:toc: macro :toclevels: 5 :figure-caption!:

Refinements

These refinements augment and enhance Ruby primitives so you can avoid https://alchemists.io/articles/ruby_antipatterns/#_monkey_patches[monkey patches]. They also allow you to develop clean and concise implementations while using less code. By refining your code, you can acquire the functionality you wish the core primitives had!

toc::[]

Features

Enhances the following objects:

Requirements

. www.ruby-lang.org[Ruby]. . A solid understanding of https://alchemists.io/articles/ruby_refinements[refinements].

Setup

To install with security, run:

source,bash

# 💡 Skip this line if you already have the public certificate installed. gem cert –add <(curl –compressed –location alchemists.io/gems.pem) gem install refinements –trust-policy HighSecurity


To install without security, run:

source,bash

gem install refinements


You can also add the gem directly to your project:

source,bash

bundle add refinements


Once the gem is installed, you only need to require it:

source,ruby

require “refinements”


Usage

Requires

If no refinements are desired, then add the following to your ‘Gemfile` instead:

source,ruby

gem “refinements”, require: false


…then require the specific refinement, as needed. Example:

source,ruby

require “refinements/array” require “refinements/binding” require “refinements/data” require “refinements/date_time” require “refinements/hash” require “refinements/io” require “refinements/log_device” require “refinements/logger” require “refinements/module” require “refinements/object” require “refinements/pathname” require “refinements/string” require “refinements/string_io” require “refinements/struct” require “refinements/symbol”


Using

Much like including/extending a module, you’ll need to modify your object(s) to use the refinement(s):

source,ruby

class Example

using Refinements::Array
using Refinements::Binding
using Refinements::Data
using Refinements::DateTime
using Refinements::Hash
using Refinements::IO
using Refinements::LogDevice
using Refinements::Logger
using Refinements::Module
using Refinements::Object
using Refinements::Pathname
using Refinements::String
using Refinements::StringIO
using Refinements::Struct
using Refinements::Symbol

end


Examples

The following sections demonstrate how each refinement enriches your objects with new capabilities.

Array

combinatorial?

Answers if an array is equal to another array when the elements are equal but in any order and/or subset.

source,ruby

example = %w[a b c]

example.combinatorial? %w[a b c] # true example.combinatorial? %w[c a b] # true example.combinatorial? %w # true example.combinatorial? %w[c b] # true example.combinatorial? %w # false example.combinatorial? %w[z b c] # false example.combinatorial? %w[a b c d] # false example.combinatorial? [] # false


compress

Removes ‘nil` and empty objects without mutating itself. Answers itself if there is nothing to remove.

source,ruby

object = Object.new example = [1, “blueberry”, nil, “”, [], {}, object]

].compress # [
1, 2].compress # [1, 2

example.compress # [1, “blueberry”, object] example # [1, “blueberry”, nil, “”, [], {}, object]


compress!

Removes ‘nil` and empty values while mutating itself. Answers `nil` if there is nothing to remove.

source,ruby

object = Object.new example = [1, “blueberry”, nil, “”, [], {}, object]

[].compress! # nil [1, 2].compress! # nil example.compress! # [1, “blueberry”, object] example # [1, “blueberry”, object]


excluding

Removes given array or elements without mutating itself.

source,ruby

1, 2, 3, 4, 5].excluding [4, 5

# [1, 2, 3]

1, 2, 3, 4, 5].excluding 4, 5 # [1, 2, 3

filter_find

⚠️ _This is deprecated and will be removed in Version 13.0.0._

Answers the first element which evaluates to true from a filtered collection.

source,ruby

handlers = [

-> object { object if object == :b },
proc { false },
-> object { object if object == :a }

]

handlers.filter_find # Enumerator::Lazy handlers.filter_find { |handler| handler.call :a } # :a handlers.filter_find { |handler| handler.call :x } # nil


including

Adds given array or elements without mutating itself.

source,ruby

1, 2, 3].including [4, 5

# [1, 2, 3, 4, 5]

1, 2, 3].including 4, 5 # [1, 2, 3, 4, 5

intersperse

Inserts additional elements, or an array, between all members of given array.

source,ruby

1, 2, 3].intersperse :a # [1, :a, 2, :a, 3
1, 2, 3].intersperse :a, :b # [1, :a, :b, 2, :a, :b, 3
1, 2, 3].intersperse %i[a b c

# [1, :a, :b, :c, 2, :a, :b, :c, 3]


many?

Answers true if an array has more than one element. Can take a block which evaluates as truthy or falsey.

source,ruby

[1, 2].many? # true [1, 2, 3].many?(&:odd?) # true [1].many? # false [].many? # false


maximum

Answers the maximum extracted value from a collection of objects.

source,ruby

Point = Data.define :x, :y points = [Point[x: 1, y: 2], Point[x: 0, y: 1], Point[x: 2, y: 3]]

points.maximum :x # 2 points.maximum :y # 3 points.maximum :z # undefined method ‘z’ for


mean

Answers mean/average all elements within an array.

source,ruby

[].mean # 0 [5].mean # 5 [1, 2, 3].mean # 2 [1.25, 1.5, 1.75].mean # 1.5


minimum

Answers the minimum extracted value from a collection of objects.

source,ruby

Point = Data.define :x, :y points = [Point[x: 1, y: 2], Point[x: 0, y: 1], Point[x: 2, y: 3]]

points.minimum :x # 0 points.minimum :y # 1 points.minimum :z # undefined method ‘z’ for


pad

Answers new array padded with given value up to a maximum size. Useful in situations where an array needs to be a specific size with padded values.

source,ruby

“a”].pad 0 # [“a”
“a”].pad “-”, 3 # [“a”, “-”, “-”

%w[a b].pad “-”, 3 # [“a”, “b”, “-”]


pick

Answers value of first element that matches given key.

source,ruby

array = [{name: “a”, label: “A”}, {name: “b”, label: “B”}, {name: “c”, label: “C”}]

array.pick :name # “a” array.pick :name, :label # [“a”, “A”] array.pick # nil [].pick :other # nil


pluck

Answers values of all elements that match given keys.

source,ruby

array = [{name: “a”, label: “A”}, {name: “b”, label: “B”}, {name: “c”, label: “C”}]

array.pluck :name # [“a”, “b”, “c”] array.pluck :name, :label # [[“a”, “A”], [“b”, “B”], [“c”, “C”]] array.pluck # []

].pluck :other # [

replace_at

Answers mutated array where an element – at a specific index – is replaced by single or multiple elements.

source,ruby

%i[a b c].replace_at 0, :x # [:x, :b, :c] %i[a b c].replace_at 1, :x # [:a, :x, :c] %i[a b c].replace_at 1, :x, :y # [:a, :x, :y, :c] %i[a b c].replace_at -1, :x # [:a, :b, :x]


ring

Answers a circular array which can enumerate before, current, after elements.

source,ruby

example = [1, 2, 3] example.ring # “#<Enumerator: …>” example.ring { |(before, current, after)| puts “#{before} #{current} #{after}” }

# [3 1 2] # [1 2 3] # [2 3 1]


supplant

Answers mutated array where first target element found is replaced by single or multiple elements.

source,ruby

%i[a b a].supplant :a, :z # [:z, :b, :a] %i[a b a].supplant :a, :z, :y # [:z, :y, :b, :a] %i[a b a].supplant :a, %i[z y] # [[:z, :y], :b, :a]


supplant_if

Answers mutated array where all target elements are replaced by single or multiple elements.

⚠️ Be aware that this can be an expensive operation on large arrays.

source,ruby

%i[a b a].supplant_if :a, :z # [:z, :b, :z] %i[a b a].supplant_if :a, :z, :y # [:z, :y, :b, :z, :y] %i[a b a].supplant_if :a, %i[z y] # [[:z, :y], :b, [:z, :y]]


to_sentence

Answers a sentence using ‘“and”` as the default conjunction and `“, ”` as the default delimiter. Useful when building documentation, answering human readable error messages, etc.

source,ruby

[].to_sentence # “” [“demo”].to_sentence # “demo” [“a”, :b].to_sentence # “a and b” [1, “a”, :b, 2.0, /w+/].to_sentence # “1, a, b, 2.0, and (?-mix:\w+)” %w[one two three].to_sentence # “one, two, and three” %w[eins zwei drei].to_sentence “und”, delimiter: “ ” # “eins zwei und drei”


💡 You can use a string or a symbol for the conjunction (i.e. ‘“and”` or `:and`).

to_usage

Further enhances ‘#to_sentence` by answering a sentence where all elements are inspected (i.e. `#inspect`) before turned into a sentence using `“and”` as the default conjunction and `“, ”` as the default delimiter. This is useful when providing detailed error messages and you need the types of all elements preserved.

source,ruby

[].to_usage # “” [“demo”].to_usage # “"demo"” [“a”, :b].to_usage # “"a" and :b” [1, “a”, :b, 2.0, /w+/].to_usage # “1, "a", :b, 2.0, and /\w+/” %w[one two three].to_usage # “"one", "two", and "three"” %w[eins zwei drei].to_usage “und”, delimiter: “ ” # “"eins" "zwei" und "drei"”


💡 You can use a string or a symbol for the conjunction (i.e. ‘“and”` or `:and`).

Binding

[]

Allows you to obtain a local variable. This is an alias to ‘#local_variable_get`.

source,ruby

a = 1 binding # 1 binding # ‘bogus’ is not defined (NameError)


[]=

Allows you to set a local variable. This is an alias to ‘#local_variable_set`.

source,ruby

a = 1 binding = 5 binding = “bad”

binding # 5 binding # # ‘bogus’ is not defined (NameError)


local?

Allows you to check if local variable is defined. This is an alias to ‘#local_variable_defined?`.

source,ruby

a = 1

binding.local? :a # true binding.local? :b # false


locals

Allows you to acquire all locally defined variables. This is an alias to ‘#local_variables`.

source,ruby

binding.locals # []

a = 1 b = 2

binding.locals # [:a, :b]


Data

diff

Allows you to obtain the differences between two objects.

source,ruby

implementation = Data.define :a, :b, :c

one = implementation.new a: 1, b: 2, c: 3 two = implementation.new a: 3, b: 2, c: 1 three = Data.define(:x, :y).new x: 1, y: 2

one.diff one # {} one.diff two # {:a=>[1, 3], :c=>[3, 1]} one.diff three # {:a=>[1, nil], :b=>[2, nil], :c=>[3, nil]} one.diff Object.new # {:a=>[1, nil], :b=>[2, nil], :c=>[3, nil]}


Any object that _is not_ the same type will have a ‘nil` value as shown in the last two examples.

DateTime

.utc

Answers new DateTime object for current UTC date/time.

source,ruby

DateTime.utc # “#<DateTime: 2019-12-31T18:17:00+00:00 ((2458849j,65820s,181867000n),+0s,2299161j)>”


Hash

.infinite

Answers new hash where missing keys, even deeply nested, answer an empty hash.

source,ruby

example = Hash.infinite example # {} example[:b] # {}


.with_default

Answers new hash where every top-level missing key has the same default value.

source,ruby

example = Hash.with_default “” example # “”

example = Hash.with_default [] example # []


compress

Removes ‘nil` and empty objects without mutating itself. Answers itself if nothing to remove.

source,ruby

object = Object.new example = {a: 1, b: “blueberry”, c: nil, d: “”, e: [], f: {}, g: object}

{}.compress # {} {a: 1, b: 2}.compress # {a: 1, b: 2} example.compress # {a: 1, b: “blueberry”, g: object} example # {a: 1, b: “blueberry”, c: nil, d: “”, e: [], f: {}, g: object}


compress!

Removes ‘nil` and empty objects while mutating itself. Answers `nil` if nothing to remove.

source,ruby

object = Object.new example = {a: 1, b: “blueberry”, c: nil, d: “”, e: [], f: {}, g: object}

{}.compress! # nil {a: 1, b: 2}.compress! # nil example.compress! # {a: 1, b: “blueberry”, g: object} example # {a: 1, b: “blueberry”, g: object}


deep_merge

Merges deeply nested hashes together without mutating itself.

source,ruby

example = {a: “A”, b: {one: “One”, two: “Two”}}

example.deep_merge b: {one: 1} # {a: “A”, b: {one: 1, two: “Two”}} example # {a: “A”, b: {one: “One”, two: “Two”}}


deep_merge!

Merges deeply nested hashes together while mutating itself.

source,ruby

example = {a: “A”, b: {one: “One”, two: “Two”}}

example.deep_merge! b: {one: 1} # {a: “A”, b: {one: 1, two: “Two”}} example # {a: “A”, b: {one: 1, two: “Two”}}


deep_stringify_keys

Answers string keys of a nested hash without mutating itself. Does not handle nested arrays, though.

source,ruby

example = {a: {b: 2}} example.deep_stringify_keys # {“a” => {“b” => 1}} example # {a: {b: 2}}


deep_stringify_keys!

Answers string keys of nested hash while mutating itself. Does not handle nested arrays, though.

source,ruby

example = {a: {b: 2}} example.deep_stringify_keys! # {“a” => {“b” => 1}} example # {“a” => {“b” => 1}}


deep_symbolize_keys

Symbolizes keys of nested hash without mutating itself. Does not handle nested arrays, though.

source,ruby

example = {“a” => {“b” => 2}} example.deep_symbolize_keys # {a: {b: 1}} example # {“a” => {“b” => 2}}


deep_symbolize_keys!

Symbolizes keys of nested hash while mutating itself. Does not handle nested arrays, though.

source,ruby

example = {“a” => {“b” => 2}} example.deep_symbolize_keys! # {a: {b: 1}} example # {a: {b: 1}}


diff

Allows you to obtain the differences between two objects.

source,ruby

one = {a: 1, b: 2, c: 3} two = {a: 3, b: 2, c: 1} three = {c: 3, b: 2, a: 1} four = Data.define(:x, :y).new x: 1, y: 2

one.diff one # {} one.diff two # {:a=>[1, 3], :c=>[3, 1]} one.diff three # {} one.diff four # {:a=>[1, nil], :b=>[2, nil], :c=>[3, nil]} one.diff Object.new # {:a=>[1, nil], :b=>[2, nil], :c=>[3, nil]}


Any object that _is not_ the same type will have a ‘nil` value as shown in the last two examples. Two hashes with the same keys but defined in different order behave as if they had the same key order.

fetch_value

Fetches value for exiting or missing key. Behavior is identical to ‘#fetch` except when the value of the key is `nil` you’ll get the default value instead. This eliminates the need for using an or expression: ‘example.fetch(:desired_key) || “default_value”`.

source,ruby

{a: “demo”}.fetch_value :a, “default” # “demo” {a: “demo”}.fetch_value :a # “demo” {a: nil}.fetch_value :a, “default” # “default” {a: nil}.fetch_value(:a) { “default” } # “default” {}.fetch_value :a # KeyError {}.fetch_value(:a) { “default” } # “default” {a: “demo”}.fetch_value # ArgumentError


flatten_keys

Flattens nested keys as top-level keys without mutating itself. Keys are converted to symbols. Does not handle nested arrays.

source,ruby

{a: {b: 1}}.flatten_keys prefix: :demo # {demo_a_b: 1} {a: {b: 1}}.flatten_keys delimiter: :| # {:“a|b” => 1}

example = {a: {b: 1}} example.flatten_keys # {a_b: 1} example # {a: {b: 1}}


flatten_keys!

Flattens nested keys as top-level keys while mutating itself. Keys are converted to symbols. Does not handle nested arrays.

source,ruby

{a: {b: 1}}.flatten_keys! prefix: :demo # {demo_a_b: 1} {a: {b: 1}}.flatten_keys! delimiter: :| # {:“a|b” => 1}

example = {a: {b: 1}} example.flatten_keys! # {a_b: 1} example # {a_b: 1}


many?

Answers true if a hash has more than one element. Can take a block which evaluates as truthy or falsey.

source,ruby

{a: 1, b: 2}.many? # true {a: 1, b: 2, c: 2}.many? { |_key, value| value == 2 } # true {a: 1}.many? # false {}.many? # false


recurse

Recursively iterates over the hash and any hash value by applying the given block to it. Does not handle nested arrays, though.

source,ruby

example = {“a” => {“b” => 1}} example.recurse(&:symbolize_keys) # {a: {b: 1}} example.recurse(&:invert) # {{“b” => 1} => “a”}


stringify_keys

Converts keys to strings without mutating itself.

source,ruby

example = {a: 1, b: 2} example.stringify_keys # {“a” => 1, “b” => 2} example # {a: 1, b: 2}


stringify_keys!

Converts keys to strings while mutating itself.

source,ruby

example = {a: 1, b: 2} example.stringify_keys! # {“a” => 1, “b” => 2} example # {“a” => 1, “b” => 2}


symbolize_keys

Converts keys to symbols without mutating itself.

source,ruby

example = {“a” => 1, “b” => 2} example.symbolize_keys # {a: 1, b: 2} example # {“a” => 1, “b” => 2}


symbolize_keys!

Converts keys to symbols while mutating itself.

source,ruby

example = {“a” => 1, “b” => 2} example.symbolize_keys! # {a: 1, b: 2} example # {a: 1, b: 2}


transform_value

Transforms a value for the specified key only if the key exists and a block is given. Otherwise, the original hash is answered. Does not mutate itself.

source,ruby

example = {a: 1, b: 2}

example.transform_value :b # {a: 1, b: 2} example.transform_value(:b) { 20 } # {a: 1, b: 20} example.transform_value(:b) { |value| value * 10 } # {a: 1, b: 20} example.transform_value :c # {a: 1, b: 2} example.transform_value(:c) { :bogus } # {a: 1, b: 2}


The original object _is not_ mutated:

source,ruby

example.transform_value(:b) { 20 } # {a: 1, b: 20} example # {a: 1, b: 2}


transform_value!

Transforms a value for the specified key only if the key exists and a block is given. Otherwise, the original hash is answered. Mutates itself.

source,ruby

example = {a: 1, b: 2}

example.transform_value! :b # {a: 1, b: 2} example.transform_value!(:b) { 20 } # {a: 1, b: 20} example.transform_value!(:b) { |value| value * 10 } # {a: 1, b: 20} example.transform_value! :c # {a: 1, b: 2} example.transform_value!(:c) { :bogus } # {a: 1, b: 2}


The original object is mutated:

source,ruby

example.transform_value!(:b) { 20 } # {a: 1, b: 20} example # {a: 1, b: 20}


transform_with

Transforms values of keys using specific operations (i.e. any object that responds to ‘#call`). Does not mutate itself and you can transform multiple values at once:

source,ruby

example = {name: “Jayne Doe”, email: “<jd@example.com>”}

example.transform_with name: -> value { value.delete_suffix “ Doe” },

email: -> value { value.tr "<>", "" }

# {name: “Jayne”, email: “jd@example.com”}


Invalid keys are ignored:

source,ruby

example.transform_with bogus: -> value { value.tr “<>”, “” } # {email: “<jd@example.com>”}


The original object _is not_ mutated:

source,ruby

example # {name: “Jayne Doe”, email: “<jd@example.com>”}


transform_with!

Transforms values of keys using specific operations (i.e. any object that responds to ‘#call`). Mutates itself and you can transform multiple values at once:

source,ruby

example = {name: “Jayne Doe”, email: “<jd@example.com>”}

example.transform_with! name: -> value { value.delete_suffix “ Doe” },

email: -> value { value.tr "<>", "" }

# {name: “Jayne”, email: “jd@example.com”}


Invalid keys are ignored:

source,ruby

example.transform_with! bogus: -> value { value.tr “<>”, “” } # {email: “<jd@example.com>”}


The original object is mutated:

source,ruby

example # {name: “Jayne”, email: “jd@example.com”}


use

Uses the hash’s keys as block arguments where the value of the block argument is equal to the value of the key found within the hash. Works best with hashes that use symbols for keys but falls back to string keys when symbol keys can’t be found.

source,ruby

example = {unit: “221B”, street: “Baker Street”, city: “London”, country: “UK”}

example.use { |unit, street| “#{unit} #{street}” } # “221B Baker Street”


IO

.void

Answers an IO stream which points to ‘/dev/null` in order to ignore any reads or writes to the stream. When given a block, the stream will automatically close upon block exit. When not given a block, you’ll need to close the stream manually.

source,ruby

io = IO.void # “#<IO:fd 20>” io = IO.void { |void| void.write “nevermore” } # “#<IO:(closed)>”


redirect

Redirects current stream to other stream when given a block. Without a block, the original stream is answered instead.

source,ruby

io = IO.new IO.sysopen(Pathname(“demo.txt”).to_s, “w+”) other = IO.new IO.sysopen(Pathname(“other.txt”).to_s, “w+”)

io.redirect other # “#<IO:fd 20>” io.redirect(other) { |stream| stream.write “demo” } # “#<IO:fd 21>”


reread

Answers full stream by rewinding to beginning of stream and reading all content.

source,ruby

io = IO.new IO.sysopen(Pathname(“demo.txt”).to_s, “w+”) io.write “This is a demo.”

io.reread # “This is a demo.” io.reread 4 # “This”

buffer = “”.dup io.reread(buffer:) # “This is a demo.” buffer # “This is a demo.”


squelch

Temporarily ignores any reads/writes for code executed within a block. Answers itself without any arguments or when given a block.

source,ruby

io = IO.new IO.sysopen(Pathname(“test.txt”).to_s, “w+”)

io.squelch # “#<IO:fd 20>” io.squelch { io.write “Test” } # “#<IO:fd 20>” io.reread # “”


LogDevice

reread

Answers previously written content by rewinding to beginning of device.

source,ruby

# With File. device = Logger::LogDevice.new “test.log” device.write “Test.” device.reread # “Test.”

# With StringIO. device = Logger::LogDevice.new StringIO.new device.write “Test.” device.reread # “Test.”

# With STDOUT. device = Logger::LogDevice.new $stdout device.write “Test.” device.reread # “”


Logger

reread

Answers previously written content by rewinding to beginning of log.

source,ruby

# With File. logger = Logger.new “test.log” logger.write “Test.” logger.reread # “Test.”

# With StringIO. logger = Logger.new StringIO.new logger.write “Test.” logger.reread # “Test.”

# With STDOUT. logger = Logger.new $stdout logger.write “Test.” logger.reread # “”


any

Allows you to log any message which is identical in behavior and functionality to the ‘Logger#unknown` method only this requires less typing and better matches the terminology used by the `#unknown` method.

source,ruby

logger = Logger.new STDOUT

logger.any “Test.” # A, [2000-01-10T09:00:00.847428 #44925] ANY – : Test.

logger.any { “Test.” } A, [2000-01-10T09:00:00.330719 #44925] ANY – : Test.


Module

pseudonym

Allows you to set a temporary name for your anonymous ‘Module` (or `Class` since `Class` inherits from `Module`) with a better default than what `#set_temporary_name` provides.

source,ruby

Module.new.pseudonym “demo” # demo-44600 Module.new.pseudonym “demo”, delimiter: “_” # demo_60900 Module.new.pseudonym “demo”, nil # demo- Module.new.pseudonym “demo”, nil, delimiter: nil # demo


The same applies for anonymous classes since classes inherit from modules:

source,ruby

Class.new.pseudonym “demo” # demo-44600 < Object Class.new.pseudonym “demo”, delimiter: “_” # demo_60900 < Object Class.new.pseudonym “demo”, nil # demo- < Object Class.new.pseudonym “demo”, nil, delimiter: nil # demo < Object


💡 While convenient, if you find yourself nullifying the suffix and/or delimiter, you’re better off using ‘#set_temporary_name`.

Object

in?

Allows you to know if ‘self` is included in, or an element of, the target object.

source,ruby

1.in? [1, 2, 3] # true 9.in? [1, 2, 3] # false

“a”.in? %w[a b c] # true “z”.in? %w[a b c] # false

:a.in? %i[a b c] # true :z.in? %i[a b c] # false

:a.in? %i[a b c].to_enum # true :z.in? %i[a b c].to_enum # false

:a.in?({a: 1, b: 2, c: 3}) # true :z.in?({a: 1, b: 2, c: 3}) # false

1.in? 1..5 # true 9.in? 1..5 # false

1.in? Set[1, 2, 3] # true 9.in? Set[1, 2, 3] # false

“a”.in? “abcde” # true “z”.in? “abcde” # false

“z”.in? Object.new # ‘String#include?` must be implemented. (NoMethodError)


to_proc

Allows you to cast any object to a proc.

source,ruby

one = Class.new { def call = :test }

.new

two = Object.new

one.to_proc #


Pathname

Pathname

Enhances the ‘Kernel` conversion function which casts `nil` into a pathname in order to avoid: `TypeError (no implicit conversion of nil into String)`. The pathname remains invalid but at least you have an instance of `Pathname`, which behaves like a _Null Object_, that can be used to construct a valid path.

source,ruby

Pathname nil # Pathname(“”)


.home

Answers user home directory.

source,ruby

Pathname.home # Pathname “/Users/demo”


.make_temp_dir

Wraps ‘Dir.mktmpdir` with the following behavior (see https://rubyapi.org/o/Dir.mktmpdir#method-c-mktmpdir[Dir.mktmpdir] for details):

  • *Without Block* - Answers a newly created Pathname instance which is not automatically cleaned up.

  • *With Block* Yields a Pathname instance, answers result of given block, and automatically cleans up temporary directory after block exits.

The following examples use truncated temporary directories for illustration purposes only. In reality, these paths will be longer depending on which operating system you are using.

source,ruby

Pathname.make_temp_dir # Pathname:/var/folders/T/temp-20200101-16940-r8 Pathname.make_temp_dir prefix: “prefix-” # Pathname:/var/folders/T/prefix-20200101-16940-r8 Pathname.make_temp_dir suffix: “-suffix” # Pathname:/var/folders/T/temp-20200101-16940-r8-suffix Pathname.make_temp_dir prefix: “prefix-”, suffix: “-suffix” # Pathname:/var/folders/T/prefix-20200101-16940-r8-suffix Pathname.make_temp_dir root: “/example” # Pathname:/example/temp-20200101-16940-r8 Pathname.make_temp_dir { “I am a block result” } # “I am a block result” Pathname.make_temp_dir { |path| path.join “sub_dir” } # Pathname:/var/folders/T/temp-20200101-16940-r8/sub_dir


.require_tree

Requires all Ruby files in given root path and corresponding nested tree structure. All files are sorted before being required to ensure consistent behavior. Example:

source,ruby

# Before Dir[File.join(__dir__, “support/shared_contexts/*/.rb”)].sort.each { |path| require path }

# After Pathname.require_tree “#{__dir__}/support/shared_contexts”


The following are further examples of potential usage:

source,ruby

# Requires all files in root directory and below. Pathname.require_tree __dir__

# Requires all files in ‘/test/*/.rb` and below. Pathname.require_tree “/test”

# Requires all files in RSpec shared examples directory structure. Pathname.require_tree SPEC_ROOT.join(“support/shared_examples”)


.root

Answers operating system root path.

source,ruby

Pathname.root # Pathname “/”


change_dir

Wraps ‘Dir.chdir` behavior by changing to directory of current path. See https://rubyapi.org/o/Dir.chdir#method-c-chdir[Dir.chdir] for details.

source,ruby

current = Pathname.pwd # “$HOME/demo” (Present Working Directory) custom = current.join(“test”).make_dir # Pathname “$HOME/demo/test” custom.change_dir # “$HOME/demo/test” (Present Working Directory) current.change_dir # “$HOME/demo” (Present Working Directory) custom.change_dir { “example” } # “example” custom.change_dir { |path| path } # Pathname “$HOME/demo/test” Pathname.pwd # “$HOME/demo” (Present Working Directory)


copy

Copies file from current location to new location while answering itself so it can be chained.

source,ruby

Pathname(“input.txt”).copy Pathname(“output.txt”) # Pathname(“input.txt”)


deep_touch

Has all of the same functionality as the ‘#touch` method while being able to create ancestor directories no matter how deeply nested the file might be.

source,ruby

Pathname(“a/b/c/d.txt”).touch # Pathname(“a/b/c/d.txt”) Pathname(“a/b/c/d.txt”).touch Time.now - 1 # Pathname(“a/b/c/d.txt”)


delete

Deletes file or directory and answers itself so it can be chained.

source,ruby

# When path exists. Pathname(“/example.txt”).touch.delete # Pathname(“/example”)

# When path doesn’t exist. Pathname(“/example.txt”).delete # Errno::ENOENT


delete_prefix

Deletes a path prefix and answers new pathname.

source,ruby

Pathname(“a/path/example-test.rb”).delete_prefix “example-” # Pathname(“a/path/test.rb”) Pathname(“example-test.rb”).delete_prefix “example-” # Pathname(“test.rb”) Pathname(“example-test.rb”).delete_prefix “miss” # Pathname(“example-test.rb”)


delete_suffix

Deletes a path suffix and answers new pathname.

source,ruby

Pathname(“a/path/test-example.rb”).delete_suffix “-example” # Pathname(“a/path/test.rb”) Pathname(“test-example.rb”).delete_suffix “-example” # Pathname(“test.rb”) Pathname(“test-example.rb”).delete_suffix “miss” # Pathname(“test-example.rb”)


directories

Answers all directories or filtered directories for current path.

source,ruby

Pathname(“/example”).directories # [Pathname(“a”), Pathname(“b”)] Pathname(“/example”).directories “a*” # [Pathname(“a”)] Pathname(“/example”).directories flag: File::FNM_DOTMATCH # [Pathname(“..”), Pathname(“.”)]


empty

Empties a directory of children (i.e. folders, nested folders, or files) or clears an existing file of contents. If a directory or file doesn’t exist, it will be created.

source,ruby

directory = Pathname(“test”).make_path file = directory.join(“test.txt”).write(“example”)

file.empty.read # “” directory.empty.children # []


extensions

Answers file extensions as an array.

source,ruby

Pathname(“example.txt.erb”).extensions # [“.txt”, “.erb”]


files

Answers all files or filtered files for current path.

source,ruby

Pathname(“/example”).files # [Pathname(“a.txt”), Pathname(“a.png”)] Pathname(“/example”).files “*.png” # [Pathname(“a.png”)] Pathname(“/example”).files flag: File::FNM_DOTMATCH # [Pathname(“.ruby-version”)]


gsub

Same behavior as ‘String#gsub` but answers a path with patterns replaced with desired substitutes.

source,ruby

Pathname(“/a/path/some/path”).gsub “path”, “test” # Pathname(“/a/test/some/test”)

Pathname(“/%placeholder%/some/%placeholder%”).gsub “%placeholder%”, “test” # Pathname(“/test/some/test”)


make_ancestors

Ensures all ancestor directories are created for a path.

source,ruby

Pathname(“/one/two”).make_ancestors # Pathname(“/one/two”) Pathname(“/one”).exist? # true Pathname(“/one/two”).exist? # false


make_dir

Provides alternative ‘#mkdir` behavior by always answering itself (even when directory exists) and not throwing errors when directory does exist in order to ensure the pathname can be chained.

source,ruby

Pathname(“/one”).make_dir # Pathname(“/one”) Pathname(“/one”).make_dir.make_dir # Pathname(“/one”)


make_path

Provides alternative ‘#mkpath` behavior by always answering itself (even when full path exists) and not throwing errors when directory does exist in order to ensure the pathname can be chained.

source,ruby

Pathname(“/one/two/three”).make_path # Pathname(“/one/two/three”) Pathname(“/one/two/three”).make_path.make_path # Pathname(“/one/two/three”)


name

Answers file name without extension.

source,ruby

Pathname(“example.txt”).name # Pathname(“example”)


puts

Wraps ‘#write` by writing content to file with new line and answering itself. Allows you to more easily swap out a `Pathname` object with similar IO objects who support `#puts`: `IO`, `StringIO`, `File`, `Kernel`, and so forth.

source,ruby

path = Pathname(“test.txt”).touch path.puts “Test.” path.read # “Test.n”

Pathname(“text.txt”).touch.puts(“Test.”).read # “Test.n”


relative_parent

Answers relative path from parent directory. This complements: ‘#relative_path_from`.

source,ruby

Pathname(“/one/two/three”).relative_parent “/one” # Pathname “two”


remove_dir

Provides alternative ‘#rmdir` behavior by always answering itself (even when full path exists) and not throwing errors when directory does exist in order to ensure the pathname can be chained.

source,ruby

Pathname(“/test”).make_dir.remove_dir.exist? # false Pathname(“/test”).remove_dir # Pathname(“/test”) Pathname(“/test”).remove_dir.remove_dir # Pathname(“/test”)


remove_tree

Provides alternative ‘#rmtree` behavior by always answering itself (even when full path exists) and not throwing errors when directory does exist in order to ensure the pathname can be chained.

source,ruby

parent_path = Pathname “/one” child_path = parent_path.join “two”

child_path.make_path parent_path.remove_tree # Pathname “/one” child_path.exist? # false parent_path.exist? # false

child_path.make_path child_path.remove_tree # Pathname “/one/two” child_path.exist? # false parent_path.exist? # true


rewrite

When given a block, it provides the contents of the recently read file for manipulation and immediate writing back to the same file.

source,ruby

Pathname(“/test.txt”).rewrite # Pathname(“/test.txt”) Pathname(“/test.txt”).rewrite { |body| body.sub “[token]”, “example” } # Pathname(“/test.txt”)


touch

Updates access and modification times for an existing path by defaulting to current time. When path doesn’t exist, it will be created as a file.

source,ruby

Pathname(“example”).touch # Pathname(“example”) Pathname(“example”).touch Time.now - 1 # Pathname(“example”) Pathname(“example.txt”).touch # Pathname(“example.txt”) Pathname(“example.txt”).touch Time.now - 1 # Pathname(“example.txt”)


write

Writes to file and answers itself so it can be chained. See ‘IO.write` for details on additional options.

source,ruby

Pathname(“example.txt”).write “test” # Pathname(“example.txt”) Pathname(“example.txt”).write “test”, offset: 1 # Pathname(“example.txt”) Pathname(“example.txt”).write “test”, mode: “a” # Pathname(“example.txt”)


String

blank?

Answers ‘true`/`false` based on whether string is blank, `<space>`, `n`, `t`, and/or `r`.

source,ruby

“ ntr”.blank? # true


camelcase

Answers a camel cased string.

source,ruby

“this_is_an_example”.camelcase # “ThisIsAnExample”


down

Answers string with only first letter down cased.

source,ruby

“EXAMPLE”.down # “eXAMPLE”


first

Answers first character of a string or first set of characters if given a number.

source,ruby

“example”.first # “e” “example”.first 4 # “exam”


indent

Answers indentation (string) which is the result of the multiplier times padding. By default, the multiplier is ‘1` and the padding is `“ ”` which equates to two spaces.

source,ruby

“example”.indent # “ example” “example”.indent 0 # “example” “example”.indent -1 # “example” “example”.indent 2 # “ example” “example”.indent 3, pad: “ ” # “ example”


last

Answers last character of a string or last set of characters if given a number.

source,ruby

“instant”.last # “t” “instant”.last 3 # “ant”


pluralize

Answers plural form of self when given a suffix to add. The plural form of the word can be dynamically calculated when given a count and a replacement pattern (i.e. string or regular expression) can be supplied for further specificity. Usage is based on https://en.wikipedia.org/wiki/English_plurals[plurals in English] which may or may not work well in other languages.

source,ruby

“apple”.pluralize “s” # apples “apple”.pluralize “s”, 0 # apples “apple”.pluralize “s”, 1 # apple “apple”.pluralize “s”, -1 # apple “apple”.pluralize “s”, 2 # apples “apple”.pluralize “s”, -2 # apples “cactus”.pluralize “i”, replace: “us” # cacti “cul-de-sac”.pluralize “ls”, replace: “l” # culs-de-sac


singularize

Answers singular form of self when given a suffix to remove (can be a string or a regular expression). The singular form of the word can be dynamically calculated when given a count and a replacement string can be supplied for further specificity. Usage is based on https://en.wikipedia.org/wiki/English_plurals[plurals in English] which may or may not work well in other languages.

source,ruby

“apples”.singularize “s” # apple “sacks”.singularize /s$/ # sack “apples”.singularize “s”, 0 # apples “apples”.singularize “s”, 1 # apple “apples”.singularize “s”, -1 # apple “apples”.singularize “s”, 2 # apples “apples”.singularize “s”, -2 # apples “cacti”.singularize “i”, replace: “us” # cactus “culs-de-sac”.singularize “ls”, replace: “l” # cul-de-sac


snakecase

Answers a snake cased string.

source,ruby

“ThisIsAnExample”.snakecase # “this_is_an_example”


squish

Removes leading, in body, and trailing whitespace, including any tabs or newlines, without mutating itself. Processes ASCII and unicode whitespace as well.

source,ruby

“one two three”.squish # “one two three” “ one two n t three ”.squish # “one two three”


titleize

Answers a title string with proper capitalization of each word.

source,ruby

“ThisIsAnExample”.titleize # “This Is An Example”


truncate

Answers a truncated, non-mutated, string for given length with optional delimiter and/or overflow.

The delimiter is the second positional parameter (optional) and is ‘nil` by default. A custom string or regular expression can be used to customize truncation behavior.

The trailer is an optional keyword parameter that is an ellipsis (i.e. ‘“…”`) by default. The trailer can be a custom or empty string. The string length of the trailer is added to the length of the string being truncated, so keep this in mind when setting truncation length.

source,ruby

demo = “It was the best of times” length = demo.length

demo.truncate 9 # “It was…” demo.truncate 12 # “It was th…” demo.truncate length # “It was the best of times” demo.truncate Float::INFINITY # “It was the best of times” demo.truncate 12, “ ” # “It was…” demo.truncate 12, /s/ # “It was…” demo.truncate 6, trailer: “” # “It was” demo.truncate 16, trailer: “… (more)” # “It was… (more)” “demo”.truncate 3 # “…”


to_bool

Answers string as a boolean.

source,ruby

“true”.to_bool # true “yes”.to_bool # true “1”.to_bool # true “”.to_bool # false “example”.to_bool # false


up

Answers string with only first letter capitalized.

source,ruby

“example”.up # “Example”


String IO

reread

Answers full string by rewinding to beginning of string and reading all content.

source,ruby

io = StringIO.new io.write “This is a test.”

io.reread # “This is a test.” io.reread 4 # “This”

buffer = “”.dup io.reread(buffer:) # “This is a test.” buffer # “This is a test.”


to_s

Answers underlying string representation for explicit conversion.

source,ruby

io = StringIO.new io.write “One” io.write “, ” io.write “Two.” io.to_s # “One, Two.”


to_str

Answers underlying string representation for implicit conversion.

source,ruby

io = StringIO.new io.write “One” io.write “, ” io.write “Two.” io.to_str # “One, Two.”


Struct

.with_positions

⚠️ _This is deprecated and will be removed in Version 13.0.0._

Answers a struct instance with given positional arguments regardless of whether the struct was constructed with positional or keyword arguments.

source,ruby

Example = Struct.new :a, :b, :c Example.with_positions 1, 2, 3 #

Example = Struct.new :a, :b, :c, keyword_init: true Example.with_positions 1, 2, 3 #


diff

Allows you to obtain the differences between two objects.

source,ruby

implementation = Struct.new :a, :b, :c

one = implementation.new a: 1, b: 2, c: 3 two = implementation.new a: 3, b: 2, c: 1 three = Struct.new(:x, :y).new x: 1, y: 2

one.diff one # {} one.diff two # {:a=>[1, 3], :c=>[3, 1]} one.diff three # {:a=>[1, nil], :b=>[2, nil], :c=>[3, nil]} one.diff Object.new # {:a=>[1, nil], :b=>[2, nil], :c=>[3, nil]}


Any object that _is not_ the same type will have a ‘nil` value as shown in the last two examples.

merge

Merges multiple attributes without mutating itself and supports any object that responds to ‘#to_h`. Works regardless of whether the struct is constructed with positional or keyword arguments.

source,ruby

example = Struct.new(“Example”, :a, :b, :c).new 1, 2, 3 other = Struct.new(“Other”, :a, :b, :c).new 7, 8, 9

example.merge a: 10 #


merge!

Merges multiple attributes while mutating itself and supports any object that responds to ‘#to_h`. Works regardless of whether the struct is constructed with positional or keyword arguments.

source,ruby

example = Struct.new(“Example”, :a, :b, :c).new 1, 2, 3 other = Struct.new(“Other”, :a, :b, :c).new 7, 8, 9

example.merge! a: 10 #


revalue

Transforms values without mutating itself. An optional hash can be supplied to target specific attributes. In the event that a block isn’t supplied, the struct will answer itself since there is nothing to operate on. Works regardless of whether the struct is constructed with positional or keyword arguments.

source,ruby

example = Struct.new(“Example”, :a, :b, :c).new 1, 2, 3

example.revalue { |value| value * 2 } #


revalue!

Transforms values while mutating itself. An optional hash can be supplied to target specific attributes. In the event that a block isn’t supplied, the struct will answer itself since there is nothing to operate on. Works regardless of whether the struct is constructed with positional or keyword arguments.

source,ruby

one = Struct.new(“One”, :a, :b, :c).new 1, 2, 3 one.revalue! { |value| value * 2 } #

two = Struct.new(“Two”, :a, :b, :c).new 1, 2, 3 two.revalue!(c: 2) { |previous, current| previous + current } #

three = Struct.new(“Three”, :a, :b, :c).new 1, 2, 3 three.revalue! c: 2 #


transmute

Transmutes given enumerable by using the foreign key map and merging those key values into the current struct while not mutating itself. Works regardless of whether the struct is constructed with positional or keyword arguments.

source,ruby

a = Struct.new(“A”, :a, :b, :c).new 1, 2, 3 b = Struct.new(“B”, :x, :y, :z).new 7, 8, 9 c = {r: 10, s: 20, t: 30}

a.transmute b, a: :x, b: :y, c: :z #


transmute!

Transmutes given enumerable by using the foreign key map and merging those key values into the current struct while mutating itself. Works regardless of whether the struct is constructed with positional or keyword arguments.

source,ruby

a = Struct.new(“A”, :a, :b, :c).new 1, 2, 3 b = Struct.new(“B”, :x, :y, :z).new 7, 8, 9 c = {r: 10, s: 20, t: 30}

a.transmute! b, a: :x, b: :y, c: :z #


with

An alias of ‘#merge` and identical in behavior (see `#merge` documentation for details). Allows you to use `Struct` and `Data` objects more interchangeably since they share the same method.

Symbol

call

Enhances symbol-to-proc functionality by allowing you to send positional, keyword, and/or a block arguments. This only works with public methods in order to not break encapsulation.

source,ruby

%w[clue crow cow].map(&:tr.call(“c”, “b”)) # [“blue”, “brow”, “bow”]

1.3, 1.5, 1.9].map(&:round.call(half: :up)) # [1, 2, 2
%w[a b c], %w[c a b]].map(&:index.call { |element| element == “b” }) # [1, 2

%w[1.out 2.in].map(&:sub.call(/./) { |bullet| bullet + “ ” }) # [“1. out”, “2. in”]

1, 2, 3].map(&:to_s.call) # [“1”, “2”, “3”

⚠️ Use of ‘#call` without any arguments should be avoided in order to not incur extra processing costs since the original symbol-to-proc call can be used instead.

Development

To contribute, run:

source,bash

git clone github.com/bkuhlmann/refinements cd refinements bin/setup


You can also use the IRB console for direct access to all objects:

source,bash

bin/console


Tests

To test, run:

source,bash

bin/rake


Credits