What is causing the issue here? - arrays

Question

when I run this code in ruby:
updated_users = []
users = [{:name => "sam" , :number => 001 }]
ids = ["aa" , "bb" , "cc"]
users.each do |user|
ids.each do |id|
user[:guid] = id
updated_users << user
end
end
p updated_users
I get:
[{:name=>"sam", :number=>1, :guid=>"cc"},
{:name=>"sam", :number=>1, :guid=>"cc"},
{:name=>"sam", :number=>1, :guid=>"cc"}]
I expected to get:
[{:name=>"sam", :number=>1, :guid=>"aa"},
{:name=>"sam", :number=>1, :guid=>"bb"},
{:name=>"sam", :number=>1, :guid=>"cc"}]
What is happening and how should I get the desired output?

Answer 1

Your question reflects a common misunderstanding held by budding Rubyists. You begin with the following:
updated_users = []
users = [{:name => "sam" , :number => 001 }]
ids = ["aa" , "bb" , "cc"]
users is an array containing a single hash. Let's note the id of that hash object:
users.first.object_id
#=> 1440
Now let's execute your code with some puts statements added to see what is going on.
users.each do |user|
puts "user = #{user}"
puts "user.object_id = #{user.object_id}"
ids.each do |id|
puts "\nid = #{id}"
user[:guid] = id
puts "user = #{user}"
puts "user.object_id = #{user.object_id}"
updated_users << user
puts "updated_users = #{updated_users}"
updated_users.size.times do |i|
puts "updated_users[#{i}].object_id = #{updated_users[i].object_id}"
end
end
end
This displays the following.
user = {:name=>"sam", :number=>1}
user.object_id = 1440
So far, so good. Now begin the ids.each loop:
id = aa
user = {:name=>"sam", :number=>1, :guid=>"aa"}
user.object_id = 1440
As user did not previously have a key :guid, the key-value pair :guid=>"aa" was added to user. Note that user's id has not changed. user is then appended to the (empty) array updated_users:
updated_users = [{:name=>"sam", :number=>1, :guid=>"aa"}]
updated_users[0].object_id = 1440
Again this is what we should expect. The next element of ids is then processed:
id = bb
user = {:name=>"sam", :number=>1, :guid=>"bb"}
user.object_id = 1440
Since user has a key :guid this merely changes the value of that key from "aa" to "bb". user is then appended to updated_users:
updated_users = [{:name=>"sam", :number=>1, :guid=>"bb"},
{:name=>"sam", :number=>1, :guid=>"bb"}]
updated_users[0].object_id = 1440
updated_users[1].object_id = 1440
The first and second elements of this array are seen to be the same object user, so changing the value of :user's key :guid affected both elements in the same way.
The same thing happens when the third and last element of ids is processed:
id = cc
user = {:name=>"sam", :number=>1, :guid=>"cc"}
user.object_id = 1440
updated_users = [{:name=>"sam", :number=>1, :guid=>"cc"},
{:name=>"sam", :number=>1, :guid=>"cc"},
{:name=>"sam", :number=>1, :guid=>"cc"}]
updated_users[0].object_id = 1440
updated_users[1].object_id = 1440
updated_users[2].object_id = 1440
Got it?
To obtain the result you want you need to append updated_users with distinct hashes derived from user:
users.each do |user|
ids.each do |id|
updated_users << user.merge(guid: id)
end
end
updated_users
#=> [{:name=>"sam", :number=>1, :guid=>"aa"},
# {:name=>"sam", :number=>1, :guid=>"bb"},
# {:name=>"sam", :number=>1, :guid=>"cc"}]
Note that users has not been mutated (changed):
users
#=> [{:name=>"sam", :number=>1}]
See Hash#merge. Note that user.merge(guid: id) is shorthand for user.merge({ guid: id }) or, equivalently, user.merge({ :guid => id }).