Copy the file RollList.csv from the moodle page to try out the following commands. The interpretation of the regular expression is given in plain English to help you gain understanding.
While grep uses a limited capability of pattern matching, egrep uses an extended list of patterns.
Lines that contain the string "MM" list all students from Meta.
egrep 'MM' RollList.csv
Lines that contain the string "MM19" will list all first years from Meta.
egrep 'MM19' RollList.csv
Lines that contain the string "ME19" will list all first years from Mech.
egrep 'ME19' RollList.csv
Lines that containt the sting ",A" will list all students whose name starts with an A.
egrep \',A\' RollList.csv
The caret symbol indicates that the string match has to be done at the beginning of the line. BTW, a dollar does the same for the end of the line.
List of first year students from Mech.
egrep '^ME19' RollList.csv
Since the command egrep picks up only those lines that match the pattern, we can combine the output of egrep command with other commands to further process the output.
Count the number of first year students from Mech.
egrep '^ME19' RollList.csv | wc -l
Count the number of second year students from Mech.
egrep '^ME18' RollList.csv | wc -l
Count the number of first year students from Meta.
egrep '^MM19' RollList.csv | wc -l
The dot symbol indicates that pattern can match at that position with any single character.
List of first year students who are from either Meta or Mech:
egrep '^M.19' RollList.csv
List of first year students from Meta whose name starts with A.
egrep 'MM19B...,A' RollList.csv
List of students from Mech whose roll numbers are within first 100.
egrep '^ME..B0..' RollList.csv
List of students from Mech whose roll numbers are above 100 and below 200.
egrep '^ME..B1..' RollList.csv
List of all kumars from Mech.
egrep '^ME.+Kumar' RollList.csv
Each pair of square brackets matches one character. The various options that be used for matching can be given inside the bracket. Ranges can also be given using a dash. Ranges work for three sets namely small alphabets, capital alphabets and numbers.
List of students who have a character "a" in their name, either capital or small. The letter can occur anywhere in the name.
egrep '[Aa]' RollList.csv
List of students whose name contains a vowel. We use the fact that names come after a comma in this file.
egrep '[aAeEiIoOuU]' RollList.csv
List of all students who have the characters "h" and "a" - either capital or small occuring in their name side by side. We are counting on the fact that roll numbers will not have this pattern.
egrep '[hH][aA]' RollList.csv
List of students who have the letter "a" occuring exactly twice and side by side in their name.
egrep '[Aa]{2}' RollList.csv
List of first year students -- from either from Mech or Meta -- whose roll numbers end with a digit between 0 and 4. This is one way to divide the class into two halves.
egrep '^M.19B..[0-4]' RollList.csv
List of students whose names start with letters in the range a to m, either small or capital.
egrep ',[a-mA-M]' RollList.csv
Special characters with a backslash indicate pattern matching with boundaries. The characters "\b" indicate word boundary.
List of students who have a variant of "jai" as a part of their name.
egrep '[jJ]a[iy]' RollList.csv
List of student who have a variant of "jai" as one of their names.
egrep '\b[jJ]a[iy]\b' RollList.csv
List of students who have a variant of "raj" as a part of their name.
egrep '[rR]aj' RollList.csv
List of students who have a variant of "raj" as one of their names.
egrep '\b[rR]aj\b' RollList.csv
If the caret symbol occurs within a pair of square brackets, it is to negate matching.
List of students who are not from first year.
egrep '1[^9]B' RollList.csv
List of students who do not have the letter "p" or "P" in their name. The plus character after the square brackets indicates at least one or more matching. Use the command without the plus and see the difference.
egrep '\b[^pP]+\b' RollList.csv
Students whose names sound like we are giving respect while pronouncing with their initial. That is, they have a G in the end as their initial. Since the name is the second and last field in each line, we can use the dollar symbol to signal that the matching should be done at the end of the line.
egrep '\b[^ ]+\b G$' RollList.csv
While square brackets match exactly one character with options given inside, parantheses and a pipe match strings of different lengths for pattern matching.
List of first year students who are from either Mech or Meta.
egrep 'M(M|E)19' RollList.csv
List of students from Mech who are from either first or second year and whose name starts with the character A.
egrep 'ME1(8|9)B...,A' RollList.csv
List of students whose name has a raj or a kumar.
egrep '(Raj|Kumar)' RollList.csv
One can also given multiple options using the pipe symbol.
List of students who have a variant of "sri" in their name.
egrep '(Shri|shri|Sri|Sri|Sree|sree|Shree|shree)' RollList.csv
List of all students
egrep '.*,.*\b.*$' RollList.csv
Use of caret symbol within square brackets to indicate negation of matching for non blank character. This will have same output as above.
egrep '.*,[^ ]+\b.*$' RollList.csv
List of students with two part names
egrep '.*,[^ ]+ [^ ]+$' RollList.csv
List of students with three part names
egrep '.*,[^ ]+ [^ ]+ [^ ]+$' RollList.csv
List of students with single initial in the end. We use the flower brackets with an integer within to state exactly how many times the pattern given in the preceding pair of square bracket should match. By asking for a non blank character that is occuring exactly once at the end of the string, we achieve this pattern.
egrep '.*,[^ ]+ [^ ]+ [^ ]{1}$' RollList.csv
List of students with single initial in the middle of their name. We use same logic as above but swap the positions of second and last part of the names.
egrep '.*,[^ ]+ [^ ]{1} [^ ]+$' RollList.csv
Regular expression matching also has capability to use the following categories of strings -- alphabetic, numeric and alphanumeric.
Match all alphabetical strings in the file. All names get matched this way.
egrep '[[:alpha:]]' RollList.csv
Match all numeric strings in the file. Only part of the roll numbers can be matched this way.
egrep '[[:digit:]]' RollList.csv
Match all alphanumeric strings in the file. You will see that roll numbers get matched as they are purely alphanumeric.
egrep '[[:alnum:]]' RollList.csv
Try and interpret the following pattens and see the output to check your understanding. Look at the contents of the file elist given below to try out.
egrep '\b[a-zA-Z]+@[a-zA-Z]+\.[a-zA-Z0-9]+\b' elist.txt
egrep '\b[a-zA-Z]+@[a-zA-Z0-9]+\.[a-zA-Z0-9]+\b' elist.txt
egrep '\b[a-zA-Z0-9]+@[a-zA-Z0-9]+\.[a-zA-Z0-9]+\b' elist.txt
egrep '\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Za-z]{2,6}\b' elist.txt
egrep '^\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Za-z]{2,6}\b$' elist.txt
Th file elist.txt should contain some emails. Contents of file elist.txt for your practice:
bogus@bogus
[email protected]
[email protected]
ThisIsMyEmailAddress
email.address
[email protected]@domain.com
Here is one email: [email protected]
[email protected]
[email protected]
Here is a complex one: [email protected]
- List processes that are being run by root or logged-in user
- List files that have write access to others
- List used space for only mounted hard disk partitions
- Find out which files in the /etc directory use the name of your machine
Output redirection using 2>~/errorfile.txt will make the errors go to a file called errofile.txt in your home directory.