In the last two posts, we talked about finding things. The result of that find command was a list of filenames or directory names… just lists of things. Which is useful if we want to know ‘where did I leave that file’. But I want to do things with these files. Suppose I want to know when all files that have an ‘s’ in them were last accessed. I could do that by finding all of those files and then doing ls -l on each one. But I’m lazy, I don’t want to do that.

As I’ve said many times, there’s more than one way of doing things in Linux. You can do things the long and complicated way or you can learn a little about the command line and do things the short and fast way. Let’s talk about that short and fast way and get to the end result. I’m lazy. I don’t want to type the same thing over and over again. Luckily, find lets me get away from that. See?

find . -name '*s*' -exec ls -l {} \;

So I’m finding all files that have an s in them, just like I’ve done before. Now, though, I’m adding a new flag, -exec. After that flag, I give the command I want. The {} means take the output of the find and put that value in those brackets. So it’s going to do ls -l on every file it finds.

One command to take care of all those commands I listed before, isn’t that neat?

Now let’s pretend I want to rid my system of all sith. Those sith infested my system with lots of files, all of them with the name sith (or some variant thereof) and I’ve got to get rid of them.

find / -iname '*sith*' -exec rm {} \;

So I find any file with the name sith or Sith or DarthSith or SithofDarkness or even sItH… and I remove them. All go away and I don’t even need to find my lightsaber.

We’ve talked about finding things in files and using grep and more to do that, let’s talk about finding files. I’ve lost my file and I can’t look through my file until I find my file.

And that’s what the command Linux® uses to find files is called. find. How’s that for imaginative naming? We’re going to find with find.

Let’s begin by cd‘ing into our home directory. Remember we can do that just by typing cd.

We want to find that file blog1.txt we created a while back. We know it exists, we know we put it somewhere, but we forget where. (It was in the home directory but let’s pretend otherwise for right now.) We tell find to start in the home directory and look for a file named blog1.txt. This is done by:

Alpha:~ computerlamp$ find . -name blog1.txt

See that ‘.’ after find? That’s really important. That tells find to start in the directory you’re in and look from there. The flag ‘-name’ tells the command ‘look for the file with this name’.

We can do other neat things. Suppose we know the file we’re looking for ends with the letters txt but we’ve forgotten if it started with blog, or blogg, or maybe bloggo, or even MyBlog. But we know it ends with txt, so that’s a start! In this case, we can do this command:

Alpha:~ computerlamp$ find . -name '*txt'

See that asterisk? That tells find to match any file that ends with txt and starts with anything else. It could be blog or bloggo, but it could even be thisfile.txt. Or turtles.txt. Or turtles-in-txt. It can match all of those and find will return every single one.

We can also find directories with find. Remember that Sewer directory we made a while back? find will find that as well:

Alpha:~ computerlamp$ find . -name Sewer

We can also use find to look into the sewer.

Alpha:~ computerlamp$ find Sewer

I didn’t give it a name flag that time. That means it’ll look into that directory and find whatever is in there.

In the last post we discussed how to find a word in the file /usr/share/dict/words using more.  This being Linux®, there’s always another way to do it.  This way uses the command grep. Jumping right to an example, try this:

Alpha:~ computerlamp$  grep way /usr/share/dict/words

This returns every single word in the file /usr/share/dict/words that contains way.  As in the last post, we’re looking for words that begin with way, not any word that contains way. We can find those words using:

Alpha:~ computerlamp$  grep ^way /usr/share/dict/words

This returns every word that begins with way.  When I tried it, I got 48 words.  How many do you get?

grep is case specific. This means that if you tell it to find way, it won’t find Way or wAy or waY. It matches precisely what you give it. As luck would have it, we don’t have to grep for every possible combination of case. There’s a flag for that. We can use the flag -i to ignore the case. That means, match every possible case when looking for the word way. For example:

Alpha:~ computerlamp$  grep -i ^way /usr/share/dict/words

That finds every possible word that begins with way, no matter what case it has in the file.

How, what happens when you look for a string in a command?  Can I grep grep?

Let’s try it:

Alpha:~ computerlamp$  grep test /usr/bin/grep

I got nothing back.  So it didn’t fail, but it didn’t work.   That’s because the command grep is a binary, which means there’s generally no text in it.

Now let’s try something else!  Suppose we want to find all strings in the file that don’t match our string.   So following our example, we want all strings in /usr/share/dict/words that don’t have the string way in them.  In this case, we do:

Alpha:~ computerlamp$  grep -v way /usr/share/dict/words

If you do this,  you get a lot of results.  An awful lot of results.