Naught is a toolkit for building Null Objects in Ruby.
Null Objects can make your code more confident.
Here's a method that's not very sure of itself.
class Geordi
def make_it_so(logger=nil)
logger && logger.info("Reversing the flux phase capacitance!")
logger && logger.info("Bounding a tachyon particle beam off of Data's cat!")
logger && logger.warn("Warning, bogon levels are rising!")
end
end
Now, observe as we give it a dash of confidence with the Null Object pattern!
class NullLogger
def debug(*); end
def info(*); end
def warn(*); end
def error(*); end
def fatal(*); end
end
class Geordi
def make_it_so(logger=NullLogger.new)
logger.info "Reversing the flux phase capacitance!"
logger.info "Bounding a tachyon particle beam off of Data's cat!"
logger.warn "Warning, bogon levels are rising!"
end
end
By providing a NullLogger
which implements [some of] the Logger
interface as no-op methods, we've gotten rid of those unsightly &&
operators.
You don't! The Null Object pattern is a very simple one at its core.
Yes. While you don't need a Null Object library, this one offers some conveniences you probably won't find elsewhere.
But there's an even more important reason I wrote this library. In the immortal last words of James T. Kirk: "It was… fun!"
Well, what would you like to do?
Sure thing!
require 'naught'
NullObject = Naught.build
null = NullObject.new
null.foo # => nil
null.bar # => nil
Naught is a toolkit for building null object classes. It is not a one-size-fits-all solution.
What else can I make for you?
OK.
require 'naught'
BlackHole = Naught.build do |config|
config.black_hole
end
null = BlackHole.new
null.foo # => <null>
null.foo.bar.baz # => <null>
null << "hello" << "world" # => <null>
That's what you use to customize the generated class to your liking. Internally, Naught uses the Builder Pattern to make this work..
Whatever. What if I want a null object that has conversions to Integer, String, etc. using sensible conversions to "zero values"?
We can do that.
require 'naught'
NullObject = Naught.build do |config|
config.define_explicit_conversions
end
null = NullObject.new
null.to_s # => ""
null.to_i # => 0
null.to_f # => 0.0
null.to_a # => []
null.to_h # => {}
null.to_c # => (0+0i)
null.to_r # => (0/1)
Ah, but what about implicit conversions such as #to_str
? Like what if I want a null object that implicitly splats the same way as an empty array?
Gotcha covered.
require 'naught'
NullObject = Naught.build do |config|
config.define_implicit_conversions
end
null = NullObject.new
null.to_str # => ""
null.to_ary # => []
a, b, c = []
a # => nil
b # => nil
c # => nil
x, y, z = null
x # => nil
y # => nil
z # => nil
That's what mimic
is for.
require 'naught'
NullIO = Naught.build do |config|
config.mimic IO
end
null_io = NullIO.new
null_io << "foo" # => nil
null_io.readline # => nil
null_io.foobar # =>
# ~> -:11:in `<main>': undefined method `foobar' for
# <null:IO>:NullIO (NoMethodError)
There is also impersonate
which takes mimic
one step further. The
generated null class will be derived from the impersonated class. This
is handy when refitting legacy code that contains type checks.
require 'naught'
NullIO = Naught.build do |config|
config.impersonate IO
end
null_io = NullIO.new
IO === null_io # => true
case null_io
when IO
puts "Yep, checks out!"
null_io << "some output"
else
raise "Hey, I expected an IO!"
end
# >> Yep, checks out!
What about predicate methods? You know, the ones that end with question marks? Shouldn't they return false
instead of nil
?
Sure, if you'd like.
require 'naught'
NullObject = Naught.build do |config|
config.predicates_return false
end
null = NullObject.new
null.foo # => nil
null.bar? # => false
null.nil? # => false
Not a problem, just define them in the .build
block.
require 'naught'
NullObject = Naught.build do |config|
config.define_explicit_conversions
config.predicates_return false
def to_path
"/dev/null"
end
# You can override methods generated by Naught
def to_s
"NOTHING TO SEE HERE MOVE ALONG"
end
def nil?
true
end
end
null = NullObject.new
null.to_path # => "/dev/null"
null.to_s # => "NOTHING TO SEE HERE MOVE ALONG"
null.nil? # => true
Well, we can make the null class a singleton, since null objects generally have no state.
require 'naught'
NullObject = Naught.build do |config|
config.singleton
end
null = NullObject.instance
null.__id__ # => 17844080
NullObject.instance.__id__ # => 17844080
NullObject.new # =>
# ~> -:11:in `<main>': private method `new' called for
# NullObject:Class (NoMethodError)
Speaking of null objects with state, we can also enable tracing. This is
handy for playing "where'd that null come from?!" Try doing that with
nil
!
require 'naught'
NullObject = Naught.build do |config|
config.traceable
end
null = NullObject.new # line 7
null.__file__ # => "example.rb"
null.__line__ # => 7
We can even conditionally enable either singleton mode (for production)
or tracing (for development). Here's an example of using the $DEBUG
global variable (set with the -d
option to ruby) to choose which one.
require 'naught'
NullObject = Naught.build do |config|
if $DEBUG
config.traceable
else
config.singleton
end
end
The only caveat is that when swapping between singleton and
non-singleton implementations, you should be careful to always
instantiate your null objects with NullObject.get
, not .new
or
.instance
. .get
will work whether the class is implemented as a
singleton or not.
NullObject.get # => <null>
Naught can make a null object behave as a pebble object.
require 'naught'
NullObject = Naught.build do |config|
if $DEBUG
config.pebble
else
config.black_hole
end
end
Now you can pass the pebble object to your code and see which messages are sent to the pebble.
null = NullObject.new
class MyConsumer < Struct.new(:producer)
def consume
producer.produce
end
end
MyConsumer.new(null).consume
# >> produce() from consume
# => <null>
Just one more thing. For maximum convenience, Naught-generated null classes also come with a full suite of conversion functions which can be included into your classes.
require 'naught'
NullObject = Naught.build
include NullObject::Conversions
# Convert nil to null objects. Everything else passes through.
Maybe(42) # => 42
Maybe(nil) # => <null>
Maybe(NullObject.get) # => <null>
Maybe{ 42 } # => 42
# Insist on a non-null (or nil) value
Just(42) # => 42
Just(nil) rescue $! # => #<ArgumentError: Null value: nil>
Just(NullObject.get) rescue $! # => #<ArgumentError: Null value: <null>>
# nils and nulls become nulls. Everything else is rejected.
Null() # => <null>
Null(42) rescue $! # => #<ArgumentError: 42 is not null!>
Null(nil) # => <null>
Null(NullObject.get) # => <null>
# Convert nulls back to nils. Everything else passes through. Useful
# for preventing null objects from "leaking" into public API return
# values.
Actual(42) # => 42
Actual(nil) # => nil
Actual(NullObject.get) # => nil
Actual { 42 } # => 42
gem install naught
- Ruby
- Fork, branch, submit PR, blah blah blah. Don't forget tests.
Naught is by Avdi Grimm.
This isn't the first Ruby Null Object library. Others to check out include:
- Null Object: Something for Nothing (PDF) by Kevlin Henney
- The Null Object Pattern (PS) by Bobby Woolf
- NullObject on WikiWiki
- Null Object pattern on Wikipedia
- Null Objects and Falsiness, by Avdi Grimm