JavaScript Regex Tester

Test a pattern in three steps

1. Enter the pattern without surrounding slashes or a trailing flags string, then enter your test text. Empty patterns and empty text are allowed.
2. Choose the flags. All matches and Unicode mode start enabled; turn All matches off when you want only the first match.
3. Select Test pattern. Inspect the matches, positions and captures, then optionally copy or download the complete current report, which includes your full input.

Read matches and captures accurately

A full match is the text accepted by the entire pattern. Parentheses can also capture values; a group that did not participate is different from one that matched an empty string.

Start positions count from zero and end positions are exclusive. Positions use JavaScript UTF-16 code units, so an emoji can occupy two positions even with Unicode mode enabled.

Multiline changes line anchors; Dot matches line breaks changes the dot. Unicode mode changes parsing and empty-match advancement, but does not make the usual word shorthand include every language's letters.

Values are shown in quoted JSON-string notation so empty, invisible and surrogate characters remain distinguishable. The exported report uses ASCII escapes and preserves the admitted pattern and text.

Examples worth checking

For pattern (\d+) and text Order 42, item 7, All matches returns 42 at [6,8) and 7 at [15,16), with each number also in capture 1. First match only returns 42.

An empty pattern on one emoji returns empty matches at positions 0 and 2 with All matches and Unicode mode. Without Unicode mode, it also returns one at position 1.

For pattern (a)?() on empty text, the full match is empty. Capture 1 is unmatched, while capture 2 participated and contains an empty string.

Scope and limits

This tool uses your browser's JavaScript regex engine with g, i, m, s and u flags. It does not offer PCRE, Python, replacement, capture positions or a guarantee of identical behavior across engines and versions.

Patterns are limited to 2,048 UTF-16 units and text to 100,000. Complete results allow at most 500 matches, 256 numbered and 256 named captures per returned match, and a 1 MiB report. Exceeding a limit produces no partial report.

A separate worker has a nominal two-second operation budget, including startup and reporting. Browser scheduling can delay cancellation, and a timeout does not prove that a pattern is invalid or unsafe.

This is a text-inspection aid, not a proof that a regex is safe, fast, secure or correct for every input. First match only reports the requested first result, not an exhaustive nonzero total.

Local handling

The tool processes your pattern and text in a browser worker without adding an upload, input history or content-bearing URL. Copy and download happen only when you choose them; the report contains your full pattern and text.