[Zope-dev] Re: Zope and Storm (SQL)

[Martijn Faassen]

Hi there,

David Pratt wrote:
[snip]
> For me, the crux of the rdb approach for legacy code is the container, 
> location and traversal. You have been very generous with your examples. 
> I am really hoping for a clearer idea of handling Container, 
> OrderedContainer, and Location which is prevalent in legacy code. 
> Overall, I can say that I quite the innovation here in getting to a 
> 'leaner' concept of Zope.

I myself wouldn't be inclined to call Container and Location only 
relevant to legacy code.

It's very convenient to have a RDB-backed object have a location, as 
that allows you to get the URL for it in the regular manner.

For Container, in the megrok.rdb prototype code we actually make 
containers be SQLAlchemy 'MappedCollection' objects. MappedCollection 
happens to implement the container API already. This way we can make 
relations be containers by writing something like:

class Departments(rdb.Container):
     # this is a MappedCollection
     rdb.key('title')

class Faculty(rdb.Model):
     # rdb.table_name('faculty') is the default
     __tablename__ = 'faculty'

     grok.traversable('departments')

     departments = relation('Department',
                            backref='faculty',
                            collection_class=Departments)


The 'departments' attribute is a container you can traverse into this 
way. You can hook up views to Departments just like you would to any 
container.

Regards,

Martijn