web2py creates a auto-incrementing primary key ID field on all tables
so it will work with any supported database. There is an ondelete
attribute on Field that defaults to CASCADE. I think you can change it
to SET NULL to get the desired behavior. It's applicable only for
reference and upload fields.
On Jun 7, 8:43 pm, Jason Lotz <jayl...@gmail.com> wrote:
> Thanks mr.freeze!! That got me in the right direction.
>
> The dept_id comes from long time with Oracle and M$ SQLServer, also I
> thought it would be nice to make the database structure easily
> compatible with other platforms. I don't know about PostgreSQL but from
> what I understand, MySQL doesn't automatically implement row_id like
> SQLite? Anyways I will just structure for SQLite and worry about other
> platforms later.
>
> Actually another quick question, can one implement triggers/procedures
> to SQLite/MySQL from web2py? I notice that web2py appadmin runs some
> procedure to delete parent records. Following my example below, using
> the appadmin, if I delete create dept_name = 'Management' & create an
> employee = 'Bob Johnson, Management'. Then delete 'Management' from
> db.department it will also delete Bob Johnson from employee (and other
> record with relation to 'Management'). It would be nice to implement a
> procedure that instead of deleting record from employee it would update
> employee.dept_id = None. Is such possible?
>
> On 06/08/2010 12:38 AM, mr.freeze wrote:
>
> > I don't think your model is not setup with the correct reference
> > fields. Try this:
>
> > db.define_table('department',
> > Field('dept_id'),
> > Field('dept_name'),format='%(name)s')
> > db.define_table('employee',
> > Field('firstName'),
> > Field('lastName'),
> > Field('dept_id', db.department))
>
> > Passing the format parameter to the department table will create the
> > default validators and represent attributes so that it shows the name
> > instead of the id.
>
> > The dept_id field on the department table seems redundant as the table
> > will already have an ID field automatically.
>
> > On Jun 7, 9:24 am, Jason Lotz<jayl...@gmail.com> wrote:
>
> >> Creating a custom crud.read form and I would like to show the
> >> field.requires=IS_IN_DB field rather than the table.id . Sorry if that
> >> doesn't make much sense. I have two related tables.
>
> >> db.define_table('employee', Field('firstName'), Field('lastName'),
> >> Field('dept_id'))
> >> db.define_table('department', Field('dept_id'), Field('dept_name'))
> >> db.employee.dept_id.requires=IS_NULL_OR(IS_IN_DB(db,'department.dept_id','department.dept_name'))
>
> >> I would like to make a custom form to show the employee name and
> >> department.
>
> >> def employee_form():
> >> form = crud.read(db.employee, request.args[0])
> >> return dict(form=form)
>
> >> ## View
> >> {{=form.custom.begin}}
> >> First Name: {{=form.custom.widget.firstName}}
> >> Last Name: {{=form.custom.widget.lastName}}
> >> Department: {{=form.custom.widget.dept_id}}
> >> {{=form.custom.end}}
>
> >> This returns the actual dept_id for the Department, but I would like
> >> to return the dept_name. Is there a better method to accomplish this
> >> or can some explain custom widgets. I have read the book for custom
> >> widgets but didn't completely understand it.
>
>
