it to everyone who needs to run
96. 3:28that program. Now, because there are
97. 3:31good things about compiled languages and
98. 3:33good things about interpreted languages,
99. 3:35there's also a third way of doing this,
100. 3:37which is a bit of both. Instead of the
101. 3:40compile model where all the work is done
102. 3:42up front, but can be a little bit
103. 3:44inflexible,
104. 3:46or the interpreted model where all the
105. 3:49work is done on the receiving end, but
106. 3:51can be a little bit slower.
107. 3:53We kind of do half and half. Up front,
108. 3:56we compile it part of the way to what's
109. 3:58called an intermediate language, which
110. 4:01takes it as far along the way to machine
111. 4:03code as it can get while still being
112. 4:05portable often across platforms.
113. 4:07You then distribute this, sending it to
114. 4:10the people who need to run it, and each
115. 4:12person who runs it takes it the last
116. 4:15step to take it to machine code on their
117. 4:17computers. This is sometimes referred to
118. 4:19as just-in-time or JIT compilation. Now,
119. 4:22this intermediate language sometimes
120. 4:24also goes by the name of bytecode. So,
121. 4:26this process has to happen somehow. It's
122. 4:29just how much of it happens on your
123. 4:31machine and how much of it happens on
124. 4:33mine.
125. 4:34Now, while theoretically all computer
126. 4:36languages could use any of these
127. 4:39methods, the normal usage of any one
128. 4:41language tends to be one or the other.
129. 4:44So, for example,
130. 4:46C, C++, and Objective-C, these are
131. 4:48typically found as compiled languages.
132. 4:50So, you need a compiler. Now, the
133. 4:52compiler can be downloaded for free, but
134. 4:54are often built into integrated
135. 4:57development environment applications.
136. 5:00Now, languages like PHP and JavaScript,
137. 5:03indeed most languages with the word
138. 5:04script at the end are usually
139. 5:06interpreted. And languages like Java,
140. 5:09C#, VB.NET, and Python use this
141. 5:12intermediate hybrid approach. Now,
142. 5:15whether a language is compiled or
143. 5:17interpreted or somewhere in between
144. 5:20is rarely a reason by itself to choose a
145. 5:22language, but it can be something that
146. 5:24you take into account. If one main
147. 5:26priority of your program is absolute
148. 5:28maximum speed running on one single
149. 5:31platform,
150. 5:32you'll probably look at a compiled
151. 5:34language. If you're more interested in
152. 5:36easily moving your code across multiple
153. 5:38platforms, you're probably more
154. 5:40interested in an interpreted one. But
155. 5:42more usually, you're driven more by what
156. 5:45you need to do. Do you need to build
157. 5:47iPhone apps or Windows desktop apps or
158. 5:49dynamic websites or in our case just
159. 5:52learn the fundamentals of programming
160. 5:54and you let that decision drive the
161. 5:56language choice and the language choice
162. 5:59will determine whether you're compiled,
163. 6:01interpreted, or somewhere in the middle.