[Zope] ZopeBook - URL to SQL

[Chris Gray]

Dieter,

Thanks, you've added a new room to my brain :)  Possibly that point should
be made in The Book since the discussion of Pluggable Brains comes at the
end of the chapter.

Cheers,
Chris


On Fri, 24 Nov 2000, Dieter Maurer wrote:

> Chris Gray writes:
>  > The Zope book discusses calling a ZSQL Method via a URL, saying that a
>  > request for a URL like:
>  > 
>  >   http://localhost:8080/zsql_method/parameter/value
>  > 
>  > "will return a result object", but what I get is the index_html document
>  > at the root, although I do get a rendered record with:
> The ZPublisher works as follows:
> 
>   It traverses the ZODB guided by the components in your URL.
>   It it arrives at the URL's end, it checks whether the
>   reached object is callable (this is the case for all
>   kinds of methods and documents). If it is, the object
>   is called.
>   If it is not, ZPublisher checks, whether the object
>   has (or can acquire) a method "index_html"
>   (this usually is the case). It then calls this method.
> 
> In your case:
> 
>   ZPublisher will reach the result object (as the Zope book
>   describes). Unless you have wrapped this object
>   into a custom class (aka brain, see "Advanced" tab),
>   this is a "Record" object.
>   A "Record" object is not callable.
>   "index_html" is looked for, found in a folder above and
>   rendered.
> 
>   That's what you see.
> 
> What can you do?
> 
>   Wrap you results into a custom brain that either
>   defines a "__call__" (that would need to be a
>   Python defined class) or an "index_html" (this
>   may be a non-persistent ZClass).
> 
> 
> Dieter
>