Regex Tester — Match, Groups and Positions in Real Time

The flags that change everything

Most 'my regex works here but not there' bugs are flag differences. Without g, you get only the first match — test() and match() disagree. Without m, ^ and $ anchor only to the string ends, not each line. Without s, the dot stops at newlines — why '.*' fails on multi-line input.

Case-insensitivity (i) is the other trap: it folds more than ASCII in Unicode mode, which is why [a-z] with i can match characters you didn't expect.

Capture groups: what the positions mean

Each match shows its index (0-based character position) and captured groups — the parts in parentheses, in order. Group 0 is always the full match. If a group is optional and didn't participate, it shows as undefined — the classic source of 'group 2 is empty' bugs.

JavaScript regex quirks worth knowing

JS supports lookahead and (in modern engines) lookbehind, named groups (?<name>…), and unicode-aware \p{…} classes. What it lacks vs PCRE: \z anchors (use $ with care), recursive patterns, and atomic groups — a pattern that works in PHP/Python may need adapting here.

Frequently asked questions

Why does my regex only match once?

You need the g flag — without it exec() and match() return just the first match. Select 'g — all matches' above.

What's the difference between match and test?

test() returns true/false; match() returns the strings (or null). With g, match() returns all matches but loses capture groups — use matchAll() or exec() in a loop for groups.

How do I match an email?

For validation that isn't insane: \b\w+@\w+\.\w+\b catches most cases. The 'perfect' RFC email regex is thousands of characters — almost nobody actually needs it.

Why does '.' not match my newline?

The dot never matches \n without the s (dotAll) flag. Use [\s\S] or enable 'gs' mode above.