Arrays of Model Objects, Intersections, Object Identificatio


#1

Hi all,

Hope you can help me with understanding how Ruby / Rails treats arrays
full
of objects.

Let’s say I have to arrays of objects. Both are the same kinds of
objects.

tomatoes = Fruit.find(:all, :conditions => [ ‘tomato = ?’, true], :limit
=>
10)
fruits = Fruit.find(:all, :limit => 10)

And I want to create an array of these objects called @my_fruits, but I
don’t want the set to contain any duplicates.

In the Ruby API documentation for arrays it shows you can use the |
operator
to perform a union between two arrays, throwing out duplicates.

OK. That sounds good. That’s what I want.

@my_fruits = tomatoes | fruits

98% of the time I’ve found this works in my tests. I don’t get
duplicates.
But I’ve had instances where I do in fact have a row repeat.

Any ideas as to why the union operator doesn’t work in this case?

In my real world ™ example, I’m sorting the database rows randomly,
and
the data changes frequently. Both arrays contain the same types of
objects.
In entirely likely that two arrays can contain the same object.

Is there a performance hit in building an each loop on the likely
smaller of
the array to test whether the other array ".include?"s the object? Would
this be considered more reliable?

Example:
@my_fruits = fruits
tomatoes.each do | fruit |
@my_fruits.push(fruit) unless @my_fruits.include?(fruit)
end

Why?

Thanks.
D. Taylor S.,
Reality Technician


#2

Hi…

Have you tried the uniq! method for arrays? (ruby 1.84)
http://corelib.rubyonrails.org/classes/Array.html#M000452

Something like

@myarray.uniq!

Hope this works for you


#3

I figured this out in a duh moment this morning. One array was formed
through traditional ActiveRecord find methods and the other was
constructed
with a find_by_sql string that included a join table in the select. The
inclusion of the join table made the objects appear as different between
the
two sets. Sanity reset.

Thanks for the suggestions.

D. Taylor S.,
Reality Technician