module Pluckers::Features::Base::HasAndBelongsToManyReflections
This module implements plucking has_and_belongs_to_many relationships in a recursive way.
The options used in this feature are:
* reflections: A hash of the reflections we will pluck recursively. The key of this hash will be the name of the reflection and the value is another hash of options. - scope: You can limit the scope of the objects plucked. E.g, you could use Author.active instead of Author.all. Notice that .all is the default. - plucker: You can use a custom plucker instead of Pluckers::Base in case you want any specific logic. Pluckers::Base is the default one. - Any other option will be passed to the plucker, so you can send any other regular option such as attributes, custom ones or even more reflections. Recursivity FTW!!
Public Instance Methods
In this method we get the reflections and for each one creates and executes a new plucker.
This pluck gives the whole process a recursive character and options for that plucker may be passed in the options hash.
# File lib/pluckers/features/base/has_and_belongs_to_many_reflections.rb, line 68 def build_results super return if @has_and_belongs_to_many_reflections.blank? build_only_ids_has_and_belongs_to_many_reflections build_complete_has_and_belongs_to_many_reflections end
Here we obtain the has_many reflections to include in the pluck operation and also include the relation foreign key in the attributes to pluck for this model.
# File lib/pluckers/features/base/has_and_belongs_to_many_reflections.rb, line 41 def configure_query super pluck_reflections = @options[:reflections] || {} return if pluck_reflections.blank? @has_and_belongs_to_many_reflections = { } # We iterate through the class reflections passed as options @klass_reflections.slice(*pluck_reflections.keys). # And select those that are HasMany select{|_, r| active_record_has_and_belongs_to_many_reflection?(r) }. # And store them in the has_many_reflection hash that will be used later each do |name, reflection| name = name.to_sym @has_and_belongs_to_many_reflections[name] = pluck_reflections[name] end end
Private Instance Methods
This method build the reflections completely, creating hashes for each record, etc.
It searches reflections that has not the :only_ids option enabled and then creates pluckers for them.
# File lib/pluckers/features/base/has_and_belongs_to_many_reflections.rb, line 83 def build_complete_has_and_belongs_to_many_reflections @has_and_belongs_to_many_reflections.reject {|_, reflection| reflection[:only_ids] }.each do |name, reflection| # As an example we will imagine that we are plucking Authors and # this relation is the :posts # We get the meta information about the reflection klass_reflection = @klass_reflections[name] # initialize some options such as the plucker or the scope of the pluck scope = reflection[:scope] || klass_reflection.klass.send(all_method) plucker = reflection[:plucker] || Pluckers::Base # We will use the _ids already fetched to check which records we should pluck ids_reflection_name = "#{name.to_s.singularize}_ids".to_sym ids_to_query = @results.map do |_, result| result[ids_reflection_name] end ids_to_query = ids_to_query.flatten # And now we create the plucker. Notice that we add a where to the # scope, so we filter the records to pluck as we only get those with # an id in the set of the _ids arrays already plucked # # In our Example we would be doing something like # Category.all.where(id: category_ids) reflection_plucker = plucker.new scope.where(id: ids_to_query), reflection # We initialize so we return an empty array if there are no record # related @results.each do |_, result| result[name] ||= [] end reflection_plucker.pluck.each do |r| @results.each do |_,result| # For each related result (category) we search those records # (BlogPost) that include the category id in its _ids array if result[ids_reflection_name].include? r[:id].to_i result[name] << r end end end # And now we get rid of duplicates @results.each do |_,result| result[name].uniq! end end end
This method build the ids for the records instead of creating the hashes.
Unlike the has_many relationships, we don't search reflections that has the :only_ids option enabled as we will need these ids also for the other relationships.
# File lib/pluckers/features/base/has_and_belongs_to_many_reflections.rb, line 146 def build_only_ids_has_and_belongs_to_many_reflections @has_and_belongs_to_many_reflections.each do |name, reflection| # As an example we will imagine that we are plucking BlogPosts and # this relation is the :categories one # We get the meta information about the reflection klass_reflection = @klass_reflections[name] # We get the ids. This query is dependant on the ActiveRecord # version, so every feature version has a different implementation join_results = has_and_belongs_to_many_ids(klass_reflection) ids_reflection_name = "#{name.to_s.singularize}_ids".to_sym # Next, we initialize the _ids array for each result @results.each do |_, result| result[ids_reflection_name] ||= [] end # And now, the foreign_keys. # In our example with BlogPost and Category they would be: # model_foreign_key = blog_post_id # related_model_foreign_key = category_id model_foreign_key = klass_reflection.foreign_key related_model_foreign_key = klass_reflection.association_foreign_key # And for each result we fill the results join_results.each do |r| @results[r[model_foreign_key].to_i][ids_reflection_name] << r[related_model_foreign_key].to_i end # And eliminate duplicates @results.each do |_,result| result[ids_reflection_name].uniq! end end end