There is a directory available to you where you can practice filename expansions.
cp -r ~eric.brauer/uli101/wordlist ~
cd ~/wordlist
less 0README
The shell not only accepts specific filenames. You can also use ‘expansions’ to capture more than one file.
touch bat bet bit bot boat bath && ls
Aside: We can combine two commands on one line using && like you see above. The second command will only run if the first completed with no errors.
If we type in ls -l bat, we will get back long list information about only that specific file, identified by three explicit characters: b, a, and t.
Ok, so here’s the first expansion symbol:
ls -l b?t
-rw-rw-r-- 1 eric eric 0 Sep 8 15:22 bat
-rw-rw-r-- 1 eric eric 0 Sep 8 15:22 bet
-rw-rw-r-- 1 eric eric 0 Sep 8 15:22 bit
-rw-rw-r-- 1 eric eric 0 Sep 8 15:22 botHere, all the matches are lower case letters. But keep in mind that character refers to any letter, number, or symbol that can be used to name a file.
That last example was very greedy, because any single character between the ‘b’ and the ‘t’ would be accepted. Let’s say that instead, we want to specify which single character we want to match. Let’s only match files with ‘a’ or ‘o’. To do this, we will use square brackets.
ls -l b[ao]t
So we have matched two filenames. Both filenames are three letters long, and both start with ‘b’ and end with ‘t’. The single character in the middle is either ‘a’ or ‘o’.
Note #1: Notice that we didn’t match the filename boat! Remember that square brackets only match a single character!
We can invert our previous match by putting an exclamation mark inside the square brackets:
ls -l b[!ao]t
This is matching any filename that is three characters long, starting with ‘b’ and ending with ‘t’ as long as the middle character isn’t an ‘a’ or an ‘o’.
Note #2: A common mistake by students is to put single characters between square brackets, ie. b[o]t. There is no difference between b[o]t and bot, except that using a single character will lose you marks on a test. However, b[!o]t will match any character that isn’t ‘o’ and is therefore acceptable.
As well specifying characters inside square brackets, we can use ranges. All of the below are acceptable:
ls -l [A-B]at ls -l [A-Za-z]at ls -l lecture[0-9] ls -l abc[a-z0-9-_]
Using either the ? or [ ] will match only a single character (and never zero characters). Our next symbol will do this:
Let’s demonstrate this by removing our old files and creating new files:
rm * touch .f f foo foot footprint && ls
Notice that .f seems to be missing. Any file or directory that begins with ‘.’ will be considered a hidden file. To view them you will have to use:
ls -a
Notice we are also seeing ‘.’ and ‘..’, which represent the current directory and the parent directory.
ls f*
We are matching any filename that starts with ‘f’, but then after that we match with zero characters, we match with two characters (oo), we match with three characters (oot), and with even more characters (ootprint). Basically, we are saying we don’t care about how many characters come after ‘f’, just as long as that ‘f’ is there at the beginning.
Note #3: Notice that we didn’t match with the hidden file .f!
ls f*t
This time, we don’t care how many characters are in between ‘f’ and ‘t’, just as long as our filename begins with ‘f’ and ends with ‘t’.
We’ve been using expansions with ls, but that doesn’t have to be the case.
cp *.jpg images/
This is going to match any filename that ends with ‘.jpg’, and copy those files into a directory called images. This is useful if you are reorganizing a web project.
rm 2018-0[1-5]-*.c
Let’s say that you’re sick of looking at your old source code from the Winter semester of 2018. (And you named all of your source code with dates!) Any date that matches between the month of 01 and 05 will get deleted.
cat lecture3?.md > lecture3.md
Maybe at some point I will decide that instead of having two lectures a week, I want to create one big markdown file for each week. cat will combine two textfiles into one.
File expansions get very useful with the find command, since usually we lost something because we typed it in wrong. The first thing we can do to expand our search is to ignore case. This is easy to do.
Instead of this command:
find ~ -name report.pdf
Use this command:
find ~ -iname report.pdf
For every alphabetic character, the find command will include all upper and lower case possibilities.
But you will also benefit sometimes from using expansions such as ’report*’ in your search.
We aren’t limited to just searching by name, either. We can combine a ‘name’ or ‘iname’ search with other conditions. For example, we might want to limit our search to only files or only directories.
find ~ -iname report* -type f
find ~ -iname oxford -type d
You can combine this with filtering your search by time. You can specify that you want to see when files were changed, and when they were accessed.
Try this. We are going to use the file called ‘friendly’ from last week. If you don’t have this file, no problem. The command below will create it:
echo "This is a test" >> friendly
find . -mmin -20
You will see testfile show up in results. -mmin stands for modification + minutes. -20 sets the search for less than 20. So this search should return all files that have had their contents changed less than 20 minutes ago. Note: Timestamps in Linux are weird. More info.
Finally, one more useful search. This is for size.
find ~ -size +1G
This will return all files larger than 1 Gigabyte. G = Gigabyte, M = Megabyte, k = kilobyte.
find is an complicated tool. By default, it will use ls to list the positive results, but you can even specify other commands to run on the results. I encourage you to search through examples and options online and through the man pages.
If you want to only see the first ten lines of a file, use head. If you want to only see the last ten lines of a file, use tail. You can also specify the number of lines with an argument:
head -3 Acceptable-Use-Policy
Please refer to http://www.senecac.on.ca/policies/itau.html for the latest and
up to date copy of the Information Technology Acceptable Use.The most common way to learn about files is to use ls -l, which you’ve seen before. But there are more options.
Let’s use an image from last week:
ls -l lecture2a-nautilus.png
Let’s learn more about this with the file command.
file lecture2a-nautilus.png
In Windows, traditionally the extension of the file was very important. If you took an image called ‘file.jpg’ and renamed it ‘file.txt’, it would probably cause notepad.exe to choke on that file. In Linux, extensions aren’t required, and we mostly keep them around by convention.
cp lecture2a-nautilus.png test.txt
file test.txt
Commands
- find, continued: type, size, amin, cmin
- file: Information about files
- head: Show first X lines
- tail: Show last X lines
- history: Show history
Expansions
- *: Zero or more characters
- ?: One character
- [ ]: Match one character of whatever’s inside the brackets
- [! ]: Exclude one character of whatever’s inside the brackets