Re: [web2py] improving DAL for normalized DB

2016-12-09 Thread Richard Vézina
Note you can do such query : db(db.table.id > 0).select("sql_field_name AS new_sql_field_name") but I don't recall how you access value of a field query like that (plain sql passed into select...) You can experiment... There is a presentation from Anthony Bastardi about these undocumented feature

Re: [web2py] improving DAL for normalized DB

2016-12-09 Thread Val K
Thank you for your reply, Richard! " a normalized way to represent web2py list:reference field with a proper many-to-many relation table " - yes! this is what I meant, 'alias_name' - just very bad example of field name - it isn't about aliasing - sorry for the mess About views I think I found

Re: [web2py] improving DAL for normalized DB

2016-12-08 Thread Richard Vézina
Hello Val K, I am not sure I understand fully what your are talking about. On one hand, I do understand that you would like a proper way to define SQL VIEW in web2py and you propose to add a switch/flag/argument to define_table() to do so, which would result in a select only table or something lik

Re: [web2py] improving DAL for normalized DB

2016-12-08 Thread Val K
It doesn't differ from reference fields. it is about automation distributed transaction. Table 'person' hasn't Field('alias_name' ) at db level, it's fake Field with reverse reference declaration ( 'storedin alias_opt.alias_name' ) that defines table.field that would be really involved in CRUD-

Re: [web2py] improving DAL for normalized DB

2016-12-06 Thread Dave S
On Tuesday, December 6, 2016 at 3:41:38 PM UTC-8, Richard wrote: > > UP, never get any answer... > How would this differ from reference fields? /dps > > On Sat, Jan 23, 2016 at 4:08 PM, Val K > > wrote: > >> Hi guys! >> I have an idea to improve DAL in scope of work with normalized DB. >> As

Re: [web2py] improving DAL for normalized DB

2016-12-06 Thread Richard Vézina
UP, never get any answer... On Sat, Jan 23, 2016 at 4:08 PM, Val K wrote: > Hi guys! > I have an idea to improve DAL in scope of work with normalized DB. > As known It's a common practice to avoid NULL value by creating separate > (option) table(s) to store non required fields. > So, it would be