Crystal: Object#in?

[jlcrochet]

Maybe this is redundant, but I have always thought that a Python-ish obj.in?(enum) would be more intuitive than the Ruby-ish enum.includes?(obj):

class Object
  def in?(e : Enumerable) : Bool
    e.includes?(self)
  end
end

1.in? 1..10  #=> true
[1, 2].in? [[1, 2], [3, 4]]  #=> true

[Blacksmoke16]

~See https://github.com/crystal-lang/crystal/wiki/FAQ#why-are-aliases-discouraged.~

EDIT: Sorry, it's not actually an alias.

[wontruefree]

I don't know about adding something like this to the root object but it would be nice to have in?

[jlcrochet]

@Blacksmoke16 Yeah, it's true that it isn't technically an alias, but I think the principle is the same: it merely provides an alternative to something the language already supports. Probably not a good idea.

[RX14]

I actually don't mind this, I've written if {a, b, c}.includes? x a bunch of times to avoid a case with a single when, and if x.in? {a, b, c} is nicer.

I'm pretty ambivalent, if others in the core team are open im happy to accept a PR.

[jlcrochet]

@RX14 Sorry, should I go ahead and reopen this? Not sure what the proper etiquette is.

[RX14]

@Marzipanzerfaust usually it ends up better to leave it a few days to make sure everyone's had their say before closing.

[straight-shoota]

@RX14 When used with a fixed set of values, x.in?(a, b, c) could also work. So this would really be a nice way to write disjunctive comparisons with a single reference than x == a || x == b || x == c.

[bew]

I actually don't mind this, I've written if {a, b, c}.includes? x a bunch of times to avoid a case with a single when, and if x.in? {a, b, c} is nicer.

Note that you'll need parentheses because the parser will think it's a block

[j8r]

Is the only advantage to write an expression the other way around?

Having two redundant methods will produce different styles: those who prefer includes?, others prefering in?, and code with both mixed.'

I thought adding methods to Object was something to avoid.

There is also Enumerable#any?, which is quite powerful.

[RX14]

@straight-shoota yeah, a

def in?(*values)
  self.in? values
end

would be neat

[RX14]

Aliases on the same object (different names for the same method) are similar but different to shortcuts on other objects.

This is on the borderline of breaking the no-aliases rule, which is why i'm open to the rest of the core team's input instead of rejecting it.

[jlcrochet]

My reasoning for this was similar to the existence of String#match versus Regex#match: one is simply the reverse of the other but is slightly more convenient in certain situations, especially in method chains.

[j8r]

In any case, the Enumerable annotation must not be put in Object, because it will make it depend on the Enumerable module (object is before enumerable in prelude)