W3C home > Mailing lists > Public > uri@w3.org > May 2005

RE: Question regarding RFC 3986, Section 3.2

From: Robert Buck <Robert.Buck@mathworks.com>
Date: Fri, 27 May 2005 13:44:01 -0400
Message-ID: <EFEB6396440FB440ADD8A753B5530B9C0601033B@MESSAGE-AH.ad.mathworks.com>
To: "uri" <uri@w3.org>


I have a question somewhat related to the prior question.

Is the operation of parsing any uri into its five major components, then
recomposing them according to section 5.3, NOT guaranteed to produce a
result identical to the input?

scheme:// -> {scheme,undefined,undefined,undefined,undefined} -> scheme:

Is this true?

As an aside...

Interestingly enough, Sun documents their source stating that only a
non-empty path may follow an empty authority, clearly made to get file:
uris working on Windows. But this would appear to be a mistake on their


>-----Original Message-----
>From: Bruce Lilly [mailto:blilly@erols.com] 
>Sent: Friday, May 27, 2005 1:05 PM
>To: uri@w3.org
>Cc: Robert Buck
>Subject: Re: Question regarding RFC 3986, Section 3.2
>On Fri May 27 2005 11:59, Robert Buck wrote:
>> Is the following a valid URI:
>>   "scheme://"
>> Ommiting the quotes of course.
>> It would seem from the grammar that everything following '//' could 
>> possibly be empty.
>Syntactically, yes.  An example would be about:// which is 
>supported by some browsers, although "about"
>is not currently a *registered* scheme.  And even the "//" is 
>not necessary with all schemes (as with the example above; see 
>"path-empty" in the syntax).
Received on Friday, 27 May 2005 17:43:44 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:25:09 UTC