[storm] get_select_expr

[Jamu Kakar]

Hi,

Well, I think we should brainstorm about the structure a bit, but as a
starting point, what do you think about having these broad sections:

 * Object relational mapper - this section describes how to use the
       ORM features of Storm.  The focus should be on describing how
       to create Storm classes, how to establish a connection to the
       database, how to implement hooks for invalidation and such,
       etc.  Basically, this broad section is focused on how to use
       Storm.

       This section should be readable from beginning to end, at the
       end of which the reader should have a very good understanding
       of how to use the various features of Storm.  It should also be
       broken into sections, such that it can be used as a reference,
       once the user knows what they want to look for.  Finally, it
       should be example-heavy, to make learning as easy as possible.

 * Framework integration - this section describes the integration
       tools for using Storm with Zope 3 and with Django.  These
       should assume knowledge of how to use Storm and focus on
       describing how the framework hooks work and include examples to
       help users quickly get Storm integrated with these frameworks.

 * Schema management - this section describes the schema management
       tools in Storm with examples about how to use them.

 * Extending Storm - this section describes the expression and
       property systems.  The audience for this section is a user that
       wants to extend Storm by providing their own expressions or
       implementing custom properties.  It should describe the
       compilation process, the way precedence is handled, that way
       values are pushed to and pulled from the database, etc.

Each of these broad sections would have subparts, but for now I think
brainstorming about the broad organization is a good first step.

Thanks,
J.

On Wed, Nov 10, 2010 at 3:57 PM, Nachiketa Mishra <nmishra at uab.edu> wrote:
> That definitely works. I will go ahead and pull the branch. How do you want
> to structure the documentation ?
>
> On Wed, Nov 10, 2010 at 8:32 AM, Jamu Kakar <jkakar at kakar.ca> wrote:
>>
>> Hi Nachiketa,
>>
>> (Please use 'Reply all' to make sure the list is included when you
>> respond to messages)
>>
>> Great, glad to hear it!  We'd would *love* help with documentation,
>> it's a real sore spot for the project.  I've started a branch:
>>
>> lp:~jkakar/storm/documentation
>>
>> With some very basic beginning work at putting a manual in place.
>> Maybe we can work together to develop and outline for it and then
>> start to fill in sections?
>>
>> Thanks,
>> J.
>>
>> On Wed, Nov 10, 2010 at 3:22 PM, Nachiketa Mishra <nmishra at uab.edu> wrote:
>> > This is great.
>> > Thanks a lot. This worked perfectly. As I love storm, can I help in
>> > documentation ?
>> > Nachiketa
>> >
>> > On Wed, Nov 10, 2010 at 8:02 AM, Jamu Kakar <jkakar at kakar.ca> wrote:
>> >>
>> >> Hi Nachiketa,
>> >>
>> >> On Wed, Nov 10, 2010 at 2:57 PM, Nachiketa Mishra <nmishra at uab.edu>
>> >> wrote:
>> >> > I have a type table with name, category_name and description and some
>> >> > audit
>> >> > columns. With Storm I am trying to just get these three columns. I am
>> >> > new to
>> >> > Storm and I am not able to figure out from the get_select_expr api
>> >> > documentation, how to just fetch these three columns.
>> >>
>> >> Unless you want to perform a subselect, you don't need to use
>> >> get_select_expr.  Given a class:
>> >>
>> >> class Thing(Storm):
>> >>
>> >>    __storm_table__ = "thing"
>> >>
>> >>    id = Int(primary=True)
>> >>    name = Unicode()
>> >>    category_name = Unicode()
>> >>    description = Unicode()
>> >>    audit_stuff = Unicode()
>> >>
>> >> You can get just the columns you want with:
>> >>
>> >> result = store.find(Thing)
>> >> result = result.values(Thing.name, Thing.category_name,
>> >> Thing.description)
>> >> for name, category_name, description in result:
>> >>    print name, category_name, description
>> >>
>> >> The ResultSet.values method is handy when you have a result set that
>> >> would normally yield objects, but for which you only want columns.  If
>> >> you know you'll only ever want columns from the result set you can
>> >> specify them in the call to find:
>> >>
>> >> result = store.find((Thing.name, Thing.category_name,
>> >> Thing.description))
>> >> for name, category_name, description in result:
>> >>    print name, category_name, description
>> >>
>> >> I hope this helps!
>> >>
>> >> Thanks,
>> >> J.
>> >
>> >
>
>