“Then the interface-makers went to work on their GUIs, and introduced a new semiotic layer between people and machines. People who use such systems have abdicated the responsibility, and surrendered the power, of sending bits directly to the chip that’s doing the arithmetic, and handed that responsibility and power over to the OS.” —Neal Stephenson, In the Beginning was the Command Line
The Command Line Interface (CLI) is a text-only interface to your computer. You can use it to access folders, files, settings—pretty much anything you would want to do on a computer. There is a version of the command line interface available on Windows, Mac OS, and [especially!] Linux/UNIX systems. For this workshop, for instance, we're going to be using Linux via a tool called Nitrous.IO.
- Control! You can type commands out and control exactly what your computer is doing, unlike on a Graphical User Interface (GUI), where your options are limited by the icons/menus some programmer has chosen to offer (and where what they do might not be what you intend).
- Scripting! You can use the command line to automate lots of things, like backing up files, editing text... the possibilities are endless!
- Programming! Most programming languages can be run from a CLI. When we do Python (yes, we'll eventually do Python), that's how we'll do it. So getting comfortable on the command line will make learning programming easier.
- And really? It just feels pretty cool to use it. :)
We'll be using Nitrous.IO. This is a really powerful tool, and we're only using a small part of it for this workshop. Definitely feel free to play with some of its functionality on your own!
You can set up your account before the workshop, if you want! Here are directions.
When you have the Nitrous console taking up the full screen, you can continue to the notes below. They can also be downloaded as a PDF.
See the white rectangle (or it may display as a blinking cursor, on some browsers)? That's the place where you enter commands, also known as the "command prompt."
In the directions below, when we say "Type a command," we want you to type the part in bold into the command prompt. Unless we add more information after the bold part, go ahead and hit "enter" after each command. There will be a few places where we have you wait, but not many.
-
pwd
Type pwd into the terminal and hit enter. (The "enter" is going to be assumed after all of your commands, from this point.)
You should see output that looks something like /home/action
This tells you where you are in the file structure. Linux, like Windows, organizes files into folders (also called directories) and subfolders (subdirectories). So if pwd returns /home/action, you're in a directory called action, which is in a directory called home
-
cd
Let's say you don't want to be in the /action directory anymore. You want to go up a level and be in /home.
Type cd ..
Type pwd
The output should be /home
You've changed directories ("cd" - get it?), up a level (that's what ".." means). But what if you didn't really want to do that? What if you think, "No, I want to be in /home/action now"? No problem!
Type cd action
Type pwd
Now, as you can see, you're back where you started, in /home/action.
-
mkdir
What if you aren't satisfied with the directories that are available to you? (Long-term, this is inevitable, right? You probably won't want to store all of your files in one place.) No problem!
Type mkdir [your name] — but don't literally type that; replace "[your name]" with your actual first name :)
Now change directory into the directory with your name. Scroll up for a reminder about changing directories, if you need it. Remember to use pwd to make sure you're in the right place.
-
nano
It's time to make a text file! There are lots of text editors available—and lots of strong opinions about which is the best one—but for today, we're going to use a simple one. It'll work great for anything you want to do, from writing to-do lists to writing code.
Type nano myfile — (go ahead and hit "enter", but don't panic when the screen changes entirely!)
You're inside an application, on a text-only interface. (How cool is that?) As you can see, there are commands along the bottom. Where you see a caret (^), it means you hold down the ctrl key along with that letter to execute that command. But first let's enter some text, before we worry about any of the commands.
Type whatever you like. I typed "Hi, my name is Coral. I like birds and chinchillas and coffee." Pretty much anything will do.
Once you've entered some text, hold down the ctrl key and hit X (shortened from now on to ctrl-x)
A prompt appears at the bottom of the screen: Save modified buffer (ANSWERING "No" WILL DESTROY CHANGES)?
Type y
The prompt now asks if this is the filename you want to write. It is, so
Hit enter
Congratulations! You just made a text file!
-
ls
How do you know you made a file, though? Not a problem. Let's have a look!
Type ls
The output should be myfile
You have just listed ("ls" - list) the contents of the current directory.
For fun, let's change directory up one level and look at what's inside that directory:
cd ..
ls
The output should be [your name] README.md workspace — the order might vary, based on where your name falls in the alphabet; that's fine
Now, change directory back into [your name], and make sure you're in the right place with pwd
One more thing: you can list the contents of directories other than the one you're in.
Type ls ..
The output should match what happened when you changed directory up a level and listed its contents, a moment ago, but you didn't have to change directory to see it! (Type pwd and confirm you're still in the same place, if you have doubts. :))
-
more / less
Let's say you want to know what's in myfile without opening it up in nano. That's totally doable.
Type more myfile
And, just to compare,
Type less myfile — you can escape the screen that comes up by typing q
It turns out, less is a bit more complex. For big files, more will let you go through it all in order, where less will let you scroll through it, both forward and backward.
If you use either more or less on a file (or any stream of data — we'll do something fancy with more later, so you can see what I mean), you can escape by hitting q (short for "quit").
-
file
Let's say you don't know what type of file myfile is. No biggie.
Type file myfile
The output should be myfile: ASCII text
-
cp
Oh, but we forgot to add an extension, showing that it's a text file, didn't we? Not a problem.
Type cp myfile myfile.txt
And then, to see how it worked, list the contents of the current directory. The output should be myfile myfile.txt
Now you have the original, myfile, plus an exact copy ("cp" - copy) with a different name, myfile.txt
-
rm
Now you have a file you don't need. No problem!
Type rm myfile — and then list the contents of the directory; the output should be myfile.txt
You've just removed ("rm" - remove) myfile.
This is a command to use carefully! Removing things is serious business.
-
mv
Copying and deleting every time you want to rename a file is a little tedious. There's a way to do the same thing in one step:
Type mv myfile.txt mynewfile.txt — and then list the contents of the directory; the output should be mynewfile.txt
You just moved ("mv" - move) a file. (You might ask why the command isn't "change name," or something. Valid question. This command is used for changing filenames, but it is also used for moving files between directories.)
If you want to try copying or moving your file out of /home/action/[your name] and into /home/action, for instance, here's the command: cp mynewfile.txt ../mynewfile.txt — you can replace 'cp' with 'mv' — and then you can move it back by typing mv ../mynewfile.txt mynewfile.txt
To show you how to find things, we're going to have to go further afield. And we're going to digress for a moment.
Type cd /usr/bin
Type ls
Whoa, right? Here, try this:
ls | more — that character in the middle is the horizontal line, usually located above the "enter" key on your keyboard
WHOA, RIGHT? — This technique is referred to as "piping output to more", which is a very useful thing to keep in your toolbox when you're faced with large files, or, in this case, large output streams from commands. (It is totally reasonable to have questions about this. Ask!) As you can see, you can hit enter to keep scrolling through the whole file list; when you're tired of looking at the list, remember, you can type q to get out of more.
One more tool for our toolset, before we move on:
Did you see the sequence of files, pygettext, pygettext2.5, pygettext2.6, etc.? (They're later in the alphabet, so you're more likely to see them on an ls than on an ls | more. No biggie if you don't go looking for them. Just believe me that they're there, OK?)
Type more pyg and then hit the tab key. — your command should now say more pygettext.
That's right; it will try to complete file names for you after you've started typing them, if you hit tab. Tab completion is AMAZING and makes every command line user's life better. But it's got limits, of course. The file you actually want is pygettext2.7; but because there are multiple files that all start with "pygettext", it completed as far as it could, and you'll have to specify further by finishing the filename yourself.
Make sure the command line says more pygettext2.7 and hit enter
You can browse the file if you want, and hit q at any point to get out of more — big file, right? It would be awful to have to look for a specific word in that file, wouldn't it?
-
grep
Type grep verbose pygettext2.7 — remember that you can use tab completion on the filename!
You'll see output something like this (but better-indented): --verbose
'style=', 'verbose', 'version', 'width=', 'exclude-file=',
verbose = 0
elif opt in ('-v', '--verbose'):
options.verbose = 1
if options.verbose:
if options.verbose:What's happening? The command grep searches within a file for instances of a particular set of characters, in this case "verbose". (It doesn't have to be a whole word. Any character string will work. And, yes, capitalization DOES matter.) The output shows you all of the places "verbose" appears within the file pygettext2.7.
It's time to learn another important tool: flags.
Type grep -n verbose pygettext2.7
Now the output has line numbers! The -n flag told the command grep that you want to know what line number each instance of the character string "verbose" appeared on.
(There are flags for ls; I usually use the flags a, h, and l, so when I list the contents of a directory, I do it by typing ls -ahl — see how the flags are combined? It's fine to go ahead and try that command in this or any directory. ls -ahl /home/action will give you a full list of all of the files in the directory we started in, for instance! I also like to use the -c tag with nano, so that I can see line and column numbers as I work.)
-
man
Want to know more about how to use grep, or any other command? Use man
Type man grep
As you can see, there are a number of flags and options available.
Feel free to scroll through, or to hit q to exit.
(Real talk: I usually Google any command I want to use, if that option is available to me, rather than using man. Different people have different preferences for finding information.)
-
up arrow
Hit the up arrow (↑)
The last command you typed (probably man grep) is now on the command line! If you keep hitting the up arrow, you can scroll through your command history. This saves a lot of time when you're typing long commands that you need to use more than once!
-
ctrl-c
You can stop the current command from running by hitting ctrl-c