picky 0.2.1 → 0.2.2

data/lib/picky/application.rb

@@ -3,6 +3,19 @@
 class Application
   class << self
     
+    # Finalize the subclass as soon as it
+    # has finished loading.
+    #
+    # Note: finalize finalizes the routes.
+    #
+    def inherited app
+      @apps ||= []
+      @apps << app
+    end
+    def finalize_apps
+      @apps.each &:finalize
+    end
+    
     # An application simply delegates to the routing to handle a request.
     #
     def call env

data/lib/picky/index/combined.rb

@@ -9,6 +9,8 @@ module Index
   #
   # TODO Need to use the right subtokens. Bake in?
   #
+  # TODO One can use it as a wrapper, and it will extract the indexes itself. Rename: ExactFirst.
+  #
   class Combined < Bundle
     
     delegate :similar,
@@ -25,6 +27,8 @@ module Index
              :load,
              :to => :@partial
     
+    # TODO initialize type_or_category # => installs itself on all full and partial
+    #
     def initialize full, partial
       @full    = full
       @partial = partial

data/lib/picky/loader.rb

@@ -70,6 +70,10 @@ module Loader
     load_user 'app/logging'
     load_user 'app/application'
     
+    # Finalize the applications.
+    #
+    Application.finalize_apps
+    
     # TODO Rewrite
     #
     Query::Qualifiers.instance.prepare

data/lib/picky/query/base.rb

@@ -32,7 +32,7 @@ module Query
       results = nil
       
       duration = timed do
-        results = execute(tokens, offset) || empty_results # TODO Does not work yet
+        results = execute(tokens, offset) || empty_results(offset) # TODO Does not work yet
       end
       results.duration = duration.round 6
       
@@ -42,21 +42,21 @@ module Query
     # Return nil if no results have been found.
     #
     def execute tokens, offset
-      results_from sorted_allocations(tokens), offset
+      results_from offset, sorted_allocations(tokens)
     end
-
+    
     # Returns an empty result with default values.
     #
-    def empty_results
-      result_type.new
+    def empty_results offset = 0
+      result_type.new offset
     end
-
+    
     # Delegates the tokenizing to the query tokenizer.
     #
     def tokenized text
       @tokenizer.tokenize text
     end
-
+    
     # Gets sorted allocations for the tokens.
     #
     # This generates the possible allocations, sorted.
@@ -111,9 +111,9 @@ module Query
     #
     # TODO Move to results. result_type.from allocations, offset
     #
-    def results_from allocations = nil, offset = 0
-      results = result_type.new allocations
-      results.prepare! offset
+    def results_from offset = 0, allocations = nil
+      results = result_type.new offset, allocations
+      results.prepare!
       results
     end
 

data/lib/picky/results/base.rb

@@ -5,16 +5,16 @@ module Results
   #
   class Base
 
-    # Duration is set by the query.
+    # Duration is set externally by the query.
     #
     attr_writer :duration
     attr_reader :allocations, :offset
 
     # Takes instances of Query::Allocations as param.
     #
-    def initialize allocations = nil
+    def initialize offset = 0, allocations = nil
+      @offset = offset
       @allocations = allocations || Query::Allocations.new
-      @offset = 0
     end
 
     def add more_results
@@ -49,9 +49,8 @@ module Results
     # Without this, the allocations are not processed,
     # and no ids are calculated.
     #
-    def prepare! offset = 0
-      @offset = offset
-      allocations.process! self.max_results, offset
+    def prepare!
+      allocations.process! self.max_results, self.offset
     end
 
     # Duration default is 0.

data/lib/picky/tokenizers/base.rb

@@ -88,6 +88,13 @@ module Tokenizers
                process tokens   # processing tokens / strings
     end
     
+    attr_accessor :substituter
+    alias substituter? substituter
+    
+    def initialize substituter = UmlautSubstituter.new
+      @substituter = substituter
+    end
+    
     # Hooks.
     #
     

data/lib/picky/tokenizers/index.rb

@@ -5,8 +5,6 @@ module Tokenizers
   #
   class Index < Base
     
-    include UmlautSubstituter
-    
     # Default handling definitions. Override in config.
     #
     removes_characters(//)
@@ -26,7 +24,7 @@ module Tokenizers
     #   5. Remove non-single stopwords. (Stopwords that occur with other words)
     #
     def preprocess text
-      text = substitute_umlauts text
+      text = substituter.substitute text if substituter?
       text.downcase!
       remove_illegals text
       contract text

data/lib/picky/tokenizers/query.rb

@@ -13,8 +13,6 @@ module Tokenizers
   #
   class Query < Base
     
-    include UmlautSubstituter
-    
     # Default query tokenizer behaviour. Override in config.
     #
     removes_characters(//)
@@ -53,10 +51,10 @@ module Tokenizers
     # TODO Perhaps move to Normalizer?
     #
     def normalize text
-      text = substitute_umlauts text # Substitute special characters TODO Move to subclass
-      text.downcase!                 # Downcase all text
-      normalize_with_patterns text   # normalize
-      text.to_sym                    # symbolize
+      text = substituter.substitute text if substituter? # Substitute special characters TODO Move to subclass
+      text.downcase!                                     # Downcase all text
+      normalize_with_patterns text                       # normalize
+      text.to_sym                                        # symbolize
     end
     
     # Returns a token for a word.

data/lib/picky/umlaut_substituter.rb

@@ -1,8 +1,20 @@
 # encoding: utf-8
 #
-module UmlautSubstituter
-  def substitute_umlauts text
-    trans = ActiveSupport::Multibyte.proxy_class.new(text).normalize(:kd)
+
+# Substitutes certain umlauts, like
+# ä, ö, ü => ae, oe, ue.
+# (and more, see specs)
+#
+class UmlautSubstituter
+  
+  attr_reader :chars
+  
+  def initialize
+    @chars = ActiveSupport::Multibyte.proxy_class
+  end
+  
+  def substitute text
+    trans = chars.new(text).normalize(:kd)
 
     # substitute special cases
     #
@@ -18,4 +30,5 @@ module UmlautSubstituter
       cp < 0x0300 || cp > 0x035F
     }.pack('U*')
   end
+  
 end

data/prototype_project/config.ru

@@ -23,8 +23,6 @@ Indexes.load_from_cache
 # Use Harakiri middleware to kill unicorn child after X requests.
 #
 # See http://vimeo.com/12614970 for more info.
-#
-# Note: Comment this.
 #
     Rack::Harakiri.after = 50
 use Rack::Harakiri
@@ -33,5 +31,4 @@ use Rack::Harakiri
 #
 # Note: Needs to be the same constant name as in app/application.rb.
 #
-    PickySearch.finalize
 run PickySearch

data/spec/ext/performant_spec.rb

@@ -3,13 +3,6 @@ require File.dirname(__FILE__) + '/../spec_helper'
 describe Performant::Array do
 
   describe "memory_efficient_intersect" do
-    before(:each) do
-      GC.disable
-    end
-    after(:each) do
-      GC.enable
-      GC.start
-    end
     it "should intersect empty arrays correctly" do
       arys = [[3,4], [1,2,3], []]
 
@@ -45,15 +38,13 @@ describe Performant::Array do
       arys = [(1..50).to_a, (10_000..20_000).to_a << 7]
 
       # brute force
-      Benchmark.realtime do
-        Performant::Array.memory_efficient_intersect(arys.sort_by(&:size))
-      end.should < 0.001
+      performance_of { Performant::Array.memory_efficient_intersect(arys.sort_by(&:size)) }.should < 0.001
     end
     it "should be optimal for 2 small arrays of 50/10_000" do
       arys = [(1..50).to_a, (10_000..20_000).to_a << 7]
 
       # &
-      Benchmark.realtime do
+      performance_of do
         arys.inject(arys.shift.dup) do |total, ary|
           total & arys
         end

data/spec/lib/application_spec.rb

@@ -24,7 +24,7 @@ describe Application do
           # Note that Picky needs the following characters to
           # pass through, as they are control characters: *"~:
           #
-          querying.removes_characters(/[^a-zA-Z0-9\s\/\-\,\&\"\~\*\:]/)
+          querying.removes_characters(/[^a-zA-Z0-9äöü\s\/\-\,\&\"\~\*\:]/)
           querying.stopwords(/\b(and|the|of|it|in|for)\b/)
           querying.splits_text_on(/[\s\/\-\,\&]+/)
           querying.normalizes_words([

data/spec/lib/cacher/partial/subtoken_spec.rb

@@ -28,7 +28,7 @@ describe Cacher::Partial::Subtoken do
         }
       end
       it "should be fast" do
-        Benchmark.realtime { @cacher.generate_from( :florian => [1], :flavia => [2] ) }.should < 0.0001
+        performance_of { @cacher.generate_from( :florian => [1], :flavia => [2] ) }.should < 0.0001
       end
       it "should handle duplicate ids" do
         @cacher.generate_from( :flo => [1], :fla => [1] ).should == {
@@ -106,7 +106,7 @@ describe Cacher::Partial::Subtoken do
           end
         end
         it "should be fast" do
-          Benchmark.realtime { @cacher.generate_from(@index) }.should < 0.005
+          performance_of { @cacher.generate_from(@index) }.should < 0.005
         end
       end
       describe "a bigger example with almost identical symbols" do
@@ -118,7 +118,7 @@ describe Cacher::Partial::Subtoken do
           end
         end
         it "should be fast" do
-          Benchmark.realtime { @cacher.generate_from(@index) }.should < 0.003
+          performance_of { @cacher.generate_from(@index) }.should < 0.003
         end
       end
     end

data/spec/lib/extensions/hash_spec.rb

@@ -39,7 +39,7 @@ describe Hash do
       lambda { @obj.to_json(:some => :option) }.should_not raise_error
     end
     it "should be fast" do
-      Benchmark.realtime { @obj.to_json }.should < 0.00006
+      performance_of { @obj.to_json }.should < 0.00006
     end
   end
 

data/spec/lib/extensions/symbol_spec.rb

@@ -6,11 +6,6 @@ describe Symbol do
     include Helpers::Measuring
     before(:each) do
       @token = (((0..9).to_a)*10).to_s.to_sym
-      GC.disable
-    end
-    after(:each) do
-      GC.enable
-      GC.start
     end
     # Note: They influence each other. each_subtoken is faster though.
     #

data/spec/lib/index/bundle_partial_generation_speed_spec.rb

@@ -29,15 +29,9 @@ describe Index::Bundle do
         random_keys  = generate_random_keys 500
         random_ids   = generate_random_ids  500
         @full.index = Hash[random_keys.zip(random_ids)]
-
-        GC.disable
-      end
-      after(:each) do
-        GC.enable
-        GC.start
       end
       it 'should be fast' do
-        Benchmark.realtime do
+        performance_of do
           @full.generate_partial
         end.should < 0.2
       end

data/spec/lib/query/allocations_spec.rb

@@ -328,7 +328,7 @@ describe Query::Allocations do
         @allocations.total.should == 110
       end
       it 'should be fast' do
-        Benchmark.realtime { @allocations.process!(20, 0) }.should < 0.0001
+        performance_of { @allocations.process!(20, 0) }.should < 0.0001
       end
     end
   end

data/spec/lib/query/combinations_spec.rb

@@ -62,21 +62,21 @@ describe 'Query::Combinations' do
       @combination2.should_receive(:ids).once.with.and_return (1..100).to_a
       @combination3.should_receive(:ids).once.with.and_return (1..10).to_a
 
-      Benchmark.realtime { @combinations.ids }.should < 0.004
+      performance_of { @combinations.ids }.should < 0.004
     end
     it "should be fast" do
       @combination1.should_receive(:ids).once.with.and_return (1..1000).to_a
       @combination2.should_receive(:ids).once.with.and_return (1..100).to_a
       @combination3.should_receive(:ids).once.with.and_return (1..10).to_a
 
-      Benchmark.realtime { @combinations.ids }.should < 0.00015
+      performance_of { @combinations.ids }.should < 0.00015
     end
     it "should be fast" do
       @combination1.should_receive(:ids).once.with.and_return (1..1000).to_a
       @combination2.should_receive(:ids).once.with.and_return (901..1000).to_a
       @combination3.should_receive(:ids).once.with.and_return (1..10).to_a
 
-      Benchmark.realtime { @combinations.ids }.should < 0.0001
+      performance_of { @combinations.ids }.should < 0.0001
     end
   end
 

data/spec/lib/query/live_spec.rb

@@ -31,7 +31,7 @@ describe Query::Live do
       allocations = stub :allocations
       @query.should_receive(:sorted_allocations).and_return allocations
 
-      @query.should_receive(:results_from).once.with(allocations, 0).and_return stub(:results, :prepare! => true)
+      @query.should_receive(:results_from).once.with(0, allocations).and_return stub(:results, :prepare! => true)
 
       @query.execute 'some query', 0
     end
@@ -53,7 +53,7 @@ describe Query::Live do
         @query.results_from(@allocations).duration.should == 0
       end
       it "should generate a result with the allocations" do
-        @query.results_from(@allocations).allocations.should == @allocations
+        @query.results_from(0, @allocations).allocations.should == @allocations
       end
     end
   end

data/spec/lib/results/base_spec.rb

@@ -20,10 +20,9 @@ describe Results do
         @allocations = stub :allocations,
                             :process! => nil, :size => 12
 
-        @results = Results::Base.new @allocations
+        @results = Results::Base.new 1234, @allocations
         @results.stub! :duration => 0.1234567890,
-                       :total    => 12345678,
-                       :offset   => 1234
+                       :total    => 12345678
       end
       it 'should output a specific log' do
         @results.to_log('some_query').should == '|0-08-16 10:07:33|0.123457|some_query                                        |12345678|1234|12|'
@@ -46,11 +45,11 @@ describe Results do
     before(:each) do
       @allocations = stub :allocations, :process! => nil, :to_result => :allocations, :total => :some_total
 
-      @results = Results::Base.new @allocations
+      @results = Results::Base.new :some_offset, @allocations
       @results.duration = :some_duration
     end
     it 'should do it correctly' do
-      @results.prepare! :some_offset
+      @results.prepare!
 
       @results.serialize.should == { :allocations => :allocations, :offset => :some_offset, :duration => :some_duration, :total => :some_total }
     end
@@ -139,7 +138,7 @@ describe Results do
           }.should_not raise_error
         end
         it 'should set the allocations to an empty array' do
-          Results::Full.new(:some_allocations).instance_variable_get(:@allocations).should == :some_allocations
+          Results::Full.new(:unimportant, :some_allocations).instance_variable_get(:@allocations).should == :some_allocations
         end
       end
       describe 'Live' do
@@ -149,7 +148,7 @@ describe Results do
           }.should_not raise_error
         end
         it 'should set the allocations to an empty array' do
-          Results::Live.new(:some_allocations).instance_variable_get(:@allocations).should == :some_allocations
+          Results::Live.new(:unimportant, :some_allocations).instance_variable_get(:@allocations).should == :some_allocations
         end
       end
     end
@@ -188,7 +187,7 @@ describe Results do
     describe 'Full' do
       it 'should delegate to allocations.total' do
         allocations = stub :allocations
-        results = Results::Full.new allocations
+        results = Results::Full.new nil, allocations
 
         allocations.should_receive(:total).once
 
@@ -198,7 +197,7 @@ describe Results do
     describe 'Live' do
       it 'should delegate to allocations.total' do
         allocations = stub :allocations
-        results = Results::Live.new allocations
+        results = Results::Live.new nil, allocations
 
         allocations.should_receive(:total).once
 

data/spec/lib/umlaut_substituter_spec.rb

@@ -2,18 +2,20 @@
 require 'spec_helper'
 
 describe UmlautSubstituter do
-  include UmlautSubstituter
+  before(:each) do
+    @substituter = UmlautSubstituter.new
+  end
 
   # A bit of metaprogramming to help with the myriads of its.
   #
   def self.it_should_substitute(special_character, normal_character)
     it "should substitute #{special_character} with #{normal_character}" do
-      substitute_umlauts(special_character).should == normal_character
+      @substituter.substitute(special_character).should == normal_character
     end
   end
   def self.it_should_not_substitute(special_character)
     it "should not substitute #{special_character}" do
-      substitute_umlauts(special_character).should == special_character
+      @substituter.substitute(special_character).should == special_character
     end
   end
 

data/spec/specific/speed_spec.rb

@@ -9,15 +9,10 @@ describe "Speccing Ruby for speed" do
         :speed => (1..5_000).to_a,
         :test => (1..1_000).to_a
       }
-      GC.disable
-    end
-    after(:each) do
-      GC.enable
-      GC.start # start the GC to minimize the chance that it will run again during the speed spec
     end
     describe "+" do
       it "should be fast" do
-        Benchmark.realtime do
+        performance_of do
           @allocs.inject([]) do |total, alloc|
             total + @ids[alloc]
           end
@@ -26,14 +21,14 @@ describe "Speccing Ruby for speed" do
     end
     describe "map and flatten!(1)" do
       it "should be fast" do
-        Benchmark.realtime do
+        performance_of do
           @allocs.map { |alloc| @ids[alloc] }.flatten!(1)
         end.should < 0.02
       end
     end
     describe "<< and flatten!(1)" do
       it "should be fast" do
-        Benchmark.realtime do
+        performance_of do
           @allocs.inject([]) do |total, alloc|
             total << @ids[alloc]
           end.flatten!(1)
@@ -42,7 +37,7 @@ describe "Speccing Ruby for speed" do
     end
     describe "<< and flatten!" do
       it "should be fast" do
-        Benchmark.realtime do
+        performance_of do
           @allocs.inject([]) do |total, alloc|
             total << @ids[alloc]
           end.flatten!

metadata

@@ -5,8 +5,8 @@ version: !ruby/object:Gem::Version
   segments: 
   - 0
   - 2
-  - 1
-  version: 0.2.1
+  - 2
+  version: 0.2.2
 platform: ruby
 authors: 
 - Florian Hanke
@@ -14,7 +14,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2010-10-12 00:00:00 +02:00
+date: 2010-10-14 00:00:00 +02:00
 default_executable: picky
 dependencies: 
 - !ruby/object:Gem::Dependency