module Sequel::Model::Associations::DatasetMethods

  1. lib/sequel/model/associations.rb

Eager loading makes it so that you can load all associated records for a set of objects in a single query, instead of a separate query for each object.

Two separate implementations are provided. eager should be used most of the time, as it loads associated records using one query per association. However, it does not allow you the ability to filter or order based on columns in associated tables. eager_graph loads all records in a single query using JOINs, allowing you to filter or order based on columns in associated tables. However, eager_graph is usually slower than eager, especially if multiple one_to_many or many_to_many associations are joined.

You can cascade the eager loading (loading associations on associated objects) with no limit to the depth of the cascades. You do this by passing a hash to eager or eager_graph with the keys being associations of the current model and values being associations of the model associated with the current model via the key.

The arguments can be symbols or hashes with symbol keys (for cascaded eager loading). Examples:

Album.eager(:artist).all
Album.eager_graph(:artist).all
Album.eager(:artist, :genre).all
Album.eager_graph(:artist, :genre).all
Album.eager(:artist).eager(:genre).all
Album.eager_graph(:artist).eager_graph(:genre).all
Artist.eager(albums: :tracks).all
Artist.eager_graph(albums: :tracks).all
Artist.eager(albums: {tracks: :genre}).all
Artist.eager_graph(albums: {tracks: :genre}).all

You can also pass a callback as a hash value in order to customize the dataset being eager loaded at query time, analogous to the way the :eager_block association option allows you to customize it at association definition time. For example, if you wanted artists with their albums since 1990:

Artist.eager(albums: proc{|ds| ds.where{year > 1990}})

Or if you needed albums and their artist's name only, using a single query:

Albums.eager_graph(artist: proc{|ds| ds.select(:name)})

To cascade eager loading while using a callback, you substitute the cascaded associations with a single entry hash that has the proc callback as the key and the cascaded associations as the value. This will load artists with their albums since 1990, and also the tracks on those albums and the genre for those tracks:

Artist.eager(albums: {proc{|ds| ds.where{year > 1990}}=>{tracks: :genre}})

Public Instance methods

as_hash(key_column=nil, value_column=nil, opts=OPTS)

If the dataset is being eagerly loaded, default to calling all instead of each.

[show source]
     # File lib/sequel/model/associations.rb
3285 def as_hash(key_column=nil, value_column=nil, opts=OPTS)
3286   if (@opts[:eager_graph] || @opts[:eager]) && !opts.has_key?(:all)
3287     opts = Hash[opts]
3288     opts[:all] = true
3289   end
3290   super
3291 end
association_join(*associations)

Adds one or more INNER JOINs to the existing dataset using the keys and conditions specified by the given association(s). Take the same arguments as eager_graph, and operates similarly, but only adds the joins as opposed to making the other changes (such as adding selected columns and setting up eager loading).

The following methods also exist for specifying a different type of JOIN:

association_full_join

FULL JOIN

association_inner_join

INNER JOIN

association_left_join

LEFT JOIN

association_right_join

RIGHT JOIN

Examples:

# For each album, association_join load the artist
Album.association_join(:artist).all
# SELECT *
# FROM albums
# INNER JOIN artists AS artist ON (artists.id = albums.artist_id)

# For each album, association_join load the artist, using a specified alias
Album.association_join(Sequel[:artist].as(:a)).all
# SELECT *
# FROM albums
# INNER JOIN artists AS a ON (a.id = albums.artist_id)

# For each album, association_join load the artist and genre
Album.association_join(:artist, :genre).all
Album.association_join(:artist).association_join(:genre).all
# SELECT *
# FROM albums
# INNER JOIN artists AS artist ON (artist.id = albums.artist_id)
# INNER JOIN genres AS genre ON (genre.id = albums.genre_id)

# For each artist, association_join load albums and tracks for each album
Artist.association_join(albums: :tracks).all
# SELECT *
# FROM artists
# INNER JOIN albums ON (albums.artist_id = artists.id)
# INNER JOIN tracks ON (tracks.album_id = albums.id)

# For each artist, association_join load albums, tracks for each album, and genre for each track
Artist.association_join(albums: {tracks: :genre}).all
# SELECT *
# FROM artists
# INNER JOIN albums ON (albums.artist_id = artists.id)
# INNER JOIN tracks ON (tracks.album_id = albums.id)
# INNER JOIN genres AS genre ON (genre.id = tracks.genre_id)

# For each artist, association_join load albums with year > 1990
Artist.association_join(albums: proc{|ds| ds.where{year > 1990}}).all
# SELECT *
# FROM artists
# INNER JOIN (
#   SELECT * FROM albums WHERE (year > 1990)
# ) AS albums ON (albums.artist_id = artists.id)

# For each artist, association_join load albums and tracks 1-10 for each album
Artist.association_join(albums: {tracks: proc{|ds| ds.where(number: 1..10)}}).all
# SELECT *
# FROM artists
# INNER JOIN albums ON (albums.artist_id = artists.id)
# INNER JOIN (
#   SELECT * FROM tracks WHERE ((number >= 1) AND (number <= 10))
# ) AS tracks ON (tracks.albums_id = albums.id)

# For each artist, association_join load albums with year > 1990, and tracks for those albums
Artist.association_join(albums: {proc{|ds| ds.where{year > 1990}}=>:tracks}).all
# SELECT *
# FROM artists
# INNER JOIN (
#   SELECT * FROM albums WHERE (year > 1990)
# ) AS albums ON (albums.artist_id = artists.id)
# INNER JOIN tracks ON (tracks.album_id = albums.id)
[show source]
     # File lib/sequel/model/associations.rb
2991 def association_join(*associations)
2992   association_inner_join(*associations)
2993 end
complex_expression_sql_append(sql, op, args)

If the expression is in the form x = y where y is a Sequel::Model instance, array of Sequel::Model instances, or a Sequel::Model dataset, assume x is an association symbol and look up the association reflection via the dataset's model. From there, return the appropriate SQL based on the type of association and the values of the foreign/primary keys of y. For most association types, this is a simple transformation, but for many_to_many associations this creates a subquery to the join table.

[show source]
     # File lib/sequel/model/associations.rb
3002 def complex_expression_sql_append(sql, op, args)
3003   r = args[1]
3004   if (((op == :'=' || op == :'!=') && r.is_a?(Sequel::Model)) ||
3005       (multiple = ((op == :IN || op == :'NOT IN') && ((is_ds = r.is_a?(Sequel::Dataset)) || (r.respond_to?(:all?) && r.all?{|x| x.is_a?(Sequel::Model)})))))
3006     l = args[0]
3007     if ar = model.association_reflections[l]
3008       raise Error, "filtering by associations is not allowed for #{ar.inspect}" if ar[:allow_filtering_by] == false
3009 
3010       if multiple
3011         klass = ar.associated_class
3012         if is_ds
3013           if r.respond_to?(:model)
3014             unless r.model <= klass
3015               # A dataset for a different model class, could be a valid regular query
3016               return super
3017             end
3018           else
3019             # Not a model dataset, could be a valid regular query
3020             return super
3021           end
3022         else
3023           unless r.all?{|x| x.is_a?(klass)}
3024             raise Sequel::Error, "invalid association class for one object for association #{l.inspect} used in dataset filter for model #{model.inspect}, expected class #{klass.inspect}"
3025           end
3026         end
3027       elsif !r.is_a?(ar.associated_class)
3028         raise Sequel::Error, "invalid association class #{r.class.inspect} for association #{l.inspect} used in dataset filter for model #{model.inspect}, expected class #{ar.associated_class.inspect}"
3029       end
3030 
3031       if exp = association_filter_expression(op, ar, r)
3032         literal_append(sql, exp)
3033       else
3034         raise Sequel::Error, "invalid association type #{ar[:type].inspect} for association #{l.inspect} used in dataset filter for model #{model.inspect}"
3035       end
3036     elsif multiple && (is_ds || r.empty?)
3037       # Not a query designed for this support, could be a valid regular query
3038       super
3039     else
3040       raise Sequel::Error, "invalid association #{l.inspect} used in dataset filter for model #{model.inspect}"
3041     end
3042   else
3043     super
3044   end
3045 end
eager(*associations)

The preferred eager loading method. Loads all associated records using one query for each association.

The basic idea for how it works is that the dataset is first loaded normally. Then it goes through all associations that have been specified via eager. It loads each of those associations separately, then associates them back to the original dataset via primary/foreign keys. Due to the necessity of all objects being present, you need to use all to use eager loading, as it can't work with each.

This implementation avoids the complexity of extracting an object graph out of a single dataset, by building the object graph out of multiple datasets, one for each association. By using a separate dataset for each association, it avoids problems such as aliasing conflicts and creating cartesian product result sets if multiple one_to_many or many_to_many eager associations are requested.

One limitation of using this method is that you cannot filter the current dataset based on values of columns in an associated table, since the associations are loaded in separate queries. To do that you need to load all associations in the same query, and extract an object graph from the results of that query. If you need to filter based on columns in associated tables, look at eager_graph or join the tables you need to filter on manually.

Each association's order, if defined, is respected. If the association uses a block or has an :eager_block argument, it is used.

To modify the associated dataset that will be used for the eager load, you should use a hash for the association, with the key being the association name symbol, and the value being a callable object that is called with the associated dataset and should return a modified dataset. If that association also has dependent associations, instead of a callable object, use a hash with the callable object being the key, and the dependent association(s) as the value.

Examples:

# For each album, eager load the artist
Album.eager(:artist).all
# SELECT * FROM albums
# SELECT * FROM artists WHERE (id IN (...))

# For each album, eager load the artist and genre
Album.eager(:artist, :genre).all
Album.eager(:artist).eager(:genre).all
# SELECT * FROM albums
# SELECT * FROM artists WHERE (id IN (...))
# SELECT * FROM genres WHERE (id IN (...))

# For each artist, eager load albums and tracks for each album
Artist.eager(albums: :tracks).all
# SELECT * FROM artists
# SELECT * FROM albums WHERE (artist_id IN (...))
# SELECT * FROM tracks WHERE (album_id IN (...))

# For each artist, eager load albums, tracks for each album, and genre for each track
Artist.eager(albums: {tracks: :genre}).all
# SELECT * FROM artists
# SELECT * FROM albums WHERE (artist_id IN (...))
# SELECT * FROM tracks WHERE (album_id IN (...))
# SELECT * FROM genre WHERE (id IN (...))

# For each artist, eager load albums with year > 1990
Artist.eager(albums: proc{|ds| ds.where{year > 1990}}).all
# SELECT * FROM artists
# SELECT * FROM albums WHERE ((year > 1990) AND (artist_id IN (...)))

# For each artist, eager load albums and tracks 1-10 for each album
Artist.eager(albums: {tracks: proc{|ds| ds.where(number: 1..10)}}).all
# SELECT * FROM artists
# SELECT * FROM albums WHERE (artist_id IN (...))
# SELECT * FROM tracks WHERE ((number >= 1) AND (number <= 10) AND (album_id IN (...)))

# For each artist, eager load albums with year > 1990, and tracks for those albums
Artist.eager(albums: {proc{|ds| ds.where{year > 1990}}=>:tracks}).all
# SELECT * FROM artists
# SELECT * FROM albums WHERE ((year > 1990) AND (artist_id IN (...)))
# SELECT * FROM albums WHERE (artist_id IN (...))
[show source]
     # File lib/sequel/model/associations.rb
3122 def eager(*associations)
3123   opts = @opts[:eager]
3124   association_opts = eager_options_for_associations(associations)
3125   opts = opts ? opts.merge(association_opts) : association_opts
3126   clone(:eager=>opts.freeze)
3127 end
eager_graph(*associations)

The secondary eager loading method. Loads all associations in a single query. This method should only be used if you need to filter or order based on columns in associated tables, or if you have done comparative benchmarking it and determined it is faster.

This method uses Dataset#graph to create appropriate aliases for columns in all the tables. Then it uses the graph's metadata to build the associations from the single hash, and finally replaces the array of hashes with an array model objects inside all.

Be very careful when using this with multiple one_to_many or many_to_many associations, as you can create large cartesian products. If you must graph multiple one_to_many and many_to_many associations, make sure your filters are narrow if the datasets are large.

Each association's order, if defined, is respected. eager_graph probably won't work correctly on a limited dataset, unless you are only graphing many_to_one, one_to_one, and one_through_one associations.

Does not use the block defined for the association, since it does a single query for all objects. You can use the :graph_* association options to modify the SQL query.

Like eager, you need to call all on the dataset for the eager loading to work. If you just call each, it will yield plain hashes, each containing all columns from all the tables.

To modify the associated dataset that will be joined to the current dataset, you should use a hash for the association, with the key being the association name symbol, and the value being a callable object that is called with the associated dataset and should return a modified dataset. If that association also has dependent associations, instead of a callable object, use a hash with the callable object being the key, and the dependent association(s) as the value.

You can specify an custom alias and/or join type on a per-association basis by providing an Sequel::SQL::AliasedExpression object instead of an a Symbol for the association name.

You cannot mix calls to eager_graph and graph on the same dataset.

Examples:

# For each album, eager_graph load the artist
Album.eager_graph(:artist).all
# SELECT ...
# FROM albums
# LEFT OUTER JOIN artists AS artist ON (artists.id = albums.artist_id)

# For each album, eager_graph load the artist, using a specified alias
Album.eager_graph(Sequel[:artist].as(:a)).all
# SELECT ...
# FROM albums
# LEFT OUTER JOIN artists AS a ON (a.id = albums.artist_id)

# For each album, eager_graph load the artist, using a specified alias
# and custom join type

Album.eager_graph(Sequel[:artist].as(:a, join_type: :inner)).all
# SELECT ...
# FROM albums
# INNER JOIN artists AS a ON (a.id = albums.artist_id)

# For each album, eager_graph load the artist and genre
Album.eager_graph(:artist, :genre).all
Album.eager_graph(:artist).eager_graph(:genre).all
# SELECT ...
# FROM albums
# LEFT OUTER JOIN artists AS artist ON (artist.id = albums.artist_id)
# LEFT OUTER JOIN genres AS genre ON (genre.id = albums.genre_id)

# For each artist, eager_graph load albums and tracks for each album
Artist.eager_graph(albums: :tracks).all
# SELECT ...
# FROM artists
# LEFT OUTER JOIN albums ON (albums.artist_id = artists.id)
# LEFT OUTER JOIN tracks ON (tracks.album_id = albums.id)

# For each artist, eager_graph load albums, tracks for each album, and genre for each track
Artist.eager_graph(albums: {tracks: :genre}).all
# SELECT ...
# FROM artists
# LEFT OUTER JOIN albums ON (albums.artist_id = artists.id)
# LEFT OUTER JOIN tracks ON (tracks.album_id = albums.id)
# LEFT OUTER JOIN genres AS genre ON (genre.id = tracks.genre_id)

# For each artist, eager_graph load albums with year > 1990
Artist.eager_graph(albums: proc{|ds| ds.where{year > 1990}}).all
# SELECT ...
# FROM artists
# LEFT OUTER JOIN (
#   SELECT * FROM albums WHERE (year > 1990)
# ) AS albums ON (albums.artist_id = artists.id)

# For each artist, eager_graph load albums and tracks 1-10 for each album
Artist.eager_graph(albums: {tracks: proc{|ds| ds.where(number: 1..10)}}).all
# SELECT ...
# FROM artists
# LEFT OUTER JOIN albums ON (albums.artist_id = artists.id)
# LEFT OUTER JOIN (
#   SELECT * FROM tracks WHERE ((number >= 1) AND (number <= 10))
# ) AS tracks ON (tracks.albums_id = albums.id)

# For each artist, eager_graph load albums with year > 1990, and tracks for those albums
Artist.eager_graph(albums: {proc{|ds| ds.where{year > 1990}}=>:tracks}).all
# SELECT ...
# FROM artists
# LEFT OUTER JOIN (
#   SELECT * FROM albums WHERE (year > 1990)
# ) AS albums ON (albums.artist_id = artists.id)
# LEFT OUTER JOIN tracks ON (tracks.album_id = albums.id)
[show source]
     # File lib/sequel/model/associations.rb
3232 def eager_graph(*associations)
3233   eager_graph_with_options(associations)
3234 end
eager_graph_with_options(associations, opts=OPTS)

Run eager_graph with some options specific to just this call. Unlike eager_graph, this takes the associations as a single argument instead of multiple arguments.

Options:

:join_type

Override the join type specified in the association

:limit_strategy

Use a strategy for handling limits on associations. Appropriate :limit_strategy values are:

true

Pick the most appropriate based on what the database supports

:distinct_on

Force use of DISTINCT ON stategy (*_one associations only)

:correlated_subquery

Force use of correlated subquery strategy (one_to_* associations only)

:window_function

Force use of window function strategy

:ruby

Don't modify the SQL, implement limits/offsets with array slicing

This can also be a hash with association name symbol keys and one of the above values, to use different strategies per association.

The default is the :ruby strategy. Choosing a different strategy can make your code significantly slower in some cases (perhaps even the majority of cases), so you should only use this if you have benchmarked that it is faster for your use cases.

[show source]
     # File lib/sequel/model/associations.rb
3256 def eager_graph_with_options(associations, opts=OPTS)
3257   return self if associations.empty?
3258 
3259   opts = opts.dup unless opts.frozen?
3260   associations = [associations] unless associations.is_a?(Array)
3261   ds = if eg = @opts[:eager_graph]
3262     eg = eg.dup
3263     [:requirements, :reflections, :reciprocals, :limits].each{|k| eg[k] = eg[k].dup}
3264     eg[:local] = opts
3265     ds = clone(:eager_graph=>eg)
3266     ds.eager_graph_associations(ds, model, ds.opts[:eager_graph][:master], [], *associations)
3267   else
3268     # Each of the following have a symbol key for the table alias, with the following values:
3269     # :reciprocals :: the reciprocal value to use for this association
3270     # :reflections :: AssociationReflection instance related to this association
3271     # :requirements :: array of requirements for this association
3272     # :limits :: Any limit/offset array slicing that need to be handled in ruby land after loading
3273     opts = {:requirements=>{}, :master=>alias_symbol(first_source), :reflections=>{}, :reciprocals=>{}, :limits=>{}, :local=>opts, :cartesian_product_number=>0, :row_proc=>row_proc}
3274     ds = clone(:eager_graph=>opts)
3275     ds = ds.eager_graph_associations(ds, model, ds.opts[:eager_graph][:master], [], *associations).naked
3276   end
3277 
3278   ds.opts[:eager_graph].freeze
3279   ds.opts[:eager_graph].each_value{|v| v.freeze if v.is_a?(Hash)}
3280   ds
3281 end
to_hash_groups(key_column, value_column=nil, opts=OPTS)

If the dataset is being eagerly loaded, default to calling all instead of each.

[show source]
     # File lib/sequel/model/associations.rb
3295 def to_hash_groups(key_column, value_column=nil, opts=OPTS)
3296   if (@opts[:eager_graph] || @opts[:eager]) && !opts.has_key?(:all)
3297     opts = Hash[opts]
3298     opts[:all] = true
3299   end
3300   super
3301 end
ungraphed()

Do not attempt to split the result set into associations, just return results as simple objects. This is useful if you want to use eager_graph as a shortcut to have all of the joins and aliasing set up, but want to do something else with the dataset.

[show source]
     # File lib/sequel/model/associations.rb
3307 def ungraphed
3308   ds = super.clone(:eager_graph=>nil)
3309   if (eg = @opts[:eager_graph]) && (rp = eg[:row_proc])
3310     ds = ds.with_row_proc(rp)
3311   end
3312   ds
3313 end

Protected Instance methods

eager_graph_association(ds, model, ta, requirements, r, *associations)

Call graph on the association with the correct arguments, update the eager_graph data structure, and recurse into eager_graph_associations if there are any passed in associations (which would be dependencies of the current association)

Arguments:

ds

Current dataset

model

Current Model

ta

table_alias used for the parent association

requirements

an array, used as a stack for requirements

r

association reflection for the current association, or an SQL::AliasedExpression with the reflection as the expression, the alias base as the alias (or nil to use the default alias), and an optional hash with a :join_type entry as the columns to use a custom join type.

*associations

any associations dependent on this one

[show source]
     # File lib/sequel/model/associations.rb
3332 def eager_graph_association(ds, model, ta, requirements, r, *associations)
3333   if r.is_a?(SQL::AliasedExpression)
3334     alias_base = r.alias
3335     if r.columns.is_a?(Hash)
3336       join_type = r.columns[:join_type]
3337     end
3338     r = r.expression
3339   else
3340     alias_base = r[:graph_alias_base]
3341   end
3342   assoc_table_alias = ds.unused_table_alias(alias_base)
3343   loader = r[:eager_grapher]
3344   if !associations.empty?
3345     if associations.first.respond_to?(:call)
3346       callback = associations.first
3347       associations = {}
3348     elsif associations.length == 1 && (assocs = associations.first).is_a?(Hash) && assocs.length == 1 && (pr_assoc = assocs.to_a.first) && pr_assoc.first.respond_to?(:call)
3349       callback, assoc = pr_assoc
3350       associations = assoc.is_a?(Array) ? assoc : [assoc]
3351     end
3352   end
3353   local_opts = ds.opts[:eager_graph][:local]
3354   limit_strategy = r.eager_graph_limit_strategy(local_opts[:limit_strategy])
3355 
3356   if r[:conditions] && !Sequel.condition_specifier?(r[:conditions]) && !r[:orig_opts].has_key?(:graph_conditions) && !r[:orig_opts].has_key?(:graph_only_conditions) && !r.has_key?(:graph_block)
3357     raise Error, "Cannot eager_graph association when :conditions specified and not a hash or an array of pairs.  Specify :graph_conditions, :graph_only_conditions, or :graph_block for the association.  Model: #{r[:model]}, association: #{r[:name]}"
3358   end
3359 
3360   ds = loader.call(:self=>ds, :table_alias=>assoc_table_alias, :implicit_qualifier=>(ta == ds.opts[:eager_graph][:master]) ? first_source : qualifier_from_alias_symbol(ta, first_source), :callback=>callback, :join_type=>join_type || local_opts[:join_type], :join_only=>local_opts[:join_only], :limit_strategy=>limit_strategy, :from_self_alias=>ds.opts[:eager_graph][:master])
3361   if r[:order_eager_graph] && (order = r.fetch(:graph_order, r[:order]))
3362     ds = ds.order_append(*qualified_expression(order, assoc_table_alias))
3363   end
3364   eager_graph = ds.opts[:eager_graph]
3365   eager_graph[:requirements][assoc_table_alias] = requirements.dup
3366   eager_graph[:reflections][assoc_table_alias] = r
3367   if limit_strategy == :ruby
3368     eager_graph[:limits][assoc_table_alias] = r.limit_and_offset 
3369   end
3370   eager_graph[:cartesian_product_number] += r[:cartesian_product_number] || 2
3371   ds = ds.eager_graph_associations(ds, r.associated_class, assoc_table_alias, requirements + [assoc_table_alias], *associations) unless associations.empty?
3372   ds
3373 end
eager_graph_associations(ds, model, ta, requirements, *associations)

Check the associations are valid for the given model. Call eager_graph_association on each association.

Arguments:

ds

Current dataset

model

Current Model

ta

table_alias used for the parent association

requirements

an array, used as a stack for requirements

*associations

the associations to add to the graph

[show source]
     # File lib/sequel/model/associations.rb
3384 def eager_graph_associations(ds, model, ta, requirements, *associations)
3385   associations.flatten.each do |association|
3386     ds = case association
3387     when Symbol, SQL::AliasedExpression
3388       ds.eager_graph_association(ds, model, ta, requirements, eager_graph_check_association(model, association))
3389     when Hash
3390       association.each do |assoc, assoc_assocs|
3391         ds = ds.eager_graph_association(ds, model, ta, requirements, eager_graph_check_association(model, assoc), assoc_assocs)
3392       end
3393       ds
3394     else
3395       raise(Sequel::Error, 'Associations must be in the form of a symbol or hash')
3396     end
3397   end
3398   ds
3399 end
eager_graph_build_associations(hashes)

Replace the array of plain hashes with an array of model objects will all eager_graphed associations set in the associations cache for each object.

[show source]
     # File lib/sequel/model/associations.rb
3403 def eager_graph_build_associations(hashes)
3404   hashes.replace(_eager_graph_build_associations(hashes, eager_graph_loader))
3405 end