searchkick 4.4.0 → 5.3.1

data/lib/searchkick/query.rb

@@ -1,5 +1,6 @@
 module Searchkick
   class Query
+    include Enumerable
     extend Forwardable
 
     @@metric_aggs = [:avg, :cardinality, :max, :min, :sum]
@@ -12,11 +13,12 @@ module Searchkick
       :took, :error, :model_name, :entry_name, :total_count, :total_entries,
       :current_page, :per_page, :limit_value, :padding, :total_pages, :num_pages,
       :offset_value, :offset, :previous_page, :prev_page, :next_page, :first_page?, :last_page?,
-      :out_of_range?, :hits, :response, :to_a, :first, :scroll
+      :out_of_range?, :hits, :response, :to_a, :first, :scroll, :highlights, :with_highlights,
+      :with_score, :misspellings?, :scroll_id, :clear_scroll, :missing_records, :with_hit
 
     def initialize(klass, term = "*", **options)
       unknown_keywords = options.keys - [:aggs, :block, :body, :body_options, :boost,
-        :boost_by, :boost_by_distance, :boost_by_recency, :boost_where, :conversions, :conversions_term, :debug, :emoji, :exclude, :execute, :explain,
+        :boost_by, :boost_by_distance, :boost_by_recency, :boost_where, :conversions, :conversions_term, :debug, :emoji, :exclude, :explain,
         :fields, :highlight, :includes, :index_name, :indices_boost, :limit, :load,
         :match, :misspellings, :models, :model_includes, :offset, :operator, :order, :padding, :page, :per_page, :profile,
         :request_params, :routing, :scope_results, :scroll, :select, :similar, :smart_aggs, :suggest, :total_entries, :track, :type, :where]
@@ -25,7 +27,7 @@ module Searchkick
       term = term.to_s
 
       if options[:emoji]
-        term = EmojiParser.parse_unicode(term) { |e| " #{e.name} " }.strip
+        term = EmojiParser.parse_unicode(term) { |e| " #{e.name.tr('_', ' ')} " }.strip
       end
 
       @klass = klass
@@ -73,7 +75,8 @@ module Searchkick
         elsif searchkick_index
           searchkick_index.name
         else
-          "_all"
+          # fixes warning about accessing system indices
+          "*,-.*"
         end
 
       params = {
@@ -109,7 +112,12 @@ module Searchkick
       request_params = query.except(:index, :type, :body)
 
       # no easy way to tell which host the client will use
-      host = Searchkick.client.transport.hosts.first
+      host =
+        if Searchkick.client.transport.respond_to?(:transport)
+          Searchkick.client.transport.transport.hosts.first
+        else
+          Searchkick.client.transport.hosts.first
+        end
       credentials = host[:user] || host[:password] ? "#{host[:user]}:#{host[:password]}@" : nil
       params = ["pretty"]
       request_params.each do |k, v|
@@ -140,9 +148,6 @@ module Searchkick
       }
 
       if options[:debug]
-        # can remove when minimum Ruby version is 2.5
-        require "pp"
-
         puts "Searchkick Version: #{Searchkick::VERSION}"
         puts "Elasticsearch Version: #{Searchkick.server_version}"
         puts
@@ -182,11 +187,11 @@ module Searchkick
       end
 
       # set execute for multi search
-      @execute = Searchkick::Results.new(searchkick_klass, response, opts)
+      @execute = Results.new(searchkick_klass, response, opts)
     end
 
     def retry_misspellings?(response)
-      @misspellings_below && Searchkick::Results.new(searchkick_klass, response).total_count < @misspellings_below
+      @misspellings_below && response["error"].nil? && Results.new(searchkick_klass, response).total_count < @misspellings_below
     end
 
     private
@@ -194,7 +199,11 @@ module Searchkick
     def handle_error(e)
       status_code = e.message[1..3].to_i
       if status_code == 404
-        raise MissingIndexError, "Index missing - run #{reindex_command}"
+        if e.message.include?("No search context found for id")
+          raise MissingIndexError, "No search context found for id"
+        else
+          raise MissingIndexError, "Index missing - run #{reindex_command}"
+        end
       elsif status_code == 500 && (
         e.message.include?("IllegalArgumentException[minimumSimilarity >= 1]") ||
         e.message.include?("No query registered for [multi_match]") ||
@@ -202,14 +211,14 @@ module Searchkick
         e.message.include?("No query registered for [function_score]")
       )
 
-        raise UnsupportedVersionError, "This version of Searchkick requires Elasticsearch 5 or greater"
+        raise UnsupportedVersionError
       elsif status_code == 400
         if (
           e.message.include?("bool query does not support [filter]") ||
           e.message.include?("[bool] filter does not support [filter]")
         )
 
-          raise UnsupportedVersionError, "This version of Searchkick requires Elasticsearch 5 or greater"
+          raise UnsupportedVersionError
         elsif e.message =~ /analyzer \[searchkick_.+\] not found/
           raise InvalidQueryError, "Bad mapping - run #{reindex_command}"
         else
@@ -225,7 +234,14 @@ module Searchkick
     end
 
     def execute_search
-      Searchkick.client.search(params)
+      name = searchkick_klass ? "#{searchkick_klass.name} Search" : "Search"
+      event = {
+        name: name,
+        query: params
+      }
+      ActiveSupport::Notifications.instrument("search.searchkick", event) do
+        Searchkick.client.search(params)
+      end
     end
 
     def prepare
@@ -239,9 +255,15 @@ module Searchkick
       default_limit = searchkick_options[:deep_paging] ? 1_000_000_000 : 10_000
       per_page = (options[:limit] || options[:per_page] || default_limit).to_i
       padding = [options[:padding].to_i, 0].max
-      offset = options[:offset] || (page - 1) * per_page + padding
+      offset = (options[:offset] || (page - 1) * per_page + padding).to_i
       scroll = options[:scroll]
 
+      max_result_window = searchkick_options[:max_result_window]
+      if max_result_window
+        offset = max_result_window if offset > max_result_window
+        per_page = max_result_window - offset if offset + per_page > max_result_window
+      end
+
       # model and eager loading
       load = options[:load].nil? ? true : options[:load]
 
@@ -260,9 +282,10 @@ module Searchkick
         should = []
 
         if options[:similar]
+          like = options[:similar] == true ? term : options[:similar]
           query = {
             more_like_this: {
-              like: term,
+              like: like,
               min_doc_freq: 1,
               min_term_freq: 1,
               analyzer: "searchkick_search2"
@@ -350,11 +373,11 @@ module Searchkick
             field_misspellings = misspellings && (!misspellings_fields || misspellings_fields.include?(base_field(field)))
 
             if field == "_all" || field.end_with?(".analyzed")
-              shared_options[:cutoff_frequency] = 0.001 unless operator.to_s == "and" || field_misspellings == false || (!below73? && !track_total_hits?)
+              shared_options[:cutoff_frequency] = 0.001 unless operator.to_s == "and" || field_misspellings == false || (!below73? && !track_total_hits?) || match_type == :match_phrase || !below80? || Searchkick.opensearch?
               qs << shared_options.merge(analyzer: "searchkick_search")
 
-              # searchkick_search and searchkick_search2 are the same for ukrainian
-              unless %w(japanese korean polish ukrainian vietnamese).include?(searchkick_options[:language])
+              # searchkick_search and searchkick_search2 are the same for some languages
+              unless %w(japanese japanese2 korean polish ukrainian vietnamese).include?(searchkick_options[:language])
                 qs << shared_options.merge(analyzer: "searchkick_search2")
               end
               exclude_analyzer = "searchkick_search2"
@@ -375,11 +398,6 @@ module Searchkick
 
             if field.start_with?("*.")
               q2 = qs.map { |q| {multi_match: q.merge(fields: [field], type: match_type == :match_phrase ? "phrase" : "best_fields")} }
-              if below61?
-                q2.each do |q|
-                  q[:multi_match].delete(:fuzzy_transpositions)
-                end
-              end
             else
               q2 = qs.map { |q| {match_type => {field => q}} }
             end
@@ -431,7 +449,7 @@ module Searchkick
         payload = {}
 
         # type when inheritance
-        where = (options[:where] || {}).dup
+        where = ensure_permitted(options[:where] || {}).dup
         if searchkick_options[:inheritance] && (options[:type] || (klass != searchkick_klass && searchkick_index))
           where[:type] = [options[:type] || klass].flatten.map { |v| searchkick_index.klass_document_type(v, true) }
         end
@@ -491,7 +509,7 @@ module Searchkick
         set_highlights(payload, fields) if options[:highlight]
 
         # timeout shortly after client times out
-        payload[:timeout] ||= "#{Searchkick.search_timeout + 1}s"
+        payload[:timeout] ||= "#{((Searchkick.search_timeout + 1) * 1000).round}ms"
 
         # An empty array will cause only the _id and _type for each hit to be returned
         # https://www.elastic.co/guide/en/elasticsearch/reference/current/search-request-source-filtering.html
@@ -688,9 +706,9 @@ module Searchkick
     def set_boost_by(multiply_filters, custom_filters)
       boost_by = options[:boost_by] || {}
       if boost_by.is_a?(Array)
-        boost_by = Hash[boost_by.map { |f| [f, {factor: 1}] }]
+        boost_by = boost_by.to_h { |f| [f, {factor: 1}] }
       elsif boost_by.is_a?(Hash)
-        multiply_by, boost_by = boost_by.partition { |_, v| v.delete(:boost_mode) == "multiply" }.map { |i| Hash[i] }
+        multiply_by, boost_by = boost_by.partition { |_, v| v.delete(:boost_mode) == "multiply" }.map(&:to_h)
       end
       boost_by[options[:boost]] = {factor: 1} if options[:boost]
 
@@ -755,7 +773,7 @@ module Searchkick
 
     def set_highlights(payload, fields)
       payload[:highlight] = {
-        fields: Hash[fields.map { |f| [f, {}] }],
+        fields: fields.to_h { |f| [f, {}] },
         fragment_size: 0
       }
 
@@ -789,7 +807,7 @@ module Searchkick
       aggs = options[:aggs]
       payload[:aggs] = {}
 
-      aggs = Hash[aggs.map { |f| [f, {}] }] if aggs.is_a?(Array) # convert to more advanced syntax
+      aggs = aggs.to_h { |f| [f, {}] } if aggs.is_a?(Array) # convert to more advanced syntax
       aggs.each do |field, agg_options|
         size = agg_options[:limit] ? agg_options[:limit] : 1_000
         shared_agg_options = agg_options.except(:limit, :field, :ranges, :date_ranges, :where)
@@ -828,8 +846,9 @@ module Searchkick
         end
 
         where = {}
-        where = (options[:where] || {}).reject { |k| k == field } unless options[:smart_aggs] == false
-        agg_filters = where_filters(where.merge(agg_options[:where] || {}))
+        where = ensure_permitted(options[:where] || {}).reject { |k| k == field } unless options[:smart_aggs] == false
+        agg_where = ensure_permitted(agg_options[:where] || {})
+        agg_filters = where_filters(where.merge(agg_where))
 
         # only do one level comparison for simplicity
         filters.select! do |filter|
@@ -864,19 +883,17 @@ module Searchkick
       }
     end
 
-    # TODO id transformation for arrays
     def set_order(payload)
-      order = options[:order].is_a?(Enumerable) ? options[:order] : {options[:order] => :asc}
-      id_field = :_id
-      payload[:sort] = order.is_a?(Array) ? order : Hash[order.map { |k, v| [k.to_s == "id" ? id_field : k, v] }]
+      payload[:sort] = options[:order].is_a?(Enumerable) ? options[:order] : {options[:order] => :asc}
     end
 
-    def where_filters(where)
-      # if where.respond_to?(:permitted?) && !where.permitted?
-      #   # TODO check in more places
-      #   Searchkick.warn("Passing unpermitted parameters will raise an exception in Searchkick 5")
-      # end
+    # provides *very* basic protection from unfiltered parameters
+    # this is not meant to be comprehensive and may be expanded in the future
+    def ensure_permitted(obj)
+      obj.to_h
+    end
 
+    def where_filters(where)
       filters = []
       (where || {}).each do |field, value|
         field = :_id if field.to_s == "id"
@@ -896,12 +913,7 @@ module Searchkick
         else
           # expand ranges
           if value.is_a?(Range)
-            # infinite? added in Ruby 2.4
-            if value.end.nil? || (value.end.respond_to?(:infinite?) && value.end.infinite?)
-              value = {gte: value.first}
-            else
-              value = {gte: value.first, (value.exclude_end? ? :lt : :lte) => value.last}
-            end
+            value = expand_range(value)
           end
 
           value = {in: value} if value.is_a?(Array)
@@ -953,7 +965,7 @@ module Searchkick
                     }
                   }
                 }
-              when :like
+              when :like, :ilike
                 # based on Postgres
                 # https://www.postgresql.org/docs/current/functions-matching.html
                 # % matches zero or more characters
@@ -961,13 +973,22 @@ module Searchkick
                 # \ is escape character
                 # escape Lucene reserved characters
                 # https://www.elastic.co/guide/en/elasticsearch/reference/current/regexp-syntax.html#regexp-optional-operators
-                reserved = %w(. ? + * | { } [ ] ( ) " \\)
+                reserved = %w(\\ . ? + * | { } [ ] ( ) ")
                 regex = op_value.dup
                 reserved.each do |v|
-                  regex.gsub!(v, "\\" + v)
+                  regex.gsub!(v, "\\\\" + v)
                 end
                 regex = regex.gsub(/(?<!\\)%/, ".*").gsub(/(?<!\\)_/, ".").gsub("\\%", "%").gsub("\\_", "_")
-                filters << {regexp: {field => {value: regex}}}
+
+                if op == :ilike
+                  if below710?
+                    raise ArgumentError, "ilike requires Elasticsearch 7.10+"
+                  else
+                    filters << {regexp: {field => {value: regex, flags: "NONE", case_insensitive: true}}}
+                  end
+                else
+                  filters << {regexp: {field => {value: regex, flags: "NONE"}}}
+                end
               when :prefix
                 filters << {prefix: {field => {value: op_value}}}
               when :regexp # support for regexp queries without using a regexp ruby object
@@ -994,7 +1015,7 @@ module Searchkick
                   when :lte
                     {to: op_value, include_upper: true}
                   else
-                    raise "Unknown where operator: #{op.inspect}"
+                    raise ArgumentError, "Unknown where operator: #{op.inspect}"
                   end
                 # issue 132
                 if (existing = filters.find { |f| f[:range] && f[:range][field] })
@@ -1022,40 +1043,37 @@ module Searchkick
       elsif value.nil?
         {bool: {must_not: {exists: {field: field}}}}
       elsif value.is_a?(Regexp)
-        if value.casefold?
-          Searchkick.warn("Case-insensitive flag does not work with Elasticsearch")
-        end
-
         source = value.source
-        unless source.start_with?("\\A") && source.end_with?("\\z")
-          # https://www.elastic.co/guide/en/elasticsearch/reference/current/query-dsl-regexp-query.html
-          Searchkick.warn("Regular expressions are always anchored in Elasticsearch")
-        end
+
+        # TODO handle other regexp options
 
         # TODO handle other anchor characters, like ^, $, \Z
         if source.start_with?("\\A")
           source = source[2..-1]
         else
-          # TODO uncomment in Searchkick 5
-          # source = ".*#{source}"
+          source = ".*#{source}"
         end
 
         if source.end_with?("\\z")
           source = source[0..-3]
         else
-          # TODO uncomment in Searchkick 5
-          # source = "#{source}.*"
+          source = "#{source}.*"
         end
 
-        {regexp: {field => {value: source, flags: "NONE"}}}
+        if below710?
+          if value.casefold?
+            raise ArgumentError, "Case-insensitive flag does not work with Elasticsearch < 7.10"
+          end
+          {regexp: {field => {value: source, flags: "NONE"}}}
+        else
+          {regexp: {field => {value: source, flags: "NONE", case_insensitive: value.casefold?}}}
+        end
       else
         # TODO add this for other values
         if value.as_json.is_a?(Enumerable)
           # query will fail, but this is better
           # same message as Active Record
-          # TODO make TypeError
-          # raise InvalidQueryError for backward compatibility
-          raise Searchkick::InvalidQueryError, "can't cast #{value.class.name}"
+          raise TypeError, "can't cast #{value.class.name}"
         end
 
         {term: {field => {value: value}}}
@@ -1118,26 +1136,29 @@ module Searchkick
       end
     end
 
+    def expand_range(range)
+      expanded = {}
+      expanded[:gte] = range.begin if range.begin
+
+      if range.end && !(range.end.respond_to?(:infinite?) && range.end.infinite?)
+        expanded[range.exclude_end? ? :lt : :lte] = range.end
+      end
+
+      expanded
+    end
+
     def base_field(k)
       k.sub(/\.(analyzed|word_start|word_middle|word_end|text_start|text_middle|text_end|exact)\z/, "")
     end
 
     def track_total_hits?
-      (searchkick_options[:deep_paging] && !below70?) || body_options[:track_total_hits]
+      searchkick_options[:deep_paging] || body_options[:track_total_hits]
     end
 
     def body_options
       options[:body_options] || {}
     end
 
-    def below61?
-      Searchkick.server_below?("6.1.0")
-    end
-
-    def below70?
-      Searchkick.server_below?("7.0.0")
-    end
-
     def below73?
       Searchkick.server_below?("7.3.0")
     end
@@ -1145,5 +1166,13 @@ module Searchkick
     def below75?
       Searchkick.server_below?("7.5.0")
     end
+
+    def below710?
+      Searchkick.server_below?("7.10.0")
+    end
+
+    def below80?
+      Searchkick.server_below?("8.0.0")
+    end
   end
 end

data/lib/searchkick/record_data.rb

@@ -25,6 +25,7 @@ module Searchkick
       {delete: record_data}
     end
 
+    # custom id can be useful for load: false
     def search_id
       id = record.respond_to?(:search_document_id) ? record.search_document_id : record.id
       id.is_a?(Numeric) ? id : id.to_s
@@ -39,7 +40,6 @@ module Searchkick
         _index: index.name,
         _id: search_id
       }
-      data[:_type] = document_type if Searchkick.server_below7?
       data[:routing] = record.search_routing if record.respond_to?(:search_routing)
       data
     end

data/lib/searchkick/record_indexer.rb

@@ -1,78 +1,163 @@
 module Searchkick
   class RecordIndexer
-    attr_reader :record, :index
+    attr_reader :index
 
-    def initialize(record)
-      @record = record
-      @index = record.class.searchkick_index
+    def initialize(index)
+      @index = index
     end
 
-    def reindex(method_name = nil, refresh: false, mode: nil)
-      unless [:inline, true, nil, :async, :queue].include?(mode)
-        raise ArgumentError, "Invalid value for mode"
-      end
-
-      mode ||= Searchkick.callbacks_value || index.options[:callbacks] || true
+    def reindex(records, mode:, method_name:, full: false, single: false)
+      # prevents exists? check if records is a relation
+      records = records.to_a
+      return if records.empty?
 
       case mode
+      when :async
+        unless defined?(ActiveJob)
+          raise Error, "Active Job not found"
+        end
+
+        # we could likely combine ReindexV2Job, BulkReindexJob, and ProcessBatchJob
+        # but keep them separate for now
+        if single
+          record = records.first
+
+          # always pass routing in case record is deleted
+          # before the async job runs
+          if record.respond_to?(:search_routing)
+            routing = record.search_routing
+          end
+
+          Searchkick::ReindexV2Job.perform_later(
+            record.class.name,
+            record.id.to_s,
+            method_name ? method_name.to_s : nil,
+            routing: routing,
+            index_name: index.name
+          )
+        else
+          Searchkick::BulkReindexJob.perform_later(
+            class_name: records.first.class.searchkick_options[:class_name],
+            record_ids: records.map { |r| r.id.to_s },
+            index_name: index.name,
+            method_name: method_name ? method_name.to_s : nil
+          )
+        end
       when :queue
         if method_name
-          raise Searchkick::Error, "Partial reindex not supported with queue option"
+          raise Error, "Partial reindex not supported with queue option"
         end
 
-        # always pass routing in case record is deleted
-        # before the queue job runs
-        if record.respond_to?(:search_routing)
-          routing = record.search_routing
-        end
+        index.reindex_queue.push_records(records)
+      when true, :inline
+        index_records, other_records = records.partition { |r| index_record?(r) }
+        import_inline(index_records, !full ? other_records : [], method_name: method_name, single: single)
+      else
+        raise ArgumentError, "Invalid value for mode"
+      end
 
-        # escape pipe with double pipe
-        value = queue_escape(record.id.to_s)
-        value = "#{value}|#{queue_escape(routing)}" if routing
-        index.reindex_queue.push(value)
-      when :async
-        unless defined?(ActiveJob)
-          raise Searchkick::Error, "Active Job not found"
-        end
+      # return true like model and relation reindex for now
+      true
+    end
 
-        # always pass routing in case record is deleted
-        # before the async job runs
-        if record.respond_to?(:search_routing)
-          routing = record.search_routing
-        end
+    def reindex_items(klass, items, method_name:, single: false)
+      routing = items.to_h { |r| [r[:id], r[:routing]] }
+      record_ids = routing.keys
 
-        Searchkick::ReindexV2Job.perform_later(
-          record.class.name,
-          record.id.to_s,
-          method_name ? method_name.to_s : nil,
-          routing: routing
-        )
-      else # bulk, inline/true/nil
-        reindex_record(method_name)
+      relation = Searchkick.load_records(klass, record_ids)
+      # call search_import even for single records for nested associations
+      relation = relation.search_import if relation.respond_to?(:search_import)
+      records = relation.select(&:should_index?)
 
-        index.refresh if refresh
-      end
+      # determine which records to delete
+      delete_ids = record_ids - records.map { |r| r.id.to_s }
+      delete_records =
+        delete_ids.map do |id|
+          construct_record(klass, id, routing[id])
+        end
+
+      import_inline(records, delete_records, method_name: method_name, single: single)
     end
 
     private
 
-    def queue_escape(value)
-      value.gsub("|", "||")
+    def index_record?(record)
+      record.persisted? && !record.destroyed? && record.should_index?
     end
 
-    def reindex_record(method_name)
-      if record.destroyed? || !record.persisted? || !record.should_index?
-        begin
-          index.remove(record)
-        rescue Elasticsearch::Transport::Transport::Errors::NotFound
-          # do nothing
+    # import in single request with retries
+    def import_inline(index_records, delete_records, method_name:, single:)
+      return if index_records.empty? && delete_records.empty?
+
+      maybe_bulk(index_records, delete_records, method_name, single) do
+        if index_records.any?
+          if method_name
+            index.bulk_update(index_records, method_name)
+          else
+            index.bulk_index(index_records)
+          end
         end
+
+        if delete_records.any?
+          index.bulk_delete(delete_records)
+        end
+      end
+    end
+
+    def maybe_bulk(index_records, delete_records, method_name, single)
+      if Searchkick.callbacks_value == :bulk
+        yield
       else
-        if method_name
-          index.update_record(record, method_name)
-        else
-          index.store(record)
+        # set action and data
+        action =
+          if single && index_records.empty?
+            "Remove"
+          elsif method_name
+            "Update"
+          else
+            single ? "Store" : "Import"
+          end
+        record = index_records.first || delete_records.first
+        name = record.class.searchkick_klass.name
+        message = lambda do |event|
+          event[:name] = "#{name} #{action}"
+          if single
+            event[:id] = index.search_id(record)
+          else
+            event[:count] = index_records.size + delete_records.size
+          end
+        end
+
+        with_retries do
+          Searchkick.callbacks(:bulk, message: message) do
+            yield
+          end
+        end
+      end
+    end
+
+    def construct_record(klass, id, routing)
+      record = klass.new
+      record.id = id
+      if routing
+        record.define_singleton_method(:search_routing) do
+          routing
+        end
+      end
+      record
+    end
+
+    def with_retries
+      retries = 0
+
+      begin
+        yield
+      rescue Faraday::ClientError => e
+        if retries < 1
+          retries += 1
+          retry
         end
+        raise e
       end
     end
   end