Using grep with Regular Expressions on Linux and Unix Systems

Linux ships with the GNU version of grep , which supports extended regular expressions and is used to search for text anywhere on a server or workstation.

1. Quick Overview of Regular Expressions

Regular expressions are patterns that match each input line. Examples include ^w1 , w1|w2 , [^ ] .

Search for the string vivek in /etc/passwd :

grep vivek /etc/passwd

Sample output:

vivek:x:1000:1000:Vivek Gite,,,:/home/vivek:/bin/bash vivekgite:x:1001:1001::/home/vivekgite:/bin/sh gitevivek:x:1002:1002::/home/gitevivek:/bin/sh

Search case‑insensitively:

grep -i -w vivek /etc/passwd

Search for either vivek or raj (case‑insensitive, whole word):

grep -E -i -w 'vivek|raj' /etc/passwd

Anchor a pattern to the start of a line:

grep ^vivek /etc/passwd

Match lines that start with vivek but not vivekgite or vivekg :

grep -w ^vivek /etc/passwd

Match lines ending with foo :

grep 'foo$' FILENAME

Search for empty lines:

grep '^$' FILENAME

2. Matching Specific Characters

Match Vivek or vivek :

grep '[vV]ivek' FILENAME

Or using a fully case‑insensitive character class:

grep '[vV][iI][Vv][Ee][kK]' FILENAME

Match a digit after the name:

grep -w '[vV]ivek[0-9]' FILENAME

Match two‑digit numbers after foo :

grep 'foo[0-9][0-9]' FILENAME

Match any alphabetic character:

grep '[A-Za-z]' FILENAME

Match lines containing either w or n :

grep [wn] FILENAME

POSIX character classes (e.g., [:digit:] , [:upper:] ) can be used inside brackets. Example – match all uppercase letters:

grep '[:upper:]' FILENAME

3. Using Wildcards

The dot . matches any single character. Example – three‑character words that start with b and end with t :

grep '\<b.t\>' FILENAME

Match exactly two characters on a line:

grep '^..$' FILENAME

Match a literal dot in an IP address (escape the dot):

grep '192\.168\.1\.254' /etc/hosts

Extended regex for an IP address:

egrep '[[:digit:]]{1,3}\.[[:digit:]]{1,3}\.[[:digit:]]{1,3}\.[[:digit:]]{1,3}' FILENAME

Case‑insensitive match for Linux or Unix :

egrep -i '^(linux|unix)' FILENAME

4. Advanced grep Patterns

Search for a pattern that begins with a hyphen using -e :

grep -e '--test--' FILENAME

OR logic with extended regex:

grep -E 'word1|word2' FILENAME

OR can also be written with escaped pipe:

grep 'word1\|word2' FILENAME

AND logic by piping two greps:

grep 'word1' FILENAME | grep 'word2'

Complex AND using a single expression:

grep 'foo.*bar\|word3.*word4' FILENAME

Quantifiers to test repetitions:

{N} – exactly N occurrences

{N,} – N or more occurrences

{min,max} – between min and max occurrences

Match strings containing two consecutive v characters:

egrep "v{2}" FILENAME

Match col or cool :

egrep 'co{1,2}l' FILENAME

Match at least three c characters:

egrep 'c{3,}' FILENAME

Match a phone number of the form 91-1234567890 (two digits, optional space or hyphen, ten digits):

grep "[[:digit:]]\{2\}[ -]\?[[:digit:]]\{10\}" FILENAME

5. Highlighting and Output Control

Highlight matches:

grep --color regex FILENAME

Show only the matching part of each line:

grep -o regex FILENAME

6. grep vs egrep

egrep is equivalent to grep -E and treats the pattern as an extended regular expression. In basic regular expressions, meta‑characters such as ? , + , { , | , ( , ) lose their special meaning unless escaped.

References: GNU grep manual, info pages, and online regex documentation.