OverviewTeaching: 25 min
Exercises: 20 minQuestions
How can I find files?
How can I find things in files?Objectives
grepto select lines from text files that match simple patterns.
findto find files and directories whose names match simple patterns.
Use the output of one command as the command-line argument(s) to another command.
Explain what is meant by ‘text’ and ‘binary’ files, and why many common tools don’t handle the latter well.
In the same way that many of us now use ‘Google’ as a verb meaning ‘to find’, Unix programmers often use the word ‘grep’. ‘grep’ is a contraction of ‘global/regular expression/print’, a common sequence of operations in early Unix text editors. It is also the name of a very useful command-line program.
grep finds and prints lines in files that match a pattern.
For our examples,
we will use a file that contains three haikus taken from a
1998 competition in Salon magazine. For this set of examples,
we’re going to be working in the writing subdirectory:
$ cd $ cd Desktop/shell-lesson-data/writing $ cat haiku.txt
The Tao that is seen Is not the true Tao, until You bring fresh toner. With searching comes loss and the presence of absence: "My Thesis" not found. Yesterday it worked Today it is not working Software is like that.
Forever, or Five Years
We haven’t linked to the original haikus because they don’t appear to be on Salon’s site any longer. As Jeff Rothenberg said, ‘Digital information lasts forever — or five years, whichever comes first.’ Luckily, popular content often has backups.
Let’s find lines that contain the word ‘not’:
$ grep not haiku.txt
Is not the true Tao, until "My Thesis" not found Today it is not working
not is the pattern we’re searching for. The grep command searches through the file, looking for matches to the pattern specified. To use it type
grep, then the pattern we’re searching for and finally the name of the file (or files) we’re searching in.
The output is the three lines in the file that contain the letters ‘not’.
By default, grep searches for a pattern in a case-sensitive way. In addition, the search pattern we have selected does not have to form a complete word, as we will see in the next example.
Let’s search for the pattern: ‘The’.
$ grep The haiku.txt
The Tao that is seen "My Thesis" not found.
This time, two lines that include the letters ‘The’ are outputted, one of which contained our search pattern within a larger word, ‘Thesis’.
To restrict matches to lines containing the word ‘The’ on its own,
we can give
grep with the
This will limit matches to word boundaries.
Later in this lesson, we will also see how we can change the search behavior of grep with respect to its case sensitivity.
$ grep -w The haiku.txt
The Tao that is seen
Note that a ‘word boundary’ includes the start and end of a line, so not
just letters surrounded by spaces.
Sometimes we don’t
want to search for a single word, but a phrase. This is also easy to do with
grep by putting the phrase in quotes.
$ grep -w "is not" haiku.txt
Today it is not working
We’ve now seen that you don’t have to have quotes around single words, but it is useful to use quotes when searching for multiple words. It also helps to make it easier to distinguish between the search term or phrase and the file being searched. We will use quotes in the remaining examples.
Another useful option is
-n, which numbers the lines that match:
$ grep -n "it" haiku.txt
5:With searching comes loss 9:Yesterday it worked 10:Today it is not working
Here, we can see that lines 5, 9, and 10 contain the letters ‘it’.
We can combine options (i.e. flags) as we do with other Unix commands.
For example, let’s find the lines that contain the word ‘the’. We can combine
-w to find the lines that contain the word ‘the’ and
-n to number the lines that match:
$ grep -n -w "the" haiku.txt
2:Is not the true Tao, until 6:and the presence of absence:
Now we want to use the option
-i to make our search case-insensitive:
$ grep -n -w -i "the" haiku.txt
1:The Tao that is seen 2:Is not the true Tao, until 6:and the presence of absence:
Now, we want to use the option
-v to invert our search, i.e., we want to output
the lines that do not contain the word ‘the’.
$ grep -n -w -v "the" haiku.txt
1:The Tao that is seen 3:You bring fresh toner. 4: 5:With searching comes loss 7:"My Thesis" not found. 8: 9:Yesterday it worked 10:Today it is not working 11:Software is like that.
If we use the
-r (recursive) option,
grep can search for a pattern recursively through a set of files in subdirectories.
Let’s search recursively for
Yesterday in the
$ grep -r Yesterday .
data/LittleWomen.txt:"Yesterday, when Aunt was asleep and I was trying to be as still as a data/LittleWomen.txt:Yesterday at dinner, when an Austrian officer stared at us and then data/LittleWomen.txt:Yesterday was a quiet day spent in teaching, sewing, and writing in my haiku.txt:Yesterday it worked
grep has lots of other options. To find out what they are, we can type:
$ grep --help
Usage: grep [OPTION]... PATTERN [FILE]... Search for PATTERN in each FILE or standard input. PATTERN is, by default, a basic regular expression (BRE). Example: grep -i 'hello world' menu.h main.c Regexp selection and interpretation: -E, --extended-regexp PATTERN is an extended regular expression (ERE) -F, --fixed-strings PATTERN is a set of newline-separated fixed strings -G, --basic-regexp PATTERN is a basic regular expression (BRE) -P, --perl-regexp PATTERN is a Perl regular expression -e, --regexp=PATTERN use PATTERN for matching -f, --file=FILE obtain PATTERN from FILE -i, --ignore-case ignore case distinctions -w, --word-regexp force PATTERN to match only whole words -x, --line-regexp force PATTERN to match only whole lines -z, --null-data a data line ends in 0 byte, not newline Miscellaneous: ... ... ...
Which command would result in the following output:
and the presence of absence:
grep "of" haiku.txt
grep -E "of" haiku.txt
grep -w "of" haiku.txt
grep -i "of" haiku.txt
The correct answer is 3, because the
-woption looks only for whole-word matches. The other options will also match ‘of’ when part of another word.
grep’s real power doesn’t come from its options, though; it comes from the fact that patterns can include wildcards. (The technical name for these is regular expressions, which is what the ‘re’ in ‘grep’ stands for.) Regular expressions are both complex and powerful; if you want to do complex searches, please look at the lesson on our website. As a taster, we can find lines that have an ‘o’ in the second position like this:
$ grep -E "^.o" haiku.txt
You bring fresh toner. Today it is not working Software is like that.
We use the
-Eoption and put the pattern in quotes to prevent the shell from trying to interpret it. (If the pattern contained a
*, for example, the shell would try to expand it before running
^in the pattern anchors the match to the start of the line. The
.matches a single character (just like
?in the shell), while the
omatches an actual ‘o’.
Tracking a Species
Leah has several hundred data files saved in one directory, each of which is formatted like this:
2013-11-05,deer,5 2013-11-05,rabbit,22 2013-11-05,raccoon,7 2013-11-06,rabbit,19 2013-11-06,deer,2
She wants to write a shell script that takes a species as the first command-line argument and a directory as the second argument. The script should return one file called
species.txtcontaining a list of dates and the number of that species seen on each date. For example using the data shown above,
Put these commands and pipes in the right order to achieve this:
cut -d : -f 2 > | grep -w $1 -r $2 | $1.txt cut -d , -f 1,3
man grepto look for how to grep text recursively in a directory and
man cutto select more than one field in a line.
An example of such a file is provided in
grep -w $1 -r $2 | cut -d : -f 2 | cut -d , -f 1,3 > $1.txt
Actually, you can swap the order of the two cut commands and it still works. At the command line, try changing the order of the cut commands, and have a look at the output from each step to see why this is the case.
You would call the script above like this:
$ bash count-species.sh bear .
You and your friend, having just finished reading Little Women by Louisa May Alcott, are in an argument. Of the four sisters in the book, Jo, Meg, Beth, and Amy, your friend thinks that Jo was the most mentioned. You, however, are certain it was Amy. Luckily, you have a file
LittleWomen.txtcontaining the full text of the novel (
shell-lesson-data/writing/data/LittleWomen.txt). Using a
forloop, how would you tabulate the number of times each of the four sisters is mentioned?
Hint: one solution might employ the commands
|, while another might utilize
grepoptions. There is often more than one way to solve a programming task, so a particular solution is usually chosen based on a combination of yielding the correct result, elegance, readability, and speed.
for sis in Jo Meg Beth Amy do echo $sis: grep -ow $sis LittleWomen.txt | wc -l done
Alternative, slightly inferior solution:
for sis in Jo Meg Beth Amy do echo $sis: grep -ocw $sis LittleWomen.txt done
This solution is inferior because
grep -conly reports the number of lines matched. The total number of matches reported by this method will be lower if there is more than one match per line.
Perceptive observers may have noticed that character names sometimes appear in all-uppercase in chapter titles (e.g. ‘MEG GOES TO VANITY FAIR’). If you wanted to count these as well, you could add the
-ioption for case-insensitivity (though in this case, it doesn’t affect the answer to which sister is mentioned most frequently).
grep finds lines in files,
find command finds files themselves.
it has a lot of options;
to show how the simplest ones work, we’ll use the directory tree shown below.
writing directory contains one file called
haiku.txt and three subdirectories:
thesis (which contains a sadly empty file,
data (which contains three files
tools directory that contains the programs
and a subdirectory called
old, with a file
For our first command,
find . (remember to run this command from the
$ find .
. ./data ./data/one.txt ./data/LittleWomen.txt ./data/two.txt ./tools ./tools/format ./tools/old ./tools/old/oldtool ./tools/stats ./haiku.txt ./thesis ./thesis/empty-draft.md
. on its own means the current working directory,
which is where we want our search to start.
find’s output is the names of every file and directory
under the current working directory.
This can seem useless at first but
find has many options
to filter the output and in this lesson we will discover some
The first option in our list is
-type d that means ‘things that are directories’.
find’s output is the names of the five directories in our little tree
$ find . -type d
./ ./data ./thesis ./tools ./tools/old
Notice that the objects
find finds are not listed in any particular order.
If we change
-type d to
we get a listing of all the files instead:
$ find . -type f
./haiku.txt ./tools/stats ./tools/old/oldtool ./tools/format ./thesis/empty-draft.md ./data/one.txt ./data/LittleWomen.txt ./data/two.txt
Now let’s try matching by name:
$ find . -name *.txt
We expected it to find all the text files,
but it only prints out
The problem is that the shell expands wildcard characters like
* before commands run.
*.txt in the current directory expands to
the command we actually ran was:
$ find . -name haiku.txt
find did what we asked; we just asked for the wrong thing.
To get what we want,
let’s do what we did with
*.txt in quotes to prevent the shell from expanding the
find actually gets the pattern
*.txt, not the expanded filename
$ find . -name "*.txt"
./data/one.txt ./data/LittleWomen.txt ./data/two.txt ./haiku.txt
Listing vs. Finding
findcan be made to do similar things given the right options, but under normal circumstances,
lslists everything it can, while
findsearches for things with certain properties and shows them.
As we said earlier,
the command line’s power lies in combining tools.
We’ve seen how to do that with pipes;
let’s look at another technique.
As we just saw,
find . -name "*.txt" gives us a list of all text files in or below the current directory.
How can we combine that with
wc -l to count the lines in all those files?
The simplest way is to put the
find command inside
$ wc -l $(find . -name "*.txt")
11 ./haiku.txt 300 ./data/two.txt 21022 ./data/LittleWomen.txt 70 ./data/one.txt 21403 total
When the shell executes this command,
the first thing it does is run whatever is inside the
It then replaces the
$() expression with that command’s output.
Since the output of
find is the four filenames
the shell constructs the command:
$ wc -l ./data/one.txt ./data/LittleWomen.txt ./data/two.txt ./haiku.txt
which is what we wanted.
This expansion is exactly what the shell does when it expands wildcards like
but lets us use any command we want as our own ‘wildcard’.
It’s very common to use
The first finds files that match a pattern;
the second looks for lines inside those files that match another pattern.
Here, for example, we can find PDB files that contain iron atoms
by looking for the string ‘FE’ in all the
.pdb files above the current directory:
$ grep "FE" $(find .. -name "*.pdb")
../data/pdb/heme.pdb:ATOM 25 FE 1 -0.924 0.535 -0.518
Matching and Subtracting
grepinverts pattern matching, so that only lines which do not match the pattern are printed. Given that, which of the following commands will find all files in
/datawhose names end in
s.txtbut whose names also do not contain the string
net? (For example,
planets.txt.) Once you have thought about your answer, you can test the commands in the
find data -name "*s.txt" | grep -v net
find data -name *s.txt | grep -v net
grep -v "net" $(find data -name "*s.txt")
- None of the above.
The correct answer is 1. Putting the match expression in quotes prevents the shell expanding it, so it gets passed to the
Option 2 is incorrect because the shell expands
*s.txtinstead of passing the wildcard expression to
Option 3 is incorrect because it searches the contents of the files for lines which do not match ‘net’, rather than searching the file names.
We have focused exclusively on finding patterns in text files. What if your data is stored as images, in databases, or in some other format?
A handful of tools extend
grepto handle a few non text formats. But a more generalizable approach is to convert the data to text, or extract the text-like elements from the data. On the one hand, it makes simple things easy to do. On the other hand, complex things are usually impossible. For example, it’s easy enough to write a program that will extract X and Y dimensions from image files for
grepto play with, but how would you write something to find values in a spreadsheet whose cells contained formulas?
A last option is to recognize that the shell and text processing have their limits, and to use another programming language. When the time comes to do this, don’t be too hard on the shell: many modern programming languages have borrowed a lot of ideas from it, and imitation is also the sincerest form of praise.
The Unix shell is older than most of the people who use it. It has survived so long because it is one of the most productive programming environments ever created — maybe even the most productive. Its syntax may be cryptic, but people who have mastered it can experiment with different commands interactively, then use what they have learned to automate their work. Graphical user interfaces may be easier to use at first, but once learned, the productivity in the shell is unbeatable. And as Alfred North Whitehead wrote in 1911, ‘Civilization advances by extending the number of important operations which we can perform without thinking about them.’
findPipeline Reading Comprehension
Write a short explanatory comment for the following shell script:
wc -l $(find . -name "*.dat") | sort -n
- Find all files with a
.datextension recursively from the current directory
- Count the number of lines each of these files contains
- Sort the output from step 2. numerically
findfinds files with specific properties that match patterns.
grepselects lines in files that match patterns.
--helpis an option supported by many bash commands, and programs that can be run from within Bash, to display more information on how to use these commands or programs.
man [command]displays the manual page for a given command.
$([command])inserts a command’s output in place.