Bug#812860: /usr/bin/uscan: [uscan] failure to download and verify package.tar.xz with package.sign

Osamu Aoki osamu at debian.org
Mon Feb 1 12:09:07 UTC 2016


Hi,

Thanks for checking.
On Sun, Jan 31, 2016 at 09:14:57PM +0100, Uwe Kleine-König wrote:
...
> I tested your change, and it works fine here now. \o/
> 
> > Script started on Sun 31 Jan 2016 05:23:24 PM JST
> > [...]
> > uscan info: Matching pattern:
> >    (?:(?:http://www.kernel.org)?\/pub\/linux\/utils\/rt\-tests\/)?rt-tests-(.*)\.tar\.xz (?:(?:https://www.kernel.org)?\/pub\/linux\/utils\/rt\-tests\/)?rt-tests-(.*)\.tar\.xz
> 
> Using (?:.....)? at the start of a regexp doesn't give any advantage,

Why.  href in a web page may be written without "http://www.kernel.org"
We certainly want to match such href string.

> and probably can be dropped without any loss. If you want to keep it:
> Why is the - quoted? You might want to make "http:" optional, because
> otherwise
> 
> 	//www.kernel.org/pub/linux/utils/rt-tests/...
> 
> isn't matched.

Do you want to match such href?  I do not understand your point here.

(Quite frankly this part of logic is inherited.  Unless someone explain
me and take responsibility of such changes, I am reluctant to change
logic around this.  base page magic etc. is a bit too complicated for
me.)

Osamu



More information about the devscripts-devel mailing list