Bash

🔍 Regular Expressions in Bash

Regular expressions (regex) allow powerful pattern matching in Bash. They’re used with tools like grep, sed, awk, and even in Bash [[ ]] conditionals.

✅ What Is a Regular Expression?

A regular expression is a pattern that describes a set of strings. You can use it to:

Match specific characters
Search or replace text
Validate inputs

🧪 Basic Regex Symbols

Symbol	Meaning	Example
`.`	Any single character	`c.t` → cat, cut
`*`	Zero or more of previous char	`lo*` → lo, loo, looo
`+`	One or more	`lo+` → loo, looo (not just “l”)
`?`	Zero or one	`lo?` → l, lo
`^`	Start of line	`^Hello`
`$`	End of line	`done$`
`[]`	Character class	`[aeiou]`
`[^]`	Negated class	`[^0-9]`
`\`	Escape special characters	`\.` matches literal `.`

🔎 Using Regex with `grep`

Basic:

grep "error" logfil

Extended (`-E`):

grep -E "^User: [A-Z][a-z]+$" users.txt

Case-insensitive:

grep -i "warning" app.log

Match any digit:

grep "[0-9]" input.txt

🧰 Regex with `[[ =~ ]]` in Bash

Bash supports regex matching inside [[ ]]:

text="user_123"
if [[ $text =~ ^user_[0-9]+$ ]]; then
  echo "Valid user ID"
fi

⚠️ Note:

No quotes around regex inside [[ ... =~ REGEX ]]
Use parentheses for groups like ([a-z]+) if needed

🔁 Regex with `sed`

Replace all numbers with `#`:

sed 's/[0-9]/#/g' file.txt

Replace multiple spaces with one:

sed -E 's/ +/ /g' text.txt

📊 Regex with `awk`

Match lines starting with “user”:

awk '/^user/ { print $0 }' users.txt

Use regex in condition:

awk '$0 ~ /error/ { print $1 }' log.txt

🧠 Advanced Patterns

Pattern	Matches
`[A-Za-z0-9_]{3,8}`	Word with 3–8 alphanumerics
`^[a-z]+\@[a-z]+\.[a-z]+$`	Basic email format
`[0-9]{3}-[0-9]{2}-[0-9]{4}`	SSN pattern (e.g. 123-45-6789)

🔒 Best Practices

Use grep -E for extended syntax
Always test with test strings before applying to files
Use [[ ]] instead of [ when doing regex in Bash

📚 References

📁 File: advance/regex-patterns.md
Author: Kashif Alam

This site is open source. Improve this page.