picky 4.11.1 → 4.11.2

data/lib/picky/category_indexed.rb

@@ -16,13 +16,40 @@ module Picky
     # Gets the weight for this token's text.
     #
     def weight token
-      bundle_for(token).weight token.text
+      bundle = bundle_for token
+      if range = token.range
+        # The math is not perfectly correct, but you
+        # get my idea. Also, we could return early.
+        #
+        # TODO Possible to speed up more?
+        #
+        range.inject(nil) do |sum, text|
+          weight = bundle.weight(text)
+          weight && (weight + (sum || 0)) || sum
+        end
+      else
+        bundle.weight token.text
+      end
     end
 
     # Gets the ids for this token's text.
     #
     def ids token
-      bundle_for(token).ids token.text
+      bundle = bundle_for token
+      if range = token.range
+        # Adding all to an array, then flattening
+        # is faster than using ary + ary.
+        #
+        range.inject([]) do |result, text|
+          # It is 30% faster using the empty check
+          # than just << [].
+          #
+          ids = bundle.ids text
+          ids.empty? ? result : result << ids
+        end.flatten
+      else
+        bundle.ids token.text
+      end
     end
 
     # Returns the right index bundle for this token.
@@ -34,6 +61,8 @@ module Picky
     # Returns a combination for the token,
     # or nil, if there is none.
     #
+    # TODO Don't throw away the weight, instead store it in the combination?
+    #
     def combination_for token
       weight(token) && Query::Combination.new(token, self)
     end

data/lib/picky/query/token.rb

@@ -43,6 +43,7 @@ module Picky
         qualify
         partialize
         similarize
+        rangify
         remove_illegals
         self
       end
@@ -186,6 +187,27 @@ module Picky
         redefine_illegals
       end
       
+      # Define a character which makes a token a range token.
+      #
+      # Default is '-'.
+      #
+      # Example:
+      #   Picky::Query::Token.range_character = "…"
+      #   try.search("year:2000…2008") # Will find results in a range.
+      #
+      @@range_character = '-'
+      def self.range_character= character
+        @@range_character = character
+      end
+      def rangify
+        if @text.include? @@range_character
+          @range = Range.new *@text.split(@@range_character, 2)
+        end
+      end
+      def range
+        @range
+      end
+      
       # Is this a "similar" character?
       #
       def similar?

data/spec/functional/custom_delimiters_spec.rb

@@ -4,11 +4,14 @@ require 'spec_helper'
 
 describe 'custom delimiters' do
   
+  # TODO Add range query delimiter.
+  #
   after(:each) do
     Picky::Query::Token.partial_character = '\*'
     Picky::Query::Token.no_partial_character = '"'
     Picky::Query::Token.similar_character = '~'
     Picky::Query::Token.no_similar_character = '"'
+    Picky::Query::Token.range_character = '-'
     Picky::Query::Token.qualifier_text_delimiter = ':'
     Picky::Query::Token.qualifiers_delimiter = ','
   end
@@ -76,4 +79,25 @@ describe 'custom delimiters' do
     Picky::Query::Token.qualifiers_delimiter = '|'
     try.search("text1|text2?hello text2?world").ids.should == [1]
   end
+  
+  it 'offers custom range characters to be set' do
+    index = Picky::Index.new :custom_range_character do
+      category :year
+    end
+
+    rangy = Struct.new :id, :year
+    
+    index.add rangy.new(1, 1977)
+    index.add rangy.new(2, 1989)
+    index.add rangy.new(3, 2001)
+    index.add rangy.new(4, 2012)
+    index.add rangy.new(4, 3000)
+
+    try = Picky::Search.new index
+    try.search("1980-2015").ids.should == [2,3,4]
+    
+    try.search("1980…2015").ids.should == []
+    Picky::Query::Token.range_character = "…"
+    try.search("1980…2015").ids.should == [2,3,4]
+  end
 end

data/spec/functional/range_queries_spec.rb

@@ -0,0 +1,89 @@
+# encoding: utf-8
+#
+require 'spec_helper'
+
+describe 'range queries' do
+  
+  let(:index) do
+    index = Picky::Index.new :range_queries do
+      category :year
+      category :alphabet
+    end
+    
+    rangy = Struct.new :id, :year, :alphabet
+    
+    index.add rangy.new(1, 2000, 'g')
+    index.add rangy.new(2, 1977, 'f')
+    index.add rangy.new(3, 1989, 'a')
+    index.add rangy.new(4, 2011, 'u')
+    index.add rangy.new(5, 3000, 'v')
+    index.add rangy.new(6, 1291, 'z')
+    index.add rangy.new(7,  881, 'm')
+    index.add rangy.new(8, 1984, 'l')
+    
+    index
+  end
+  let(:try) { Picky::Search.new index }
+
+  it 'still works with exact queries' do
+    try.search('1980').ids.should == []
+    try.search('1989').ids.should == [3]
+  end
+  
+  it 'works with misses' do
+    try.search('900-1200').ids.should == []
+  end
+  
+  it 'survives huge ranges' do
+    try.search('0-10000 a').ids.should == [3]
+    
+    # Quote to make it non-partial.
+    #
+    try.search('0-3000"').ids.should == [7,6,2,8,3,1,4,5]
+  end
+  it 'is semi-reasonably fast with huge ranges' do
+    # Quote to make it non-partial.
+    #
+    performance_of { try.search('0-3000"') }.should < 0.21
+    
+    # Note it is much much faster with an additional token.
+    #
+    performance_of { try.search('0-3000 a') }.should < 0.0085
+  end
+    
+  it 'handles basic range queries' do
+    try.search('1980-2001').ids.should == [8,3,1]
+    try.search('f-u').ids.should == [2,1,8,7,4]
+  end
+  
+  it 'can handle qualifiers' do
+    try.search('year:1980-2001').ids.should == [8,3,1]
+    try.search('alphabet:f-u').ids.should == [2,1,8,7,4]
+  end
+  
+  it 'can be combined with other search words' do
+    try.search('1980-2001 a').ids.should == [3]
+    try.search('f-u 881').ids.should == [7]
+  end
+  
+  it 'can handle multiple range queries' do
+    try.search('1980-2001 a-h').ids.should == [3,1]
+    try.search('f-u 881-1977').ids.should == [2,7]
+  end
+  
+  it 'can be combined with partial queries' do
+    try.search('198* a-h').ids.should == [3]
+    try.search('a-h 198').ids.should == [3]
+  end
+  
+  it 'works with nonsensical ranges' do
+    try.search('h-a').ids.should == []
+  end
+  
+  # it 'handles combined range/partial queries' do
+  #   # TODO This still needs to be refined. It is madness.
+  #   #
+  #   try.search('198-200*').ids.should == [8,3,1,4,5,7,6,2]
+  # end
+  
+end

data/spec/lib/category_indexed_spec.rb

@@ -64,53 +64,120 @@ describe Picky::Category do
   end
 
   describe 'weight' do
-    before(:each) do
-      @token = stub :token, :text => :some_text
-    end
-    context 'partial bundle' do
-      before(:each) do
-        @category.stub! :bundle_for => @partial
+    let(:token) { stub :token, :text => :some_text }
+    context 'without range' do
+      before :each do
+        token.stub! :range => nil
+      end
+      context 'partial bundle' do
+        before(:each) do
+          @category.stub! :bundle_for => @partial
+        end
+        it 'should receive weight with the token text' do
+          @partial.should_receive(:weight).once.with :some_text
+
+          @category.weight token
+        end
       end
-      it 'should receive weight with the token text' do
-        @partial.should_receive(:weight).once.with :some_text
+      context 'exact bundle' do
+        before(:each) do
+          @category.stub! :bundle_for => @exact
+        end
+        it 'should receive weight with the token text' do
+          @exact.should_receive(:weight).once.with :some_text
 
-        @category.weight @token
+          @category.weight token
+        end
       end
     end
-    context 'exact bundle' do
-      before(:each) do
-        @category.stub! :bundle_for => @exact
+    context 'with range' do
+      before :each do
+        token.stub! :range => (1..3)
+      end
+      context 'partial bundle' do
+        before(:each) do
+          @category.stub! :bundle_for => @partial
+        end
+        it 'should receive weight with the token text' do
+          @partial.should_receive(:weight).once.times.with(1).and_return(1)
+          @partial.should_receive(:weight).once.times.with(2).and_return(2)
+          @partial.should_receive(:weight).once.times.with(3).and_return(3)
+
+          @category.weight(token).should == 6
+        end
+        it 'returns nil if none hit' do
+          @partial.should_receive(:weight).once.times.with(1).and_return(nil)
+          @partial.should_receive(:weight).once.times.with(2).and_return(nil)
+          @partial.should_receive(:weight).once.times.with(3).and_return(nil)
+
+          @category.weight(token).should == nil
+        end
       end
-      it 'should receive weight with the token text' do
-        @exact.should_receive(:weight).once.with :some_text
+      context 'exact bundle' do
+        before(:each) do
+          @category.stub! :bundle_for => @exact
+        end
+        it 'should receive weight with the token text' do
+          @exact.should_receive(:weight).once.times.with(1).and_return(1)
+          @exact.should_receive(:weight).once.times.with(2).and_return(2)
+          @exact.should_receive(:weight).once.times.with(3).and_return(3)
 
-        @category.weight @token
+          @category.weight(token).should == 6
+        end
       end
     end
   end
 
   describe 'ids' do
-    before(:each) do
-      @token = stub :token, :text => :some_text
-    end
-    context 'partial bundle' do
-      before(:each) do
-        @category.stub! :bundle_for => @partial
+    let(:token) { stub :token, :text => :some_text }
+    context 'without range' do
+      before(:each) { token.stub! :range => nil }
+      context 'partial bundle' do
+        before(:each) do
+          @category.stub! :bundle_for => @partial
+        end
+        it 'should receive ids with the token text' do
+          @partial.should_receive(:ids).once.with :some_text
+
+          @category.ids token
+        end
       end
-      it 'should receive ids with the token text' do
-        @partial.should_receive(:ids).once.with :some_text
+      context 'exact bundle' do
+        before(:each) do
+          @category.stub! :bundle_for => @exact
+        end
+        it 'should receive ids with the token text' do
+          @exact.should_receive(:ids).once.with :some_text
 
-        @category.ids @token
+          @category.ids token
+        end
       end
     end
-    context 'exact bundle' do
-      before(:each) do
-        @category.stub! :bundle_for => @exact
+    context 'with range' do
+      before(:each) { token.stub! :range => (1..3) }
+      context 'partial bundle' do
+        before(:each) do
+          @category.stub! :bundle_for => @partial
+        end
+        it 'should receive ids with the token text' do
+          @partial.should_receive(:ids).once.with(1).and_return [1]
+          @partial.should_receive(:ids).once.with(2).and_return [2]
+          @partial.should_receive(:ids).once.with(3).and_return [3]
+
+          @category.ids(token).should == [1,2,3]
+        end
       end
-      it 'should receive ids with the token text' do
-        @exact.should_receive(:ids).once.with :some_text
+      context 'exact bundle' do
+        before(:each) do
+          @category.stub! :bundle_for => @exact
+        end
+        it 'should receive ids with the token text' do
+          @exact.should_receive(:ids).once.with(1).and_return [1]
+          @exact.should_receive(:ids).once.with(2).and_return [2]
+          @exact.should_receive(:ids).once.with(3).and_return [3]
 
-        @category.ids @token
+          @category.ids(token).should == [1,2,3]
+        end
       end
     end
   end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: picky
 version: !ruby/object:Gem::Version
-  version: 4.11.1
+  version: 4.11.2
   prerelease: 
 platform: ruby
 authors:
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-11-03 00:00:00.000000000 Z
+date: 2012-11-06 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rspec
@@ -34,7 +34,7 @@ dependencies:
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
-        version: 4.11.1
+        version: 4.11.2
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
@@ -42,7 +42,7 @@ dependencies:
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
-        version: 4.11.1
+        version: 4.11.2
 - !ruby/object:Gem::Dependency
   name: text
   requirement: !ruby/object:Gem::Requirement
@@ -282,6 +282,7 @@ files:
 - spec/functional/non_specific_ids_larger_than_20_spec.rb
 - spec/functional/only_spec.rb
 - spec/functional/pool_spec.rb
+- spec/functional/range_queries_spec.rb
 - spec/functional/realtime_spec.rb
 - spec/functional/regression_spec.rb
 - spec/functional/reloading_spec.rb
@@ -445,6 +446,7 @@ test_files:
 - spec/functional/non_specific_ids_larger_than_20_spec.rb
 - spec/functional/only_spec.rb
 - spec/functional/pool_spec.rb
+- spec/functional/range_queries_spec.rb
 - spec/functional/realtime_spec.rb
 - spec/functional/regression_spec.rb
 - spec/functional/reloading_spec.rb