searchkick 2.3.2 → 5.2.1

data/lib/searchkick/indexer.rb

@@ -1,3 +1,5 @@
+# thread-local (technically fiber-local) indexer
+# used to aggregate bulk callbacks across models
 module Searchkick
   class Indexer
     attr_reader :queued_items
@@ -14,15 +16,20 @@ module Searchkick
     def perform
       items = @queued_items
       @queued_items = []
-      if items.any?
-        response = Searchkick.client.bulk(body: items)
-        if response["errors"]
-          first_with_error = response["items"].map do |item|
-            (item["index"] || item["delete"] || item["update"])
-          end.find { |item| item["error"] }
-          raise Searchkick::ImportError, "#{first_with_error["error"]} on item with id '#{first_with_error["_id"]}'"
-        end
+
+      return if items.empty?
+
+      response = Searchkick.client.bulk(body: items)
+      if response["errors"]
+        # note: delete does not set error when item not found
+        first_with_error = response["items"].map do |item|
+          (item["index"] || item["delete"] || item["update"])
+        end.find { |item| item["error"] }
+        raise ImportError, "#{first_with_error["error"]} on item with id '#{first_with_error["_id"]}'"
       end
+
+      # maybe return response in future
+      nil
     end
   end
 end

data/lib/searchkick/log_subscriber.rb

@@ -0,0 +1,57 @@
+# based on https://gist.github.com/mnutt/566725
+module Searchkick
+  class LogSubscriber < ActiveSupport::LogSubscriber
+    def self.runtime=(value)
+      Thread.current[:searchkick_runtime] = value
+    end
+
+    def self.runtime
+      Thread.current[:searchkick_runtime] ||= 0
+    end
+
+    def self.reset_runtime
+      rt = runtime
+      self.runtime = 0
+      rt
+    end
+
+    def search(event)
+      self.class.runtime += event.duration
+      return unless logger.debug?
+
+      payload = event.payload
+      name = "#{payload[:name]} (#{event.duration.round(1)}ms)"
+
+      index = payload[:query][:index].is_a?(Array) ? payload[:query][:index].join(",") : payload[:query][:index]
+      type = payload[:query][:type]
+      request_params = payload[:query].except(:index, :type, :body)
+
+      params = []
+      request_params.each do |k, v|
+        params << "#{CGI.escape(k.to_s)}=#{CGI.escape(v.to_s)}"
+      end
+
+      debug "  #{color(name, YELLOW, true)}  #{index}#{type ? "/#{type.join(',')}" : ''}/_search#{params.any? ? '?' + params.join('&') : nil} #{payload[:query][:body].to_json}"
+    end
+
+    def request(event)
+      self.class.runtime += event.duration
+      return unless logger.debug?
+
+      payload = event.payload
+      name = "#{payload[:name]} (#{event.duration.round(1)}ms)"
+
+      debug "  #{color(name, YELLOW, true)}  #{payload.except(:name).to_json}"
+    end
+
+    def multi_search(event)
+      self.class.runtime += event.duration
+      return unless logger.debug?
+
+      payload = event.payload
+      name = "#{payload[:name]} (#{event.duration.round(1)}ms)"
+
+      debug "  #{color(name, YELLOW, true)}  _msearch #{payload[:body]}"
+    end
+  end
+end

data/lib/searchkick/middleware.rb

@@ -1,10 +1,17 @@
-require "faraday/middleware"
+require "faraday"
 
 module Searchkick
   class Middleware < Faraday::Middleware
     def call(env)
-      if env[:method] == :get && env[:url].path.to_s.end_with?("/_search")
+      path = env[:url].path.to_s
+      if path.end_with?("/_search")
         env[:request][:timeout] = Searchkick.search_timeout
+      elsif path.end_with?("/_msearch")
+        # assume no concurrent searches for timeout for now
+        searches = env[:request_body].count("\n") / 2
+        # do not allow timeout to exceed Searchkick.timeout
+        timeout = [Searchkick.search_timeout * searches, Searchkick.timeout].min
+        env[:request][:timeout] = timeout
       end
       @app.call(env)
     end

data/lib/searchkick/model.rb

@@ -1,154 +1,108 @@
 module Searchkick
   module Model
     def searchkick(**options)
-      unknown_keywords = options.keys - [:_all, :batch_size, :callbacks, :conversions, :default_fields,
-        :filterable, :geo_shape, :highlight, :ignore_above, :index_name, :index_prefix, :language,
-        :locations, :mappings, :match, :merge_mappings, :routing, :searchable, :settings, :similarity,
-        :special_characters, :stem_conversions, :suggest, :synonyms, :text_end,
-        :text_middle, :text_start, :word, :wordnet, :word_end, :word_middle, :word_start]
+      options = Searchkick.model_options.merge(options)
+
+      unknown_keywords = options.keys - [:_all, :_type, :batch_size, :callbacks, :case_sensitive, :conversions, :deep_paging, :default_fields,
+        :filterable, :geo_shape, :highlight, :ignore_above, :index_name, :index_prefix, :inheritance, :language,
+        :locations, :mappings, :match, :max_result_window, :merge_mappings, :routing, :searchable, :search_synonyms, :settings, :similarity,
+        :special_characters, :stem, :stemmer, :stem_conversions, :stem_exclusion, :stemmer_override, :suggest, :synonyms, :text_end,
+        :text_middle, :text_start, :unscope, :word, :word_end, :word_middle, :word_start]
       raise ArgumentError, "unknown keywords: #{unknown_keywords.join(", ")}" if unknown_keywords.any?
 
       raise "Only call searchkick once per model" if respond_to?(:searchkick_index)
 
       Searchkick.models << self
 
-      class_eval do
-        cattr_reader :searchkick_options, :searchkick_klass
+      options[:_type] ||= -> { searchkick_index.klass_document_type(self, true) }
+      options[:class_name] = model_name.name
+
+      callbacks = options.key?(:callbacks) ? options[:callbacks] : :inline
+      unless [:inline, true, false, :async, :queue].include?(callbacks)
+        raise ArgumentError, "Invalid value for callbacks"
+      end
+
+      base = self
+
+      mod = Module.new
+      include(mod)
+      mod.module_eval do
+        def reindex(method_name = nil, mode: nil, refresh: false)
+          self.class.searchkick_index.reindex([self], method_name: method_name, mode: mode, refresh: refresh, single: true)
+        end unless base.method_defined?(:reindex)
+
+        def similar(**options)
+          self.class.searchkick_index.similar_record(self, **options)
+        end unless base.method_defined?(:similar)
+
+        def search_data
+          data = respond_to?(:to_hash) ? to_hash : serializable_hash
+          data.delete("id")
+          data.delete("_id")
+          data.delete("_type")
+          data
+        end unless base.method_defined?(:search_data)
+
+        def should_index?
+          true
+        end unless base.method_defined?(:should_index?)
+      end
 
-        callbacks = options.key?(:callbacks) ? options[:callbacks] : true
+      class_eval do
+        cattr_reader :searchkick_options, :searchkick_klass, instance_reader: false
 
         class_variable_set :@@searchkick_options, options.dup
         class_variable_set :@@searchkick_klass, self
-        class_variable_set :@@searchkick_callbacks, callbacks
-        class_variable_set :@@searchkick_index, options[:index_name] ||
-          (options[:index_prefix].respond_to?(:call) && proc { [options[:index_prefix].call, model_name.plural, Searchkick.env, Searchkick.index_suffix].compact.join("_") }) ||
-          [options.key?(:index_prefix) ? options[:index_prefix] : Searchkick.index_prefix, model_name.plural, Searchkick.env, Searchkick.index_suffix].compact.join("_")
+        class_variable_set :@@searchkick_index_cache, Searchkick::IndexCache.new
 
         class << self
           def searchkick_search(term = "*", **options, &block)
-            Searchkick.search(term, {model: self}.merge(options), &block)
+            if Searchkick.relation?(self)
+              raise Searchkick::Error, "search must be called on model, not relation"
+            end
+
+            Searchkick.search(term, model: self, **options, &block)
           end
           alias_method Searchkick.search_method_name, :searchkick_search if Searchkick.search_method_name
 
-          def searchkick_index
-            index = class_variable_get :@@searchkick_index
-            index = index.call if index.respond_to? :call
-            Searchkick::Index.new(index, searchkick_options)
+          def searchkick_index(name: nil)
+            index_name = name || searchkick_klass.searchkick_index_name
+            index_name = index_name.call if index_name.respond_to?(:call)
+            index_cache = class_variable_get(:@@searchkick_index_cache)
+            index_cache.fetch(index_name) { Searchkick::Index.new(index_name, searchkick_options) }
           end
           alias_method :search_index, :searchkick_index unless method_defined?(:search_index)
 
-          def enable_search_callbacks
-            class_variable_set :@@searchkick_callbacks, true
-          end
-
-          def disable_search_callbacks
-            class_variable_set :@@searchkick_callbacks, false
-          end
-
-          def search_callbacks?
-            class_variable_get(:@@searchkick_callbacks) && Searchkick.callbacks?
-          end
-
-          def searchkick_reindex(method_name = nil, full: false, **options)
-            scoped = (respond_to?(:current_scope) && respond_to?(:default_scoped) && current_scope && current_scope.to_sql != default_scoped.to_sql) ||
-              (respond_to?(:queryable) && queryable != unscoped.with_default_scope)
-
-            refresh = options.fetch(:refresh, !scoped)
-
-            if method_name
-              # update
-              searchkick_index.import_scope(searchkick_klass, method_name: method_name)
-              searchkick_index.refresh if refresh
-              true
-            elsif scoped && !full
-              # reindex association
-              searchkick_index.import_scope(searchkick_klass)
-              searchkick_index.refresh if refresh
-              true
-            else
-              # full reindex
-              searchkick_index.reindex_scope(searchkick_klass, options)
-            end
+          def searchkick_reindex(method_name = nil, **options)
+            searchkick_index.reindex(self, method_name: method_name, **options)
           end
           alias_method :reindex, :searchkick_reindex unless method_defined?(:reindex)
 
           def searchkick_index_options
             searchkick_index.index_options
           end
-        end
-
-        callback_name = callbacks == :async ? :reindex_async : :reindex
-        if respond_to?(:after_commit)
-          after_commit callback_name, if: proc { self.class.search_callbacks? }
-        elsif respond_to?(:after_save)
-          after_save callback_name, if: proc { self.class.search_callbacks? }
-          after_destroy callback_name, if: proc { self.class.search_callbacks? }
-        end
 
-        def reindex(method_name = nil, refresh: false, async: false, mode: nil)
-          klass_options = self.class.searchkick_index.options
-
-          if mode.nil?
-            mode =
-              if async
-                :async
-              elsif Searchkick.callbacks_value
-                Searchkick.callbacks_value
-              elsif klass_options.key?(:callbacks) && klass_options[:callbacks] != :async
-                # TODO remove 2nd condition in next major version
-                klass_options[:callbacks]
+          def searchkick_index_name
+            @searchkick_index_name ||= begin
+              options = class_variable_get(:@@searchkick_options)
+              if options[:index_name]
+                options[:index_name]
+              elsif options[:index_prefix].respond_to?(:call)
+                -> { [options[:index_prefix].call, model_name.plural, Searchkick.env, Searchkick.index_suffix].compact.join("_") }
+              else
+                [options.key?(:index_prefix) ? options[:index_prefix] : Searchkick.index_prefix, model_name.plural, Searchkick.env, Searchkick.index_suffix].compact.join("_")
               end
-          end
-
-          case mode
-          when :queue
-            if method_name
-              raise Searchkick::Error, "Partial reindex not supported with queue option"
-            else
-              self.class.searchkick_index.reindex_queue.push(id.to_s)
-            end
-          when :async
-            if method_name
-              # TODO support Mongoid and NoBrainer and non-id primary keys
-              Searchkick::BulkReindexJob.perform_later(
-                class_name: self.class.name,
-                record_ids: [id.to_s],
-                method_name: method_name ? method_name.to_s : nil
-              )
-            else
-              self.class.searchkick_index.reindex_record_async(self)
             end
-          else
-            if method_name
-              self.class.searchkick_index.update_record(self, method_name)
-            else
-              self.class.searchkick_index.reindex_record(self)
-            end
-            self.class.searchkick_index.refresh if refresh
           end
-        end unless method_defined?(:reindex)
-
-        # TODO remove this method in next major version
-        def reindex_async
-          reindex(async: true)
-        end unless method_defined?(:reindex_async)
-
-        def similar(options = {})
-          self.class.searchkick_index.similar_record(self, options)
-        end unless method_defined?(:similar)
-
-        def search_data
-          respond_to?(:to_hash) ? to_hash : serializable_hash
-        end unless method_defined?(:search_data)
-
-        def should_index?
-          true
-        end unless method_defined?(:should_index?)
+        end
 
-        if defined?(Cequel) && self < Cequel::Record && !method_defined?(:destroyed?)
-          def destroyed?
-            transient?
-          end
+        # always add callbacks, even when callbacks is false
+        # so Model.callbacks block can be used
+        if respond_to?(:after_commit)
+          after_commit :reindex, if: -> { Searchkick.callbacks?(default: callbacks) }
+        elsif respond_to?(:after_save)
+          after_save :reindex, if: -> { Searchkick.callbacks?(default: callbacks) }
+          after_destroy :reindex, if: -> { Searchkick.callbacks?(default: callbacks) }
         end
       end
     end

data/lib/searchkick/multi_search.rb

@@ -2,25 +2,24 @@ module Searchkick
   class MultiSearch
     attr_reader :queries
 
-    def initialize(queries, retry_misspellings: false)
+    def initialize(queries)
       @queries = queries
-      @retry_misspellings = retry_misspellings
     end
 
     def perform
       if queries.any?
-        perform_search(queries, retry_misspellings: @retry_misspellings)
+        perform_search(queries)
       end
     end
 
     private
 
-    def perform_search(queries, retry_misspellings: true)
-      responses = client.msearch(body: queries.flat_map { |q| [q.params.except(:body), q.body] })["responses"]
+    def perform_search(search_queries, perform_retry: true)
+      responses = client.msearch(body: search_queries.flat_map { |q| [q.params.except(:body), q.body] })["responses"]
 
       retry_queries = []
-      queries.each_with_index do |query, i|
-        if retry_misspellings && query.retry_misspellings?(responses[i])
+      search_queries.each_with_index do |query, i|
+        if perform_retry && query.retry_misspellings?(responses[i])
           query.send(:prepare) # okay, since we don't want to expose this method outside Searchkick
           retry_queries << query
         else
@@ -28,11 +27,11 @@ module Searchkick
         end
       end
 
-      if retry_misspellings && retry_queries.any?
-        perform_search(retry_queries, retry_misspellings: false)
+      if retry_queries.any?
+        perform_search(retry_queries, perform_retry: false)
       end
 
-      queries
+      search_queries
     end
 
     def client

data/lib/searchkick/process_batch_job.rb

@@ -1,23 +1,20 @@
 module Searchkick
   class ProcessBatchJob < ActiveJob::Base
-    queue_as :searchkick
+    queue_as { Searchkick.queue_name }
 
-    def perform(class_name:, record_ids:)
-      klass = class_name.constantize
-      scope = Searchkick.load_records(klass, record_ids)
-      scope = scope.search_import if scope.respond_to?(:search_import)
-      records = scope.select(&:should_index?)
+    def perform(class_name:, record_ids:, index_name: nil)
+      model = Searchkick.load_model(class_name)
+      index = model.searchkick_index(name: index_name)
 
-      # determine which records to delete
-      delete_ids = record_ids - records.map { |r| r.id.to_s }
-      delete_records = delete_ids.map { |id| m = klass.new; m.id = id; m }
+      items =
+        record_ids.map do |r|
+          parts = r.split(/(?<!\|)\|(?!\|)/, 2)
+            .map { |v| v.gsub("||", "|") }
+          {id: parts[0], routing: parts[1]}
+        end
 
-      # bulk reindex
-      index = klass.searchkick_index
-      Searchkick.callbacks(:bulk) do
-        index.bulk_index(records) if records.any?
-        index.bulk_delete(delete_records) if delete_records.any?
-      end
+      relation = Searchkick.scope(model)
+      RecordIndexer.new(index).reindex_items(relation, items, method_name: nil)
     end
   end
 end

data/lib/searchkick/process_queue_job.rb

@@ -1,22 +1,31 @@
 module Searchkick
   class ProcessQueueJob < ActiveJob::Base
-    queue_as :searchkick
+    queue_as { Searchkick.queue_name }
 
-    def perform(class_name:)
-      model = class_name.constantize
+    def perform(class_name:, index_name: nil, inline: false)
+      model = Searchkick.load_model(class_name)
+      index = model.searchkick_index(name: index_name)
+      limit = model.searchkick_options[:batch_size] || 1000
 
-      limit = model.searchkick_index.options[:batch_size] || 1000
-      record_ids = model.searchkick_index.reindex_queue.reserve(limit: limit)
-      if record_ids.any?
-        Searchkick::ProcessBatchJob.perform_later(
-          class_name: model.name,
-          record_ids: record_ids
-        )
-        # TODO when moving to reliable queuing, mark as complete
+      loop do
+        record_ids = index.reindex_queue.reserve(limit: limit)
+        if record_ids.any?
+          batch_options = {
+            class_name: class_name,
+            record_ids: record_ids.uniq,
+            index_name: index_name
+          }
 
-        if record_ids.size == limit
-          Searchkick::ProcessQueueJob.perform_later(class_name: class_name)
+          if inline
+            # use new.perform to avoid excessive logging
+            Searchkick::ProcessBatchJob.new.perform(**batch_options)
+          else
+            Searchkick::ProcessBatchJob.perform_later(**batch_options)
+          end
+
+          # TODO when moving to reliable queuing, mark as complete
         end
+        break unless record_ids.size == limit
       end
     end
   end