Skip to main content
In Linux, regular expressions (regex) empower you to locate and filter text in files far beyond simple string matching. In this first part of our series, you will learn basic regex operators and how to craft simple yet powerful patterns with grep. Imagine you need to extract every IPv4 address (e.g., 203.102.3.5) from hundreds of configuration files. A naive approach that searches for “digit–dot–digit” ([0-9]\.[0-9]) would also match incomplete fragments like 1.2. By combining regex operators, you can define precise constraints and avoid false positives:
The image shows a grid of ten dark-themed code editor windows displaying configuration files with highlighted IP addresses. The text appears to be related to system or network configurations.
In mathematics, we often define a variable x by specifying its domain and constraints:
Let x be an integer, x > 3, and x < 8.
Then x ∈ .
The image shows a mathematical puzzle with conditions for an integer ( x ) where ( x > 3 ) and ( x < 8 ), with a sequence of numbers starting at 3 and ending at 8.
Regular expressions use a similar principle: you combine operators to specify exactly what a match should look like. Below is a quick reference of common regex metacharacters:
The image displays a set of regex operators, including symbols like ^, $, ., *, +, {}, ?, |, [], (), and [^].

​
1. Matching at the Beginning of a Line (^)

Consider a file names.txt containing:
A plain search for sam returns any line with that substring:
To match only lines that start with sam, prefix your pattern with ^:

​
2. Matching at the End of a Line ($)

Similarly, to find entries ending in sam, append $ to the pattern:
You can also apply this to system files, for example, looking for lines in /etc/login.defs that end with the digit 7:

​
3. The Dot (.) — Any Single Character

A period in regex matches exactly one character. For instance, c.t matches cat, cut, or c1t, but not ct.
To restrict matches to whole words, use the -w flag:

​
3.1 Escaping the Dot

If your goal is to match a literal dot (.), escape it with a backslash:

​
4. The Asterisk (*) — Zero or More

The asterisk applies to the element immediately before it, allowing that element to repeat zero or more times. For example, let* matches le, let, lett, letttt, etc.:
You can combine . and * to match any sequence of characters. For instance, extracting paths enclosed by slashes:

​
5. The Plus (+) — One or More

In basic grep, the + is not treated specially unless escaped. Use \+ to indicate “one or more” of the preceding item:
Basic grep requires escaping ?, +, |, (, and ). To leverage these metacharacters without backslashes, switch to extended regex mode using grep -E or the egrep command.
By combining these core operators, you can design precise search patterns to sift through logs, configuration files, or any text data on your Linux system.

​
Further Reading and References

Watch Video