How to match URLs with a colon in their name

[vlukashov]

Colon (:) is a special character for path-to-regexp, so it's unclear how to write a route spec that allows using colons literally. For example, Wikipedia (and other media-wiki sites) uses colons in category page URLs: https://en.wikipedia.org/wiki/Category:JavaScript_libraries

How would I write a route spec for such URLs with path-to-regexp?

• My first assumption was that a colon can be escaped with a \, like in a regexp: /wiki/Category\::category. That did not work with 6.2.0.
• My second assumption was that a colon can be repeated for its literal meaning, like % in URL-encoding: /wiki/Category:::category. That did not work with 6.2.0 either.

By now I assume that this class of URLs cannot be handled with path-to-regexp, and I'd be happy to learn that I'm wrong.

---

NOTE: A colon (:) might legitimately be used in the path part of a URL according to the RFC 3986. It is actually used in the wild (Wikipedia).

"Common" characters that are NOT allowed in URL pathnames (according to my reading of the RFC) are ", #, <, >, ?, [, ], \, ^, {, }, ​`​. It would be great if path-to-regexp allowed to optionally change the character it uses in the path parameters syntax from the current : to one from the above list.

Am I overthinking this?

[dougwilson]

You can use backslash to escape the colon. The backslash needs to be in the string, ans since Javascript itself treats backslash as a special characher, to get a backslash in a string requires escaping it, like so: '\\'. This means a string of your path, witten as a Javascript literal, would need to be '/wiki/Category\\::category'

[vlukashov]

Indeed! Thanks for pointing out my mistake with a \: vs. \\:!