picky 2.0.0.pre2 → 2.0.0.pre3

data/lib/picky/application.rb

@@ -130,7 +130,7 @@
 #     books.define_category :title,
 #                           qualifiers: [:t, :title, :titre],
 #                           partial:    Partial::Substring.new(:from => 1),
-#                           similarity: Similarity::Phonetic.new(2)
+#                           similarity: Similarity::DoubleMetaphone.new(2)
 #     books.define_category :author,
 #                           partial: Partial::Substring.new(:from => -2)
 #     books.define_category :isbn

data/lib/picky/cli.rb

@@ -71,7 +71,7 @@ module Picky
     end
     class Generate < Base
       def execute name, args, params
-        system "picky-generate #{args.join(' ')}"
+        Kernel.system "picky-generate #{args.join(' ')}"
       end
     end
     class Help < Base
@@ -83,7 +83,7 @@ module Picky
           "  picky #{command} #{params_to_s(params)}"
         end.join(?\n)
 
-        puts "Possible commands:\n" + commands
+        Kernel.puts "Possible commands:\n#{commands}\n"
       end
     end
 

data/lib/picky/index/base.rb

@@ -79,7 +79,7 @@ INDEX
     #
     # === Options
     # * partial: Partial::None.new or Partial::Substring.new(from: starting_char, to: ending_char). Default is Partial::Substring.new(from: -3, to: -1).
-    # * similarity: Similarity::None.new or Similarity::Phonetic.new(similar_words_searched). Default is Similarity::None.new.
+    # * similarity: Similarity::None.new or Similarity::DoubleMetaphone.new(similar_words_searched). Default is Similarity::None.new.
     # * qualifiers: An array of qualifiers with which you can define which category you’d like to search, for example “title:hobbit” will search for hobbit in just title categories. Example: qualifiers: [:t, :titre, :title] (use it for example with multiple languages). Default is the name of the category.
     # * qualifier: Convenience options if you just need a single qualifier, see above. Example: qualifiers => :title. Default is the name of the category.
     # * source: Use a different source than the index uses. If you think you need that, there might be a better solution to your problem. Please post to the mailing list first with your application.rb :)

data/lib/picky/internals/generators/similarity/double_metaphone.rb

@@ -0,0 +1,32 @@
+# encoding: utf-8
+#
+module Internals
+
+  module Generators
+
+    module Similarity
+
+      # It's actually a combination of double metaphone
+      # and Levenshtein.
+      #
+      # It uses the double metaphone to get similar words
+      # and ranks them using the levenshtein.
+      #
+      class DoubleMetaphone < Phonetic
+
+        # Encodes the given symbol.
+        #
+        # Returns a symbol.
+        #
+        def encoded sym
+          codes = Text::Metaphone.double_metaphone sym.to_s
+          codes.first.to_sym unless codes.empty?
+        end
+
+      end
+
+    end
+
+  end
+
+end

data/lib/picky/internals/generators/similarity/metaphone.rb

@@ -0,0 +1,32 @@
+# encoding: utf-8
+#
+module Internals
+
+  module Generators
+
+    module Similarity
+
+      # It's actually a combination of metaphone
+      # and Levenshtein.
+      #
+      # It uses the metaphone to get similar words
+      # and ranks them using the levenshtein.
+      #
+      class Metaphone < Phonetic
+
+        # Encodes the given symbol.
+        #
+        # Returns a symbol.
+        #
+        def encoded sym
+          code = Text::Metaphone.metaphone sym.to_s
+          code.to_sym if code
+        end
+
+      end
+
+    end
+
+  end
+
+end

data/lib/picky/internals/generators/similarity/{double_levenshtone.rb → phonetic.rb}

@@ -6,31 +6,23 @@ module Internals
 
     module Similarity
 
-      # DoubleLevensthone means that it's a combination of
-      # * DoubleMetaphone
-      # and
-      # * Levenshtein
-      # :)
+      # It's actually a combination of double metaphone
+      # and Levenshtein.
       #
-      class DoubleLevenshtone < Strategy
+      # It uses the double metaphone to get similar words
+      # and ranks them using the levenshtein.
+      #
+      class Phonetic < Strategy
 
         attr_reader :amount
 
         #
         #
         def initialize amount = 10
+          raise "From Picky 2.0 on you need to use the DoubleMetaphone similarity instead of the Phonetic similarity." if self.class == Phonetic
           @amount = amount
         end
 
-        # Encodes the given symbol.
-        #
-        # Returns a symbol.
-        #
-        def encoded sym
-          codes = Text::Metaphone.double_metaphone sym.to_s
-          codes.first.to_sym unless codes.empty?
-        end
-
         # Generates an index for the given index (in exact index style).
         #
         # In the following form:
@@ -41,7 +33,7 @@ module Internals
           sort hash
         end
 
-        private
+        protected
 
           # Sorts the index values in place.
           #
@@ -69,13 +61,9 @@ module Internals
           end
 
       end
-    
-      # ... aka Phonetic.
-      #
-      Phonetic = DoubleLevenshtone
 
     end
 
   end
-  
+
 end

data/lib/picky/internals/generators/similarity/soundex.rb

@@ -0,0 +1,32 @@
+# encoding: utf-8
+#
+module Internals
+
+  module Generators
+
+    module Similarity
+
+      # It's actually a combination of soundex
+      # and Levenshtein.
+      #
+      # It uses the soundex to get similar words
+      # and ranks them using the levenshtein.
+      #
+      class Soundex < Phonetic
+
+        # Encodes the given symbol.
+        #
+        # Returns a symbol.
+        #
+        def encoded sym
+          code = Text::Soundex.soundex sym.to_s
+          code.to_sym if code
+        end
+
+      end
+
+    end
+
+  end
+
+end

data/lib/picky/internals/index/redis/basic.rb

@@ -1,9 +1,9 @@
 module Internals
 
   module Index
-  
+
     class Redis
-    
+
       # Redis Backend Accessor.
       #
       # Provides necessary helper methods for its
@@ -12,23 +12,23 @@ module Internals
       # dump/load methods.
       #
       class Basic
-      
+
         attr_reader :namespace, :backend
-      
+
         # An index cache takes a path, without file extension,
         # which will be provided by the subclasses.
         #
         def initialize namespace
           @namespace = namespace
-          
+
           # TODO Turn this inside out such that people can pass in
           # their own Redis instance.
-          # 
+          #
           # TODO Make the :db a real option.
           #
           @backend = ::Redis.new :db => 15
         end
-      
+
         # Does nothing.
         #
         def load
@@ -39,13 +39,13 @@ module Internals
         def retrieve
           # Nothing.
         end
-      
+
         # Redis does not backup.
         #
         def backup
           # Nothing.
         end
-      
+
         # Deletes the Redis index namespace.
         #
         def delete
@@ -54,10 +54,10 @@ module Internals
           #       but since we cannot delete by key pattern,
           #       we don't do anything.
         end
-      
+
         # Checks.
         #
-      
+
         # Is this cache suspiciously small?
         #
         def cache_small?
@@ -79,11 +79,11 @@ module Internals
         def size
           backend.dbsize
         end
-      
+
       end
-    
+
     end
-  
+
   end
-  
+
 end

data/lib/picky/internals/index/redis/list_hash.rb

@@ -1,11 +1,11 @@
 module Internals
 
   module Index
-  
+
     class Redis
-    
+
       class ListHash < Basic
-      
+
         # Writes the hash into Redis.
         #
         # TODO Performance: rpush as you get the values instead of putting it together in an array first.
@@ -14,33 +14,33 @@ module Internals
           hash.each_pair do |key, values|
             redis_key = "#{namespace}:#{key}"
             i = 0
-            @backend.multi do
-              @backend.del redis_key
-            
+            backend.multi do
+              backend.del redis_key
+
               values.each do |value|
                 i += 1
-                @backend.zadd redis_key, i, value
+                backend.zadd redis_key, i, value
               end
             end
           end
         end
-      
+
         # Get a collection.
         #
         def collection sym
-          @backend.lrange "#{namespace}:#{sym}", 0, -1
+          backend.lrange "#{namespace}:#{sym}", 0, -1
         end
-      
+
         # Get a single value.
         #
         def member sym
           raise "Can't retrieve a single value from a Redis ListHash. Use Index::Redis::StringHash."
         end
-      
+
       end
-    
+
     end
-  
+
   end
 
 end

data/lib/picky/internals/index/redis/string_hash.rb

@@ -1,36 +1,38 @@
 module Internals
 
   module Index
-  
+
     class Redis
-    
+
       class StringHash < Basic
-      
+
         # Writes the hash into Redis.
         #
+        # TODO Could we use multi?
+        #
         def dump hash
           redis = backend
           hash.each_pair do |key, value|
             redis.hset namespace, key, value
           end
         end
-      
+
         # Get a collection.
         #
         def collection sym
           raise "Can't retrieve a collection from a StringHash. Use Index::Redis::ListHash."
         end
-      
+
         # Get a single value.
         #
         def member sym
           backend.hget namespace, sym
         end
-      
+
       end
-    
+
     end
-  
+
   end
-  
+
 end

data/lib/picky/internals/indexers/serial.rb

@@ -1,25 +1,25 @@
 # encoding: utf-8
 #
 module Indexers
-  
+
   # The indexer defines the control flow.
   #
   class Serial
-    
+
     attr_accessor :tokenizer, :source
-    
+
     def initialize configuration, source, tokenizer
       @configuration = configuration
       @source        = source || raise_no_source
       @tokenizer     = tokenizer
     end
-    
+
     # Raise a no source exception.
     #
     def raise_no_source
       raise NoSourceSpecifiedException.new("No source given for #{@configuration}.")
     end
-    
+
     # Delegates the key format to the source.
     #
     # Default is to_i.
@@ -27,7 +27,7 @@ module Indexers
     def key_format
       @source.key_format || :to_i
     end
-    
+
     # Selects the original id (indexed id) and a column to process. The column data is called "token".
     #
     # Note: Puts together the parts first in an array, then releasing the array from time to time by joining.
@@ -39,7 +39,7 @@ module Indexers
     def process
       comma   = ?,
       newline = ?\n
-      
+
       # TODO Move open to config?
       #
       # @category.prepared_index do |file|
@@ -67,6 +67,6 @@ module Indexers
     def indexing_message
       timed_exclaim %Q{"#{@configuration.identifier}": Starting indexing.}
     end
-    
+
   end
 end

data/lib/picky/internals/indexing/bundle/base.rb

@@ -180,7 +180,7 @@ module Internals
         # Raises an appropriate error message for the given cache.
         #
         def raise_cache_missing what
-          raise "#{what} cache for #{identifier} missing."
+          raise "Error: The #{what} cache for #{identifier} is missing."
         end
 
         # Warns the user if the similarity index is small.

data/lib/picky/internals/indexing/bundle/memory.rb

@@ -17,10 +17,7 @@ module Internals
         alias backend files
 
         def to_s
-          <<-MEMORY
-Memory
-#{@backend.indented_to_s}
-MEMORY
+          "Memory\n#{@backend.indented_to_s}"
         end
 
       end

data/lib/picky/internals/indexing/category.rb

@@ -14,7 +14,7 @@ module Internals
       #  * index: Index to which this category is attached to.
       # Options:
       #  * partial: Partial::None.new, Partial::Substring.new(from:start_char, to:up_to_char) (defaults from:-3, to:-1)
-      #  * similarity: Similarity::None.new (default), Similarity::Phonetic.new(amount_of_similarly_linked_words)
+      #  * similarity: Similarity::None.new (default), Similarity::DoubleMetaphone.new(amount_of_similarly_linked_words)
       #  * source: Use if the category should use a different source.
       #  * from: The source category identifier to take the data from.
       #
@@ -23,6 +23,8 @@ module Internals
       #  * weights:
       #  * tokenizer:
       #
+      # TODO Should source be not optional, or taken from the index?
+      #
       def initialize name, index, options = {}
         @name = name
         @from = options[:from]
@@ -59,8 +61,6 @@ Category(#{name} from #{from}):
         @from || name
       end
 
-      # TODO Spec.
-      #
       def backup_caches
         timed_exclaim "Backing up #{identifier}."
         exact.backup

data/lib/picky/internals/query/combinations/base.rb

@@ -8,11 +8,11 @@ module Internals
     # An allocation consists of a number of combinations.
     #
     module Combinations # :nodoc:all
-    
+
       # Base Combinations contain methods for calculating score and ids.
       #
       class Base
-      
+
         attr_reader :combinations
 
         delegate :empty?, :to => :@combinations
@@ -36,13 +36,7 @@ module Internals
         def weighted_score weights
           weights.score @combinations
         end
-    
-        # Wrap the combinations into an allocation with the result_identifier.
-        #
-        def pack_into_allocation result_identifier
-          Allocation.new self, result_identifier
-        end
-    
+
         # Filters the tokens and identifiers such that only identifiers
         # that are passed in, remain, including their tokens.
         #
@@ -70,11 +64,11 @@ module Internals
         def to_result
           @combinations.map &:to_result
         end
-    
+
       end
 
     end
 
   end
-  
+
 end

data/lib/picky/internals/query/combinations/redis.rb

@@ -8,63 +8,83 @@ module Internals
     # An allocation consists of a number of combinations.
     #
     module Combinations # :nodoc:all
-    
+
       # Redis Combinations contain specific methods for
       # calculating score and ids in memory.
       #
       class Redis < Base
-      
-        # TODO Err… yeah. Wrap in Picky specific wrapper.
+
+        # Connect to the backend.
+        #
+        # TODO Use specific Picky Redis wrapper.
+        #
+        def self.redis
+          @redis ||= ::Redis.new :db => 15
+        end
+
+        attr_reader :redis
+
+        #
         #
         def initialize combinations
           super combinations
-        
-          @@redis ||= ::Redis.new :db => 15
+
+          @redis = self.class.redis
         end
-      
+
         # Returns the result ids for the allocation.
         #
         def ids amount, offset
           return [] if @combinations.empty?
-        
+
           identifiers = @combinations.inject([]) do |identifiers, combination|
             identifiers << "#{combination.identifier}"
           end
-        
+
           result_id = generate_intermediate_result_id
-          
+
           # Intersect and store.
           #
-          @@redis.zinterstore result_id, identifiers
-        
+          redis.zinterstore result_id, identifiers
+
           # Get the stored result.
           #
-          results = @@redis.zrange result_id, offset, (offset + amount)
-          
+          results = redis.zrange result_id, offset, (offset + amount)
+
           # Delete the stored result as it was only for temporary purposes.
           #
-          @@redis.del result_id
-          
+          # Note: I could also not delete it, but that would not be clean at all.
+          #
+          redis.del result_id
+
           results
         end
-      
+
         # Generate a multiple host/process safe result id.
         #
+        # Note: Generated when this class loads.
+        #
         require 'socket'
-        @@host = Socket.gethostname
-        define_method :host do
-          @@host
+        def self.extract_host
+          @host ||= Socket.gethostname
+        end
+        def host
+          self.class.extract_host
+        end
+        extract_host
+        def pid
+          @pid ||= Process.pid
         end
         # Use the host and pid (generated lazily in child processes) for the result.
         #
         def generate_intermediate_result_id
-          :"#{host}:#{@pid ||= Process.pid}:picky:result"
+          :"#{host}:#{pid}:picky:result"
         end
-      
+
       end
-    
+
     end
-  
+
   end
-  
+
 end