module Sequel::Model::ClassMethods

  1. lib/sequel/model/base.rb

Class methods for Sequel::Model that implement basic model functionality.

  • All of the method names in Model::DATASET_METHODS have class methods created that call the Model's dataset with the method of the same name with the given arguments.


FINDER_TYPES = [:first, :all, :each, :get].freeze  


allowed_columns [R]

Which columns should be the only columns allowed in a call to a mass assignment method (e.g. set) (default: not set, so all columns not otherwise restricted are allowed).

cache_anonymous_models [RW]

Whether to cache the anonymous models created by Sequel::Model(). This is required for reloading them correctly (avoiding the superclass mismatch). True by default for backwards compatibility.

dataset_method_modules [R]

Array of modules that extend this model's dataset. Stored so that if the model's dataset is changed, it will be extended with all of these modules.

default_set_fields_options [RW]

The default options to use for Model#set_fields. These are merged with the options given to set_fields.

fast_instance_delete_sql [R]

SQL string fragment used for faster DELETE statement creation when deleting/destroying model instances, or nil if the optimization should not be used. For internal use only.

fast_pk_lookup_sql [R]

SQL string fragment used for faster lookups by primary key, or nil if the optimization should not be used. For internal use only.

instance_dataset [R]

The dataset that instance datasets (#this) are based on. Generally a naked version of the model's dataset limited to one row. For internal use only.

plugins [R]

Array of plugin modules loaded by this class

# => [Sequel::Model, Sequel::Model::Associations]
primary_key [R]

The primary key for the class. Sequel can determine this automatically for many databases, but not all, so you may need to set it manually. If not determined automatically, the default is :id.

raise_on_save_failure [RW]

Whether to raise an error instead of returning nil on a failure to save/create/save_changes/update/destroy due to a validation failure or a before_* hook returning false (default: true).

raise_on_typecast_failure [RW]

Whether to raise an error when unable to typecast data for a column (default: false). This should be set to true if you want to have model setter methods raise errors if the argument cannot be typecast properly.

require_modification [RW]

Whether to raise an error if an UPDATE or DELETE query related to a model instance does not modify exactly 1 row. If set to false, Sequel will not check the number of rows modified (default: true).

require_valid_table [RW]

Requires that all models have valid tables, raising exceptions if creating a model without a valid table backing it. Enabling this will break code like:

class Foo < Sequel::Model
  set_dataset :my_foo

As when Sequel::Model is subclassed, before #set_dataset is executed, it will try to get the schema for the foos table, which will raise an exception. You would need to switch to using:

class Foo < Sequel::Model(:my_foo)


Foo = Sequel::Model()
Foo.set_dataset :my_foo
simple_pk [R]

Should be the literal primary key column name if this Model's table has a simple primary key, or nil if the model has a compound primary key or no primary key.

simple_table [R]

Should be the literal table name if this Model's dataset is a simple table (no select, order, join, etc.), or nil otherwise. This and #simple_pk are used for an optimization in Model.[].

strict_param_setting [RW]

Whether new/set/update and their variants should raise an error if an invalid key is used. A key is invalid if no setter method exists for that key or the access to the setter method is restricted (e.g. due to it being a primary key field). If set to false, silently skip any key where the setter method doesn't exist or access to it is restricted.

typecast_empty_string_to_nil [RW]

Whether to typecast the empty string ('') to nil for columns that are not string or blob. In most cases the empty string would be the way to specify a NULL SQL value in string form (nil.to_s == ''), and an empty string would not usually be typecast correctly for other types, so the default is true.

typecast_on_assignment [RW]

Whether to typecast attribute values on assignment (default: true). If set to false, no typecasting is done, so it will be left up to the database to typecast the value correctly.

use_after_commit_rollback [RW]

Whether to enable the after_commit and after_rollback hooks when saving/destroying instances. On by default, can be turned off for performance reasons or when using prepared transactions (which aren't compatible with after commit/rollback).

use_transactions [RW]

Whether to use a transaction by default when saving/deleting records (default: true). If you are sending database queries in before_* or after_* hooks, you shouldn't change the default setting without a good reason.

Public Instance methods

Model (source)

Lets you create a Model subclass with its dataset already set. source should be an instance of one of the following classes:


Sets the database for this model to source. Generally only useful when subclassing directly from the returned class, where the name of the subclass sets the table name (which is combined with the Database in source to create the dataset to use)


Sets the dataset for this model to source.


Sets the table name for this model to source. The class will use the default database for model classes in order to create the dataset.

The purpose of this method is to set the dataset/database automatically for a model class, if the table name doesn't match the implicit name. This is neater than using #set_dataset inside the class, doesn't require a bogus query for the schema.

When creating subclasses of Sequel::Model itself, this method is usually called on Sequel itself, using Sequel::Model(:something).

# Using a symbol
class Comment < Sequel::Model(:something)
  table_name # => :something

# Using a dataset
class Comment < Sequel::Model(DB1[:something])
  dataset # => DB1[:something]

# Using a database
class Comment < Sequel::Model(DB1)
  dataset # => DB1[:comments]
[show source]
# File lib/sequel/model/base.rb, line 188
def Model(source)
  if cache_anonymous_models
    mutex = @Model_mutex
    cache = mutex.synchronize{@Model_cache ||= {}}

    if klass = mutex.synchronize{cache[source]}
      return klass

  klass =

  if source.is_a?(::Sequel::Database)
    klass.db = source

  if cache_anonymous_models
    mutex.synchronize{cache[source] = klass}

[] (*args)

Returns the first record from the database matching the conditions. If a hash is given, it is used as the conditions. If another object is given, it finds the first record whose primary key(s) match the given argument(s). If no object is returned by the dataset, returns nil.

Artist[1] # SELECT * FROM artists WHERE id = 1
# => #<Artist {:id=>1, ...}>

Artist[:name=>'Bob'] # SELECT * FROM artists WHERE (name = 'Bob') LIMIT 1
# => #<Artist {:name=>'Bob', ...}>
[show source]
# File lib/sequel/model/base.rb, line 223
def [](*args)
  args = args.first if args.size <= 1
  args.is_a?(Hash) ? first_where(args) : (primary_key_lookup(args) unless args.nil?)
call (values)

Initializes a model instance as an existing record. This constructor is used by Sequel to initialize model instances when fetching records. Requires that values be a hash where all keys are symbols. It probably should not be used by external code.

[show source]
# File lib/sequel/model/base.rb, line 232
def call(values)
  o = allocate
  o.instance_variable_set(:@values, values)
clear_setter_methods_cache ()

Clear the #setter_methods cache

[show source]
# File lib/sequel/model/base.rb, line 239
def clear_setter_methods_cache
  @setter_methods = nil unless frozen?
columns ()

Returns the columns in the result set in their original order. Generally, this will use the columns determined via the database schema, but in certain cases (e.g. models that are based on a joined dataset) it will use Dataset#columns to find the columns.

# => [:id, :name]
[show source]
# File lib/sequel/model/base.rb, line 250
def columns
  return @columns if @columns
  return nil if frozen?
create (values = {}, &block)

Creates instance using new with the given values and block, and saves it.

# INSERT INTO artists (name) VALUES ('Bob')

Artist.create do |a| = 'Jim'
end # INSERT INTO artists (name) VALUES ('Jim')
[show source]
# File lib/sequel/model/base.rb, line 264
def create(values = {}, &block)
  new(values, &block).save
dataset ()

Returns the dataset associated with the Model class. Raises an Error if there is no associated dataset for this class. In most cases, you don't need to call this directly, as Model proxies many dataset methods to the underlying dataset.

Artist.dataset.all # SELECT * FROM artists
[show source]
# File lib/sequel/model/base.rb, line 274
def dataset
  @dataset || raise(Error, "No dataset associated with #{self}")
dataset= (ds)

Alias of #set_dataset

[show source]
# File lib/sequel/model/base.rb, line 279
def dataset=(ds)
dataset_module (mod = nil)

Extend the dataset with a module, similar to adding a plugin with the methods defined in DatasetMethods. This is the recommended way to add methods to model datasets.

If an argument, it should be a module, and is used to extend the underlying dataset. Otherwise an anonymous module is created, and if a block is given, it is module_evaled, allowing you do define dataset methods directly using the standard ruby def syntax. Returns the module given or the anonymous module created.

# Usage with existing module
Album.dataset_module Sequel::ColumnsIntrospection

# Usage with anonymous module
Album.dataset_module do
  def foo
# => :bar
# => :bar

Any anonymous modules created are actually instances of Sequel::Model::DatasetModule (a Module subclass), which allows you to call the subset method on them, which defines a dataset method that adds a filter. There are also a number of other methods with the same names as the dataset methods, which can use to define named dataset methods:

Album.dataset_module do
  subset :released, Sequel.identifier(release_date) <= Sequel::CURRENT_DATE
  order :by_release_date, :release_date
  select :for_select_options, :id, :name, :release_date
# => "SELECT * FROM artists WHERE (release_date <= CURRENT_DATE)"
# => "SELECT * FROM artists ORDER BY release_date"
# => "SELECT id, name, release_date FROM artists"
# => "SELECT id, name, release_date FROM artists WHERE (release_date <= CURRENT_DATE) ORDER BY release_date"

The following methods are supported: distinct, eager, exclude, exclude_having, grep, group, group_and_count, group_append, having, limit, offset, order, order_append, order_prepend, select, select_all, select_append, select_group, where, and server.

Any public methods in the dataset module will have class methods created that call the method on the dataset, assuming that the class method is not already defined.

[show source]
# File lib/sequel/model/base.rb, line 334
def dataset_module(mod = nil)
  if mod
    raise Error, "can't provide both argument and block to Model.dataset_module" if block_given?
    @dataset_module ||=
    @dataset_module.module_eval(& if block_given?
db ()

Returns the database associated with the Model class. If this model doesn't have a database associated with it, assumes the superclass's database, or the first object in Sequel::DATABASES. If no Sequel::Database object has been created, raises an error.

Artist.db.transaction do # BEGIN
  # INSERT INTO artists (name) VALUES ('Bob')
end # COMMIT
[show source]
# File lib/sequel/model/base.rb, line 357
def db
  return @db if @db
  @db = self == Model ? Sequel.synchronize{DATABASES.first} : superclass.db
  raise(Error, "No database associated with #{self}: have you called Sequel.connect or #{self}.db= ?") unless @db
db= (db)

Sets the database associated with the Model class. If the model has an associated dataset, sets the model's dataset to a dataset on the new database with the same options used by the current dataset. This can be used directly on Sequel::Model to set the default database to be used by subclasses, or to override the database used for specific models:

Sequel::Model.db = DB1
Artist.db = DB2

Note that you should not use this to change the model's database at runtime. If you have that need, you should look into Sequel's sharding support.

[show source]
# File lib/sequel/model/base.rb, line 378
def db=(db)
  @db = db
  set_dataset(db.dataset.clone(@dataset.opts)) if @dataset
db_schema ()

Returns the cached schema information if available or gets it from the database. This is a hash where keys are column symbols and values are hashes of information related to the column. See Database#schema.

# {:id=>{:type=>:integer, :primary_key=>true, ...},
#  :name=>{:type=>:string, :primary_key=>false, ...}}
[show source]
# File lib/sequel/model/base.rb, line 391
def db_schema
  return @db_schema if @db_schema
  return nil if frozen?
  @db_schema = get_db_schema
def_Model (mod)

Define a Model method on the given module that calls the Model method on the receiver. This is how the Sequel::Model() method is defined, and allows you to define Model() methods on other modules, making it easier to have custom model settings for all models under a namespace. Example:

module Foo
  Model =
  DB = Model.db = Sequel.connect(ENV['FOO_DATABASE_URL'])
  Model.plugin :prepared_statements

  class Bar < Model
    # Uses Foo::DB[:bars]

  class Baz < Model(:my_baz)
    # Uses Foo::DB[:my_baz]
[show source]
# File lib/sequel/model/base.rb, line 145
def def_Model(mod)
  model = self
  (class << mod; self; end).send(:define_method, :Model) do |source|
def_column_alias (meth, column)

Create a column alias, where the column methods have one name, but the underlying storage uses a different name.

[show source]
# File lib/sequel/model/base.rb, line 399
def def_column_alias(meth, column)
  overridable_methods_module.module_eval do
    define_method("#{meth}="){|v| self[column] = v}
def_dataset_method (*args, &block)

If a block is given, define a method on the dataset (if the model currently has an dataset) with the given argument name using the given block. Also define a class method on the model that calls the dataset method. Stores the method name and block so that it can be reapplied if the model's dataset changes.

If a block is not given, just define a class method on the model for each argument that calls the dataset method of the same argument name.

Using #dataset_module is recommended over using this method. In addition to allowing more natural ruby syntax for defining methods manually, it also offers numerous helper methods that make defining common dataset methods more easily, as well as supporting dataset caching (assuming the arguments allow it).

# Add new dataset method and class method that calls it

# Just add a class method that calls an existing dataset method
Artist.paginate(2, 10)
[show source]
# File lib/sequel/model/base.rb, line 428
def def_dataset_method(*args, &block)
  raise(Error, "No arguments given") if args.empty?

  if block
    raise(Error, "Defining a dataset method using a block requires only one argument") if args.length > 1
    dataset_module{define_method(args.first, &block)}
    args.each{|arg| def_model_dataset_method(arg)}
find (*args, &block)

Finds a single record according to the supplied filter. You are encouraged to use Model.[] or Model.first instead of this method.

# SELECT * FROM artists WHERE (name = 'Bob') LIMIT 1

Artist.find{name > 'M'}
# SELECT * FROM artists WHERE (name > 'M') LIMIT 1
[show source]
# File lib/sequel/model/base.rb, line 447
def find(*args, &block)
  if args.length == 1 && !block
    # Use optimized finder
    where(*args, &block).first
find_or_create (cond, &block)

Like find but invokes create with given conditions when record does not exist. Unlike find in that the block used in this method is not passed to find, but instead is passed to create only if find does not return an object.

# SELECT * FROM artists WHERE (name = 'Bob') LIMIT 1
# INSERT INTO artists (name) VALUES ('Bob')

Artist.find_or_create(:name=>'Jim'){|a| a.hometown = 'Sactown'}
# SELECT * FROM artists WHERE (name = 'Jim') LIMIT 1
# INSERT INTO artists (name, hometown) VALUES ('Jim', 'Sactown')
[show source]
# File lib/sequel/model/base.rb, line 468
def find_or_create(cond, &block)
  find(cond) || create(cond, &block)
finder (meth=OPTS, opts=OPTS, &block)

Create an optimized finder method using a dataset placeholder literalizer. This pre-computes the SQL to use for the query, except for given arguments.

There are two ways to use this. The recommended way is to pass a symbol that represents a model class method that returns a dataset:

def Artist.by_name(name)

Artist.finder :by_name

This creates an optimized first_by_name method, which you can call normally:


The alternative way to use this to pass your own block:

Artist.finder(:name=>:first_by_name){|pl, ds| ds.where(:name=>pl.arg).limit(1)}

Note that if you pass your own block, you are responsible for manually setting limits if necessary (as shown above).



When using a symbol method name, this specifies the arity of the method. This should be used if if the method accepts an arbitrary number of arguments, or the method has default argument values. Note that if the method is defined as a dataset method, the class method Sequel creates accepts an arbitrary number of arguments, so you should use this option in that case. If you want to handle multiple possible arities, you need to call the finder method multiple times with unique :arity and :name methods each time.


The name of the method to create. This must be given if you pass a block. If you use a symbol, this defaults to the symbol prefixed by the type.


The module in which to create the finder method. Defaults to the singleton class of the model.


The type of query to run. Can be :first, :each, :all, or :get, defaults to :first.


This doesn't handle all possible cases. For example, if you have a method such as:

def Artist.by_name(name)
  name ? where(:name=>name) : exclude(:name=>nil)

Then calling a finder without an argument will not work as you expect.

Artist.finder :by_name
# WHERE (name IS NULL)

See Dataset::PlaceholderLiteralizer for additional caveats.

[show source]
# File lib/sequel/model/base.rb, line 530
def finder(meth=OPTS, opts=OPTS, &block)
  if block
    raise Error, "cannot pass both a method name argument and a block of Model.finder" unless meth.is_a?(Hash)
    raise Error, "cannot pass two option hashes to Model.finder" unless opts.equal?(OPTS)
    opts = meth
    raise Error, "must provide method name via :name option when passing block to Model.finder" unless meth_name = opts[:name]

  type = opts.fetch(:type, :first)
  unless prepare = opts[:prepare]
    raise Error, ":type option to Model.finder must be :first, :all, :each, or :get" unless FINDER_TYPES.include?(type)
  limit1 = type == :first || type == :get
  meth_name ||= opts[:name] || :"#{type}_#{meth}"

  argn = lambda do |model|
    if arity = opts[:arity]
      method = block || model.method(meth)
      (method.arity < 0 ? method.arity.abs - 1 : method.arity)

  loader_proc = if prepare
    proc do |model|
      args = prepare_method_args('$a',
      ds = if block
        model.instance_exec(*args, &block)
        model.send(meth, *args)
      ds = ds.limit(1) if limit1
      model_name =
      if model_name.to_s.empty?
        model_name = model.object_id
        model_name = model_name.gsub(/\W/, '_')
      ds.prepare(type, :"#{model_name}_#{meth_name}")
    proc do |model|
      n =
      block ||= lambda do |pl, model2|
        args = (0...n).map{pl.arg}
        ds = model2.send(meth, *args)
        ds = ds.limit(1) if limit1

      Sequel::Dataset::PlaceholderLiteralizer.loader(model, &block) 

  @finder_loaders[meth_name] = loader_proc
  mod = opts[:mod] || (class << self; self; end)
  if prepare
    def_prepare_method(mod, meth_name)
    def_finder_method(mod, meth_name, type)
first (*args, &block)

An alias for calling first on the model's dataset, but with optimized handling of the single argument case.

[show source]
# File lib/sequel/model/base.rb, line 596
def first(*args, &block)
  if args.length == 1 && !block && !args.first.is_a?(Integer)
    # Use optimized finder
    dataset.first(*args, &block)
first! (*args, &block)

An alias for calling first! on the model's dataset, but with optimized handling of the single argument case.

[show source]
# File lib/sequel/model/base.rb, line 607
def first!(*args, &block)
  first(*args, &block) || raise(
freeze ()

Freeze a model class, disallowing any further changes to it.

[show source]
# File lib/sequel/model/base.rb, line 612
def freeze


  if @dataset
    @finder_loaders.each_key{|k| finder_for(k)}
    @setter_methods = [].freeze

  @allowed_columns.freeze if @allowed_columns

implicit_table_name ()

Returns the implicit table name for the model class, which is the demodulized, underscored, pluralized name of the class.

Artist.implicit_table_name # => :artists
Foo::ArtistAlias.implicit_table_name # => :artist_aliases
[show source]
# File lib/sequel/model/base.rb, line 691
def implicit_table_name
include (*mods)

Clear the #setter_methods cache when a module is included, as it may contain setter methods.

[show source]
# File lib/sequel/model/base.rb, line 640
def include(*mods)
inherited (subclass)

If possible, set the dataset for the model subclass as soon as it is created. Also, make sure the inherited class instance variables are copied into the subclass.

Sequel queries the database to get schema information as soon as a model class is created:

class Artist < Sequel::Model # Causes schema query
[show source]
# File lib/sequel/model/base.rb, line 654
def inherited(subclass)
  ivs = subclass.instance_variables.collect(&:to_s)
  inherited_instance_variables.each do |iv, dup|
    next if ivs.include?(iv.to_s)
    if (sup_class_value = instance_variable_get(iv)) && dup
      sup_class_value = case dup
      when :dup
      when :hash_dup
        h = {}
        sup_class_value.each{|k,v| h[k] = v.dup}
      when Proc
        raise Error, "bad inherited instance variable type: #{dup.inspect}"
    subclass.instance_variable_set(iv, sup_class_value)

  unless ivs.include?("@dataset")
    if @dataset && self != Model
      subclass.set_dataset(@dataset.clone, :inherited=>true)
    elsif (n = && !n.to_s.empty?
load (values)

Calls call with the values hash. Only for backwards compatibility.

[show source]
# File lib/sequel/model/base.rb, line 696
def load(values)
method_added (meth)

Clear the #setter_methods cache when a setter method is added

[show source]
# File lib/sequel/model/base.rb, line 701
def method_added(meth)
  clear_setter_methods_cache if meth.to_s =~ SETTER_METHOD_REGEXP
no_primary_key ()

Mark the model as not having a primary key. Not having a primary key can cause issues, among which is that you won't be able to update records.

Artist.primary_key # => :id
Artist.primary_key # => nil
[show source]
# File lib/sequel/model/base.rb, line 712
def no_primary_key
  self.simple_pk = @primary_key = nil
plugin (plugin, *args, &block)

Loads a plugin for use with the model class, passing optional arguments to the plugin. If the plugin is a module, load it directly. Otherwise, require the plugin from either sequel/plugins/#{plugin} or sequel_#{plugin}, and then attempt to load the module using a the camelized plugin name under Sequel::Plugins.

[show source]
# File lib/sequel/model/base.rb, line 722
def plugin(plugin, *args, &block)
  m = plugin.is_a?(Module) ? plugin : plugin_module(plugin)
  unless @plugins.include?(m)
    @plugins << m
    m.apply(self, *args, &block) if m.respond_to?(:apply)
    extend(m::ClassMethods) if plugin_module_defined?(m, :ClassMethods)
    include(m::InstanceMethods) if plugin_module_defined?(m, :InstanceMethods)
    if plugin_module_defined?(m, :DatasetMethods)
      dataset_extend(m::DatasetMethods, :create_class_methods=>false)
  m.configure(self, *args, &block) if m.respond_to?(:configure)
prepared_finder (meth=OPTS, opts=OPTS, &block)

Similar to finder, but uses a prepared statement instead of a placeholder literalizer. This makes the SQL used static (cannot vary per call), but allows binding argument values instead of literalizing them into the SQL query string.

If a block is used with this method, it is instance_execed by the model, and should accept the desired number of placeholder arguments.

The options are the same as the options for finder, with the following exception:


Specifies the type of prepared statement to create

[show source]
# File lib/sequel/model/base.rb, line 788
def prepared_finder(meth=OPTS, opts=OPTS, &block)
  if block
    raise Error, "cannot pass both a method name argument and a block of Model.finder" unless meth.is_a?(Hash)
    meth = meth.merge(:prepare=>true)
    opts = opts.merge(:prepare=>true)
  finder(meth, opts, &block)
primary_key_hash (value)

Returns primary key attribute hash. If using a composite primary key value such be an array with values for each primary key in the correct order. For a standard primary key, value should be an object with a compatible type for the key. If the model does not have a primary key, raises an Error.

Artist.primary_key_hash(1) # => {:id=>1}
Artist.primary_key_hash([1, 2]) # => {:id1=>1, :id2=>2}
[show source]
# File lib/sequel/model/base.rb, line 744
def primary_key_hash(value)
  case key = @primary_key
  when Symbol
    {key => value}
  when Array
    hash = {}{|k,v| hash[k] = v}
    raise(Error, "#{self} does not have a primary key")
qualified_primary_key_hash (value, qualifier=table_name)

Return a hash where the keys are qualified column references. Uses the given qualifier if provided, or the #table_name otherwise. This is useful if you plan to join other tables to this table and you want the column references to be qualified.

# SELECT * FROM artists WHERE ( = 1)
[show source]
# File lib/sequel/model/base.rb, line 764
def qualified_primary_key_hash(value, qualifier=table_name)
  case key = @primary_key
  when Symbol
    {, key) => value}
  when Array
    hash = {}{|k,v| hash[, k)] = v}
    raise(Error, "#{self} does not have a primary key")
restrict_primary_key ()

Restrict the setting of the primary key(s) when using mass assignment (e.g. set). Because this is the default, this only make sense to use in a subclass where the parent class has used unrestrict_primary_key.

[show source]
# File lib/sequel/model/base.rb, line 801
def restrict_primary_key
  @restrict_primary_key = true
restrict_primary_key? ()

Whether or not setting the primary key(s) when using mass assignment (e.g. set) is restricted, true by default.

[show source]
# File lib/sequel/model/base.rb, line 808
def restrict_primary_key?
set_allowed_columns (*cols)

Set the columns to allow when using mass assignment (e.g. set). Using this means that any columns not listed here will not be modified. If you have any virtual setter methods (methods that end in =) that you want to be used during mass assignment, they need to be listed here as well (without the =).

It may be better to use a method such as set_only or set_fields that lets you specify the allowed fields per call.

Artist.set_allowed_columns(:name, :hometown)
Artist.set(:name=>'Bob', :hometown=>'Sactown') # No Error
Artist.set(:name=>'Bob', :records_sold=>30000) # Error
[show source]
# File lib/sequel/model/base.rb, line 823
def set_allowed_columns(*cols)
  @allowed_columns = cols
set_dataset (ds, opts=OPTS)

Sets the dataset associated with the Model class. ds can be a Symbol, LiteralString, SQL::Identifier, SQL::QualifiedIdentifier, SQL::AliasedExpression (all specifying a table name in the current database), or a Dataset. If a dataset is used, the model's database is changed to the database of the given dataset. If a dataset is not used, a dataset is created from the current database with the table name given. Other arguments raise an Error. Returns self.

This changes the row_proc of the dataset to return model objects and extends the dataset with the dataset_method_modules. It also attempts to determine the database schema for the model, based on the given dataset.


Note that you should not use this to change the model's dataset at runtime. If you have that need, you should look into Sequel's sharding support.

[show source]
# File lib/sequel/model/base.rb, line 848
def set_dataset(ds, opts=OPTS)
  inherited = opts[:inherited]
  @dataset = convert_input_dataset(ds)
  @require_modification = Sequel::Model.require_modification.nil? ? @dataset.provides_accurate_rows_matched? : Sequel::Model.require_modification
  if inherited
    self.simple_table = superclass.simple_table
    @columns = superclass.instance_variable_get(:@columns)
    @db_schema = superclass.instance_variable_get(:@db_schema)
    @dataset = @dataset.with_extend(*@dataset_method_modules.reverse) if @dataset_method_modules
    @db_schema = get_db_schema

set_primary_key (key)

Sets the primary key for this model. You can use either a regular or a composite primary key. To not use a primary key, set to nil or use no_primary_key. On most adapters, Sequel can automatically determine the primary key to use, so this method is not needed often.

class Person < Sequel::Model
  # regular key
  set_primary_key :person_id

class Tagging < Sequel::Model
  # composite key
  set_primary_key [:taggable_id, :tag_id]
[show source]
# File lib/sequel/model/base.rb, line 879
def set_primary_key(key)
  if key.is_a?(Array)
    if key.length < 2
      key = key.first
      key = key.dup.freeze
  self.simple_pk = if key && !key.is_a?(Array)
    (@dataset || db).literal(key).freeze
  @primary_key = key
setter_methods ()

Cache of setter methods to allow by default, in order to speed up new/set/update instance methods.

[show source]
# File lib/sequel/model/base.rb, line 895
def setter_methods
  return @setter_methods if @setter_methods
  raise if frozen?
  @setter_methods = get_setter_methods
subset (name, *args, &block)

Sets up a dataset method that returns a filtered dataset. Sometimes thought of as a scope, and like most dataset methods, they can be chained. For example:

Topic.subset(:popular){num_posts > 100}
Topic.subset(:recent){created_on > - 7}

Allows you to do:


to get topics with a username that includes joe that have more than 100 posts and were created less than 7 days ago.

Both the args given and the block are passed to Dataset#filter.

This method creates dataset methods that do not accept arguments. To create dataset methods that accept arguments, you should use define a method directly inside a dataset_module block.

[show source]
# File lib/sequel/model/base.rb, line 923
def subset(name, *args, &block)
  if block || args.flatten.any?{|arg| arg.is_a?(Proc)}
    def_dataset_method(name){filter(*args, &block)}
    key = :"_subset_#{name}_ds"
    def_dataset_method(name) do
table_name ()

Returns name of primary table for the dataset. If the table for the dataset is aliased, returns the aliased name.

Artist.table_name # => :artists
Sequel::Model(:foo).table_name # => :foo
Sequel::Model(:foo___bar).table_name # => :bar
[show source]
# File lib/sequel/model/base.rb, line 940
def table_name
unrestrict_primary_key ()

Allow the setting of the primary key(s) when using the mass assignment methods. Using this method can open up security issues, be very careful before using it.

Artist.set(:id=>1) # Error
Artist.set(:id=>1) # No Error
[show source]
# File lib/sequel/model/base.rb, line 950
def unrestrict_primary_key
  @restrict_primary_key = false
with_pk (pk)

Return the model instance with the primary key, or nil if there is no matching record.

[show source]
# File lib/sequel/model/base.rb, line 956
def with_pk(pk)
with_pk! (pk)

Return the model instance with the primary key, or raise NoMatchingRow if there is no matching record.

[show source]
# File lib/sequel/model/base.rb, line 961
def with_pk!(pk)
  with_pk(pk) || raise(