Feature #14982

Improve namespace system in ruby to avoiding top-level names chaos

Added by jjyr (Jinyang Jiang) 4 days ago. Updated 2 days ago.

Target version:




Ruby has evaluation all class/module names in top-level context(aka TOPLEVEL_BINDING).
As a user we basically hard to know how many names in the current context, is causing chaos in some cases. For example:

case 1:

Put common used errors class in a single file, like below

# utils/errors.rb

class FooError

class BarError

In other files under 'utils' we want to use those errors, so the best practice is to use require_relative 'errors' in each file we need.

# utils/binary_helper.rb

# we forget require errors

module BinaryHelper
# ...
  raise BarError
# ...

But sometime we may forget to require dependencies in a file, it's hard to notice because
if RubyVM already execute the requires we still can access the name BarError,

but if user directly to require 'utils/binary_helper', he/she will got an NameError.

case 2:

Two gems use same top-level module name, so we can't use them together

The Reason of The Problem

The reason is we let module author to decision which module user can use. ('require' is basically evaluation, highly dependent on the module author's design)

But we should let users control which names to use and available in context. As many other popular languages dose(Rust, Python..)

I think the solution is basically the same philosophy compares to refinement feature.

The Design

I propose an improved namespace to Ruby, to solve the problems and still compatible with the current Ruby module system.

class Foo

# introduce Kernel#namespace
namespace :Hello do
  # avoiding namespace chaos
  # Foo -> NameError, can't access TOPLEVEL_BINDING directly

  # Kernel#import method, introduce Foo name from TOPLEVEL_BINDING
  import :Foo

  # in a namespace user can only access imported name

  # import constant to another alias name
  # can avoid writing nested module/class names
  import :"A::B::C::D", as: :E

  # require then import, for convenient 
  import :"A::B::C::D", as: :E, from: 'some_rb_file'

  # import same name from two gems
  import :"Foo", as: :Foo_A, from: 'foo_a'
  import :"Foo", as: :Foo_B, from: 'foo_b'

  # import names in batch
  import %i{"A::B::C::D", "AnotherClass"}, from: 'some_rb_file'

  # import and alias in batch
  import {:"A::B::C::D" => :E, :Foo => Foo2}, from: 'some_rb_file'

  class Bar
    def xxx
      # can access all names in namespace scope
      [Foo, Foo_A, Foo_B]

Hello.class #  -> module. namespace is just a module
Hello::Bar # so we do not broken current ruby module design

# namespace system is intent to let user to control names in context
# So user can choose use the old require way

require 'hello'


# Or user can use namespace system as we do in hello.rb

namespace :Example do
  import :"Hello::Bar", as: :Bar
  Bar # ok
  Foo # name error, cause we do not import Foo in :Example namespace

Foo # ok, cause Foo is loaded in TOPLEVEL_BINDING

# define nested namespace

# more clear syntax than “module Example::NestedExample”
namespace :NestedExample, under: Example do

namespace :Example2 do
  namespace :NestedExample do


  • Completely compatible with the current module system, a gem user can completely ignore whether a gem is write in Namespace or not.
  • User can completely control which names in current context/scope.
  • May solve the top module name conflict issue(depends on VM implementation).
  • Avoid introducing new keyword and syntax.
  • Type hint or name hint can be more accuracy under namespace(not sure).


  • Need to modify Ruby VM to support the feature.


#1 Updated by jjyr (Jinyang Jiang) 4 days ago

  • Description updated (diff)

#2 Updated by jjyr (Jinyang Jiang) 4 days ago

  • Description updated (diff)

#3 Updated by jjyr (Jinyang Jiang) 4 days ago

  • Description updated (diff)
  • Subject changed from Introduce new namespace system to ruby to avoiding top-level names chaos to Improve namespace system in ruby to avoiding top-level names chaos

#4 [ruby-core:88449] Updated by shevegen (Robert A. Heiler) 4 days ago

I think this has come up before in other issue requests, at the least
in one variant or the other (if I recall correctly, Hiroshi Shibata
also suggested some variant that is a bit similar to your "import"
example, but I do not remember the full issue's content; may have
been some months ago or perhaps a few years).

I also had a few ideas; e. g. to be able to attach meta-information
to class/modules (that way we can find out who is the original author,
when and what changes may have been made etc...).

I agree with this comment here a lot by the way:

import :"A::B::C::D", as: :E

Not necessarily about using the name "import", but with the ability
to re-define namespaces at require time. We can of course already
do so by including modules and removing (older) constants, but I
always thought it may be more elegant to be able to do so the moment
we require ruby code.

I am not so sure about the rest of the suggestion. I don't have any
particularly strong pro or con opinion, although I am a bit wary.

Part of the suggestions all are a bit complicated, API-wise and from
the scope. I understand that, if we want more flexibility, we may
need to be able to have a way to add code which requires more characters
and such. But one thing that is great in ruby, even if we say that
"having no namespaces is a disadvantage", is that using modules and
classes on the toplevel space, is very, very simple. People very
quickly understand that concept.

class Cat
  def meow
    puts 'The cat meows.'

With namespaces as suggested here, we may add another layer of
complexity; and while I do agree with some stronger form of
control possible over "namespaces" in ruby, I am not sure if
the proposal in this form is having a good trade-off. But as I
wrote, it's not that I have a big opinion either way - I think
the biggest concern I have had in regards to namespaces was
when ruby were to use PHP's "solution" and syntax for
namespaces ... :P

As for refinements - the odd thing is that I agree behind the
proposal and ideas, but the syntax and API is so weird to me
and it feels ... strange to use them. I also have no alternative
suggestion, so this is not good; best way would be to have
both namespaces, namespace scopes and refinements in a single
issue with a great, beautiful syntax. :D

(We should however had also consider whether the status quo
is actually better than the proposed changes. And to some
extent I'd rather use a status quo than want to transition
into changes that do not seem to be as worthwhile to be
had - even though I actually agree with a LOT on what is
said about namespaces, refinements etc...)

As for requiring ruby code, I agree. In particular for larger
projects written in ruby, it may be useful to not only have
more control, but make managing that ruby code simpler. In
your example, the author who wrote the code must have forgotten
to require some other files; but I understand that this may
be tedious if one has a large project with lots of .rb files.
Then there are also circular warnings which are no fun at all.

I am confident that this may improve in the long run - matz
always said that ruby is for humans rather than computers
and that the core team will listen to (and prioritize on)
"real problems" and painpoints people have when writing
ruby code. And personally I think that a lot of these problems
emerge when one writes a lot of ruby code and has lots of
ruby files, too.

#5 [ruby-core:88454] Updated by jjyr (Jinyang Jiang) 3 days ago

I believe namespace can reduce the complexity of organizing codes in large projects.

Recently I write rust code in a large project. Our code base is dependent on other several complex projects. With rust use syntax( we can import the names which we need from other projects without chaotic the global names. It's easy to manage the complexity at the module level.

I can’t image how to handle those complexities under the current ruby requiring system.

I understand the concern, is this a “real problem” or just because we saw the feature in other languages so we want it?

I am currently working on a medium-level ruby project( and I find it's easy to forget to add some unnecessary requiring.
So in this situation, a user can't directly require this file, He must handle the dependencies manually or require the whole gem.

I do not have real experiences working in millions of lines ruby project, but I believe namespace can help to control complexity, from my other languages experiences.

So I paste my proposal, let the community to discuss the problem whether is real or fake.

I realized the essence of the proposal is to allow user to manipulate Binding. Currently, Ruby only has the TOPLEVEL_BINDING, we can't isolated Binding from top-level.

So the proposal essentially requests several primitive to control Binding.

1 isolated requiring: allows to evaluation required files under an isolated Binding(not polluting TOPLEVEL_BINDING).

# requiring into isolated Binding
require ‘foo’, into: :IsolatedBindingModule1
IsolatedBindingModule1.class # Module
IsolatedBindingModule1::Foo # access names

# the old way should still work
# requiring and polluting TOPLEVEL_BINDING
require ‘foo’

2 namespace: allows users to create an isolated Binding scope.

namespace do # Create an isolated binding.The name maybe not accurate, can be discussed.
    Foo # NameError

3 import: allow user import names from a Binding into another Binding.

namespace do
  # import name from a ruby file
  # import primitive is a convenient way to use isolated requiring
  import :Foo, from: “foo”

Then we can extend those methods to more conveniently be used, for example:

#6 Updated by jjyr (Jinyang Jiang) 3 days ago

  • Description updated (diff)

#7 [ruby-core:88458] Updated by shyouhei (Shyouhei Urabe) 2 days ago

I like this idea in general. I too want to have "requiring into isolated Binding".

One thing I would like to add, "namespace" shall be a keyword rather than a normal method taking a block. Blocks can be passed around:

namespace :Foo do
  import :Foo from: 'foo'
    return lamnda do

namespace(:Bar, & # => Error, or ...?

We should forbid this kind of headache.

#8 [ruby-core:88459] Updated by jeremyevans0 (Jeremy Evans) 2 days ago

In the Why? section, case 1 is just a programming error on the library developer's or library user's part, depending on how the library is documented.

In terms of case 2 (multiple gems define overlapping constants in the top level namespace), that could be a issue in theory, but it doesn't generally present a problem in practice as library authors in most cases take care to choose non-overlapping names.

It will probably be challenging to make import work with ruby's constant lookup if the top level namespace is not actually modified. Libraries may expect that their constants names are available in the top level namespace:

# foo_a.rb
class Foo
    # or Foo
    # or Object.const_get(:Foo)

# foo_b.rb
class Foo

# main
namespace :Bar do
  import :Foo, as: :Foo_A, from: 'foo_a'
  import :Foo, as: :Foo_B, from: 'foo_b'

I suppose it is possible, but CREF handling in the VM would have to be made significantly more complex to implement it correctly (so that the code works the same both via require and import :as).

You can sort of get what you want in terms of an isolated namespace with BasicObject subclasses:

class Foo

class Hello < BasicObject

class Hello
  Foo # NameError

Hello::Foo = ::Foo

class Hello
  Foo # no NameError

module A
  module B
    module C
      module D

Hello::E = ::A::B::C::D

def; 1; end
Object.send(:remove_const, :Foo)

require 'foo_a'
# in foo_a.rb
# class Foo; end
# def; 2; end
Hello::Foo_A = Object.send(:remove_const, :Foo)

require 'foo_b'
# in foo_b.rb
# class Foo; end
# def; 3; end
Hello::Foo_B = Object.send(:remove_const, :Foo)

class Hello
  class Bar
    def xxx
      [Foo, Foo_A, Foo_B].map{|s| [s,]}
# [[Foo, 1], [Foo, 2], [Foo, 3]]

Such an approach certainly has it's own issues, though. Also, it doesn't really address the issue of trying to handle overlapping top level constants in separate libraries.

I see the benefits of this proposal, as require into isolated binding is a nice to have assuming everything continues to work. However, I don't think the benefits of this proposal would exceed the implementation and maintenance cost.

#9 [ruby-core:88461] Updated by jjyr (Jinyang Jiang) 2 days ago

shyouhei (Shyouhei Urabe) wrote:

I like this idea in general. I too want to have "requiring into isolated Binding".

One thing I would like to add, "namespace" shall be a keyword rather than a normal method taking a block. Blocks can be passed around:

namespace :Foo do
  import :Foo from: 'foo'
    return lamnda do

namespace(:Bar, & # => Error, or ...?

We should forbid this kind of headache.

Totally agreed!

#10 [ruby-core:88462] Updated by jjyr (Jinyang Jiang) 2 days ago

I can imagine how hard to implement the "requiring into isolated binding" correct in VM.

If we want to maintain the compatible we need to maintain the reference of the name cross different require/import.

Think of a situation:

# foo.rb
class Foo
  @count = 0
  def self.count
    @count += 1

# another_file.rb
namespace do
  import :Foo, from: 'foo'
  Foo.count # should return 1

require 'foo', into: :FooSpace
FooSpace::Foo.count # should return 2

# require 'foo'

Foo.count # should return 3

The name Foo from path 'foo' should always be the same reference whether how the name imported.
And imaging in case we have path 'foo/bar' and 'foo' require 'foo/bar', the name from 'foo/bar' should also be the same reference across files.

It's mean each time we require or require into a file, we put it in a Binding associated with the file path. (so we can think the requiring in ruby VM is like to find a file and Binding by absolute path)

In VM level, we may also need to implement a nested Binding structure to support this feature: find a name from each required bindings. (like examples above, RubyVm howto lookup the name Foo?)

So it is hard to support this feature, and maybe cause VM performance issue on name lookup.

Also available in: Atom PDF