Message-Id: <9406290046.AA10012@nada.kth.se>
To: "Daniel W. Connolly" <connolly@hal.com>
Subject: Re: Treatment of trailing slash after directory name
In-Reply-To: <9406282217.AA03820@ulua.hal.com>
from ""Daniel W. Connolly" <connolly@hal.com> "
Date: Wed, 29 Jun 1994 02:46:56 +0200
From: Peter Svanberg <psv@nada.kth.se>
Quoting: "Daniel W. Connolly" <connolly@hal.com>
>
> Since you're talking about relative URI's, I assume you're using the
> WWW world-view.
OK, I guess I am... ;-)
> In that case, my interpretation of the spec agrees with your
> intuition and the HTParse() code in the WWWLibrary:
Does it? The crucial problem was the case when "c" is a
directory ("non-terminal path element"?). The test (below)
doesn't agree with my intuition in that case.
> connolly@ulua {** NONE **}../url_test[620] test_htparse <foo.test
>
> ***********
> here: magic://a/b/c/
> there: d
> scheme: magic:
> hostport: //a
> path: /b/c/d
> id:
>
> ***********
> here: magic://a/b/c
> there: d
> scheme: magic:
> hostport: //a
> path: /b/d
> id:
--- Peter Svanberg, NADA, KTH, <psv@nada.kth.se>