Re: [Sqlalchemy-tickets] [sqlalchemy] #2823: Wrong sqlalchemy.sql.expression.false() / true() compilation when using SQL Server dialect

[sqlalchemy <mi...@zz...>]

#2823: Wrong sqlalchemy.sql.expression.false() / true() compilation when using SQL
Server dialect
-------------------------------------------+-------------------------------
      Reporter:  leavittx                  |      Owner:
          Type:  defect                    |     Status:  new
      Priority:  medium                    |  Milestone:
     Component:  mssql                     |   Severity:  no triage
    Resolution:                            |  selected yet
Progress State:  needs questions answered  |   Keywords:
-------------------------------------------+-------------------------------

Comment (by leavittx):

 I'd not use such expressions at all, but that's the filter code which
 sqlalchemy-orm-tree extension produces (https://github.com/monetizeio
 /sqlalchemy-orm-tree). So I see two possible solutions for the problem
 with true()/false() in isolation of a comparison. The first one is to
 generate hacky constructions like you said (I don't think they will slow
 the things so much). The second one is to do expression analysys &
 simplifying in some cases (i.e.  'false OR someexpr' becomes 'someexpr',
 'true AND someexpr' becomes 'someexpr' too, 'true OR someexpr' becomes
 '1=1 or someexpr', 'false AND someexpr' becomes '1=0'). I think the
 current behavior (generation of invalid sql code for sql server) is
 unacceptable!

--
Ticket URL: <http://www.sqlalchemy.org/trac/ticket/2823#comment:2>
sqlalchemy <http://www.sqlalchemy.org/>
The Database Toolkit for Python