Rails - Oct 2007 - Edge rails: associations in sanitize_sql_hash_for_conditions hack/feature

Hi,

a ticket entry in TRAC (http://dev.rubyonrails.org/ticket/9733) put me
onto a small feature that I would like to implement in ActiveRecord
but since it involves a change to a private ActiveRecord::Base method
I''m not sure how best to go about it.

The change is trivial and verrrry lightly tested but it allows me to
put the name of an association into the :condition hash. ActiveRecord
recognises it as such and so it allows me to have table names that are
different to the association name.

An example:

class FormalProof < AR
 belongs_to :formal_result
end

class FormalResult < AR
  has_many :formal_proofs
end

For legacy reasons the tables are DM19_FORMAL and DM19_FORMAL_PROOFS
for FormalResult and FormalProof respectively.

I''d like to write:

FormalProof.find(:first,:conditions =>
{''formal_result.module_name'' =>
''module1''}, :include => :formal_result)

and have it emit:

SELECT DM19_FORMALPROOFS.`id` AS t0_r0, ...,  FROM DM19_FORMALPROOFS
LEFT OUTER JOIN DM19_FORMAL ON DM19_FORMAL.id DM19_FORMALPROOFS.formal_result_id
WHERE (`DM19_FORMAL`.`module_name`
= ''module1'')

The edge release doesn''t know that the :conditions hash refers to an
association, it considers it a table.column_name reference and the
generated SQL contains the wrong (for me) table name.

Anywaaaay, my hack looks like:

In base.rb, method sanitize_sql_hash_for_conditions

             # Extract table name from qualified attribute names.
             if attr.include?(''.'')
               table_name, attr = attr.split(''.'', 2)
-              table_name = connection.quote_table_name(table_name)
+              if reflect_on_association(table_name.to_sym)
+                table_name connection.quote_table_name(reflect_on_association
(table_name.to_sym).klass.table_name)
+              else
+                table_name = connection.quote_table_name(table_name)
+              end
             else
               table_name = quoted_table_name
             end

which seems to do the right thing for me. I don''t think it''s
too bad a
hack either ...

How do I get this into my application?!

Allan


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Talk" group.
To post to this group, send email to
rubyonrails-talk-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-talk-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-talk?hl=en
-~----------~----~----~----~------~----~------~--~---

On 10/17/07, Allan
<Allan.Cochrane-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
wrote:
> I''d like to write:
>
> FormalProof.find(:first,:conditions =>
{''formal_result.module_name'' =>
> ''module1''}, :include => :formal_result)
>
> and have it emit:
>
> SELECT DM19_FORMALPROOFS.`id` AS t0_r0, ...,  FROM DM19_FORMALPROOFS
> LEFT OUTER JOIN DM19_FORMAL ON DM19_FORMAL.id >
DM19_FORMALPROOFS.formal_result_id WHERE (`DM19_FORMAL`.`module_name`
> = ''module1'')
>
> The edge release doesn''t know that the :conditions hash refers to
an
> association, it considers it a table.column_name reference and the
> generated SQL contains the wrong (for me) table name.
String conditions are always sql fragments and shouldn''t be conflated
with associations, in my opinion.

For associations, I would love to see
  FormalProof.find(:first, :include => :formal_result,
    :conditions => { :formal_result => { :module_name =>
''module1'' }
similar to the way deep association :include are specified.

See http://dev.rubyonrails.org/ticket/9907 for the beginning of that approach.

jeremy

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Talk" group.
To post to this group, send email to
rubyonrails-talk-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-talk-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-talk?hl=en
-~----------~----~----~----~------~----~------~--~---

On Oct 17, 8:34 pm, "Jeremy Kemper"
<jer...-w7CzD/W5Ocjk1uMJSBkQmQ@public.gmane.org>
wrote:
> On 10/17/07, Allan
<Allan.Cochr...-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> String conditions are always sql fragments and shouldn''t be
conflated
> with associations, in my opinion.
Perhaps, but this feature would be useful to me, in particular. Hence
my question is really about how best to hack my world or maybe offer a
plugin given the very low level nature of the hack.

> For associations, I would love to see
>   FormalProof.find(:first, :include => :formal_result,
>     :conditions => { :formal_result => { :module_name =>
''module1'' }
> similar to the way deep association :include are specified.
>
> See http://dev.rubyonrails.org/ticket/9907for the beginning of that
approach.
That would indeed be very useful, but until then, can anyone point me
in the right direction? Maybe follow the approach here
http://rlucas.net/blog/bugfix/ruby_active_record_makes_raw_sql_a_royal_pain.html

Thanks,

Allan


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Talk" group.
To post to this group, send email to
rubyonrails-talk-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-talk-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-talk?hl=en
-~----------~----~----~----~------~----~------~--~---

On 10/17/07, Allan
<Allan.Cochrane-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
wrote:
> That would indeed be very useful, but until then, can anyone point me
> in the right direction? Maybe follow the approach here
>
http://rlucas.net/blog/bugfix/ruby_active_record_makes_raw_sql_a_royal_pain.html
I was hoping you were volunteering to move ''until then'' closer
to ''now'' ;)

jeremy

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Talk" group.
To post to this group, send email to
rubyonrails-talk-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-talk-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-talk?hl=en
-~----------~----~----~----~------~----~------~--~---