Allow setting stopper_strategy

The stopper is only used to stop stemmed words by default. It can stop all words
or be disabled.

.gitignore

@@ -1,4 +1,5 @@
 html/
 doc/
 *.gem
-Gemfile.lock
+Gemfile.lock
+vendor/

lib/xapian_fu/xapian_db.rb

@@ -62,6 +62,10 @@ class DocNotFound < XapianFuError ; end
   #
   #   db = XapianDb.new(:language => :italian, :stopper => false)
   #
+  # The <tt>:stopper_strategy</tt> option specifies the default stop strategy
+  # that will be used when indexing and can be: <tt>:none</tt>, <tt>:all</tt> or
+  # <tt>:stemmed</tt>. Defaults to <tt>:stemmed</tt>
+  #
   # == Spelling suggestions
   #
   # The <tt>:spelling</tt> option controls generation of a spelling
@@ -172,6 +176,8 @@ class XapianDb # :nonew:
     attr_reader :field_options
     attr_accessor :weights_function
     attr :field_weights
+    # The default stopper strategy
+    attr_accessor :stopper_strategy
 
     def initialize( options = { } )
       @options = { :index_positions => true, :spelling => true }.merge(options)
@@ -196,6 +202,7 @@ def initialize( options = { } )
       @language = @options.fetch(:language, :english)
       @stemmer = @options.fetch(:stemmer, @language)
       @stopper = @options.fetch(:stopper, @language)
+      @stopper_strategy = @options.fetch(:stopper_strategy, :stemmed)
       @field_options = {}
       setup_fields(@options[:fields])
       @store_values << @options[:store]

lib/xapian_fu/xapian_doc.rb

@@ -227,6 +227,27 @@ def stopper
       end
     end
 
+    STOPPER_STRATEGIES = {
+      :none    => 0,
+      :all     => 1,
+      :stemmed => 2
+    }
+
+    def stopper_strategy
+      if @stopper_strategy
+        @stopper_strategy
+      else
+        @stopper_strategy =
+          if ! @options[:stopper_strategy].nil?
+            @options[:stopper_strategy]
+          elsif db
+            db.stopper_strategy
+          else
+            :stemmed
+          end
+      end
+    end
+
     # Return this document's language which is set on initialize, inherited
     # from the database or defaults to :english
     def language
@@ -276,6 +297,7 @@ def generate_terms
       tg.document = xapian_document
       tg.stopper = stopper if stopper
       tg.stemmer = stemmer
+      tg.set_stopper_strategy(XapianDoc::STOPPER_STRATEGIES.fetch(stopper_strategy, 2))
       tg.set_flags Xapian::TermGenerator::FLAG_SPELLING if db.spelling
       index_method = db.index_positions ? :index_text : :index_text_without_positions
       fields.each do |k,o|

spec/xapian_doc_spec.rb

@@ -261,4 +261,38 @@
     end
   end
 
+  describe "stopper_strategy" do
+    it "should stop all stop words when stopper_strategy is set to :all " do
+      xdb = XapianDb.new(:stopper_strategy => :all )
+      xdoc = xdb.documents.new("She fished for fish").to_xapian_document
+      terms = xdoc.terms.collect { |t| t.term }
+      terms.should_not include "for"
+      terms.should include "fish"
+    end
+
+    it "should stop stemmed words by when stopper_strategy is set to :stemmed " do
+      xdb = XapianDb.new(:stopper_strategy => :stemmed)
+      xdoc = xdb.documents.new("She fished for fish").to_xapian_document
+      terms = xdoc.terms.collect { |t| t.term }
+      terms.should_not include "Zfor"
+      terms.should include "fish"
+    end
+
+    it "should stop no words by when stopper_strategy is set to :none " do
+      xdb = XapianDb.new(:stopper_strategy => :none)
+      xdoc = xdb.documents.new("She fished for fish").to_xapian_document
+      terms = xdoc.terms.collect { |t| t.term }
+      terms.should include "Zfor"
+      terms.should include "for"
+      terms.should include "fish"
+    end
+
+    it "should stop stemmed words by default " do
+      xdb = XapianDb.new
+      xdoc = xdb.documents.new("She fished for fish").to_xapian_document
+      terms = xdoc.terms.collect { |t| t.term }
+      terms.should_not include "Zfor"
+      terms.should include "fish"
+    end
+  end
 end

xapian-fu.gemspec

@@ -25,6 +25,7 @@ Gem::Specification.new do |s|
 
   s.add_development_dependency("rspec", "~> 2.7")
   s.add_development_dependency("rake", "~> 0")
+  s.add_development_dependency("irb", "~> 0")
   s.add_development_dependency("rdoc", "~> 4")
 
   s.requirements << "libxapian-dev, or the xapian-ruby gem"