Regular Expressions (Regex)#

Regular Expressions (regex) are patterns used to match and filter text. In Linux, they are commonly used with tools like grep, sed, and awk.

Practical note: Regex is dense but composable—focus on common patterns and reuse.

Core Constructs#

grep Modes#

Exercises#

1. Exclude Lines Containing ##

cat /etc/ssh/sshd_config | grep -v "#"

• -v → invert match (exclude comments)

2. Words Starting with Permit#

grep -E '\bPermit\w*' /etc/ssh/sshd_config

• \bPermit → word starts with “Permit”
• \w* → remaining characters

3. Words Ending with Authentication#

grep -E '\w*Authentication\b' /etc/ssh/sshd_config

• \w* → prefix
• Authentication\b → word ends with target

4. Lines Containing Key#

grep 'Key' /etc/ssh/sshd_config

• Simple substring match (no regex needed)

5. Lines Starting with Password and Containing yes#

⚠️ Original pattern is incorrect (^Password|yes = OR logic)

Correct:#

grep -E '^Password.*yes' /etc/ssh/sshd_config

• ^Password → start of line
• .* → any characters
• yes → must appear later in line

6. Lines Ending with yes#

grep -E 'yes$' /etc/ssh/sshd_config

• $ → end of line anchor

Common Patterns#

Best Practices#

• Prefer grep -E (extended regex)

• Avoid unnecessary cat:

  grep pattern file

• Build patterns incrementally

• Validate assumptions (anchors, grouping, precedence)

Mental Model#

• Regex = pattern → match set
• Anchors (^, $) constrain position
• Quantifiers (*, +) control repetition
• Pipes (|) introduce branching logic
• Most errors come from incorrect grouping or precedence