Consider this reduced example from Exercism. Given that binding[1] only happens from right of the match operator to left of it, why does the binding work in the function head here?
defmodule Example do
def named_function(:a = variable_a) do
{variable_a, 1}
end
end
Example.named_function(:a)
# => {:a, 1}
So the parameter value is first matched with variable_a. The match operator always returns the right hand side value, so afterwards :a = :a is evaluated, which doesnât raise and again return the right hand side.
Given those behaviours order in the function head essentially doesnât matter much, though having more constraining match values more to the right could be a micro optimization.
Great, thank you. I tried reversing the expression, (ie def named_function(variable_a = :a) do), and it seemed to work in the same way. Reasoning it through, variable_a = (:a = parameter_value) seems to make more sense from a âfunction selectionâ point of view, so why is the example in the original post preferred (if at all)?
âMatchingâ in patterns is not really the same as matching in a body. It is actually based on âaliasesâ in Erlang where the rationale is that both sides of the = must match the argument and that they really have nothing to do with each other. Also all variables in the patterns are usable afterwards. That is why a common use is to both test what an argument should look and get a reference to the whole structure:
def foo({;a, :b, x, y, z} = all_of_it) when x == y + z do use(all_of_it) %We know all_of_it has the right format end
Or to be a bit fun
def foo({a,b,c} = {x,y,z}) do use_1(a, b, c) == use_1(x, y, z) end
The Erlang docs helped me clarify some nuance, assuming these things map onto Elixir:
The = character is used to denote two similar but distinct operators: the match operator and the compound pattern operator. Which one is meant is determined by context.
If the matching succeeds, any unbound variable in the pattern becomes bound and the value of Expr is returned. If multiple match operators are applied in sequence, they will be evaluated from right to left.
A pattern is just a term that can contain unbound variables. A compound pattern is of the form Pattern = Pattern:
The compound pattern operator is used to construct a compound pattern from two patterns. Compound patterns are accepted everywhere a pattern is accepted. A compound pattern matches if all of its constituent patterns match. It is not legal for a pattern that is part of a compound pattern to use variables (as keys in map patterns or sizes in binary patterns) bound in other sub patterns of the same compound pattern.
The compound pattern operator does not imply that its operands are matched in any particular order. That means that it is not legal to bind a variable in Pattern1 and use it in Pattern2, or vice versa.
When we declare a function, the arguments are actually treated as patterns (or compound patterns) that will be matched against the expression supplied when calling the function.
I donât know if âdoes not imply that its operands are matched in any particular orderâ holds when invoking functions in Elixir and matching the supplied expression to the declared pattern, though.
The distinction between a pattern (a term that can contain unbound variables) and other types of expressions and understanding where each can be used made it click for me I think.
When you specify an argument in a function declaration or write the left-hand side of a case, youâre writing patterns and compound patterns, not match expressions.
Matching and binding happens in match expressions. When you use = where a pattern is expected, youâre just writing a compound pattern.