Regex Builder

Continue in the tool

More pages for this tool

url regex

URL regex

A URL has a scheme, host, optional port, path, query and fragment. This preset detects HTTP and HTTPS inside free text, but its restricted character class cuts the match short as soon as it hits a port or fragment, two common pieces of a real URL.

The six parts of a URL and where this pattern stops

A full URL has a scheme (`https://`), a host (`example.com`), an optional port (`:8080`), a path (`/path`), a query (`?ok=1`) and a fragment (`#section`). The pattern `https?:\/\/[\w.-]+(?:\/[\w./?%&=-]*)?` covers scheme, host, path and query, because the path character class includes `?`, `%`, `&` and `=`, but it covers neither port nor fragment, because `:` and `#` never appear in either character class used here.
The character class is restricted on purpose, so the pattern does not turn greedy and swallow unrelated text after the URL. The cost of that caution is that the match stops at the first character outside the class: `https://example.com:8080/path` only matches up to `https://example.com`, because the port `:` ends the host segment before the path is ever read.
To reliably pull apart a URL, the JavaScript `new URL()` constructor is the right tool, not regex: it exposes `protocol`, `hostname`, `port`, `pathname`, `search` and `hash` already separated, understands port and fragment natively, and throws an error instead of returning a silently wrong fragment when the URL is invalid. Regex is for spotting a loose URL inside text; `new URL()` is for taking apart a URL you have already isolated.

Three URLs, three different results

URL with a path

Input: https://jkit.tools/pt/ferramentas/regex-builder
Expected output: match completo: https://jkit.tools/pt/ferramentas/regex-builder

With no port or fragment, every host and path character sits inside the accepted classes, so the whole URL ends up in the result.

URL with port and query

Input: https://example.com:8080/path?ok=1
Expected output: match parcial: https://example.com

The `:8080` cuts the match right after the host, because `:` is not in the `[\w.-]` class; the path and query, even though valid, are never reached.

URL with a fragment

Input: https://example.com/path?ok=1#section
Expected output: match parcial: https://example.com/path?ok=1

Here the query does make it into the result, but `#section` is left out because `#` is not in the path class, only `?`, `%`, `&` and `=` are.

Full tool FAQ

Does the tool test regex patterns with flags and examples?

Yes. The idea is to let you build the expression, apply flags and inspect matches in real time, which is especially useful for debugging, learning and fine-tuning patterns.

What is regex?

What is regex used for in practice?

What are regex flags?

Can regex validate any complex format?

Why do some regex patterns become too slow?

Is there such a thing as a perfect email regex?

What is the most common mistake when learning regex?

Frequently asked questions

Does this regex recognize a port, like :8080, in the URL?

No. The host class `[\w.-]+` does not include `:`, so as soon as the pattern hits the port colon, it stops capturing right there. `https://example.com:8080/path` only matches as `https://example.com`, with no port and no path.

Does the regex capture the fragment (the anchor after #) of the URL?

No. The path class accepts `?`, `%`, `&` and `=`, but not `#`, so everything after the fragment symbol is left out of the result, even when the query and path before it match normally.

Why use new URL() instead of this regex to process a URL?

`new URL()` splits the string into `protocol`, `hostname`, `port`, `pathname`, `search` and `hash`, already ready to use, understands port and fragment natively, and throws on an invalid URL. This regex is for spotting a loose URL inside text, not for safely taking it apart.

Next steps

Continue in the tool

Open full tool

url regex

URL regex

The six parts of a URL and where this pattern stops

A full URL has a scheme (`https://`), a host (`example.com`), an optional port (`:8080`), a path (`/path`), a query (`?ok=1`) and a fragment (`#section`). The pattern `https?:\/\/[\w.-]+(?:\/[\w./?%&=-]*)?` covers scheme, host, path and query, because the path character class includes `?`, `%`, `&` and `=`, but it covers neither port nor fragment, because `:` and `#` never appear in either character class used here.

The character class is restricted on purpose, so the pattern does not turn greedy and swallow unrelated text after the URL. The cost of that caution is that the match stops at the first character outside the class: `https://example.com:8080/path` only matches up to `https://example.com`, because the port `:` ends the host segment before the path is ever read.

To reliably pull apart a URL, the JavaScript `new URL()` constructor is the right tool, not regex: it exposes `protocol`, `hostname`, `port`, `pathname`, `search` and `hash` already separated, understands port and fragment natively, and throws an error instead of returning a silently wrong fragment when the URL is invalid. Regex is for spotting a loose URL inside text; `new URL()` is for taking apart a URL you have already isolated.

Three URLs, three different results

URL with a path

Input: https://jkit.tools/pt/ferramentas/regex-builder
Expected output: match completo: https://jkit.tools/pt/ferramentas/regex-builder

With no port or fragment, every host and path character sits inside the accepted classes, so the whole URL ends up in the result.

URL with port and query

Input: https://example.com:8080/path?ok=1
Expected output: match parcial: https://example.com

The `:8080` cuts the match right after the host, because `:` is not in the `[\w.-]` class; the path and query, even though valid, are never reached.

URL with a fragment

Input: https://example.com/path?ok=1#section
Expected output: match parcial: https://example.com/path?ok=1

Here the query does make it into the result, but `#section` is left out because `#` is not in the path class, only `?`, `%`, `&` and `=` are.

Full tool FAQ

Does the tool test regex patterns with flags and examples?

Yes. The idea is to let you build the expression, apply flags and inspect matches in real time, which is especially useful for debugging, learning and fine-tuning patterns.

What is regex?

What is regex used for in practice?

What are regex flags?

Can regex validate any complex format?

Why do some regex patterns become too slow?

Is there such a thing as a perfect email regex?

What is the most common mistake when learning regex?

Frequently asked questions

Does this regex recognize a port, like :8080, in the URL?

Does the regex capture the fragment (the anchor after #) of the URL?

No. The path class accepts `?`, `%`, `&` and `=`, but not `#`, so everything after the fragment symbol is left out of the result, even when the query and path before it match normally.