On 8/12/05, Romero, Ron <[EMAIL PROTECTED]> wrote:
> 
> On August 12, 2005, Craig McClanahan [mailto:[EMAIL PROTECTED] wrote:
> 
> > On 8/12/05, Romero, Ron <[EMAIL PROTECTED]> wrote:
> >
> > > I'm working with the Shale Validator.  I would like a new, simpler
> > > "regexValidator" tag that better supports my situation (and
> hopefully
> > > supports other people's situation).
> >
> > I'm definitely +1 on having support for regex validation ... but
> > doesn't the "mask" validator already do that for you?  If not, it
> > would seem the right answer would be adding a new Commons Validator
> > implementation so that all the frameworks using C-V would benefit.
> >
> 
> Actually, what I'm going for is more of a "repackaging" of the mask
> validator, a convenience tag.  The tag would resolve into a
> commonsValidator, but it would look up the mask and arg first.  You
> get the same functionality as the mask validator, but you can
> administer it from a single config file.  It's a cleaner separation of
> concern.  The JSF developer concentrates on what is being input, a
> name, a phone number, a social security number.  And the regex
> developer (who rights the regex config file) concentrates on what
> those types look like (ten digit phone number codes, no punctuation
> Characters in names, etc.)
> 
> This wouldn't go into C-V because it's not another type of validation,
> it's about making a cleaner interface for faces developers.
> 
> > This would also mean you don't need a new tag ... the existing Shale
> > tags for integrating validators would work.
> 
> Ah, but I want a new tag! :-)  Seriously, a new tag would make it
> easier to use --- it's simpler to code and to understand.  True, it
> would hide the flexibility and power of the commonsValidator tag, but
> we don't need or want that additional flexibility, and we'd like to
> hide it from everyone else.
> 

And here's where the suggested approach isn't as appealing to me. 
Once you open the floodgates of creating a new tag for *this* sort of
a validator, you likely end up where you need a tag for *every* sort
of validator.  To the maximum degree possible, I'd prefer to stick
with just one tag <s:commonsValidator> that can call any of them.  We
only need to ensure that there is a way to set all the needed
configurable properties such a validator would need.

At the same time, having predefined patterns with easy to use names is
exactly the sort of enhancement that *should* be made at the C-V
level, so that everyone can benefit from it, rather than hiding it
inside Shale :-).

Craig

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to