- From: Wilfred Springer <wilfredspringer@gmail.com>
- Date: Sun, 06 Apr 2008 15:13:12 -0400
- To: "Ben Ramsey" <benramsey.lists@gmail.com>
- Cc: URI <uri@w3.org>
- Message-ID: <4c75af00804030830i75727fa4h24a8f682064db1a0@mail.gmail.com>
Hi Ben, (Even though I have been subscribed to this list for ages, I haven't kept track of all of the discussions, so I can't help you directly.) I also started to wonder about this today. The grammar suggests that you can do something like this with all operators: {operator|arg|var=val,var=val,var=val} .... with val being the default value. Now the spec doesn't include any examples doing this. I started to wonder what it would be like for the list operator: {-list|&|nodes=...} The spec is inconclusive on how the default value should be represented. But since the list operator only accepts list type of variables, it must be a value. I would say, we should either drop default values from the variable section of the expansion expression, or introduce a way to represent list values. Wilfred Springer 2008/3/10, Ben Ramsey <benramsey.lists@gmail.com>: > > > I've combed through the list and can't find whether this has been > previously discussed, so forgive me if I'm repeating something. > > I know that you can use a URI template to define a default value for a > parameter (i.e. {foo=bar}), but has anyone discussed the use of a list of > acceptable values for a parameter? > > Perhaps something like: > > {foo=[bar,baz,qux]} > > In this case, the only acceptable values for foo are bar, baz, and qux. > Implementors would determine how to handle unacceptable values. > > If this has been discussed and decided against, what was the reasoning for > rejecting this idea? > > -- > Ben Ramsey > http://benramsey.com/ > > >
Received on Sunday, 6 April 2008 19:13:43 UTC