Ruby Hash initialize confusion - ruby

I have two classes.
class Sky
attr_accessor :args
def initialize(args)
#args = args
puts 'Initializing sky'
end
end
class ShadowMask
attr_accessor :sky
def initialize(args)
args.each{|k, v| p "#{k}: #{v.to_s}"}
#sky = args.fetch(:sky, Sky.new({}))
end
end
ShadowMask can be created either with a default Sky:
sm_default = ShadowMask.new({})
# Initializing sky
# => #<ShadowMask:0x007fa215230eb0 #sky=#<Sky:0x007fa215230e60 #args={}>>
sm_default.sky
# => #<Sky:0x007fa215230e60 #args={}>
sm_default.sky.args
# => {}
or with a Sky that has been previously created:
skyobj = Sky.new("Sky Object")
# Initializing sky
# => #<Sky:0x007fa21481a020 #args="Sky Object">
sm = ShadowMask.new(:sky => skyobj)
# "sky: #<Sky:0x007fa21481a020>"
# Initializing sky
# => #<ShadowMask:0x007fa21521ae80 #sky=#<Sky:0x007fa21481a020 #args="Sky Object">>
In this second case, the instance of Sky already exists, and I do not want to see the output Initializing sky from the Sky initialization.
The problem with my actual code is that
puts 'Initializing sky'
is a call to a method that performs several calculations to complete the initialization and set several attributes. And this is repeated without need every time a ShadowMask is created.
Interestingly, if I replace
#sky = args.fetch(:sky, Sky.new({}))
with something like
#sky = args.fetch(:sky, 'AnyString')
it works fine, but I would loose the possibility of creating a new Sky if needed.
I am not sure if the problem is in the syntax or I am making a conceptual mistake.

I think you need to pass block to fetch in order to not see Initializing sky:
#sky = args.fetch(:sky) {Sky.new({})}
the idea behind this is that when you call any method, initially its params (in this case Sky.new({})) will be invoked. When you pass block - it will be invoked after inside method fetch, not before.

If all you want is to provide a default Sky object when key :sky is omitted, then fetch is a sub-optimal choice. This will work better:
#sky = args[:sky] || Sky.new({})
The "problem" with fetch is that this will result in a nil sm.sky:
sm = ShadowMask.new(sky: nil)
If this is desired behaviour for you, then use fetch. If not, use ||.

Related

How do I tack a string onto a variable and evaluated the entire thing as a variable in Ruby?

I have the following Ruby code:
module BigTime
FOO1_MONEY_PIT = 500
FOO2_MONEY_PIT = 501
class LoseMoney
##SiteName = 'FOO1'
#site_num = ##SiteName_MONEY_PIT
def other_unimportant_stuff
whatever
end
end
end
So, what I'm trying to do here is set the SiteName and then use SiteName and combine it with the string _MONEY_PIT so I can access FOO1_MONEY_PIT and store its contents (500 in this case) in #site_num. Of course, the above code doesn't work, but there must be a way I can do this?
Thanks!!
If you want to dynamically get the value of a constant, you can use Module#const_get:
module BigTime
FOO1_MONEY_PIT = 500
FOO2_MONEY_PIT = 501
class LoseMoney
##SiteName = 'FOO1'
#site_num = BigTime.const_get(:"#{##SiteName}_MONEY_PIT")
end
end
Do not, under any circumstance, use Kernel#eval for this. Kernel#eval is extremely dangerous in any context where there is even the slightest possibility that an attacker may be able to control parts of the argument.
For example, if a user can choose the name of the site, and they name their site require 'fileutils'; FileUtils.rm_rf('/'), then Ruby will happily evaluate that code, just like you told it to!
Kernel#eval is very dangerous and you should not get into the habit of just throwing an eval at a problem. It is a very specialized tool that should only be employed when there is no other option (spoiler alert: there almost always is another option), and only after a thorough security review.
Please note that dynamically constructing variable names is already a code smell by itself, regardless of whether you use eval or not. It pretty much always points to a design flaw somewhere. In general, you can almost guaranteed replace the multiple variables with a data structure. E.g. in this case something like this:
module BigTime
MONEY_PITS = {
'FOO1' => 500,
'FOO2' => 501,
}.freeze
class LoseMoney
##SiteName = 'FOO1'
#site_num = MONEY_PITS[##SiteName]
end
end
You can refactor this as to use a Hash for your name lookups, and a getter method to retrieve it for easy testing/validation. For example:
module BigTime
MONEY_PITS = { FOO1: 500, FOO2: 501 }
MONEY_PIT_SUFFIX = '_MONEY_PIT'
class LoseMoney
##site = :FOO1
def initialize
site_name
end
def site_name
#site_name ||= '%d%s' % [MONEY_PITS[##site], MONEY_PIT_SUFFIX]
end
end
end
BigTime::LoseMoney.new.site_name
#=> "500_MONEY_PIT"

Assignment method created using define_singleton_method returns the wrong value

Background
The Entity class is a base class that gets inherited by several subclasses that holds entities received over a REST API. The entity classes are immutable and should return a new instance of themselves whenever a change is attempted.
The Entity class has an .update() method that takes a hash of values to update, if the changes aren't really changes it returns itself and if there are real changes it returns a new instance of itself with the changes effected before instantiation.
To be user friendly Entity also allows for direct assignment to properties (so that if a subclass of Entity has a name attribute you can do instance.name = 'New Name') that also returns a new instance of the class. This is implemented in terms of update using dynamic methods that are created when the class is instantiated.
And they are the problem.
Problem
The code in the Entity class looks, in part, like this (for a complete code listing and tests check out the Github repo: https://github.com/my-codeworks/fortnox-api.git):
require "virtus"
require "ice_nine"
class Entity
extend Forwardable
include Virtus.model
def initialize( hash = {} )
super
create_attribute_setter_methods
IceNine.deep_freeze( self )
end
def update( hash )
attributes = self.to_hash.merge( hash )
return self if attributes == self.to_hash
self.class.new( attributes )
end
private
def create_attribute_setter_methods
attribute_set.each do |attribute|
name = attribute.options[ :name ]
create_attribute_setter_method( name )
end
end
def create_attribute_setter_method( name )
self.define_singleton_method "#{name}=" do | value |
self.update( name => value )
end
end
end
Doing this:
instance.update( name: 'New Name' )
and this:
instance.name = 'New Name'
Should be the same, literally since one is implemented in terms of the other.
While .update() works perfectly the .attr=() methods return the value you assign.
So in the above example .update() returns a new instance of the Entity subclass but .attr=() returns 'New Name' ...
I have tries capturing the output inside the .attr=() method and log it before returning so that I have this:
self.define_singleton_method "#{name}=" do | value |
p "Called as :#{name}=, redirecting to update( #{name}: #{value} )"
r = self.update( name => value )
p "Got #{r} back from update"
return r
end
And the log lines say:
"Called as :name=, redirecting to update( name: 'New Name' )"
"Got #<TestEntity:0x007ffedbd0ad18> back from update"
But all I get is the string 'New Name'...
My forehead is bloody and no posts I find show anything close to this. I bet I'm doing something wrong but I can't find it.
Getting dirty
The Github repo has tests in rspec that you can run, the failing ones are focused right now and some extra logging is in the Entity class to capture the different internal steps.
Comments, links and/or pull requests are welcome.
Turns out that the = methods always return the value being assigned.
o = Struct.new(:key).new(1)
o.define_singleton_method("something") { #something }
o.define_singleton_method("something=") do |v|
#something = v
return 6
end
As you can see, I've 'fixed' the return value to 6 each time something= is called. Let's see if it works:
o.something = 1 #=> outputs 1, not 6
o.something #=> outputs 1, so the method did indeed run
Conclusion? My guess is that an = method will return the value that you are assigning through it. And IMO it's better this way; one reason would be to ensure proper functioning of assignment chains:
new_val = o.something = some_val

Ruby method calls without defining variables

I am beyond confused on where the :find is coming from line 17, as well as :findcity... is that how you call a fucntion within a predefined method call from ruby???
cities = {'CA' => 'San Francisco',
'MI' => 'Detroit',
'FL' => 'Jacksonville'}
cities['NY'] = 'New York'
cities['OR'] = 'Portland'
def find_city(map, state)
if map.include? state
return map[state]
else
return "Not found."
end
end
# ok pay attention!
cities[:find] = method(:find_city)
while true
print "State? (ENTER to quit) "
state = gets.chomp
break if state.empty?
# this line is the most important ever! study!
puts cities[:find].call(cities, state)
end
For starters if you are a beginner in Ruby just don't bother trying to understand it. This is not the usual way of doing things in Ruby.
But here are some explanations:
:find is a Symbol and it could be :search or something else in this example.
You could actually use a different variable to store the method instead of storing inside the cities Hash. Like so:
# Instead of doing this
hash = {} # => {}
hash[:puts_method] = method(:puts)
hash[:puts_method].call("Foo")
# Foo
# You can just
puts_method = method(:puts)
puts_method.call("Foo")
# Foo
The find_city is the method defined in your code. Passing the symbol :find_city to the method method returns you an object representing that method (very meta uh?) of the class Method.
So like in the example above we can have an object representing the method puts with which we can send the method call to call it.
the_puts = method(:puts)
# => #<Method: Object(Kernel)#puts>
the_puts.call("Hey!")
# Hey!
# => nil
# Which is the same as simply doing
puts("Hey!")
# Hey!
# => nil

Testing a lambda

I am creating an import feature that imports CSV files into several tables. I made a module called CsvParser which parses a CSV file and creates records. My models that receive the create actions extends theCsvParser. They make a call to CsvParser.create and pass the correct attribute order and an optional lambda called value_parser. This lambda transforms values in a hash to a preffered format.
class Mutation < ActiveRecord::Base
extend CsvParser
def self.import_csv(csv_file)
attribute_order = %w[reg_nr receipt_date reference_number book_date is_credit sum balance description]
value_parser = lambda do |h|
h["is_credit"] = ((h["is_credit"] == 'B') if h["is_credit"].present?)
h["sum"] = -1 * h["sum"].to_f unless h["is_credit"]
return [h]
end
CsvParser.create(csv_file, self, attribute_order, value_parser)
end
end
The reason that I'm using a lambda instead of checks inside the CsvParser.create method is because the lambda is like a business rule that belongs to this model.
My question is how i should test this lambda. Should i test it in the model or the CsvParser? Should i test the lambda itself or the result of an array of the self.import method? Maybe i should make another code structure?
My CsvParser looks as follows:
require "csv"
module CsvParser
def self.create(csv_file, klass, attribute_order, value_parser = nil)
parsed_csv = CSV.parse(csv_file, col_sep: "|")
records = []
ActiveRecord::Base.transaction do
parsed_csv.each do |row|
record = Hash.new {|h, k| h[k] = []}
row.each_with_index do |value, index|
record[attribute_order[index]] = value
end
if value_parser.blank?
records << klass.create(record)
else
value_parser.call(record).each do |parsed_record|
records << klass.create(parsed_record)
end
end
end
end
return records
end
end
I'm testing the module itself:
require 'spec_helper'
describe CsvParser do
it "should create relations" do
file = File.new(Rails.root.join('spec/fixtures/files/importrelaties.txt'))
Relation.should_receive(:create).at_least(:once)
Relation.import_csv(file).should be_kind_of Array
end
it "should create mutations" do
file = File.new(Rails.root.join('spec/fixtures/files/importmutaties.txt'))
Mutation.should_receive(:create).at_least(:once)
Mutation.import_csv(file).should be_kind_of Array
end
it "should create strategies" do
file = File.new(Rails.root.join('spec/fixtures/files/importplan.txt'))
Strategy.should_receive(:create).at_least(:once)
Strategy.import_csv(file).should be_kind_of Array
end
it "should create reservations" do
file = File.new(Rails.root.join('spec/fixtures/files/importreservering.txt'))
Reservation.should_receive(:create).at_least(:once)
Reservation.import_csv(file).should be_kind_of Array
end
end
Some interesting questions. A couple of notes:
You probably shouldn't have a return within the lambda. Just make the last statement [h].
If I understand the code correctly, the first and second lines of your lambda are overcomplicated. Reduce them to make them more readable and easier to refactor:
h["is_credit"] = (h['is_credit'] == 'B') # I *think* that will do the same
h['sum'] = h['sum'].to_f # Your original code would have left this a string
h['sum'] *= -1 unless h['is_credit']
It looks like your lambda doesn't depend on anything external (aside from h), so I would test it separately. You could even make it a constant:
class Mutation < ActiveRecord::Base
extend CsvParser # <== See point 5 below
PARSE_CREDIT_AND_SUM = lambda do |h|
h["is_credit"] = (h['is_credit'] == 'B')
h['sum'] = h['sum'].to_f
h['sum'] *= -1 unless h['is_credit']
[h]
end
Without knowing the rationale, it's hard to say where you should put this code. My gut instinct is that it is not the job of the CSV parser (although a good parser may detect floating point numbers and convert them from strings?) Keep your CSV parser reusable. (Note: Re-reading, I think you've answered this question yourself - it is business logic, tied to the model. Go with your gut!)
Lastly, you are defining and the method CsvParser.create. You don't need to extend CsvParser to get access to it, although if you have other facilities in CsvParser, consider making CsvParser.create a normal module method called something like create_from_csv_file

How to execute some procedure in all ruby files in current directory?

I'm new to Ruby - I'm having troubles on every step...
Imagine a Ruby script main.rb and a lot of unknown script files script1.rb ... scriptN.rb.
Each scriptX.rb contains unique module with one procedure needs to be executed:
Module X
def some_procedure(i)
puts "{#i} Module X procedure executed successfully!"
end
end
All I need is to:
iterate over all files in current directory
if current file has name like /^script.*?\.rb$/
then load it and execute some_procedure
How can I do it in main.rb ?
Thank you in advance!
Choose from these great answers in SO on loading the files: Best way to require all files from a directory in ruby?
Then in your files, just have them execute on load, rather than on a method call.
The problem might be that, when a file is required, it doesn't return the list of modules (or, in general, constants) which it defines. So, unless you don't know which module a script has defined, you will not know where to pass your some_procedure message.
As a workaround, you may try getting the list of defined constants before and after the script was required, find a difference, i.e. list of constants during require, and iterate through all of them, checking which one implements the method you need.
First, we need to put some restriction:
Every file script_my1.rb will have the module named Script_my1. I.e. first letter capitalized, all other letters - lowercase.
Create two files script_my1.rb and script_my2.rb as follows:
---script_my1.rb:
module Script_my1
#value = 0
def self.some_procedure(i)
puts "#{i} my1 executed!"
#value = i
end
def self.another_procedure()
return #value
end
end
---script_my2.rb:
module Script_my2
#value = 0
def self.some_procedure(i)
puts "#{i} my2 executed!"
#value = i
end
def self.another_procedure()
return #value
end
end
Now the main script, that loads and executes some_procedure() in each module, and then another_procedure().
Please notice, that each module can have separated variables with the same name #value.
Moreover, I think every module can be executed in a separate thread and have access to global variables, but I have not tested it yet.
---main.rb:
# Load all files from the current directory
# with name like script_xxx.rb
i = 1
result = nil
Dir['./script_*.rb'].each { |f|
next if File.directory?(f)
require (f)
moduleName = f[2,f.length].rpartition('.rb')[0].capitalize
eval ( "#{moduleName}.some_procedure(%d)" % i )
eval ( "result = #{moduleName}.another_procedure()" )
puts result
i = i + 1
}
Output of this program is:
1 my1 executed!
1
2 my2 executed!
2
That is all!
Some improvement to previous solution can be made. If we want to avoid special naming, we can use global hash to store procedure's names. Each loaded script_xx.rb file would register it's own procedures in this global hash.
Please notice, that in this case we make two cycles:
first we load all files script_xx.b
every file while loading will register it's procedures in $global_procs array.
then iterate over all entries in $global_procs to execute all registered procedures via eval()
Hope, this is a more 'ruby-like' solution!
---script_my1.rb
module My1
#value = 0
def self.some_procedure(i)
puts "#{i} my1 executed!"
#value = i
end
def self.another_procedure()
return #value
end
end
$global_procs << { 'module' => 'My1',
'some_procedure' => 'My1.some_procedure',
'another_procedure' => 'My1.another_procedure' }
---script_my2.rb
module MMM2
#value = 0
def self.some_procedure(i)
puts "#{i} MMM2 executed!"
#value = i
end
def self.another_procedure()
return #value
end
end
$global_procs << { 'module' => 'MMM2',
'some_procedure' => 'MMM2.some_procedure',
'another_procedure' => 'MMM2.another_procedure' }
---main.rb
# Create global array for holding module's info
$global_procs = []
Dir['./script_*.rb'].each { |f|
next if File.directory?(f)
require (f)
}
i = 1
result = nil
$global_procs.each { |p|
puts "Module name: " + p['module']
eval(p['some_procedure']+'(i)')
result = eval(p['another_procedure']+'()')
puts result
i = i + 1
}

Resources