from sqlalchemy.sql import column, tuple_
a, b = column('a'), column('b')
print tuple_(a, b).op("OVERLAPS")(tuple_(1, 2))
"(a, b) OVERLAPS (:param_1, :param_2)"
On May 8, 2013, at 6:39 AM, gio <giovanni.lagr...@auditq.com> wrote:
> Hi all
>
> Searching how to solve the
>
> (start,end).op('OVERLAPS')(,start1,end1)
> Problem with postgres and salqalchemy 0.8
>
> Some hints ?
>
>
> Note:
> _CompareMixin
> is not presen in SQLalchem 0.8
>
>
> Am Freitag, 21. August 2009 09:50:10 UTC+2 schrieb David Bolen:
> Has anyone generated ORM queries using the OVERLAPS SQL operator that
> reference columns in the tables in the query? I've been experimenting
> with various approaches and can't seem to cleanly get the column names
> (with their appropriate alias based on the rest of the query) into the
> overlaps clause.
> I'm basically issuing an ORM query and want to check that the date
> range given by two columns in one of the objects being queried is
> overlaps with a computed date range. In some cases the object whose
> columns I am checking is the primary target of the query whereas in
> others it's a joined class.
>
> I found an older post from March where Michael suggested the form
>
> <somexpression>.op('OVERLAPS', <someotherexpression>)
>
> but I can't figure out how to apply that, and in particular what sort
> of expression will produce a tuple at the SQL layer, yet still support
> the "op" method? Namely the output needs to be of the form:
>
> (start, stop) OVERLAPS (start, stop)
>
> So I figured I'd try straight text, and was attempting something like:
>
> query(MappedClass).
> filter('(:c_start, :c_end) overlaps (:start, :end)').
> params(c_start=MappedClass.start_col, c_end=MappedClass.end_col,
> start=<datetimevalue>, end=<datetimevalue>)
>
> but I'm having trouble identifying an appropriate value for the
> c_start/c_end params to generate the column names in the resulting
> SQL. The above gives can't adapt the InstrumentedAttribute references
> in the params.
>
> In the meantime I can fall back to a pure textual filter which will
> have to assume how the mapped class will be aliased, but that feels
> fragile and it'd be nice if I could let SQLAlchemy generate the column
> names somehow. I get the feeling though that OVERLAPS is a bit
> unusual in terms of the necessary support since it has non-scalar left
> and right values for the operator.
>
> Thanks for any help.
>
> -- David
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "sqlalchemy" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sqlalchemy+unsubscr...@googlegroups.com.
> To post to this group, send email to sqlalchemy@googlegroups.com.
> Visit this group at http://groups.google.com/group/sqlalchemy?hl=en.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>
--
You received this message because you are subscribed to the Google Groups
"sqlalchemy" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to sqlalchemy+unsubscr...@googlegroups.com.
To post to this group, send email to sqlalchemy@googlegroups.com.
Visit this group at http://groups.google.com/group/sqlalchemy?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.
