Regex Find and Replace

Test a regex replacement before running it on production data. The replacement field supports $1, $2 for capture-group references and $& for the full match.

Replacement syntax

Examples

Anonymize email domains

Pattern:     (\b\w+@)\w+\.\w+
Replacement: $1example.com
Input:       Email alice@gmail.com or bob@yahoo.co.uk
Output:      Email alice@example.com or bob@example.com

Reformat dates (US → ISO)

Pattern:     (\d{1,2})/(\d{1,2})/(\d{4})
Replacement: $3-$1-$2
Input:       Meeting on 4/26/2026 and follow-up 5/1/2026.
Output:      Meeting on 2026-4-26 and follow-up 2026-5-1.

Wrap log levels

Pattern:     \b(ERROR|WARN|INFO)\b
Replacement: [$1]
Input:       INFO request started; ERROR upstream timeout
Output:      [INFO] request started; [ERROR] upstream timeout

Surround code with backticks (markdown)

Pattern:     \b(uuid|json|base64|regex|jwt)\b
Replacement: `$&`
Input:       The uuid is a json field, base64-encoded.
Output:      The `uuid` is a `json` field, `base64`-encoded.

Programmatic equivalents

// JavaScript
text.replace(/(\b\w+@)\w+\.\w+/g, "$1example.com");

// Python
import re
re.sub(r"(\b\w+@)\w+\.\w+", r"\1example.com", text)

// sed (POSIX)
sed -E 's/(\b[a-z]+@)[a-z]+\.[a-z]+/\1example.com/g'

// Go
re := regexp.MustCompile("(\\b\\w+@)\\w+\\.\\w+")
re.ReplaceAllString(text, "${1}example.com")

Common pitfalls

• Forgetting the g flag. Without it, only the first match is replaced. The flag-toggle checkboxes above show the flags clearly.
• Special characters in the replacement. Literal $ requires $$. A literal \\ (in JavaScript source) requires \\\\ — but the UI shows you the rendered replacement directly, so usually no escaping needed.
• Capture-group numbering. Groups are numbered by their opening parenthesis order, including non-capturing wrappers. Use named groups (?<name>...) for clarity.
• Lookbehind / lookahead syntax differs. JavaScript supports (?=...), (?!...), (?<=...), (?<!...) — but very old browsers may not. Use modern targets in production.