On 1/13/06, Max Ischenko <[EMAIL PROTECTED]> wrote:
>
> Here is how what I ended up with, in case of anyone interested:
>
>
> @turbogears.expose(content_type="text/html; charset=UTF-8",
> html="bookswap.templates.addbook")
> def addbook(self, **kw):
> ### XXX: why this is needed?
> for k,v in kw.items():
> if isinstance(v, str): kw[k] = unicode(v, 'utf-8')
> if kw: # feed data if any
> addBookForm.input(kw)
> cherrypy.request.input_values = kw # to display form with
> data
> if cherrypy.request.method == 'POST':
> errors = cherrypy.request.form_errors
> cherrypy.request.form_processed = True
> if errors:
> turbogears.flash(_('please correct errors below'))
> else:
> log.info("Adding book from %s", kw)
> turbogears.flash(_('book has been added'))
> turbogears.redirect('/addbook')
> return dict(form=addBookForm)
>
> Note the absense of inputform parameter in expose() and the need to set
> cherrypy.request.input_values in order the form to be displayed filled
> with user data.
Yeah, that's pretty ugly. You don't *have* to put the data in
input_values. You can alternatively pass a dictionary (or instance) to
the form from your template.
> A weird thing: yesterdays unicode data were handled OK by this
> particular method but after today's SVN update it broke and I had to
> add this hack in order to make it work again:
Hmm... We did have the UnicodeDecodeFilter included. I wonder if that
is not included any longer for some reason.
Kevin
