on_define and IS_IN_DB()

[Annet]

I am not sure I understand the following paragraph in the book:

http://web2py.com/books/default/chapter/29/06/the-database-abstraction-layer#on_define

I have lazy_table = true and a

Field('vertex_id', 'reference vertex', requires=IS_IN_DB(db, 'vtx_vertex.id', '%(id)s'))

In the book it says:

The simple requires values could be added to the Field definitions and the table would still be lazy.

However, requires which take a Set object as the first argument, such as IS_IN_DB, will make a query like

db.sometable.somefield == some_value

which would cause sometable to be defined early. This is the situation saved by on_define.

Does this mean that I should put IS_IN_DB() in the on_define keyword argument?

on_define = lambda table: [table.vertex_id.set_attributes(requires=IS_IN_DB(db, 'vtx_vertex.id', '%(id)s')]

Kind regards,

Annet