A better URL::isWellFormed()

General JUCE discussion
1

URL::isWellFormed() only checks if the url isn’t empty at the moment.

could it at least checks also that the url doesn’t contain spaces or other forbidden characters?

2

any chance this gets improved at some point?
It’s been 18years (no joke) that the implementation is marked ‘todo’ :slight_smile:

bool URL::isWellFormed() const
{
    //xxx TODO
    return url.isNotEmpty();
}
3

I had a sneaking suspicion that URLs are as rule-less as email addresses, but it looks like there are at elast some standards that could be followed:

Whether or not that’s suitable to use in a cross-platform framework, idk - I wouldn’t be surprised if different platforms/browsers/etc. have different rules. E.g. you say a URL shouldn’t have spaces, but URLs can point to files which on most OSs can have spaces in the name.

This regex seems to do a good job of validating website URLs: RegExr: Learn, Build, & Test RegEx - maybe a more generic one is needed for file URLs?

4

They’re not, URIs are very well defined, and most modern languages have standards-compliant ways of dealing with them.

Operating systems/browsers/etc respect the rules for parsing them. It’s a trivial problem to solve. Representing a resource as a URI is slightly non trivial.

E.g. you say a URL shouldn’t have spaces, but URLs can point to files which on most OSs can have spaces in the name.

non-URI safe characters must be escaped in URIs.