FRIGN writes:
> That's definitely a good point. However, fortifying the regexes to
> strictly match URIs solves this problem instantly (Hell, just check for
> spaces!).
That also doesn't really work, as a basic example, "&" is a perfectly
valid character in a URI without encoding, but it has other meaning to
most shells (it is a backgrounding operator).
I just think there are too many potential pitfalls.
- application/pgp-signature attachment: stored
Received on Sun May 04 2014 - 13:23:11 CEST