Todo lo que necesita saber por referencia vs por valor

Cuando se trata de ingeniería de software, hay bastantes conceptos y términos mal entendidos. Por referencia vs por valor es definitivamente uno de ellos.

Recuerdo el día en que leí sobre el tema y cada fuente que revisé parecía contradecir la anterior. Me llevó algún tiempo comprenderlo bien. No tuve elección ya que es un tema fundamental si eres ingeniero de software.

Me encontré con un error desagradable hace unas semanas y decidí escribir un artículo para que otras personas pudieran tener más facilidad para resolver todo esto.

Codifico en Ruby a diario. También utilizo JavaScript con bastante frecuencia, así que he elegido estos dos lenguajes para esta presentación.

Sin embargo, para comprender todos los conceptos, también usaremos algunos ejemplos de Go y Perl.

Para comprender todo el tema, debe comprender 3 cosas diferentes:

  • Cómo se implementan las estructuras de datos subyacentes en el lenguaje (objetos, tipos primitivos, mutabilidad,).
  • Cómo funciona la asignación / copia / reasignación / comparación de variables
  • Cómo se pasan las variables a las funciones

Tipos de datos subyacentes

En Ruby no hay tipos primitivos y todo es un objeto, incluidos enteros y booleanos.

Y sí, hay uno TrueClassen Ruby.

true.is_a?(TrueClass) => true3.is_a?(Integer) => truetrue.is_a?(Object) => true3.is_a?(Object) => trueTrueClass.is_a?(Object) => trueInteger.is_a?(Object) => true

Estos objetos pueden ser mutables o inmutables.

Inmutable significa que no hay forma de que pueda cambiar el objeto una vez creado. Solo hay una instancia para un valor dado con uno object_idy permanece igual sin importar lo que haga.

Por defecto en Ruby los tipos de objetos inmutables son: Boolean, Numeric, nil, y Symbol.

En MRI, el object_id de un objeto es el mismo que el VALUE que representa el objeto en el nivel C. Para la mayoría de los tipos de objetos, VALUE es un puntero a una ubicación en la memoria donde se almacenan los datos reales del objeto.

De ahora en adelante usaremos object_ide memory addressindistintamente.

Ejecutemos algo de código Ruby en MRI para un símbolo inmutable y una cadena mutable:

:symbol.object_id => 808668:symbol.object_id => 808668'string'.object_id => 70137215233780'string'.object_id => 70137215215120

Como puede ver, mientras que la versión del símbolo mantiene el mismo object_id para el mismo valor, los valores de cadena pertenecen a diferentes direcciones de memoria.

A diferencia de Ruby, JavaScript tiene tipos primitivos.

Ellos son - Boolean, null, undefined, String, y Number.

El resto de los tipos de datos de ir bajo el paraguas de objetos ( Array, Function, y Object). No hay nada especial aquí es mucho más sencilla que la de Ruby.

[] instanceof Array => true[] instanceof Object => true3 instanceof Object => false

Asignación, copia, reasignación y comparación de variables

En Ruby, cada variable es solo una referencia a un objeto (ya que todo es un objeto).

a = 'string'b = a
# If you reassign a with the same value
a = 'string'puts b => 'string'puts a == b => true # values are the sameputs a.object_id == b.object_id => false # memory adr-s. differ
# If you reassign a with another value
a = 'new string'puts a => 'new string'puts b => 'string'puts a == b => false # values are differentputs a.object_id == b.object_id => false # memory adr-s. differ too

Cuando asigna una variable, es una referencia a un objeto, no al objeto en sí. Cuando copia un objeto, b = aambas variables apuntarán a la misma dirección.

Este comportamiento se denomina copiar por valor de referencia .

Estrictamente hablando, en Ruby y JavaScript, todo se copia por valor.

Sin embargo, cuando se trata de objetos, los valores resultan ser las direcciones de memoria de esos objetos. Gracias a esto podemos modificar los valores que se encuentran en esas direcciones de memoria. Nuevamente, esto se llama copia por valor de referencia, pero la mayoría de la gente se refiere a esto como copia por referencia.

Sería una copia por referencia si después de reasignar aa 'nueva cadena', btambién apuntara a la misma dirección y tuviera el mismo valor de 'nueva cadena'.

Lo mismo con un tipo inmutable como Integer:

a = 1b = a
a = 1puts b => 1puts a == b => true # comparison by valueputs a.object_id == b.object_id => true # comparison by memory adr.

Al reasignar una con el mismo número entero, las estancias de direcciones de memoria del mismo desde un determinado número entero siempre tiene el mismo object_id.

Como puede ver cuando compara cualquier objeto con otro, se compara por valor. Si quieres comprobar si son el mismo objeto que tienes que usarobject_id.

Veamos la versión de JavaScript:

var a = 'string';var b = a;a = 'string'; # a is reassigned to the same value
console.log(a); => 'string'console.log(b); => 'string'console.log(a === b); => true // comparison by value
var a = [];var b = a;
console.log(a === b); => true
a = [];
console.log(a); => []console.log(b); => []console.log(a === b); => false // comparison by memory address

Excepto la comparación: JavaScript utiliza por valor para tipos primitivos y por referencia para objetos. El comportamiento parece ser el mismo que en Ruby.

Bueno, no del todo.

Primitive values in JavaScript will not be shared between multiple variables . Even if you set the variables equal to each other. Every variable representing a primitive value is guaranteed to belong to a unique memory location.

This means none of the variables will ever point to the same memory address. It is also important that the value itself is stored in a physical memory location.

In our example when we declare b = a, b will point to a different memory address with the same ‘string’ value right away. So you don’t need to reassign a to point to a different memory address.

This is called copied by value since you have no access to the memory address only to the value.

Let’s see a better example where all this matters.

In Ruby if we modify the value that sits in the memory address then all the references that point to the address will have the same updated value:

a = 'x'b = a
a.concat('y')puts a => 'xy'puts b => 'xy'
b.concat('z')puts a => 'xyz'puts b => 'xyz'
a = 'z'puts a => 'z'puts b => 'xyz'
a[0] = 'y'puts a => 'y'puts b => 'xyz'

You might think in JavaScript only the value of a would change but no. You can’t even change the original value as you don’t have direct access to the memory address.

You could say you assigned ‘x’ to a but it was assigned by value so a’s memory address holds the value ‘x’, but you can’t change it as you have no reference to it.

var a = 'x';var b = a;
a.concat('y');console.log(a); => 'x'console.log(b); => 'x'
a[0] = 'z';console.log(a); => 'x';

The behavior of JavaScript objects and implementation are the same like Ruby’s mutable objects. Both copy be reference value.

JavaScript primitive types are copied by value. The behavior is the same like Ruby’s immutable objects which are copied by reference value .

Huh?

Again, when you copy something by value it means you can’t change (mutate) the original value since there is no reference to the memory address. From the perspective of the writing code this is the same thing like having immutable entities that you can’t mutate.

If you compare Ruby and JavaScript the only data type that ‘behaves’ differently by default is String (that’s why we used String in the examples above).

In Ruby it is a mutable object and it is copied/passed by reference value while in JavaScript it is a primitive type and copied/passed by value.

When you wanna clone (not copy) an object you have to do it explicitly in both languages so you can make sure the original object won’t be modified:

a = { 'name': 'Kate' }b = a.cloneb['name'] = 'Anna'puts a => {:name=>"Kate"}
var a = { 'name': 'Kate' };var b = {...a}; // with the new ES6 syntaxb['name'] = 'Anna';console.log(a); => {name: "Kate"}

It is crucial to remember this otherwise you will run into some nasty bugs when you invoke your code more than once. A good example would be a recursive function where you use the object as argument.

Another one is React (JavaScript front-end framework) where you always have to pass a new object for updating state as the comparison works based on object id.

This is faster because you don’t have to go through the object line by line to see if it has been changed.

How variables are passed to functions

Passing variables to functions is working the same way like copying for the same data types in most of the languages.

In JavaScript primitive types are copied and passed by value and objects are copied and passed by reference value.

I think this is the reason why people only talk about pass by value or pass by reference and never seem to mention copying. I guess they assume copying works the same way.

a = 'b'
def output(string) # passed by reference value string = 'c' # reassigned so no reference to the original puts stringend
output(a) => 'c'puts a => 'b'
def output2(string) # passed by reference value string.concat('c') # we change the value that sits in the address puts stringend
output(a) => 'bc'puts a => 'bc'

Now in JavaScript:

var a = 'b';
function output (string) { // passed by value string = 'c'; // reassigned to another value console.log(string);}
output(a); => 'c'console.log(a); => 'b'
function output2 (string) { // passed by value string.concat('c'); // we can't modify it without reference console.log(string);}
output2(a); => 'b'console.log(a); => 'b'

If you pass an object (not a primitive type like we did) in JavaScript to the function it works the same way like the Ruby example.

Other languages

We have already seen how copy/pass by value and copy/pass by reference value work. Now we will see what pass by reference is about and we also discover how we can change objects if we pass by value.

As I looked for pass by reference languages I couldn’t find too many and I ended up choosing Perl. Let’s see how copying works in Perl:

my $x = 'string';my $y = $x;$x = 'new string';
print "$x"; => 'new string'print "$y"; => 'string'
my $a = {data => "string"};my $b = $a;$a->{data} = "new string";
print "$a->{data}\n"; => 'new string'print "$b->{data}\n"; => 'new string'

Well this seems to be the same just like in Ruby. I haven’t found any proof but I would say Perl is copied by reference value for String.

Now let’s check what pass by reference means:

my $x = 'string';print "$x"; => 'string'
sub foo { $_[0] = 'new string'; print "$_[0]"; => 'new string'}
foo($x);
print "$x"; => 'new string'

Since Perl is passed by reference if you do a reassignment within the function it will change the original value of the memory address as well.

For pass by value language I have chosen Go as I intend to deepen my Go knowledge in the foreseeable future:

package mainimport "fmt"
func changeAddress(a *int) { fmt.Println(a) *a = 0 // setting the value of the memory address to 0}
func changeValue(a int) { fmt.Println(a) a = 0 // we change the value within the function fmt.Println(a)}
func main() { a := 5 fmt.Println(a) fmt.Println(&a) changeValue(a) // a is passed by value fmt.Println(a) changeAddress(&a) // memory address of a is passed by value fmt.Println(a)}
When you compile and run the code you will get the following:
0xc42000e32855050xc42000e3280

If you wanna change the value of a memory address you have to use pointers and pass around memory addresses by value. A pointer holds the memory address of a value.

The & operator generates a pointer to its operand and the * operator denotes the pointer’s underlying value. This basically means that you pass the memory address of a value with & and you set the value of a memory address with *.

Conclusion

How to evaluate a language:

  1. Understand the underlying data types in the language. Read some specifications and play around with them. It usually boils down to primitive types and objects. Then check if those objects are mutable or immutable. Some languages use different copying/passing tactics for different data types.
  2. Next step is the variable assignment, copying, reassignment and comparison. This is the most crucial part I think. Once you get this you will be able to figure out what’s going on. It helps a lot if you check the memory addresses when playing around.
  3. Passing variables to functions usually is not special. It usually works the same way like copying in most languages. Once you you know how the variables are copied and reassigned you already know how they are passed to functions.

The languages we used here:

  • Go: Copied and passed by value
  • JavaScript: Primitive types are copied/passed by value, objects are copied/passed by reference value
  • Ruby: Copied and passed by reference value + mutable/immutable objects
  • Perl: Copied by reference value and passed by reference

When people say passed by reference they usually mean passed by reference value. Passing by reference value means that variables are passed around by value but those values are references to the objects.

As you saw Ruby only uses pass by reference value while JavaScript uses a mixed strategy. Still, the behavior is the same for almost all the data types due to the different implementation of the data structures.

Most of the mainstream languages are either copied and passed by value or copied and passed by reference value. For the last time: Pass by reference value is usually called pass by reference.

In general pass by value is safer as you won’t run into issues since you can’t accidentally change the original value. It is also slower to write because you have to use pointers if you want to change the objects.

It’s the same idea like with static typing vs dynamic typing — development speed at the cost of safety. As you guessed pass by value is usually a feature of lower level languages like C, Java or Go.

Pass by reference or reference value are usually used by higher level languages like JavaScript, Ruby and Python.

When you discover a new language go through the process like we did here and you will understand how it works.

This is not an easy topic and I am not sure everything is correct what I wrote here. If you think I have made some mistakes in this article, please let me know in the comments.