immosquare-yaml 0.1.1 → 0.1.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: aef05b6cde6a978367813b253fe4dcc378674aa2b6e0192efe967caaf5babd21
-  data.tar.gz: 1bea9842a643d06132b023cf10c8e1f3624c94d83fdff9d62509017fc6873064
+  metadata.gz: 7ace65cc1c3b599449f0c7af622e242507a010262e536484c1424bb19eeb9ba3
+  data.tar.gz: b10ce8e8239464b58ff87be76512db3fd2731c38aabf256d4f917d26e5cc8ede
 SHA512:
-  metadata.gz: b067dc205dd3e76680c7b76cc851a7676244fcbfb224865f8fdd7368ea185b7fc7612e282c621144b44c7edde5edb13e0dd7db43baa4d6a62ec9a5aa58bfbd1d
-  data.tar.gz: 226772ab6581607c3d41ba8e3fe0bdb360d12e900a4440eba6554d0eb646d7b518f91794dae055f6ff234a798f74b1f0678e5e908dd36e41170441e28f43884a
+  metadata.gz: '048cdac2661a8ab810c12547f5e398612b5a6f94788340055cf8e46b004f3a357b9c1269f210bf29e75fa329248d86f4a92108ad097aabcf39fe91b722af5d12'
+  data.tar.gz: f4eddc1f73762f31a84b6f2b30e5dbc2bcc7c63ce8f5cb83a8ad5668dbde11d71d3a7aed5b37d0c7bf7c8fef02e0e74743bb0f992ccc9ae304b27adf37ff5da8

data/lib/immosquare-yaml/configuration.rb

@@ -0,0 +1,12 @@
+module ImmosquareYaml
+  class Configuration
+
+    attr_accessor :openai_api_key, :openai_model
+
+    def initialize
+      @openai_api_key = nil
+      @openai_model   = nil
+    end
+
+  end
+end

data/lib/immosquare-yaml/railtie.rb

@@ -0,0 +1,9 @@
+module ImmosquareYaml
+  class Railtie < Rails::Railtie
+
+    rake_tasks do
+      load "tasks/immosquare-yaml.rake"
+    end
+
+  end
+end

data/lib/immosquare-yaml/shared_methods.rb

@@ -0,0 +1,45 @@
+module ImmosquareYaml
+  module SharedMethods
+    INDENT_SIZE         = 2
+    NOTHING             = "".freeze
+    SPACE               = " ".freeze
+    NEWLINE             = "\n".freeze
+    SIMPLE_QUOTE        = "'".freeze
+    DOUBLE_QUOTE        = '"'.freeze
+    DOUBLE_SIMPLE_QUOTE = "''".freeze
+    WEIRD_QUOTES_REGEX  = /‘|’|“|”|‛|‚|„|‟|#{Regexp.quote(DOUBLE_SIMPLE_QUOTE)}/.freeze
+    YML_SPECIAL_CHARS   = ["-", "`", "{", "}", "|", "[", "]", ">", ":", "\"", "'", "*", "=", "%", ",", "!", "?", "&", "#", "@"].freeze
+    RESERVED_KEYS       = [
+      "yes", "no", "on", "off", "true", "false",
+      "Yes", "No", "On", "Off", "True", "False",
+      "YES", "NO", "ON", "OFF", "TRUE", "FALSE"
+    ].freeze
+
+    
+    ##============================================================##
+    ## Deep transform values resursively
+    ##============================================================##
+    def deep_transform_values(hash, &block)
+      hash.transform_values do |value|
+        if value.is_a?(Hash)
+          deep_transform_values(value, &block)
+        else
+          block.call(value)
+        end
+      end
+    end
+
+    ##============================================================##
+    ## sort_by_key Function
+    ## Purpose: Sort a hash by its keys, optionally recursively, with 
+    ## case-insensitive comparison and stripping of double quotes.
+    ## ============================================================ #
+    def sort_by_key(hash, recursive = false, &block)
+      block ||= proc {|a, b| a.to_s.downcase.gsub("\"", "") <=> b.to_s.downcase.gsub("\"", "") }
+      hash.keys.sort(&block).each_with_object({}) do |key, seed|
+        seed[key] = hash[key]
+        seed[key] = sort_by_key(seed[key], true, &block) if recursive && seed[key].is_a?(Hash)
+      end
+    end
+  end
+end

data/lib/immosquare-yaml/translate.rb

@@ -0,0 +1,331 @@
+require "iso-639"
+require "httparty"
+
+
+module ImmosquareYaml
+  
+  module Translate
+    extend SharedMethods
+
+    class << self
+
+      def translate(file_path, locale_to,  options = {})
+        begin
+          ##=============================================================##
+          ## options
+          ##=============================================================##
+          options = {
+            :reset_translations => false
+          }.merge(options)
+          options[:reset_translations] = false  if ![true, false].include?(options[:reset_translations])
+
+
+          ##=============================================================##
+          ## Load config keys from config_dev.yml
+          ##=============================================================##
+          raise("Error: openai_api_key not found in config_dev.yml") if ImmosquareYaml.configuration.openai_api_key.nil? 
+          raise("Error: File #{file_path} not found")                if !File.exist?(file_path)
+          raise("Error: locale is not a locale")                     if !locale_to.is_a?(String) || locale_to.size != 2
+  
+          ##============================================================##
+          ## We clean the file before translation
+          ##============================================================##
+          ImmosquareYaml.clean(file_path)
+
+          ##============================================================##
+          ## We parse the clean input file
+          ##============================================================##
+          hash_from = ImmosquareYaml.parse(file_path)
+          raise("#{file_path} is not a correct yml translation file") if !hash_from.is_a?(Hash) && hash_from.keys.size > 1
+
+          ##============================================================##
+          ## Check if the locale is present in the file
+          ##============================================================##
+          locale_from = hash_from.keys.first.to_s
+          raise("Error: The destination file (#{locale_to}) is the same as the source file (#{locale_from}).")    if locale_from == locale_to
+          raise("Error: Expected the source file (#{file_path}) to end with '#{locale_from}.yml' but it didn't.") if !file_path.end_with?("#{locale_from}.yml")
+
+
+          ##============================================================##
+          ## Prepare the output file
+          ##============================================================##
+          file_basename        = File.basename(file_path)
+          file_dirname         = File.dirname(file_path)
+          translated_file_path = "#{file_dirname}/#{file_basename.gsub("#{locale_from}.yml", "#{locale_to}.yml")}"
+
+          ##============================================================##
+          ## We create a hash with all keys from the source file
+          ##============================================================##
+          hash_to = {locale_to => hash_from.delete(locale_from)}
+
+          ##============================================================##
+          ## We create a array with all keys from the source file
+          ##============================================================##
+          array_to = translatable_array(hash_to)
+          array_to = array_to.map {|k, v| [k, v, nil] }
+          
+          ##============================================================##
+          ## If we already have a translation file for the language
+          ## we get the values in it and put it in our
+          ## file... You have to do well with !nil?
+          ## to retrieve the values "" and " "...
+          ##============================================================##
+          if File.exist?(translated_file_path) && options[:reset_translations] == false
+            temp_hash = ImmosquareYaml.parse(translated_file_path)
+            raise("#{translated_file_path} is not a correct yml translation file") if !temp_hash.is_a?(Hash) && temp_hash.keys.size > 1
+
+            ##============================================================##
+            ## t can be nil if the key is not present in the source file
+            ##============================================================##
+            translatable_array(temp_hash).each do |key, value|
+              t    = array_to.find {|k, _v| k == key }
+              t[2] = value if !t.nil? && !value.nil?
+            end
+          end
+
+          ##============================================================##
+          ## Here we have to do all the translation logic...
+          ## For the moment we use the OPENAI API, but we can imagine
+          ## using other translation APIs in the future.
+          ##============================================================##
+          translated_array = translate_with_open_ai(array_to, locale_from, locale_to)
+
+          ##============================================================##
+          ## Then we have to reformat the output yml file
+          ##============================================================##
+          final_array = translated_array.map {|k, _from, to| [k, to] }
+          final_hash  = translatable_hash(final_array)
+
+          ##============================================================##
+          ## We write the output file and clean it
+          ##============================================================##
+          File.write(translated_file_path, ImmosquareYaml.dump(final_hash))
+          ImmosquareYaml.clean(translated_file_path)
+        rescue StandardError => e
+          puts(e.message)
+          puts(e.backtrace)
+          false
+        end
+      end
+
+
+      private
+
+      ##============================================================##
+      ## To translatable hash to array
+      ## opitons are :
+      ## :format    => "string" or "array"
+      ## :keys_only => true or false
+      ## {:fr=>{"demo1"=>"demo1", "demo2"=>{"demo2-1"=>"demo2-1"}}}
+      ## format = "string" and keys_only = false => [["fr.demo1", "demo1"], ["fr.demo2.demo2-1", "demo2-1"]]
+      ## format = "string" and keys_only = true  => ["fr.demo1", "fr.demo2.demo2-1"]
+      ## format = "array"  and keys_only = false => [[["fr", "demo1"], "demo1"], [["fr", "demo2", "demo2-1"], "demo2-1"]]
+      ## format = "array"  and keys_only = true  => [["fr", "demo1"], ["fr", "demo2", "demo2-1"]]
+      ## ============================================================ 
+      def translatable_array(hash, key = nil, result = [], **options)
+        options = {
+          :format    => "string",
+          :keys_only => false
+        }.merge(options)
+        options[:keys_only] = false    if ![true, false].include?(options[:keys_only])
+        options[:format]    = "string" if !["string", "array"].include?(options[:format])
+
+
+        if hash.is_a?(Hash)
+          hash.each_key do |k|
+            translatable_array(hash[k], "#{key}#{":" if !key.nil?}#{k}", result, **options)
+          end
+        else
+          r2 = options[:format] == "string" ? key.split(":").join(".") : key.split(":")
+          result << (options[:keys_only] ? r2 : [r2, hash])
+        end
+        result
+        
+      end
+
+      ##============================================================##
+      ## We can do the inverse of the previous function
+      ##============================================================##
+      def translatable_hash(array)
+        data_hash = array.to_h
+        final     = {}
+        data_hash.each do |key, value|
+          key_parts     = key.split(".")
+          leaf          = key_parts.pop
+          parent        = key_parts.inject(final) {|h, k| h[k] ||= {} }
+          parent[leaf]  = value
+        end
+        final
+      end      
+
+      ##============================================================##
+      ## Translate with OpenAI
+      ## 
+      ## [
+      ##  ["en.mlsconnect.contact_us", "Nous contacter", "Contact us"], 
+      ##  ["en.mlsconnect.description", "Description", nil],
+      ##  ...
+      ## ]
+      ##============================================================##
+      def translate_with_open_ai(array, from, to)
+        ##============================================================##
+        ## https://platform.openai.com/docs/models/
+        ## No all models are available for all users.
+        ## The model `gpt-4-32k` does not exist or you do not have access to it.
+        ## Learn more: https://help.openai.com/en/articles/7102672-how-can-i-access-gpt-4.
+        ##============================================================##
+        model_name = ImmosquareYaml.configuration.openai_model
+        models     = [
+          {:name => "gpt-3.5-turbo",      :tokens => 4097,    :input => 0.0015, :output => 0.002, :group_size => 75},
+          {:name => "gpt-3.5-turbo-16k",  :tokens => 16_385,  :input => 0.0030, :output => 0.004, :group_size => 300},
+          {:name => "gpt-4",              :tokens => 8192,    :input => 0.0300, :output => 0.060, :group_size => 150},
+          {:name => "gpt-4-32k",          :tokens => 32_769,  :input => 0.0600, :output => 0.120, :group_size => 600}
+        ]
+        model = models.find {|m| m[:name] == model_name }
+        model = models.find {|m| m[:name] == "gpt-3.5-turbo-16k" } if model.nil?
+
+        ##============================================================##
+        ## Manage blank values
+        ##============================================================##
+        blank_values       = [NOTHING, SPACE, "\"\"", "\"#{SPACE}\""]
+        cant_be_translated = "CANNOT-BE-TRANSLATED"
+        array              = array.map do |key, from, to|
+          [key, from, blank_values.include?(from) ? from : to] 
+        end 
+
+
+        ##============================================================##
+        ## we want to send as little data as possible to openAI because
+        ## we pay for the volume of data sent. So we're going to send. We put
+        ## a number rather than a string for the translations to be made.
+        ## We take the 16k model to have 16,000k tokens per request
+        ## (around 16,000/4 = 4000 characters).
+        ## ==
+        ## Remove the translations that have already been made
+        ##============================================================##
+        data_open_ai = array.clone
+        data_open_ai = data_open_ai.map.with_index {|(_k, from, to), index| [index, from, to] }
+        data_open_ai = data_open_ai.select {|_index, from, to| !from.nil? && to.nil? }
+
+        ##============================================================##
+        ## Remove quotes surrounding the value if they are present.
+        ## and remove to to avoid error in translation
+        ##============================================================##
+        data_open_ai = data_open_ai.map do |index, from, _to| 
+          from = from.to_s
+          from = from[1..-2] while (from.start_with?(DOUBLE_QUOTE) && from.end_with?(DOUBLE_QUOTE)) || (from.start_with?(SIMPLE_QUOTE) && from.end_with?(SIMPLE_QUOTE))
+          [index, from] 
+        end
+
+        return array if data_open_ai.empty?
+        
+        ##============================================================##
+        ## Call OpenAI API
+        ##============================================================##
+        index         = 0
+        group_size    = model[:group_size]
+        from_iso      = ISO_639.find_by_code(from).english_name.split(";").first
+        to_iso        = ISO_639.find_by_code(to).english_name.split(";").first
+        ai_resuslts   = []
+        prompt_system = "You are a translation tool from #{from_iso} to #{to_iso}\n" \
+                        "The input is an array of pairs, where each pair contains an index and a string to translate, formatted as [index, string_to_translate]\n" \
+                        "Your task is to create an output array where each element is a pair consisting of the index and the translated string, formatted as [index, 'string_translated']\n" \
+                        "\nRules to respect:\n" \
+                        "- Do not escape apostrophes in translated strings; leave them as they are.\n" \
+                        "- Special characters, except apostrophes, that need to be escaped in translated strings should be escaped using a single backslash (\\), not double (\\\\).\n" \
+                        "- If a string cannot be translated use the string '#{cant_be_translated}' translated as the translation value witouth quote (simple or double) quote, just the string\n" \
+                        "- If you dont know the correct translatation but the original word seems to make sense in the original language use it for the translated field otherwise use the #{cant_be_translated} strategy of the preceding point\n" \
+                        "- Use only doubles quotes (\") to enclose translated strings and avoid using single quotes (').\n" \
+                        "- Your output must ONLY be an array with the same number of pairs as the input, without any additional text or explanation.\n" \
+                        "- You need to check that the globle array is correctly closed at the end of the response. (the response must therefore end with ]] to to be consistent)"
+        prompt_init   = "Please proceed with translating the following array:"
+        headers       = {
+          "Content-Type"  => "application/json",
+          "Authorization" => "Bearer #{ImmosquareYaml.configuration.openai_api_key}"
+        }
+        
+        
+        ##============================================================##
+        ## Loop
+        ##============================================================##
+        puts("fields to translate : #{data_open_ai.size}#{" by group of #{group_size}" if data_open_ai.size > group_size}")
+        while index < data_open_ai.size
+          data_group = data_open_ai[index, group_size]
+
+
+          begin
+            puts("call OPENAI Api (with model #{model[:name]}) #{" for #{data_group.size} fields" if data_open_ai.size > group_size}")
+            prompt = "#{prompt_init}:\n\n#{data_group.inspect}\n\n"
+            body   = {
+              :model       => model[:name],
+              :messages    => [
+                {:role => "system", :content => prompt_system},
+                {:role => "user",   :content => prompt}
+              ],
+              :temperature => 0.0
+            }
+            t0   = Time.now
+            call = HTTParty.post("https://api.openai.com/v1/chat/completions", :body => body.to_json, :headers => headers, :timeout => 240)
+            
+            puts("responded in #{(Time.now - t0).round(2)} seconds")
+            raise(call["error"]["message"]) if call.code != 200 
+
+            
+            ##============================================================##
+            ## We check that the result is complete
+            ##============================================================##
+            response  = JSON.parse(call.body)
+            choice    = response["choices"][0]
+            raise("Result is not complete") if choice["finish_reason"] != "stop"
+
+            
+            ##============================================================##
+            ## We calculate the estimate price of the call
+            ##============================================================##
+            input_price   = (response["usage"]["prompt_tokens"] / 1000.0) * model[:input]
+            output_price  = (response["usage"]["completion_tokens"] / 1000.0) * model[:output]
+            price         = input_price + output_price
+            puts("Estimate price => #{input_price.round(3)} + #{output_price.round(3)} = #{price.round(3)} USD")
+
+            ##============================================================##
+            ## We check that the result is an array
+            ##============================================================##
+            content = eval(choice["message"]["content"])
+            raise("Is not an array") if !content.is_a?(Array)
+
+            ##============================================================##
+            ## We save the result
+            ##============================================================##
+            content.each do |index, translation|
+              ai_resuslts << [index, translation == cant_be_translated ? nil : translation]
+            end
+          rescue StandardError => e
+            puts("error OPEN AI API => #{e.message}")
+            puts(e.message)
+            puts(e.backtrace)
+          end
+          index += group_size
+        end
+
+
+        ##============================================================##
+        ## We put the translations in the original array
+        ##============================================================##
+        ai_resuslts.each do |index, translation|
+          array[index][2] = translation
+        end
+
+        ##============================================================##
+        ## We return the modified array
+        ##============================================================##
+        array.map.with_index do |(k, from, to), index| 
+          from = from.to_s
+          to   = "#{DOUBLE_QUOTE}#{to}#{DOUBLE_QUOTE}" if ai_resuslts.find {|i, _t| i == index } && ((from.start_with?(DOUBLE_QUOTE) && from.end_with?(DOUBLE_QUOTE)) || (from.start_with?(SIMPLE_QUOTE) && from.end_with?(SIMPLE_QUOTE)))
+          [k, from, to] 
+        end
+      end
+
+
+    end
+  end
+end

data/lib/immosquare-yaml/version.rb

@@ -0,0 +1,3 @@
+module ImmosquareYaml
+  VERSION = "0.1.2".freeze
+end

data/lib/immosquare-yaml.rb

@@ -1,19 +1,29 @@
+require_relative "immosquare-yaml/configuration"
+require_relative "immosquare-yaml/shared_methods"
+require_relative "immosquare-yaml/translate"
+require_relative "immosquare-yaml/railtie" if defined?(Rails)
+
+
 module ImmosquareYaml
+  extend SharedMethods
+  
   class << self
 
-    INDENT_SIZE         = 2
-    SPACE               = " ".freeze
-    NEWLINE             = "\n".freeze
-    SIMPLE_QUOTE        = "'".freeze
-    DOUBLE_QUOTE        = '"'.freeze
-    DOUBLE_SIMPLE_QUOTE = "''".freeze
-    WEIRD_QUOTES_REGEX  = /‘|’|“|”|‛|‚|„|‟|#{Regexp.quote(DOUBLE_SIMPLE_QUOTE)}/.freeze
-    YML_SPECIAL_CHARS   = ["-", "`", "{", "}", "|", "[", "]", ">", ":", "\"", "'", "*", "=", "%", ",", "!", "?", "&", "#", "@"].freeze
-    RESERVED_KEYS       = [
-      "yes", "no", "on", "off", "true", "false",
-      "Yes", "No", "On", "Off", "True", "False",
-      "YES", "NO", "ON", "OFF", "TRUE", "FALSE"
-    ].freeze
+    
+
+    
+    ##===========================================================================##
+    ## Gem configuration
+    ##===========================================================================##
+    attr_writer :configuration
+
+    def configuration
+      @configuration ||= Configuration.new
+    end
+
+    def config
+      yield(configuration)
+    end
 
     ##===========================================================================##
     ## This method cleans a specified YAML file by processing it line by line.
@@ -167,7 +177,8 @@ module ImmosquareYaml
             ## Remove quotes surrounding the value if they are present.
             ## They are not necessary in this case after | or |-
             ##============================================================##
-            value = value[1..-2] if (value.start_with?(DOUBLE_QUOTE) && value.end_with?(DOUBLE_QUOTE)) || (value.start_with?(SIMPLE_QUOTE) && value.end_with?(SIMPLE_QUOTE))
+            value = value[1..-2] while (value.start_with?(DOUBLE_QUOTE) && value.end_with?(DOUBLE_QUOTE)) || (value.start_with?(SIMPLE_QUOTE) && value.end_with?(SIMPLE_QUOTE))
+
 
             ##=============================================================##
             ## We parse on the 2 types of line breaks
@@ -189,8 +200,8 @@ module ImmosquareYaml
       ## Finalizing the construction by adding a newline at the end and 
       ## removing whitespace from empty lines.
       ##===========================================================================##
-      lines += [""]
-      lines = lines.map {|l| l.strip.empty? ? "" : l }
+      lines += [NOTHING]
+      lines = lines.map {|l| l.strip.empty? ? NOTHING : l }
       lines.join("\n")
     end
 
@@ -280,7 +291,7 @@ module ImmosquareYaml
         ## Detecting blank lines to specially handle the last line within a block; 
         ## if we are inside a block or it's the last line, we avoid skipping
         ##===================================================================================#
-        blank_line = current_line.gsub(NEWLINE, "").empty?
+        blank_line = current_line.gsub(NEWLINE, NOTHING).empty?
         next if !(last_line || inblock || !blank_line)
 
         ##============================================================##
@@ -391,14 +402,14 @@ module ImmosquareYaml
         ## If the line is commented out, we keep and we remove newlines
         ##============================================================##
         if current_line.lstrip.start_with?("#")
-          lines << current_line.gsub(NEWLINE, "")
+          lines << current_line.gsub(NEWLINE, NOTHING)
         ##================================================= ============##
         ## If is in a block (multiline > | or |-), we clean
         ## the line because it can start with spaces tabs etc.
         ## and put it with the block indenter
         ##================================================= ============##
         elsif inblock == true
-          current_line = current_line.gsub(NEWLINE, "").strip
+          current_line = current_line.gsub(NEWLINE, NOTHING).strip
           lines << "#{SPACE * (inblock_indent + INDENT_SIZE)}#{current_line}"
         ##================================================= ============##
         ## if the line ends with a multi-line character and we have a key.
@@ -413,7 +424,7 @@ module ImmosquareYaml
         ## $      : Matches the end of the line/string.
         ##================================================= ============##
         elsif current_line.rstrip.match?(/\S+: [>|](\d*)[-+]?$/)
-          lines << current_line.gsub(NEWLINE, "")
+          lines << current_line.gsub(NEWLINE, NOTHING)
           inblock_indent = indent_level
           inblock        = true
         ##============================================================##
@@ -427,7 +438,7 @@ module ImmosquareYaml
         ## my key: line1 line2 line3
         ##============================================================##
         elsif split.size < 2
-          lines[-1] = (lines[-1] + " #{current_line.lstrip}").gsub(NEWLINE, "")
+          lines[-1] = (lines[-1] + " #{current_line.lstrip}").gsub(NEWLINE, NOTHING)
         ##============================================================##
         ## Otherwise we are in the case of a classic line
         ## key: value or key: without value
@@ -470,8 +481,8 @@ module ImmosquareYaml
       ## spaces on "empty" lines + double spaces
       ## with the same technique as above
       ##============================================================##
-      lines += [""]
-      lines = lines.map {|l| (l.strip.empty? ? "" : l).to_s.gsub(/(?<=\S)\s+/, SPACE) }
+      lines += [NOTHING]
+      lines = lines.map {|l| (l.strip.empty? ? NOTHING : l).to_s.gsub(/(?<=\S)\s+/, SPACE) }
       File.write(file_path, lines.join(NEWLINE))
     end    
 
@@ -481,35 +492,32 @@ module ImmosquareYaml
     ##============================================================##
     ## Strategy:
     ## 1. Forcefully convert the key to a string to handle gsub operations, especially if it's an integer.
-    ## 2. Check if the key is an integer.
-    ## 3. Remove quotes if they are present.
+    ## 2. Remove quotes if they are present.
+    ## 3. Check if the key is an integer.
     ## 4. Re-add quotes if the key is a reserved word or an integer.
-    #
-    ## Regular Expression Explanation:
-    ## /\A(['“‘”’"])(.*)\1\z/
-    ## \A:         Matches the start of the string, ensuring our pattern begins at the very start of the string.
-    ## (['“‘”’"]): Captures a single quote character. It matches any of the characters specified within the brackets.
-    ##             This includes various types of single and double quotes.
-    ## (.*) :      Captures zero or more of any character. It "captures" the entirety of the string between the quotes.
-    ## \1:         Refers back to the first captured group, ensuring the same type of quote character is found at the end.
-    ## \z:         Matches the end of the string, ensuring our pattern matches up to the very end.
-    #
-    ## In the second argument of gsub, we use '\2' to refer back to the content captured by the second capture group.
+    ##
     ## This allows us to fetch the string without the surrounding quotes.
     ##============================================================##
     def clean_key(key)
       ##============================================================##
       ## Convert key to string to avoid issues with gsub operations
-      ## + Check if the key is an integer 
       ##============================================================##
-      key    = key.to_s
-      is_int = key =~ /\A[-+]?\d+\z/
+      key = key.to_s
 
       ##============================================================##
       ## Remove surrounding quotes from the key
+      ##============================================================##
+      key = key[1..-2] if (key.start_with?(DOUBLE_QUOTE) && key.end_with?(DOUBLE_QUOTE)) || (key.start_with?(SIMPLE_QUOTE) && key.end_with?(SIMPLE_QUOTE))
+
+      ##============================================================##
+      ## Check if the key is an integer 
+      ##============================================================##
+      is_int = key =~ /\A[-+]?\d+\z/
+
+      ##============================================================##
+      ## 
       ## Re-add quotes if the key is in the list of reserved keys or is an integer
       ##============================================================##
-      key = key.gsub(/\A(['“”‘’"]?)(.*)\1\z/, '\2')
       key = "\"#{key}\"" if RESERVED_KEYS.include?(key) || is_int
       key
     end
@@ -543,7 +551,7 @@ module ImmosquareYaml
       ## \v: corresponds to a vertical tab
       ## We keep the \n
       ##============================================================##
-      value = value.gsub(/[\t\r\f\v]+/, "")
+      value = value.gsub(/[\t\r\f\v]+/, NOTHING)
       
       ##============================================================##
       ## Replace multiple spaces with a single space.
@@ -561,10 +569,11 @@ module ImmosquareYaml
       value = value.gsub(WEIRD_QUOTES_REGEX, SIMPLE_QUOTE)
 
       ##============================================================##
-      ## Remove quotes surrounding the value if they are present.
+      ## Remove all quotes surrounding the value if they are present.
       ## They will be re-added later if necessary.
+      ## """"value"""" => value
       ##============================================================##
-      value = value[1..-2] if (value.start_with?(DOUBLE_QUOTE) && value.end_with?(DOUBLE_QUOTE)) || (value.start_with?(SIMPLE_QUOTE) && value.end_with?(SIMPLE_QUOTE))
+      value = value[1..-2] while (value.start_with?(DOUBLE_QUOTE) && value.end_with?(DOUBLE_QUOTE)) || (value.start_with?(SIMPLE_QUOTE) && value.end_with?(SIMPLE_QUOTE))
 
       ##============================================================##
       ## Convert emoji representations such as \U0001F600 to their respective emojis.
@@ -603,32 +612,6 @@ module ImmosquareYaml
       value
     end
 
-    ##============================================================##
-    ## Deep transform values resursively
-    ##============================================================##
-    def deep_transform_values(hash, &block)
-      hash.transform_values do |value|
-        if value.is_a?(Hash)
-          deep_transform_values(value, &block)
-        else
-          block.call(value)
-        end
-      end
-    end
-
-    ##============================================================##
-    ## sort_by_key Function
-    ## Purpose: Sort a hash by its keys, optionally recursively, with 
-    ## case-insensitive comparison and stripping of double quotes.
-    ## ============================================================ #
-    def sort_by_key(hash, recursive = false, &block)
-      block ||= proc {|a, b| a.to_s.downcase.gsub(DOUBLE_QUOTE, "") <=> b.to_s.downcase.gsub(DOUBLE_QUOTE, "") }
-      hash.keys.sort(&block).each_with_object({}) do |key, seed|
-        seed[key] = hash[key]
-        seed[key] = sort_by_key(seed[key], true, &block) if recursive && seed[key].is_a?(Hash)
-      end
-    end
-
     ##============================================================##
     ## parse_xml Function
     ## Purpose: Parse an XML file into a nested hash representation.
@@ -657,7 +640,7 @@ module ImmosquareYaml
         ##============================================================##
         ## Check for blank lines (which can be present within multi-line blocks)
         ##============================================================##
-        blank_line = line.gsub(NEWLINE, "").empty?
+        blank_line = line.gsub(NEWLINE, NOTHING).empty?
         
         ##============================================================##
         ## Split the line into key and value.
@@ -685,9 +668,9 @@ module ImmosquareYaml
         ## We no longer have the >
         ## because it is transformed in the clean_xml into | 
         ##============================================================##
-        elsif line.gsub("#{key}:", "").strip.start_with?("|")
+        elsif line.gsub("#{key}:", NOTHING).strip.start_with?("|")
           inblock     = indent_level
-          block_type  = line.gsub("#{key}:", "").strip
+          block_type  = line.gsub("#{key}:", NOTHING).strip
           result      = last_keys.reduce(nested_hash) {|hash, k| hash[k] }
           result[key] = [block_type, []]
           last_keys << key

data/lib/tasks/immosquare-yaml.rake

@@ -0,0 +1,27 @@
+namespace :immosquare_yaml do
+  
+  ##============================================================##
+  ## Function to translate translation files in rails app
+  ##============================================================##
+  desc "Translate translation files in rails app"
+  task :translate => :environment do
+    source_locale  = "fr"
+    locales        = I18n.available_locales.map(&:to_s).reject {|l| l == source_locale }
+    Dir.glob("#{Rails.root}/config/locales/**/*#{source_locale}.yml").each do |file|
+      locales.each do |locale|
+        ImmosquareYaml::Translate.translate(file, locale) 
+      end
+    end
+  end
+
+  ##============================================================##
+  ## Function to clean translation files in rails app
+  ##============================================================##
+  desc "Clean translation files in rails app"
+  task :clean => :environment do
+    Dir.glob("#{Rails.root}/config/locales/**/*.yml").each do |file|
+      ImmosquareYaml.clean(file)
+    end
+  end
+  
+end

metadata

@@ -1,29 +1,43 @@
 --- !ruby/object:Gem::Specification
 name: immosquare-yaml
 version: !ruby/object:Gem::Version
-  version: 0.1.1
+  version: 0.1.2
 platform: ruby
 authors:
 - IMMO SQUARE
 autorequire:
-bindir: exe
+bindir: bin
 cert_chain: []
-date: 2023-09-21 00:00:00.000000000 Z
+date: 2023-09-25 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: iso-639
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ">"
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: 0.2.5
+        version: 0.3.6
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ">"
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: 0.2.5
+        version: 0.3.6
+- !ruby/object:Gem::Dependency
+  name: httparty
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.21.0
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.21.0
 description: IMMOSQUARE-YAML is a specialized Ruby gem tailored primarily for parsing
   and dumping YML translation files, addressing challenges faced with other parsers
   like interpreting translation keys as booleans, multi-line strings, and more.
@@ -34,7 +48,12 @@ extensions: []
 extra_rdoc_files: []
 files:
 - lib/immosquare-yaml.rb
-- lib/version.rb
+- lib/immosquare-yaml/configuration.rb
+- lib/immosquare-yaml/railtie.rb
+- lib/immosquare-yaml/shared_methods.rb
+- lib/immosquare-yaml/translate.rb
+- lib/immosquare-yaml/version.rb
+- lib/tasks/immosquare-yaml.rake
 homepage: https://github.com/IMMOSQUARE/immosquare-yaml
 licenses:
 - MIT
@@ -47,7 +66,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
-      version: 2.6.0
+      version: 2.7.2
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="

data/lib/version.rb

@@ -1,5 +0,0 @@
-# frozen_string_literal: true
-
-module ImmosquareYaml
-  VERSION = "0.1.1"
-end