Right.  And you could easily do additional validation without having
to wrap it up in a FancyValidator or whatever.  I may go this way
myself.

On Jan 7, 3:11 pm, "Brian O'Connor" <[email protected]> wrote:
> I meant to reply to this a little bit sooner.  I had a solution that isn't
> amazingly elegant, but works, and supports javascript addition of arbitrary
> number of elements.
>
> I do not use the @validate decorator.
>
> This was originally designed for a newspaper website to add n number of
> authors to an article via form.
>
> The controller:
>
>     def create(self):
>         params = dict(request.params)
>         params['staff_authors'] = request.params.getall('staff_authors')
>         schema = ArticleForm()
>
>         # ...
> Schema:
>
> class ArticleForm(formencode.Schema):
>     allow_extra_fields = True
>     filter_extra_fields = True
>     # other fields....
>     staff_authors = formencode.ForEach(ExistingStaffMember())
>
> Then, I could just run
>
>         try:
>             self.form_result = schema.to_python(params, c)
>         except formencode.Invalid, error:
>             # handle errors here
>
> I hope that helps.  Like I said, not the most elegant, but worked for me.
>
> On Thu, Jan 7, 2010 at 8:50 AM, Mike Burrows (asplake) 
> <[email protected]>wrote:
>
>
>
>
>
> > Perhaps you're right - maybe @validate is more trouble than it is
> > worth.  I have discovered that it doesn't handle repeating elements
> > properly: it negelects to flatten the errors dict.  I worked around
> > this issue in a fill_render() helper which I use to render my forms:
>
> > def fill_render(template_name, values):
> >    if isinstance(c.form_errors, dict):
> >        # UGH! Modify c.form_errors in place, relying on the fact that
> > it is
> >        # aliased to the errors variable in
> > pylons.decorators.validate.
> >        # The validate decorator neglects to flatten any repeating
> > groups.
> >        errors = dict((k, v)
> >                      for k, v in variabledecode.variable_encode
> > (c.form_errors).items()
> >                      if not k.endswith('--repetitions'))
> >        c.form_errors.clear()
> >        c.form_errors.update(errors)
> >    return htmlfill.render(render(template_name),
> >                           variabledecode.variable_encode(values))
>
> > I appreciate that Pylons isn't 1.0 yet but it concerns me a bit that
> > this stuff doesn't work out of the box; makes one wonder that it's not
> > used much.  As per the start of this thread, it would be cool if
> > regular application code didn't need to call
> > formencode.variabledecode.variable_encode() (twice!) - all it would
> > take is a better render() and a fix to @validate.  But is @validate
> > definitely the way forward, as opposed to (say) something that might
> > be called from within a controller action?
>
> > I hate to whine - otherwise I'm happy :-)
>
> > Mike
> > [email protected]
> >http://positiveincline.com
> >http://twitter.com/asplake
>
> > On Dec 24 2009, 7:20 am, Ian Wilson <[email protected]> wrote:
> > > On Wed, Dec 23, 2009 at 8:26 PM, Mike Orr <[email protected]> wrote:
> > > > On Wed, Dec 23, 2009 at 3:58 AM, Mike Burrows (asplake)
> >  > > <[email protected]> wrote:
> > > >> I'm far too new to Pylons to feel confident enough to contribute
> > > >> documentation but I have linked to a few of my blog posts here.  I
> > > >> don't see too many others doing the same though, so I have to wonder
> > > >> about etiquette. Is this to be encouraged?
>
> > > > It's fine etiquette-wise, as long as it's a few important posts and
> > > > not every single one.  But from the perspective of somebody looking
> > > > for reference material in the future, they'd find it easier if it's
> > > > linked in a topic page in the Pylons Cookbook:  The developers
> > > > periodically go through the Cookbook and put the best pieces into the
> > > > official docs, although there hasn't been a sweep recently.
>
> > > >http://wiki.pylonshq.com/display/pylonscookbook/Home
>
> > > > For especially short pieces, the Pylons FAQ is a good place.
> > > >http://wiki.pylonshq.com/display/pylonsfaq/Home
>
> > > > There is also the Snippets section on the website, but I've never used
> > > > it and I'm not exactly sure what it's for compared to the other two.
> > > >http://pylonshq.com/snippets
>
> > > >> And who is blogging
> > > >> regularly about Pylons?  The most recent post on Planet Pylons dates
> > > >> back to March and I don't know where else to look.
>
> > > > Ian Bicking (blog.ianbicking.org) and Ben Bangert (groovie.org) have
> > > > blogs where they post about Pylons-related software.  The Pylons
> > > > community as a whole doesn't blog as much as others do, I think
> > > > because we're too busy working.  The developers are focusing on
> > > > finishing Pylons 1.0, and the marketing push has been waiting for
> > > > that.
>
> > > >> On a related topic, how about an occasional post here (perhaps I'm too
> > > >> new here to have seen one) and a prominent link on the pylonshq front
> > > >> page about how to contribute?  Better to confront problems than
> > > >> perpetually working around them, don't you think?
>
> > > > I suppose. I'm not sure what it would say though beyond the usual
> > > > open-source stuff: testers and documenters always welcome. It's kind
> > > > of been, if you want to contribute, be active on the list answering
> > > > questions until you find a task to do, or ask the list or one of the
> > > > developers what needs to be done.
>
> > > > The website has a Contributing menu link but it's broken, hmm.  The
> > > > there's also a Community section on the wiki although it's not the
> > > > easiest to find.
> > > >http://wiki.pylonshq.com/display/pylonscommunity/Home
>
> > > > --
> > > > Mike Orr <[email protected]>
>
> > > > --
>
> > > > You received this message because you are subscribed to the Google
> > Groups "pylons-discuss" group.
> > > > To post to this group, send email to [email protected].
> > > > To unsubscribe from this group, send email to
> > [email protected]<pylons-discuss%[email protected]>
> > .
> > > > For more options, visit this group athttp://
> > groups.google.com/group/pylons-discuss?hl=en.
>
> > > Hey everybody,
> > > I made a library to wrap up formencode and htmlfill usage in the
> > > context of web form submissions that satisfies my use cases:
>
> > >http://bitbucket.org/ianjosephwilson/formprocess/
>
> > > Here is a demo of it with tons of javascript to handle dynamic
> > > repeating elements, in the party controller(its a pretty DRY party),
> > > you can ignore that if you just want to consider static forms which I
> > > demonstrate with the login controller:
>
> > >http://bitbucket.org/ianjosephwilson/demoformprocess/
>
> > > I think some of the repeater stuff might need to go back in the
> > > library and the library might still need some more functionality but
> > > check it out.  It uses formencode and htmlfill.
>
> > > If someone has a better solution to dynamic repeating fields I would
> > > _LOVE_ to hear about it because my solution tastes like spaghetti in
> > > vomit sauce.  Someone has to be doing it out there somewhere. My
> > > solution has always been to relabel inputs and labels so that their
> > > order is strictly maintained.  This allows elements to be removed or
> > > dragged and dropped and still the submission has the correct order.
> > > It seems that in the puff(php) you would just do name="name[]" and
> > > then the server would get the order of the fields in the POST, is that
> > > not reliable?  Do we _really_ need name-0, name-1, name-2?  Was that
> > > just created for something like GET submissions where its an unordered
> > > dictionary?  I know that formencode wasn't meant to be tied to the web
> > > environment directly but maybe something could be done specifically
> > > for ordered dictionaries?
>
> > > -Ian
>
> > --
> > You received this message because you are subscribed to the Google Groups
> > "pylons-discuss" group.
> > To post to this group, send email to [email protected].
> > To unsubscribe from this group, send email to
> > [email protected]<pylons-discuss%[email protected]>
> > .
> > For more options, visit this group at
> >http://groups.google.com/group/pylons-discuss?hl=en.
>
> --
> Brian O'Connor
-- 
You received this message because you are subscribed to the Google Groups 
"pylons-discuss" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/pylons-discuss?hl=en.


Reply via email to