pg_search 0.7.1 → 0.7.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: cd010cfa10f28bb6d953f4fe09dc9c19f10c6239
-  data.tar.gz: c8ece97ff1b44ed62d98272a0df15ff42813dc50
+  metadata.gz: dc5ae6e460c82b93f20a957e4c5a14454fca1f84
+  data.tar.gz: 2a262bd65fb27b8efa689dde40c289f3a3ebc4fb
 SHA512:
-  metadata.gz: 8d1358647c4e451ddd3f837cfd51b9a4bd6c363f075fe8965128c30bfe65d279ef9795ab50282284641cdce0264e381823dfc8e847859a84980c6e4b439ae2cb
-  data.tar.gz: b1cf6a6c1efc8e6b3227a3c87b5877c42501d5ca963f96a94d74ae809a39292f78359004cb93f67e73aee49b5ed124439dce8ce12c0ac9f78bc5eeddb5ca6b5a
+  metadata.gz: d9133994756acc700d4cd09e7b9904f1ab89fcc0e3d79e9aff06017a64b96f4ec7239d5267fc1c146f072846de60bf9d36399bfd4171e9a733180a8f7139cfc5
+  data.tar.gz: efbe3aaab68b0495499a8709b0a5a50454beea08b13f645439d4263fd0df6c538080069665b36e769938d4993414ee9d568c3c8f00cf5099900142bbc3014e5c

data/CHANGELOG.md

@@ -1,5 +1,9 @@
 # pg_search changelog
 
+## 0.7.2
+
+*   Add :threshold option for configuring how permissive trigram searches are.
+
 ## 0.7.1
 
 *   Fix issue with {:using => :trigram, :ignoring => :accents} that generated
@@ -7,15 +11,15 @@
 
 ## 0.7.0
 
-*   Start requiring Ruby 1.9.2 or later
+*   Start requiring Ruby 1.9.2 or later.
 
 ## 0.6.4
 
-*   Fix issue with using more than two features in the same scope
+*   Fix issue with using more than two features in the same scope.
 
 ## 0.6.3
 
-*   Fix issues and deprecations for Active Record 4.0.0.rc1
+*   Fix issues and deprecations for Active Record 4.0.0.rc1.
 
 ## 0.6.2
 

data/README.md

@@ -680,6 +680,42 @@ be installed before this feature can be used.
 
     Website.kinda_spelled_like("Yahoo!") # => [yahooo, yohoo]
 
+##### :threshold
+
+By default, trigram searches find records which have a similarity of at least 0.3
+using pg_trgm's calculations. You may specify a custom threshold if you prefer.
+Higher numbers match more strictly, and thus return fewer results. Lower numbers
+match more permissively, letting in more results.
+
+    class Vegetable < ActiveRecord::Base
+      include PgSearch
+
+      pg_search_scope :strictly_spelled_like,
+                      :against => :name,
+                      :using => {
+                        :trigram => {
+                          :threshold => 0.5
+                        }
+                      }
+
+      pg_search_scope :roughly_spelled_like,
+                      :against => :name,
+                      :using => {
+                        :trigram => {
+                          :threshold => 0.1
+                        }
+                      }
+    end
+
+    cauliflower = Vegetable.create! :name => "cauliflower"
+
+    Vegetable.roughly_spelled_like("couliflower") # => [cauliflower]
+    Vegetable.strictly_spelled_like("couliflower") # => [cauliflower]
+
+    Vegetable.roughly_spelled_like("collyflower") # => [cauliflower]
+    Vegetable.strictly_spelled_like("collyflower") # => []
+
+
 ### Ignoring accent marks (PostgreSQL 9.0 and newer only)
 
 Most of the time you will want to ignore accent marks when searching. This

data/lib/pg_search/features/trigram.rb

@@ -2,25 +2,33 @@ module PgSearch
   module Features
     class Trigram < Feature
       def conditions
-        Arel::Nodes::Grouping.new(
-          Arel::Nodes::InfixOperation.new("%", normalized_document, normalized_query)
-        )
+        if options[:threshold]
+          Arel::Nodes::Grouping.new(
+            similarity.gteq(options[:threshold])
+          )
+        else
+          Arel::Nodes::Grouping.new(
+            Arel::Nodes::InfixOperation.new("%", normalized_document, normalized_query)
+          )
+        end
       end
 
       def rank
-        Arel::Nodes::Grouping.new(
-          Arel::Nodes::NamedFunction.new(
-            "similarity",
-            [
-              normalized_document,
-              normalized_query
-            ]
-          )
-        )
+        Arel::Nodes::Grouping.new(similarity)
       end
 
       private
 
+      def similarity
+        Arel::Nodes::NamedFunction.new(
+          "similarity",
+          [
+            normalized_document,
+            normalized_query
+          ]
+        )
+      end
+
       def normalized_document
         Arel::Nodes::Grouping.new(Arel.sql(normalize(document)))
       end

data/lib/pg_search/version.rb

@@ -1,3 +1,3 @@
 module PgSearch
-  VERSION = "0.7.1"
+  VERSION = "0.7.2"
 end

data/spec/integration/pg_search_spec.rb

@@ -326,6 +326,30 @@ describe "an Active Record model which includes PgSearch" do
         results = ModelWithPgSearch.with_trigrams('cdefhijkl')
         results.should == [included]
       end
+
+      context "when a threshold is specified" do
+        before do
+          ModelWithPgSearch.pg_search_scope :with_strict_trigrams, :against => [:title, :content], :using => {trigram: {threshold: 0.5}}
+          ModelWithPgSearch.pg_search_scope :with_permissive_trigrams, :against => [:title, :content], :using => {trigram: {threshold: 0.1}}
+        end
+
+        it "uses the threshold in the trigram expression" do
+          low_similarity = ModelWithPgSearch.create!(:title => "a")
+          medium_similarity = ModelWithPgSearch.create!(:title => "abc")
+          high_similarity = ModelWithPgSearch.create!(:title => "abcdefghijkl")
+
+          results = ModelWithPgSearch.with_strict_trigrams("abcdefg")
+          expect(results).to include(high_similarity)
+          expect(results).not_to include(medium_similarity, low_similarity)
+
+          results = ModelWithPgSearch.with_trigrams("abcdefg")
+          expect(results).to include(high_similarity, medium_similarity)
+          expect(results).not_to include(low_similarity)
+
+          results = ModelWithPgSearch.with_permissive_trigrams("abcdefg")
+          expect(results).to include(high_similarity, medium_similarity, low_similarity)
+        end
+      end
     end
 
     context "using tsearch" do

data/spec/lib/pg_search/features/trigram_spec.rb

@@ -11,8 +11,12 @@ describe PgSearch::Features::Trigram do
   ]}
   let(:normalizer) { PgSearch::Normalizer.new(config) }
   let(:config) { OpenStruct.new(:ignore => [], :postgresql_version => 90000) }
-  let(:coalesced_colums) { %Q{coalesce(#{Model.quoted_table_name}."name"::text, '') || ' ' || \
-                           coalesce(#{Model.quoted_table_name}."content"::text, '')}.squeeze(' ') }
+
+  let(:coalesced_columns) do
+    <<-SQL.strip_heredoc.chomp
+      coalesce(#{Model.quoted_table_name}."name"::text, '') || ' ' || coalesce(#{Model.quoted_table_name}."content"::text, '')
+    SQL
+  end
 
   with_model :Model do
     table do |t|
@@ -22,24 +26,35 @@ describe PgSearch::Features::Trigram do
   end
 
   describe 'conditions' do
-    context 'paying attention to accents' do
-      it 'escapes the search document and query' do
-        config.ignore = []
-        expect(feature.conditions.to_sql).to eq("((#{coalesced_colums}) % '#{query}')")
-      end
+    it 'escapes the search document and query' do
+      config.ignore = []
+      expect(feature.conditions.to_sql).to eq("((#{coalesced_columns}) % '#{query}')")
     end
 
-    context 'ignoring to accents' do
+    context 'ignoring accents' do
       it 'escapes the search document and query, but not the accent function' do
         config.ignore = [:accents]
-        expect(feature.conditions.to_sql).to eq("((unaccent(#{coalesced_colums})) % unaccent('#{query}'))")
+        expect(feature.conditions.to_sql).to eq("((unaccent(#{coalesced_columns})) % unaccent('#{query}'))")
       end
     end
+
+    context 'when a threshold is specified' do
+      let(:options) do
+        { threshold: 0.5 }
+      end
+
+      it 'uses a minimum similarity expression instead of the "%" operator' do
+        expect(feature.conditions.to_sql).to eq(
+          "(similarity((#{coalesced_columns}), '#{query}') >= 0.5)"
+        )
+      end
+    end
+
   end
 
   describe '#rank' do
     it 'returns an expression using the similarity() function' do
-      expect(feature.rank.to_sql).to eq("(similarity((#{coalesced_colums}), '#{query}'))")
+      expect(feature.rank.to_sql).to eq("(similarity((#{coalesced_columns}), '#{query}'))")
     end
   end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: pg_search
 version: !ruby/object:Gem::Version
-  version: 0.7.1
+  version: 0.7.2
 platform: ruby
 authors:
 - Grant Hutchins