gojee-sunspot 2.0.3 → 2.0.4

data/lib/sunspot/dsl/field_query.rb

@@ -0,0 +1,327 @@
+module Sunspot
+  module DSL
+    # 
+    # Provides an API for areas of the query DSL that operate on specific
+    # fields. This functionality is provided by the query DSL and the dynamic
+    # query DSL.
+    #
+    class FieldQuery < Scope
+      def initialize(search, query, setup) #:nodoc:
+        @search, @query = search, query
+        super(query.scope, setup)
+      end
+
+      # Specify the order that results should be returned in. This method can
+      # be called multiple times; precedence will be in the order given.
+      #
+      # ==== Parameters
+      #
+      # field_name<Symbol>:: the field to use for ordering
+      # direction<Symbol>:: :asc or :desc (default :asc)
+      #
+      def order_by(field_name, direction = nil)
+        sort =
+          if special = Sunspot::Query::Sort.special(field_name)
+            special.new(direction)
+          else
+            Sunspot::Query::Sort::FieldSort.new(
+              @setup.field(field_name), direction
+            )
+          end
+        @query.add_sort(sort)
+      end
+
+      #
+      # Specify that the results should be ordered based on their
+      # distance from a given point.
+      #
+      # ==== Parameters
+      #
+      # field_name<Symbol>::
+      #   the field that stores the location (declared as `latlon`)
+      # lat<Numeric>::
+      #   the reference latitude
+      # lon<Numeric>::
+      #   the reference longitude
+      # direction<Symbol>::
+      #   :asc or :desc (default :asc)
+      # 
+      def order_by_geodist(field_name, lat, lon, direction = nil)
+        @query.add_sort(
+          Sunspot::Query::Sort::GeodistSort.new(@setup.field(field_name), lat, lon, direction)
+        )
+      end
+
+      # 
+      # DEPRECATED Use <code>order_by(:random)</code>
+      #
+      def order_by_random
+        order_by(:random)
+      end
+
+      # Specify a field for result grouping. Grouping groups documents
+      # with a common field value, return only the top document per
+      # group.
+      #
+      # More information in the Solr documentation:
+      # <http://wiki.apache.org/solr/FieldCollapsing>
+      #
+      # ==== Parameters
+      #
+      # field_name<Symbol>:: the field to use for grouping
+      def group(*field_names, &block)
+        options = Sunspot::Util.extract_options_from(field_names)
+
+        field_names.each do |field_name|
+          field = @setup.field(field_name)
+          group = @query.add_group(Sunspot::Query::FieldGroup.new(field))
+          @search.add_field_group(field)
+
+          if block
+            Sunspot::Util.instance_eval_or_call(
+              FieldGroup.new(@query, @setup, group),
+              &block
+            )
+          end
+        end
+      end
+
+      #
+      # Request a facet on the search query. A facet is a feature of Solr that
+      # determines the number of documents that match the existing search *and*
+      # an additional criterion. This allows you to build powerful drill-down
+      # interfaces for search, at each step presenting the searcher with a set
+      # of refinements that are known to return results.
+      #
+      # In Sunspot, each facet returns zero or more rows, each of which
+      # represents a particular criterion conjoined with the actual query being
+      # performed. For _field_ _facets_, each row represents a particular value
+      # for a given field. For _query_ _facets_, each row represents an
+      # arbitrary scope; the facet itself is just a means of logically grouping
+      # the scopes.
+      #
+      # === Examples
+      #
+      # ==== Field Facets
+      #
+      # A field facet is specified by passing one or more Symbol arguments to
+      # this method:
+      #
+      #   Sunspot.search(Post) do
+      #     with(:blog_id, 1)
+      #     facet(:category_id)
+      #   end
+      #   
+      # The facet specified above will have a row for each category_id that is
+      # present in a document which also has a blog_id of 1.
+      #
+      # ==== Multiselect Facets
+      #
+      # In certain circumstances, it is beneficial to exclude certain query
+      # scopes from a facet; the most common example is multi-select faceting,
+      # where the user has selected a certain value, but the facet should still
+      # show all options that would be available if they had not:
+      #
+      #   Sunspot.search(Post) do
+      #     with(:blog_id, 1)
+      #     category_filter = with(:category_id, 2)
+      #     facet(:category_id, :exclude => category_filter)
+      #   end
+      # 
+      # Although the results of the above search will be restricted to those
+      # with a category_id of 2, the category_id facet will operate as if a
+      # category had not been selected, allowing the user to select additional
+      # categories (which will presumably be ORed together).
+      # 
+      # It possible to exclude multiple filters by passing an array:
+      #
+      #   Sunspot.search(Post) do
+      #     with(:blog_id, 1)
+      #     category_filter = with(:category_id, 2)
+      #     author_filter = with(:author_id, 3)
+      #     facet(:category_id,
+      #           :exclude => [category_filter, author_filter].compact)
+      #   end
+      #
+      # You should consider using +.compact+ to ensure that the array does not 
+      # contain any nil values.
+      #
+      # <strong>As far as I can tell, Solr only supports multi-select with
+      # field facets; if +:exclude+ is passed to a query facet, this method will
+      # raise an error. Also, the +:only+ and +:extra+ options use query
+      # faceting under the hood, so these can't be used with +:extra+ either.
+      # </strong>
+      #
+      # ==== Query Facets
+      #
+      # A query facet is a collection of arbitrary scopes, each of which
+      # represents a row. This is specified by passing a block into the #facet
+      # method; the block then contains one or more +row+ blocks, each of which
+      # creates a query facet row. The +row+ blocks follow the usual Sunspot
+      # scope DSL.
+      #
+      # For example, a query facet can be used to facet over a set of ranges:
+      #
+      #   Sunspot.search(Post) do
+      #     facet(:average_rating) do
+      #       row(1.0..2.0) do
+      #         with(:average_rating, 1.0..2.0)
+      #       end
+      #       row(2.0..3.0) do
+      #         with(:average_rating, 2.0..3.0)
+      #       end
+      #       row(3.0..4.0) do
+      #         with(:average_rating, 3.0..4.0)
+      #       end
+      #       row(4.0..5.0) do
+      #         with(:average_rating, 4.0..5.0)
+      #       end
+      #     end
+      #   end
+      #
+      # Note that the arguments to the +facet+ and +row+ methods simply provide
+      # labels for the facet and its rows, so that they can be retrieved and
+      # identified from the Search object. They are not passed to Solr and no
+      # semantic meaning is attached to them. The label for +facet+ should be
+      # a symbol; the label for +row+ can be whatever you'd like.
+      #
+      # ==== Parameters
+      #
+      # field_names...<Symbol>:: fields for which to return field facets
+      #
+      # ==== Options
+      #
+      # :sort<Symbol>::
+      #   Either :count (values matching the most terms first) or :index (lexical)
+      # :limit<Integer>::
+      #   The maximum number of facet rows to return
+      # :minimum_count<Integer>::
+      #   The minimum count a facet row must have to be returned
+      # :zeros<Boolean>::
+      #   Return facet rows for which there are no matches (equivalent to
+      #   :minimum_count => 0). Default is false.
+      # :exclude<Object,Array>::
+      #   Exclude one or more filters when performing the faceting (see
+      #   Multiselect Faceting above). The object given for this argument should
+      #   be the return value(s) of a scoping method (+with+, +any_of+,
+      #   +all_of+, etc.). <strong>Only can be used for field facets that do not
+      #   use the +:extra+ or +:only+ options.</strong>
+      # :name<Symbol>::
+      #   Give a custom name to a field facet. The main use case for this option
+      #   is for requesting the same field facet multiple times, using different
+      #   filter exclusions (see Multiselect Faceting above). If you pass this
+      #   option, it is also the argument that should be passed to Search#facet
+      #   when retrieving the facet result.
+      # :only<Array>::
+      #   Only return facet rows for the given values. Useful if you are only
+      #   interested in faceting on a subset of values for a given field.
+      #   <strong>Only applies to field facets.</strong>
+      # :extra<Symbol,Array>::
+      #   One or more of :any and :none. :any returns a facet row with a count
+      #   of all matching documents that have some value for this field. :none
+      #   returns a facet row with a count of all matching documents that have
+      #   no value for this field. The facet row(s) corresponding to the extras
+      #   have a value of the symbol passed. <strong>Only applies to field
+      #   facets.</strong>
+      #
+      def facet(*field_names, &block)
+        options = Sunspot::Util.extract_options_from(field_names)
+
+        if block
+          if field_names.length != 1
+            raise(
+              ArgumentError,
+              "wrong number of arguments (#{field_names.length} for 1)"
+            )
+          end
+          if options.has_key?(:exclude)
+            raise(
+              ArgumentError,
+              "can't use :exclude with query facets"
+            )
+          end
+          search_facet = @search.add_query_facet(field_names.first, options)
+          Sunspot::Util.instance_eval_or_call(
+            QueryFacet.new(@query, @setup, search_facet),
+            &block
+          )
+        elsif options[:only]
+          if options.has_key?(:exclude)
+            raise(
+              ArgumentError,
+              "can't use :exclude with :only (see documentation)"
+            )
+          end
+          field_names.each do |field_name|
+            field = @setup.field(field_name)
+            search_facet = @search.add_field_facet(field, options)
+            Util.Array(options[:only]).each do |value|
+              facet = Sunspot::Query::QueryFacet.new
+              facet.add_positive_restriction(field, Sunspot::Query::Restriction::EqualTo, value)
+              @query.add_query_facet(facet)
+              search_facet.add_row(value, facet.to_boolean_phrase)
+            end
+          end
+        else
+          field_names.each do |field_name|
+            search_facet = nil
+            field = @setup.field(field_name)
+            facet =
+              if options[:time_range]
+                unless field.type.is_a?(Sunspot::Type::TimeType)
+                  raise(
+                    ArgumentError,
+                    ':time_range can only be specified for Date or Time fields'
+                  )
+                end
+                search_facet = @search.add_date_facet(field, options)
+                Sunspot::Query::DateFieldFacet.new(field, options)
+              else
+                search_facet = @search.add_field_facet(field, options)
+                Sunspot::Query::FieldFacet.new(field, options)
+              end
+            @query.add_field_facet(facet)
+            Util.Array(options[:extra]).each do |extra|
+              if options.has_key?(:exclude)
+                raise(
+                  ArgumentError,
+                  "can't use :exclude with :extra (see documentation)"
+                )
+              end
+              extra_facet = Sunspot::Query::QueryFacet.new
+              case extra
+              when :any
+                extra_facet.add_negated_restriction(
+                  field,
+                  Sunspot::Query::Restriction::EqualTo,
+                  nil
+                )
+              when :none
+                extra_facet.add_positive_restriction(
+                  field,
+                  Sunspot::Query::Restriction::EqualTo,
+                  nil
+                )
+              else
+                raise(
+                  ArgumentError,
+                  "Allowed values for :extra are :any and :none"
+                )
+              end
+              search_facet.add_row(extra, extra_facet.to_boolean_phrase)
+              @query.add_query_facet(extra_facet)
+            end
+          end
+        end
+      end
+
+      def dynamic(base_name, &block)
+        dynamic_field_factory = @setup.dynamic_field_factory(base_name)
+        Sunspot::Util.instance_eval_or_call(
+          FieldQuery.new(@search, @query, dynamic_field_factory),
+          &block
+        )
+      end
+    end
+  end
+end

data/lib/sunspot/dsl/fields.rb

@@ -0,0 +1,103 @@
+module Sunspot
+  module DSL #:nodoc:
+    # The Fields class provides a DSL for specifying field definitions in the
+    # Sunspot.setup block. As well as the #text method, which creates fulltext
+    # fields, uses #method_missing to allow definition of typed fields. The
+    # available methods are determined by the constants defined in 
+    # Sunspot::Type - in theory (though this is untested), plugin developers
+    # should be able to add support for new types simply by creating new
+    # implementations in Sunspot::Type
+    #
+    class Fields
+      def initialize(setup) #:nodoc:
+        @setup = setup
+      end
+
+      # Add a text field. Text fields are tokenized before indexing and are
+      # the only fields searched in fulltext searches. If a block is passed,
+      # create a virtual field; otherwise create an attribute field.
+      #
+      # If options are passed, they will be applied to all the given fields.
+      #
+      # ==== Parameters
+      #
+      # names...<Symbol>:: One or more field names
+      #
+      # ==== Options
+      #
+      # :boost<Float>::
+      #   Index-time boost that should be applied to this field for keyword search
+      # :default_boost<Float>::
+      #   Default search-time boost to apply to this field during keyword
+      #   search. Can be overriden with DSL::Fulltext#fields or
+      #   DSL::Fulltext#boost_fields method.
+      #
+      def text(*names, &block)
+        options = names.pop if names.last.is_a?(Hash)
+        names.each do |name|
+          @setup.add_text_field_factory(
+            name,
+            options || {},
+            &block
+          )
+        end
+      end
+
+      # 
+      # Specify a document-level boost. As with fields, you have the option of
+      # passing an attribute name which will be called on each model, or a block
+      # to be evaluated in the model's context. As well as these two options,
+      # this method can also take a constant number, meaning that all indexed
+      # documents of this class will have the specified boost.
+      #
+      # ==== Parameters
+      #
+      # attr_name<Symbol,~.to_f>:: Attribute name to call or a numeric constant
+      #
+      def boost(attr_name = nil, &block)
+        @setup.add_document_boost(attr_name, &block)
+      end
+
+      # method_missing is used to provide access to typed fields, because
+      # developers should be able to add new Sunspot::Type implementations
+      # dynamically and have them recognized inside the Fields DSL. Like #text,
+      # these methods will create a VirtualField if a block is passed, or an
+      # AttributeField if not.
+      #
+      # ==== Example
+      #
+      #   Sunspot.setup(File) do
+      #     time :mtime
+      #   end
+      #
+      # The call to +time+ will create a field of type Sunspot::Types::TimeType
+      #
+      def method_missing(method, *args, &block)
+        options = Util.extract_options_from(args)
+        type_const_name = "#{Util.camel_case(method.to_s.sub(/^dynamic_/, ''))}Type"
+        trie = options.delete(:trie)
+        type_const_name = "Trie#{type_const_name}" if trie
+        begin
+          type_class = Type.const_get(type_const_name)
+        rescue(NameError)
+          if trie
+            raise ArgumentError, "Trie fields are only valid for numeric and time types"
+          else
+            super(method, *args, &block)
+          end
+        end
+        type = type_class.instance
+        name = args.shift
+        if method.to_s =~ /^dynamic_/
+          if type.accepts_dynamic?
+            @setup.add_dynamic_field_factory(name, type, options, &block)
+          else
+            super(method, *args, &block)
+          end
+        else
+          @setup.add_field_factory(name, type, options, &block)
+        end
+      end
+    end
+  end
+end

data/lib/sunspot/dsl/fulltext.rb

@@ -0,0 +1,243 @@
+module Sunspot
+  module DSL
+    # 
+    # This DSL exposes the functionality provided by Solr's fulltext Dismax
+    # handler.
+    #
+    class Fulltext
+      attr_reader :exclude_fields #:nodoc:
+
+      # accept function in boost
+      include Functional
+
+      def initialize(query, setup) #:nodoc:
+        @query, @setup = query, setup
+        @fields_added = false
+        @exclude_fields = []
+      end
+
+      # 
+      # Specify which fields to search. Field names specified as arguments are
+      # given default boost; field boosts can be specified by passing a hash of
+      # field names keyed to boost values as the last argument.
+      #
+      # If you wish to boost certain fields without restricting which fields are
+      # searched, use #boost_fields
+      #
+      # === Example
+      #
+      #   Sunspot.search(Post) do
+      #     keywords 'search is cool' do
+      #       fields(:body, :title => 2.0)
+      #     end
+      #   end
+      #
+      # This would search the :body field with default boost (1.0), and the :title
+      # field with a boost of 2.0
+      #
+      def fields(*field_names)
+        @fields_added = true
+        boosted_fields = field_names.pop if field_names.last.is_a?(Hash)
+        field_names.each do |field_name|
+          @setup.text_fields(field_name).each do |field|
+            @query.add_fulltext_field(field, field.default_boost)
+          end
+        end
+        if boosted_fields
+          boosted_fields.each_pair do |field_name, boost|
+            @setup.text_fields(field_name).each do |field|
+              @query.add_fulltext_field(field, boost)
+            end
+          end
+        end
+      end
+
+      # 
+      # Exclude the given fields from the search. All fields that are configured
+      # for the types under search and not listed here will be searched.
+      #
+      def exclude_fields(*field_names)
+        @exclude_fields.concat(field_names)
+      end
+
+      # 
+      # Enable keyword highlighting for this search. By default, the fields 
+      # under search will be highlighted; you may also may pass one or more
+      # symbol arguments indicating fields to be highlighted (they don't even
+      # have to be the same fields you're searching).
+      #
+      # === Example
+      #
+      #   Sunspot.search(Post) do
+      #     keywords 'show me the highlighting' do
+      #       highlight :title, :body
+      #     end
+      #   end
+      #
+      # You may also pass a hash of options as the last argument. Options are
+      # the following:
+      #
+      # Full disclosure: I barely understand what these options actually do;
+      # this documentation is pretty much just copied from the
+      # (http://wiki.apache.org/solr/HighlightingParameters#head-23ecd5061bc2c86a561f85dc1303979fe614b956)[Solr Wiki]
+      # 
+      # :max_snippets::
+      #   The maximum number of highlighted snippets to generate per field
+      # :fragment_size::
+      #   The number of characters to consider for a highlighted fragment
+      # :merge_continuous_fragments::
+      #   Collapse continuous fragments into a single fragment
+      # :phrase_highlighter::
+      #   Highlight phrase terms only when they appear within the query phrase
+      #   in the document
+      # :require_field_match::
+      #   If true, a field will only be highlighted if the query matched in
+      #   this particular field (only has an effect if :phrase_highlighter is
+      #   true as well)
+      #
+      def highlight(*args)
+        options = args.last.kind_of?(Hash) ? args.pop : {}
+        fields = []
+        args.each { |field_name| fields.concat(@setup.text_fields(field_name)) }
+
+        @query.add_highlight(fields, options)
+      end
+
+      # 
+      # Phrase fields are an awesome dismax feature that adds extra boost to
+      # documents for which all the fulltext keywords appear in close proximity
+      # in one of the given fields. Excellent for titles, headlines, etc.
+      #
+      # Boosted fields are specified in a hash of field names to a boost, as
+      # with the #fields and #boost_fields methods.
+      #
+      # === Example
+      #
+      #   Sunspot.search(Post) do
+      #     keywords 'nothing reveals like relevance' do
+      #       phrase_fields :title => 2.0
+      #     end
+      #   end
+      #
+      def phrase_fields(boosted_fields)
+        if boosted_fields
+          boosted_fields.each_pair do |field_name, boost|
+            @setup.text_fields(field_name).each do |field|
+              @query.add_phrase_field(field, boost)
+            end
+          end
+        end
+      end
+
+      # 
+      # The maximum number of words that can appear between search terms for a
+      # field to qualify for phrase field boost. See #query_phrase_slop for
+      # examples. Phrase slop is only meaningful if phrase fields are specified
+      # (see #phrase_fields), and it does not have an effect on which results
+      # are returned; only on what their respective boosts are.
+      #
+      def phrase_slop(slop)
+        @query.phrase_slop = slop
+      end
+
+      # 
+      # Boost queries allow specification of an arbitrary scope for which
+      # matching documents should receive an extra boost. You can either specify 
+      # a boost factor and a block, or a boost function. The block is evaluated
+      # in the usual scope DSL, and field names are attribute fields, not text
+      # fields, as in other scope.
+      #
+      # The boost function can be a constant (numeric or string literal), 
+      # a field name or another function. You can build arbitrarily complex 
+      # functions, which are passed transparently to solr.
+      #
+      # This method can be called more than once for different boost queries
+      # with different boosts.
+      #
+      # === Example
+      #
+      #   Sunspot.search(Post) do
+      #     keywords 'super fan' do
+      #       boost(2.0) do
+      #         with(:featured, true)
+      #       end
+      #
+      #       boost(function { sum(:average_rating, product(:popularity, 10)) })
+      #     end
+      #   end
+      #
+      # In the above search, featured posts will receive a boost of 2.0 and all posts 
+      # will be boosted by (average_rating + popularity * 10).
+      #
+      def boost(factor_or_function, &block)
+        if factor_or_function.is_a?(Sunspot::Query::FunctionQuery)
+          @query.add_boost_function(factor_or_function)
+        else
+          Sunspot::Util.instance_eval_or_call(
+            Scope.new(@query.create_boost_query(factor_or_function), @setup),
+            &block
+          )
+        end
+      end
+
+      #
+      # Add boost to certain fields, without restricting which fields are
+      # searched.
+      #
+      # === Example
+      #
+      #   Sunspot.search(Post) do
+      #     keywords('pork sandwich') do
+      #       boost_fields :title => 1.5
+      #     end
+      #   end
+      #
+      def boost_fields(boosts)
+        boosts.each_pair do |field_name, boost|
+          begin
+            @setup.text_fields(field_name).each do |field|
+              @query.add_fulltext_field(field, boost)
+            end
+          rescue Sunspot::UnrecognizedFieldError
+            # We'll let this one slide.
+          end
+        end
+      end
+      
+      #
+      # The minimum number of search terms that a result must match. By
+      # default, all search terms must match; if the number of search terms
+      # is less than this number, the default behavior applies.
+      #
+      def minimum_match(minimum_match)
+        @query.minimum_match = minimum_match
+      end
+
+      #
+      # The number of words that can appear between the words in a
+      # user-entered phrase (i.e., keywords in quotes) and still match. For
+      # instance, in a search for "\"great pizza\"" with a query phrase slop of
+      # 1, "great pizza" and "great big pizza" will match, but "great monster of
+      # a pizza" will not. Default behavior is a query phrase slop of zero.
+      #
+      def query_phrase_slop(slop)
+        @query.query_phrase_slop = slop
+      end
+
+      #
+      # A tiebreaker coefficient for scores derived from subqueries that are
+      # lower-scoring than the maximum score subquery. Typically a near-zero
+      # value is useful. See
+      # http://wiki.apache.org/solr/DisMaxRequestHandler#tie_.28Tie_breaker.29
+      # for more information.
+      #
+      def tie(tie)
+        @query.tie = tie
+      end
+
+      def fields_added? #:nodoc:
+        @fields_added
+      end
+    end
+  end
+end

data/lib/sunspot/dsl/function.rb

@@ -0,0 +1,27 @@
+module Sunspot
+  module DSL
+    class Function #:nodoc:
+      def initialize(functional) #:nodoc:
+        @functional = functional
+      end
+
+      # Special case to handle <http://wiki.apache.org/solr/FunctionQuery#sub>
+      # because `Kernel#sub` exists so `method_missing` will not be called
+      # for this function.
+      def sub(*args) #:nodoc:
+        create_function_query(:sub, *args)
+      end
+
+      def method_missing(method, *args, &block)
+        create_function_query(method, *args)
+      end
+
+      private
+
+      def create_function_query(method, *args)
+        function_args = args.map { |arg| @functional.create_function_query(arg) }
+        Sunspot::Query::FunctionalFunctionQuery.new(method, function_args)
+      end
+    end
+  end
+end