Yes I was looking for the inverse operation. My first inclination was the
fblock 1, lookup solution, but thought I'd troll here to see if there was
some built-in I was unaware of. The change cascade is interesting, although
somewhat less elegant than I'd hoped for.

On Feb 5, 2008 3:59 AM, Rob van der Heij <[EMAIL PROTECTED]> wrote:

> On Feb 5, 2008 4:38 AM, Bruce Hayden <[EMAIL PROTECTED]> wrote:
>
> > There is the URLDECODE stage, which may also found under its old name
> > of URLDEBLOCK.  I find the help for it in the author's help library.
>
> But I think Bob was actually looking for the inverse operation.
> Browsing the RFC gives at least 22 code points that would need to be
> escaped. Alternatively, it offers 11 characters plus letters and
> digits that would not need to be encoded.
> Since the RFC allows for that, would it be an option to escape *all*
> characters in the URL (except those that need to have their special
> meaning)?
>
> I see no options beyond a cascade of a lot of "change" stages. Make
> sure you put the % first and the most likely ones last. And it might
> be attractive to wrap part of the change stages in a "verify" to avoid
> needless passing through all stages.
> Even when you generate a change for all ~ 180 remaining code points,
> it still would be cheaper than a fblock 1 with a lookup stage.
>
> Rob
>

Reply via email to