From: Rob Biedenharn <Rob@...>
Date: 2006-11-02T23:35:36+09:00
Subject: Re: variable pointer


On Nov 2, 2006, at 8:55 AM, Rob Sanheim wrote:

> On 11/1/06, dblack@wobblini.net <dblack@wobblini.net> wrote:
>> Hi --
>>
>> On Thu, 2 Nov 2006, ara.t.howard@noaa.gov wrote:
>>
>> > On Thu, 2 Nov 2006 dblack@wobblini.net wrote:
>> >
>> >> It was recently noted on the Rails mailing list that something  
>> like
>> >> that happens in ActiveRecord.  Given Person and Address models,  
>> where
>> >> Person has_one Address:
>> >>
>> >>  a = person.address
>> >>  person.address = Address.find(some_other_address)
>> >>  puts a       # a has changed!
>> >>
>> >> I find it un-Rubyish and disconcerting (though I'm still  
>> waiting for
>> >> someone to explain the rationale; there may be something I'm  
>> missing).
>> >
>> > can you show an example?  are you saying that activerecord  
>> avoids apparent
>> > copy-on-write semantics with some trickery?
>>
>> Here's an example from a program which models Ideas; each Idea  
>> belongs
>> to a Page.
>>
>> >> i = Idea.find(6)
>> => #<Idea:0xb79eec2c @attributes={"page_id"=>"6", "id"=>"6",
>> "content"=>"Go to Paris"}>
>>
>> # Save the idea's page in p
>> >> p = i.page
>> => #<Page:0xb79e4a88 @attributes={"scratchpad_id"=>"6",
>> "title"=>"Domestic travel", "id"=>"6"}>
>>
>> # Change the idea's page to a different one
>> >> i.page = Page.find(7)
>> => #<Page:0xb79d1b18 @attributes={"scratchpad_id"=>"1",
>> "title"=>"Foreign travel", "id"=>"7"}>
>>
>> # Examine the saved page -- it's the new one
>> >> p
>> => #<Page:0xb79d1b18 @attributes={"scratchpad_id"=>"1",
>> "title"=>"Foreign travel", "id"=>"7"}>
>>
>> I don't know the rationale.  When I (think I'm) assign(ing) an object
>> to a variable, I really don't want it to turn out that it's some
>> special arrangement whereby the variable is subject to reassignment
>> without notice.  I expect:
>>
>>    p = i.page
>>
>> to be exactly equivalent to:
>>
>>    p = Idea.find(i.page.id)
>>
>> with respect to p.
>>
>
> Wow, that looks nasty.  Can you point me to the releveant thread in
> the rails group where this was discussed?
>
> - Rob
>
> -- 
> http://www.robsanheim.com
> http://www.seekingalpha.com
> http://www.ajaxian.com
>

Why is this so hard to understand?  Rails isn't really doing anything  
magic or really even tricky here -- just "unexpected".

You are just confusing references and objects.

 >> require 'ostruct'
=> true
 >> o = OpenStruct.new('page' => "original", 'cover' => "hard")
=> #<OpenStruct cover="hard", page="original">
 >> p = o.page
=> "original"

No surprise there.  Let's change the contents of the page.

 >> o.page.replace("forgery")
=> "forgery"

...and check in with our variable (aka, object reference):

 >> p
=> "forgery"

Yup. Still refers to the same String object as o.page.  Now let's do  
an assignment:

 >> o.page = "restoration"
=> "restoration"
 >> p
=> "forgery"

o.page now refers to a new object.  But o.page = "restoration" is  
really just sugar for o.page=("restoration"), right?

 >> def o.page=(kind); self.page.replace(kind); end
=> nil

Make p refer to the same String object as o.page again:

 >> p = o.page
=> "restoration"
 >> p
=> "restoration"
 >> o.page
=> "restoration"

Now this "assignment" calls the newly defined page= method:

 >> o.page = "imitation"
=> "imitation"
 >> p
=> "imitation"
 >> p.replace("faux")
=> "faux"
 >> o.page
=> "faux"

So the way that variables are references to objects and the syntactic  
sugar that makes 'o.page = foo' look like assignment even though it's  
sending foo to the page= method of the o object are just colliding in  
your brain.

-Rob

Rob Biedenharn		http://agileconsultingllc.com
Rob@AgileConsultingLLC.com