Shell by Example: Regular Expressions

Shell by Example: Regular Expressions POSIX

	`#!/bin/sh`
Regular expressions are powerful patterns for matching text. Shell uses regex through commands like grep, sed, and awk.
Create sample data for demonstrations:	`cat >/tmp/sample.txt <<'EOF' hello world Hello World HELLO WORLD hello123 hello-world hello_world hello.world test@example.com user123@domain.org 192.168.1.1 10.0.0.255 2024-03-15 03/15/2024 phone: 555-1234 phone: (555) 123-4567 EOF`
Basic grep - find lines matching a pattern:	`echo "Lines containing 'hello':" grep "hello" /tmp/sample.txt`
Case-insensitive search with -i:	`echo "Case-insensitive 'hello':" grep -i "hello" /tmp/sample.txt`
Basic regex metacharacters:
. matches any single character	`echo "Pattern 'h.llo':" grep "h.llo" /tmp/sample.txt`
matches zero or more of previous character	`echo "Pattern 'helo' (zero or more 'l'):" grep "helo" /tmp/sample.txt`
^ matches start of line	`echo "Lines starting with 'hello':" grep "^hello" /tmp/sample.txt`
$ matches end of line	`echo "Lines ending with 'world':" grep "world$" /tmp/sample.txt`
Character classes with []:	`echo "Pattern 'hello[0-9]' (hello + digit):" grep "hello[0-9]" /tmp/sample.txt echo "Pattern 'hello[._-]' (hello + separator):" grep "hello[._-]" /tmp/sample.txt`
Negated character class with [^]:	`echo "Lines with 'hello' NOT followed by digit:" grep "hello[^0-9]" /tmp/sample.txt`
Extended regex with -E (or egrep):	`echo "Extended regex patterns:" # + matches one or more echo "One or more digits:" grep -E "[0-9]+" /tmp/sample.txt # ? matches zero or one echo "Optional 's' (phones?):" grep -E "phones?" /tmp/sample.txt # {n,m} matches n to m occurrences echo "Exactly 3 digits:" grep -E "[0-9]{3}" /tmp/sample.txt # \| for alternation echo "Lines with 'hello' or 'test':" grep -E "hello\|test" /tmp/sample.txt`
Grouping with ():	`echo "Grouped pattern ((hello\|test)@):" grep -E "(hello\|test)@" /tmp/sample.txt`
Word boundaries:	`echo "Word 'hello' (not part of larger word):" grep -w "hello" /tmp/sample.txt`
grep -o shows only matching part:	`echo "Extract email addresses:" grep -oE "[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,}" /tmp/sample.txt`
grep -v inverts match (show non-matching):	`echo "Lines NOT containing 'hello':" grep -v "hello" /tmp/sample.txt \| head -5`
grep -c counts matches:	`echo "Count of lines with 'hello': $(grep -c "hello" /tmp/sample.txt)"`
grep -n shows line numbers:	`echo "With line numbers:" grep -n "world" /tmp/sample.txt \| head -3`
Using sed for search and replace:	`echo "sed substitution examples:" # Replace first occurrence echo "hello world world" \| sed 's/world/universe/' # Replace all occurrences (g flag) echo "hello world world" \| sed 's/world/universe/g' # Case-insensitive (I flag, GNU sed) echo "Hello HELLO hello" \| sed 's/hello/hi/gi' # Delete lines matching pattern echo "Delete lines with 'HELLO':" sed '/HELLO/d' /tmp/sample.txt \| head -3 # Print only matching lines (like grep) echo "sed -n with /p:" sed -n '/^hello/p' /tmp/sample.txt`
sed with capture groups:	`echo "Capture groups (swap words):" echo "foo bar" \| sed 's/$foo$ $bar$/\2 \1/' # Extended regex with -E echo "Extract domain from email:" echo "user@example.com" \| sed -E 's/.*@(.+)/\1/'`
Using awk for regex:	`echo "awk regex examples:" # Match lines echo "Lines matching /hello/:" awk '/hello/' /tmp/sample.txt \| head -3 # Negate match echo "Lines NOT matching /hello/:" awk '!/hello/' /tmp/sample.txt \| head -3 # Match specific field echo "Field matching:" echo "alice 30" \| awk '$1 ~ /^a/'`
Common regex patterns:	`echo "Common patterns:" # IP address (simplified) echo "IP addresses:" grep -E "^[0-9]+\.[0-9]+\.[0-9]+\.[0-9]+$" /tmp/sample.txt # Date YYYY-MM-DD echo "ISO dates:" grep -E "^[0-9]{4}-[0-9]{2}-[0-9]{2}$" /tmp/sample.txt # Phone numbers echo "Phone numbers:" grep -E "[0-9]{3}.*[0-9]{4}" /tmp/sample.txt`
Bash Bash’s =~ operator for regex matching:	`if [[ "hello123" =~ ^hello[0-9]+$ ]]; then echo "Matches!" echo "Full match: ${BASH_REMATCH[0]}" fi`
Capture groups:	`if [[ "user@example.com" =~ ^(.+)@(.+)$ ]]; then echo "User: ${BASH_REMATCH[1]}" echo "Domain: ${BASH_REMATCH[2]}" fi`
Validate input with regex:	`validate_email() { echo "$1" \| grep -qE "^[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,}$" } if validate_email "test@example.com"; then echo "Valid email" fi if ! validate_email "invalid-email"; then echo "Invalid email" fi`
Cleanup	`rm /tmp/sample.txt echo "Regular expressions examples complete"`

« Printf | Line Filters »