encoded_id 0.3.0 → 0.4.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: b69898cce6c9514fc7269f139e4280673b319cf2c879cbb925fdc5a8c03abc27
-  data.tar.gz: b0f17c21e5baa1a171df12629d27ebae511c647a66c0dc82dd8690e573c94754
+  metadata.gz: 5cc2ba3de20b77f871eacb0820fb0ab151281bf6ff6281c992f9e47dc8657138
+  data.tar.gz: 22e9cc3a0dcaecaf34aea73b25681651efe8a4528b4cb70d01ea0aa5fdb73b57
 SHA512:
-  metadata.gz: 133e6d70f9a3613e9009e0d5fe65dd3fdebd095c79533d7976924ed2d9589a47d4854a19d0a9f718caf864673a066f07d92b6ba5947ae043ecb170ed2aab6175
-  data.tar.gz: a5368e9bc5eb8f49438cff06c5c3fbd22d054b386d31ae194ccdea2f8ed148bc9f54492eb93da2e88893056615c6ef73dc24eaef1e22194feac2e160660f0f10
+  metadata.gz: 04f9083352c9559df93ef8418397f43b9360b070025ba75e1ad4ba116fced9635f479939ca9cc55f9b9e594d380e8c4fa503834c85e1ef1994c187de0f5ce3a0
+  data.tar.gz: b2527e2301ad455b22d8ce1a5c652a6814485a703b032af635834ba145b904050ea3b15f3cb9d189b0e3a3a1361e5dce4c4fdbf662f819c56bbbcd212c8f7008

data/README.md

@@ -1,65 +1,77 @@
 # EncodedId
 
-Encode your numerical IDs (eg record primary keys) into obfuscated strings that can be used in URLs. 
+Encode numerical or hex IDs into obfuscated strings that can be used in URLs. 
 
-`::EncodedId::ReversibleId.new(salt: my_salt).encode(123)` => `"p5w9-z27j"`
+```ruby
+coder = ::EncodedId::ReversibleId.new(salt: my_salt)
+coder.encode(123)
+# => "p5w9-z27j"
+coder.encode_hex("10f8c")
+# => "w72a-y0az"
+```
 
-The obfuscated strings are reversible, so you can decode them back into the original numerical IDs. Also supports 
-encoding multiple IDs at once.
+The obfuscated strings are reversible (they decode them back into the original IDs). 
 
-```
-reversibles = ::EncodedId::ReversibleId.new(salt: my_salt)
-reversibles.encode([78, 45])  # "7aq6-0zqw"
-reversibles.decode("7aq6-0zqw")  # [78, 45]
-```
+Also supports encoding multiple IDs at once.
 
-Length of the ID, the alphabet used, and the number of characters per group can be configured.
+```ruby
+my_salt = "salt!"
+coder = ::EncodedId::ReversibleId.new(salt: my_salt)
 
-The custom alphabet (at least 16 characters needed) and character group sizes is to make the IDs easier to read or share.
-Easily confused characters (eg `i` and `j`, `0` and `O`, `1` and `I` etc) are mapped to counterpart characters, to help 
-common mistakes when sharing (eg customer over phone to customer service agent).
+# One of more values can be encoded
+coder.encode([78, 45])
+# => "z2j7-0dmw"
 
-Also supports UUIDs if needed
+# The encoded string can then be reversed back into the original IDs
+coder.decode("z2j7-0dmw")
+# => [78, 45]
 
-```
-::EncodedId::ReversibleId.new(salt: my_salt).encode_hex("9a566b8b-8618-42ab-8db7-a5a0276401fd")
-=> "rppv-tg8a-cx8q-gu9e-zq15-jxes-4gpr-06xk-wfk8-aw"
+# The decoder can be resilient to easily confused characters
+coder.decode("z2j7-Odmw") # (note the capital 'o' instead of zero)
+# => [78, 45]
 ```
 
 ## Features
 
-Build with https://hashids.org
+* encoded IDs are reversible (uses with https://hashids.org)
+* supports slugged IDs (eg `beef-tenderloins-prime--p5w9-z27j`)
+* supports multiple IDs encoded in one encoded string (eg `7aq6-0zqw` decodes to `[78, 45]`)
+* supports encoding of hex strings (eg UUIDs), including multiple IDs encoded in one string **(experimental)**
+* supports custom alphabets for the encoded string (at least 16 characters needed)
+  - by default uses a variation of the Crockford reduced character set (https://www.crockford.com/base32.html)
+  - easily confused characters (eg `i` and `j`, `0` and `O`, `1` and `I` etc) are mapped to counterpart characters, to help
+      avoid common readability mistakes when reading/sharing
+  - build in profanity limitation
+* encoded string can be split into groups of letters to improve human-readability
+  - eg `nft9hr834htu` as `nft9-hr83-4htu`
 
-* Hashids are reversible, no need to persist the generated Id
-* supports slugged IDs (eg 'beef-tenderloins-prime--p5w9-z27j')
-* supports multiple IDs encoded in one `EncodedId` (eg '7aq6-0zqw' decodes to `[78, 45]`)
-* supports encoding of hex strings (eg UUIDs), including mutliple IDs encoded in one `EncodedId`
-* uses a reduced character set (Crockford alphabet) & ids split into groups of letters, ie 'human-readability'
-* profanity limitation
+### Rails support `encoded_id-rails`
 
-To use with **Rails** check out the `encoded_id-rails` gem.
+To use with **Rails** check out the [`encoded_id-rails`](https://github.com/stevegeek/encoded_id-rails) gem.
 
-## Note on security of encoded IDs (hashids)
+```ruby
+class User < ApplicationRecord
+  include EncodedId::WithEncodedId
+end
+
+User.find_by_encoded_id("p5w9-z27j")
+# => #<User id: 78>
+```
+
+### Note on security of encoded IDs (hashids)
 
 **Encoded IDs are not secure**. It maybe possible to reverse them via brute-force. They are meant to be used in URLs as 
 an obfuscation. The algorithm is not an encryption.
 
 Please read more on https://hashids.org/
 
-
-## Compared to alternate Gems
+## Compare to alternate Gems
 
 - https://github.com/excid3/prefixed_ids
 - https://github.com/namick/obfuscate_id
 - https://github.com/norman/friendly_id
 - https://github.com/SPBTV/with_uid
 
-## See also
-
-- https://hashids.org
-- https://www.crockford.com/wrmg/base32.html
-
-
 ## Installation
 
 Install the gem and add to the application's Gemfile by executing:
@@ -70,20 +82,189 @@ If bundler is not being used to manage dependencies, install the gem by executin
 
     $ gem install encoded_id
 
-## Usage
+## `EncodedId::ReversibleId.new`
+
+To create an instance of the encoder/decoder use `.new` with the `salt` option:
+
+```ruby
+coder = EncodedId::ReversibleId.new(
+  # The salt is required
+  salt: ...,
+  # And then the following options are optional
+  length: 8, 
+  split_at: 4, 
+  split_with: "-",
+  alphabet: EncodedId::Alphabet.modified_crockford,
+  hex_digit_encoding_group_size: 4 # Experimental
+)
+```
+
+Note the `salt` value is required and should be a string of some length (greater than 3 characters). This is used to generate the encoded string. 
+
+It will need to be the same value when decoding the string back into the original ID. If the salt is changed, the encoded
+strings will be different and possibly decode to different IDs.
+
+### Options
 
-TODO: Write usage instructions here
+The encoded ID is configurable. The following can be changed:
 
-### Rails
+- the length, eg 8 characters for `p5w9-z27j`
+- the alphabet used in it (min 16 characters)
+- and the number of characters to split the output into and the separator
 
-To use with rails try the `encoded_id-rails` gem.
+### `length`
+
+`length`: the minimum length of the encoded string. The default is 8 characters.
+
+The actual length of the encoded string can be longer if the inputs cannot be represented in the minimum length.
+
+### `alphabet`
+
+`alphabet`: the alphabet used in the encoded string. By default it uses a variation of the Crockford reduced character set (https://www.crockford.com/base32.html).
+
+`alphabet` must be an instance of `EncodedId::Alphabet`.
+
+The default alphabet is `EncodedId::Alphabet.modified_crockford`.
+
+To create a new alphabet, use `EncodedId::Alphabet.new`:
 
 ```ruby
-    class User < ApplicationRecord
-      include EncodedId::WithEncodedId
-    end
+alphabet = EncodedId::Alphabet.new("0123456789abcdef")
+```
+
+`EncodedId::Alphabet.new(characters, equivalences)`
+
+**characters**
+
+`characters`: the characters of the alphabet. Can be a string or array of strings.
+
+Note that the `characters` of the alphabet must be at least 16 _unique_ characters long.
 
-    User.find_by_encoded_id("p5w9-z27j")  # => #<User id: 78>
+
+```ruby
+alphabet = EncodedId::Alphabet.new("ςερτυθιοπλκξηγφδσαζχψωβνμ")
+coder = ::EncodedId::ReversibleId.new(salt: my_salt, alphabet: alphabet)
+coder.encode(123)
+# => "πφλχ-ψησω"
+```
+
+Note that larger alphabets can result in shorter encoded strings (but remember that `length` specifies the minimum length
+of the encoded string).
+
+**equivalences**
+
+You can optionally pass an appropriate character `equivalences` mapping. This is used to map easily confused characters 
+to their counterpart. 
+
+`equivalences`: a hash of characters keys, with their equivalent alphabet character mapped to in the values. 
+
+Note that the characters to be mapped:
+- must not be in the alphabet, 
+- must map to a character that is in the alphabet.
+
+`nil` is the default value which means no equivalences are used.
+
+```ruby
+alphabet = EncodedId::Alphabet.new("!@#$%^&*()+-={}", {"_" => "-"})
+coder = ::EncodedId::ReversibleId.new(salt: my_salt, alphabet: alphabet)
+coder.encode(123)
+# => "}*^(-^}*="
+```
+
+### `split_at` and `split_with`
+
+For readability, the encoded string can be split into groups of characters. 
+
+`split_at`: specifies the number of characters to split the encoded string into. Defaults to 4. 
+
+`split_with`: specifies the separator to use between the groups. Default is `-`.
+
+### `hex_digit_encoding_group_size`
+
+**Experimental**
+
+`hex_digit_encoding_group_size`: specifies the number of hex digits to encode in a group. Defaults to 4. Can be
+between 1 and 32.
+
+Can be used to control the size of the encoded string when encoding hex strings. Larger values will result in shorter
+encoded strings for long inputs, and shorter values will result in shorter encoded strings for smaller inputs.
+
+But note that bigger values will also result in larger markers that separate the groups so could end up increasing
+the encoded string length undesirably.
+
+See below section `Using with hex strings` for more details.
+
+## `EncodedId::ReversibleId#encode`
+
+`#encode(id)`: where `id` is an integer or array of integers to encode.
+
+```ruby
+coder.encode(123)
+# => "p5w9-z27j"
+
+# One of more values can be encoded
+coder.encode([78, 45])
+# => "z2j7-0dmw"
+```
+
+## `EncodedId::ReversibleId#decode`
+
+`#decode(encoded_id)`: where `encoded_id` is a string to decode.
+
+```ruby
+# The encoded string can then be reversed back into the original IDs
+coder.decode("z2j7-0dmw")
+# => [78, 45]
+```
+
+## Using with hex strings
+
+**Experimental** (subject to incompatible changes in future versions)
+
+```ruby
+# Hex input strings are also supported
+coder.encode_hex("10f8c")
+# => "w72a-y0az"
+```
+
+When encoding hex strings, the input is split into groups of hex digits, and each group is encoded separately as its
+integer equivalent. In other words the input is converted into an array of integers and encoded as normal with the
+`encode` method.
+
+eg with `hex_digit_encoding_group_size=1` and inpu `f1`, is split into `f` and `1`, and then encoded as `15` and `1` 
+respectively, ie `encode` is called with `[15, 1]`.
+
+To encode multiple hex inputs the encoded string contains markers to indicate the start of a new hex input. This
+marker is equal to an integer value which is 1 larger than the maximum value the hex digit encoding group size can
+represent (ie it is `2^(hex_digit_encoding_group_size * 4)`).
+
+So for a hex digit encoding group size of 4 (ie group max value is `0xFFFF`), the marker is `65536`
+
+For example with `hex_digit_encoding_group_size=1` for the inputs `f1` and `e2` encoded together, the 
+actual encoded integer array is `[15, 1, 16, 14, 2]`.
+
+### `EncodedId::ReversibleId#encode_hex`
+
+`encode_hex(hex_string)` , where `hex_string` is a string of hex digits or an array of hex strings.
+
+```ruby
+# UUIDs will result in long output strings...
+coder.encode_hex("9a566b8b-8618-42ab-8db7-a5a0276401fd")
+# => "5jjy-c8d9-hxp2-qsve-rgh9-rxnt-7nb5-tve7-bf84-vr"
+# 
+# but there is an option to help reduce this... 
+coder = ::EncodedId::ReversibleId.new(salt: my_salt, hex_digit_encoding_group_size: 32)
+coder.encode_hex("9a566b8b-8618-42ab-8db7-a5a0276401fd")
+# => "vr7m-qra8-m5y6-dkgj-5rqr-q44e-gp4a-52"
+```
+
+### `EncodedId::ReversibleId#decode_hex`
+
+`decode_hex(encoded_id)` , where the output is an array of hex strings.
+
+```ruby
+coder.decode_hex("5jjy-c8d9-hxp2-qsve-rgh9-rxnt-7nb5-tve7-bf84-vr")
+# => ["9a566b8b-8618-42ab-8db7-a5a0276401fd"]
 ```
 
 ## Development
@@ -95,9 +276,14 @@ To install this gem onto your local machine, run `bundle exec rake install`. To
 number in `version.rb`, and then run `bundle exec rake release`, which will create a git tag for the version, push git 
 commits and the created tag, and push the `.gem` file to [rubygems.org](https://rubygems.org).
 
+## See also
+
+- https://hashids.org
+- https://www.crockford.com/base32.html
+
 ## Contributing
 
-Bug reports and pull requests are welcome on GitHub at https://github.com/[USERNAME]/encoded_id.
+Bug reports and pull requests are welcome on GitHub at https://github.com/stevegeek/encoded_id.
 
 ## License
 

data/lib/encoded_id/alphabet.rb

@@ -0,0 +1,39 @@
+# frozen_string_literal: true
+
+module EncodedId
+  class Alphabet
+    def initialize(characters, equivalences = nil)
+      unless (characters.is_a?(Array) || characters.is_a?(String)) && characters.size > 0
+        raise InvalidAlphabetError, "Alphabet must be a string or array"
+      end
+      unique_alphabet = (characters.is_a?(Array) ? characters : characters.chars).uniq
+      raise InvalidAlphabetError, "Alphabet must be at least 16 unique characters" if unique_alphabet.size < 16
+      @characters = unique_alphabet.join
+
+      # Equivalences is a hash of characters to their equivalent character.
+      # Characters to be mapped must not be in the alphabet, and must map to a character that is in the alphabet.
+      raise InvalidConfigurationError, "Character equivalences must be a hash or nil" unless equivalences.nil? || equivalences.is_a?(Hash)
+      valid_equivalences = equivalences.nil? || (unique_alphabet & equivalences.keys).empty? && (equivalences.values - unique_alphabet).empty?
+      raise InvalidConfigurationError unless valid_equivalences
+      @equivalences = equivalences
+    end
+
+    attr_reader :characters, :equivalences
+
+    class << self
+      def modified_crockford
+        # Note we downcase first, so mappings are only for lower case. Also Crockford suggests i==1,
+        # but here i==j is used.
+
+        new(
+          "0123456789abcdefghjkmnpqrstuvwxyz",
+          {
+            "o" => "0",
+            "i" => "j",
+            "l" => "1"
+          }
+        )
+      end
+    end
+  end
+end

data/lib/encoded_id/reversible_id.rb

@@ -8,18 +8,29 @@ require "hashids"
 # Note hashIds already has a built in profanity limitation algorithm
 module EncodedId
   class ReversibleId
-    ALPHABET = "0123456789abcdefghjkmnpqrstuvwxyz"
+    def initialize(salt:, length: 8, split_at: 4, split_with: "-", alphabet: Alphabet.modified_crockford, hex_digit_encoding_group_size: 4)
+      raise InvalidAlphabetError, "alphabet must be an instance of Alphabet" unless alphabet.is_a?(Alphabet)
+      @alphabet = alphabet
 
-    def initialize(salt:, length: 8, split_at: 4, alphabet: ALPHABET, hex_digit_encoding_group_size: 4)
-      unique_alphabet = alphabet.chars.uniq
-      raise InvalidAlphabetError, "Alphabet must be at least 16 characters" if unique_alphabet.size < 16
-
-      @human_friendly_alphabet = unique_alphabet.join
+      raise InvalidConfigurationError, "Salt must be a string and longer that 3 characters" unless salt.is_a?(String) && salt.size > 3
       @salt = salt
+      # Target length of the encoded string (the minimum but not maximum length)
+      raise InvalidConfigurationError, "Length must be an integer greater than 0" unless length.is_a?(Integer) && length > 0
       @length = length
+      # Split the encoded string into groups of this size
+      unless (split_at.is_a?(Integer) && split_at > 0) || split_at.nil?
+        raise InvalidConfigurationError, "Split at must be an integer greater than 0 or nil"
+      end
       @split_at = split_at
+      unless split_with.is_a?(String) && !alphabet.characters.include?(split_with)
+        raise InvalidConfigurationError, "Split with must be a string and not part of the alphabet"
+      end
+      @split_with = split_with
       # Number of hex digits to encode in each group, larger values will result in shorter hashes for longer inputs.
       # Vice versa for smaller values, ie a smaller value will result in smaller hashes for small inputs.
+      if hex_digit_encoding_group_size < 1 || hex_digit_encoding_group_size > 32
+        raise InvalidConfigurationError, "hex_digit_encoding_group_size must be > 0 and <= 32"
+      end
       @hex_digit_encoding_group_size = hex_digit_encoding_group_size
     end
 
@@ -51,7 +62,12 @@ module EncodedId
 
     private
 
-    attr_reader :salt, :length, :human_friendly_alphabet, :split_at, :hex_digit_encoding_group_size
+    attr_reader :salt,
+      :length,
+      :alphabet,
+      :split_at,
+      :split_with,
+      :hex_digit_encoding_group_size
 
     def prepare_input(value)
       inputs = value.is_a?(Array) ? value.map(&:to_i) : [value.to_i]
@@ -61,7 +77,7 @@ module EncodedId
     end
 
     def encoded_id_generator
-      @encoded_id_generator ||= ::Hashids.new(salt, length, human_friendly_alphabet)
+      @encoded_id_generator ||= ::Hashids.new(salt, length, alphabet.characters)
     end
 
     def split_regex
@@ -69,17 +85,19 @@ module EncodedId
     end
 
     def humanize_length(hash)
-      hash.gsub(split_regex, '\0-')
+      hash.gsub(split_regex, "\\0#{split_with}")
     end
 
     def convert_to_hash(str)
-      map_crockford_set(str.delete("-").downcase)
+      clean = str.delete(split_with).downcase
+      alphabet.equivalences.nil? ? clean : map_equivalent_characters(clean)
     end
 
-    def map_crockford_set(str)
-      # Crockford suggest i==1 , but I think i==j is more appropriate as we
-      # only use lowercase
-      str.tr("o", "0").tr("l", "1").tr("i", "j")
+    def map_equivalent_characters(str)
+      alphabet.equivalences.reduce(str) do |cleaned, ceq|
+        from, to = ceq
+        cleaned.tr(from, to)
+      end
     end
 
     # TODO: optimize this
@@ -107,7 +125,7 @@ module EncodedId
 
     # Marker to separate hex strings, must be greater than largest value encoded
     def hex_string_separator
-      @hex_string_separator ||= 2.pow(hex_digit_encoding_group_size * 4) + 1
+      @hex_string_separator ||= 2.pow(hex_digit_encoding_group_size * 4)
     end
 
     # TODO: optimize this

data/lib/encoded_id/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module EncodedId
-  VERSION = "0.3.0"
+  VERSION = "0.4.0"
 end

data/lib/encoded_id.rb

@@ -1,12 +1,15 @@
 # frozen_string_literal: true
 
 require_relative "encoded_id/version"
+require_relative "encoded_id/alphabet"
 require_relative "encoded_id/reversible_id"
 
 module EncodedId
-  class EncodedIdFormatError < ArgumentError; end
+  class InvalidConfigurationError < StandardError; end
 
   class InvalidAlphabetError < ArgumentError; end
 
+  class EncodedIdFormatError < ArgumentError; end
+
   class InvalidInputError < ArgumentError; end
 end

data/sig/encoded_id.rbs

@@ -3,10 +3,17 @@ module EncodedId
   InvalidAlphabetError: ::ArgumentError
   InvalidInputError: ::ArgumentError
 
-  class ReversibleId
-    ALPHABET: ::String
+  class Alphabet
+    def initialize: (String, ?::Hash[::String, ::String]) -> void
+
+    attr_reader alphabet: String
+    attr_reader equivalences: ::Hash[::String, ::String]
 
-    def initialize: (salt: ::String, ?length: ::Integer, ?split_at: ::Integer, ?alphabet: ::String, ?hex_digit_encoding_group_size: ::Integer) -> void
+    def self.modified_crockford: () -> Alphabet
+  end
+
+  class ReversibleId
+    def initialize: (salt: ::String, ?length: ::Integer, ?split_at: ::Integer, ?split_with: ::String, ?alphabet: Alphabet, ?hex_digit_encoding_group_size: ::Integer) -> void
 
     # Encode the input values into a hash
     def encode: (untyped values) -> ::String
@@ -30,9 +37,10 @@ module EncodedId
 
     attr_reader length: ::Integer
 
-    attr_reader human_friendly_alphabet: ::String
+    attr_reader alphabet: Alphabet
 
     attr_reader split_at: ::Integer | nil
+    attr_reader split_with: ::String
 
     attr_reader hex_digit_encoding_group_size: ::Integer
 
@@ -46,7 +54,7 @@ module EncodedId
 
     def convert_to_hash: (::String str) -> ::String
 
-    def map_crockford_set: (::String str) -> ::String
+    def map_equivalent_characters: (::String str) -> ::String
 
     def integer_representation: (untyped hexs) -> ::Array[::Integer]
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: encoded_id
 version: !ruby/object:Gem::Version
-  version: 0.3.0
+  version: 0.4.0
 platform: ruby
 authors:
 - Stephen Ierodiaconou
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2022-10-12 00:00:00.000000000 Z
+date: 2022-12-04 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: hashids
@@ -42,6 +42,7 @@ files:
 - Rakefile
 - Steepfile
 - lib/encoded_id.rb
+- lib/encoded_id/alphabet.rb
 - lib/encoded_id/reversible_id.rb
 - lib/encoded_id/version.rb
 - rbs_collection.yaml