URIs validated

[m-alreeni]

No description provided.

[kenwenzel]

Thank you for this PR.
Does the regular expression cover all possible URIs or does it somehow restrict the range of valid URIs?

[m-alreeni]

@kenwenzel it covers the RFC 3986 path characters (absolute and relative)

[kenwenzel]

@kenwenzel it covers the RFC 3986 path characters (absolute and relative)

OK. Where is this regex defined? Is it contained in the RFC somewhere?

[m-alreeni]

OK. Where is this regex defined? Is it contained in the RFC somewhere?

it is just a simplification of the ABNF rules in the RFC , turned into something i can drop into Java/Scala quickly.
It matches: Absolute URIs with a scheme (scheme:[^\s]+) and Relative paths with RFC 3986–allowed characters ([a-zA-Z0-9._~!$&'()*+,;=:@/-]+)
i tested it in online tool and wrote a test around your example for JSON and a simple csv file with space in header.
Should i make it more strict or tolerant or cover addtional cases?

[kenwenzel]

I just want to make sure it is tolerant enough to capture all possible URIs.
Does it also accept URIs like:
this%20is%relative or http://example.org/contains%20space?

[m-alreeni]

I just want to make sure it is tolerant enough to capture all possible URIs. Does it also accept URIs like: this%20is%relative or http://example.org/contains%20space?

no this case is not coverd. Should i update it to cover percent-encoding (%20, %2F, etc.)?

[kenwenzel]

Yes - but we need to make sure that ALL valid URIs are covered. Maybe we find some other regex that is more permissive?

[m-alreeni]

ered. Maybe we find some other regex that is more permissive?

OK i'll search first if something like this exists (better than write or update one by myself).
For Example https://mathiasbynens.be/demo/url-regex