1
0
Fork 0
mirror of https://github.com/rails/rails.git synced 2022-11-09 12:12:34 -05:00
rails--rails/activerecord/test/cases/attributes_test.rb

254 lines
7.8 KiB
Ruby
Raw Normal View History

require "cases/helper"
class OverloadedType < ActiveRecord::Base
attribute :overloaded_float, :integer
attribute :overloaded_string_with_limit, :string, limit: 50
attribute :non_existent_decimal, :decimal
attribute :string_with_default, :string, default: "the overloaded default"
end
class ChildOfOverloadedType < OverloadedType
end
class GrandchildOfOverloadedType < ChildOfOverloadedType
attribute :overloaded_float, :float
end
class UnoverloadedType < ActiveRecord::Base
self.table_name = "overloaded_types"
end
module ActiveRecord
class CustomPropertiesTest < ActiveRecord::TestCase
test "overloading types" do
data = OverloadedType.new
data.overloaded_float = "1.1"
data.unoverloaded_float = "1.1"
assert_equal 1, data.overloaded_float
assert_equal 1.1, data.unoverloaded_float
end
test "overloaded properties save" do
data = OverloadedType.new
data.overloaded_float = "2.2"
data.save!
data.reload
assert_equal 2, data.overloaded_float
assert_kind_of Integer, OverloadedType.last.overloaded_float
assert_equal 2.0, UnoverloadedType.last.overloaded_float
assert_kind_of Float, UnoverloadedType.last.overloaded_float
end
test "properties assigned in constructor" do
data = OverloadedType.new(overloaded_float: "3.3")
assert_equal 3, data.overloaded_float
end
test "overloaded properties with limit" do
assert_equal 50, OverloadedType.type_for_attribute("overloaded_string_with_limit").limit
assert_equal 255, UnoverloadedType.type_for_attribute("overloaded_string_with_limit").limit
end
test "nonexistent attribute" do
data = OverloadedType.new(non_existent_decimal: 1)
assert_equal BigDecimal.new(1), data.non_existent_decimal
assert_raise ActiveRecord::UnknownAttributeError do
UnoverloadedType.new(non_existent_decimal: 1)
end
end
test "model with nonexistent attribute with default value can be saved" do
klass = Class.new(OverloadedType) do
attribute :non_existent_string_with_default, :string, default: "nonexistent"
end
model = klass.new
assert model.save
end
test "changing defaults" do
data = OverloadedType.new
unoverloaded_data = UnoverloadedType.new
assert_equal "the overloaded default", data.string_with_default
assert_equal "the original default", unoverloaded_data.string_with_default
end
test "defaults are not touched on the columns" do
assert_equal "the original default", OverloadedType.columns_hash["string_with_default"].default
end
test "children inherit custom properties" do
data = ChildOfOverloadedType.new(overloaded_float: "4.4")
assert_equal 4, data.overloaded_float
end
test "children can override parents" do
data = GrandchildOfOverloadedType.new(overloaded_float: "4.4")
assert_equal 4.4, data.overloaded_float
end
Attribute assignment and type casting has nothing to do with columns It's finally finished!!!!!!! The reason the Attributes API was kept private in 4.2 was due to some publicly visible implementation details. It was previously implemented by overloading `columns` and `columns_hash`, to make them return column objects which were modified with the attribute information. This meant that those methods LIED! We didn't change the database schema. We changed the attribute information on the class. That is wrong! It should be the other way around, where schema loading just calls the attributes API for you. And now it does! Yes, this means that there is nothing that happens in automatic schema loading that you couldn't manually do yourself. (There's still some funky cases where we hit the connection adapter that I need to handle, before we can turn off automatic schema detection entirely.) There were a few weird test failures caused by this that had to be fixed. The main source came from the fact that the attribute methods are now defined in terms of `attribute_names`, which has a clause like `return [] unless table_exists?`. I don't *think* this is an issue, since the only place this caused failures were in a fake adapter which didn't override `table_exists?`. Additionally, there were a few cases where tests were failing because a migration was run, but the model was not reloaded. I'm not sure why these started failing from this change, I might need to clear an additional cache in `reload_schema_from_cache`. Again, since this is not normal usage, and it's expected that `reset_column_information` will be called after the table is modified, I don't think it's a problem. Still, test failures that were unrelated to the change are worrying, and I need to dig into them further. Finally, I spent a lot of time debugging issues with the mutex used in `define_attribute_methods`. I think we can just remove that method entirely, and define the attribute methods *manually* in the call to `define_attribute`, which would simplify the code *tremendously*. Ok. now to make this damn thing public, and work on moving it up to Active Model.
2015-01-30 16:03:36 -05:00
test "overloading properties does not attribute method order" do
attribute_names = OverloadedType.attribute_names
assert_equal %w(id overloaded_float unoverloaded_float overloaded_string_with_limit string_with_default non_existent_decimal), attribute_names
end
test "caches are cleared" do
klass = Class.new(OverloadedType)
Attribute assignment and type casting has nothing to do with columns It's finally finished!!!!!!! The reason the Attributes API was kept private in 4.2 was due to some publicly visible implementation details. It was previously implemented by overloading `columns` and `columns_hash`, to make them return column objects which were modified with the attribute information. This meant that those methods LIED! We didn't change the database schema. We changed the attribute information on the class. That is wrong! It should be the other way around, where schema loading just calls the attributes API for you. And now it does! Yes, this means that there is nothing that happens in automatic schema loading that you couldn't manually do yourself. (There's still some funky cases where we hit the connection adapter that I need to handle, before we can turn off automatic schema detection entirely.) There were a few weird test failures caused by this that had to be fixed. The main source came from the fact that the attribute methods are now defined in terms of `attribute_names`, which has a clause like `return [] unless table_exists?`. I don't *think* this is an issue, since the only place this caused failures were in a fake adapter which didn't override `table_exists?`. Additionally, there were a few cases where tests were failing because a migration was run, but the model was not reloaded. I'm not sure why these started failing from this change, I might need to clear an additional cache in `reload_schema_from_cache`. Again, since this is not normal usage, and it's expected that `reset_column_information` will be called after the table is modified, I don't think it's a problem. Still, test failures that were unrelated to the change are worrying, and I need to dig into them further. Finally, I spent a lot of time debugging issues with the mutex used in `define_attribute_methods`. I think we can just remove that method entirely, and define the attribute methods *manually* in the call to `define_attribute`, which would simplify the code *tremendously*. Ok. now to make this damn thing public, and work on moving it up to Active Model.
2015-01-30 16:03:36 -05:00
assert_equal 6, klass.attribute_types.length
assert_equal 6, klass.column_defaults.length
assert_not klass.attribute_types.include?("wibble")
klass.attribute :wibble, Type::Value.new
Attribute assignment and type casting has nothing to do with columns It's finally finished!!!!!!! The reason the Attributes API was kept private in 4.2 was due to some publicly visible implementation details. It was previously implemented by overloading `columns` and `columns_hash`, to make them return column objects which were modified with the attribute information. This meant that those methods LIED! We didn't change the database schema. We changed the attribute information on the class. That is wrong! It should be the other way around, where schema loading just calls the attributes API for you. And now it does! Yes, this means that there is nothing that happens in automatic schema loading that you couldn't manually do yourself. (There's still some funky cases where we hit the connection adapter that I need to handle, before we can turn off automatic schema detection entirely.) There were a few weird test failures caused by this that had to be fixed. The main source came from the fact that the attribute methods are now defined in terms of `attribute_names`, which has a clause like `return [] unless table_exists?`. I don't *think* this is an issue, since the only place this caused failures were in a fake adapter which didn't override `table_exists?`. Additionally, there were a few cases where tests were failing because a migration was run, but the model was not reloaded. I'm not sure why these started failing from this change, I might need to clear an additional cache in `reload_schema_from_cache`. Again, since this is not normal usage, and it's expected that `reset_column_information` will be called after the table is modified, I don't think it's a problem. Still, test failures that were unrelated to the change are worrying, and I need to dig into them further. Finally, I spent a lot of time debugging issues with the mutex used in `define_attribute_methods`. I think we can just remove that method entirely, and define the attribute methods *manually* in the call to `define_attribute`, which would simplify the code *tremendously*. Ok. now to make this damn thing public, and work on moving it up to Active Model.
2015-01-30 16:03:36 -05:00
assert_equal 7, klass.attribute_types.length
assert_equal 7, klass.column_defaults.length
assert klass.attribute_types.include?("wibble")
Attribute assignment and type casting has nothing to do with columns It's finally finished!!!!!!! The reason the Attributes API was kept private in 4.2 was due to some publicly visible implementation details. It was previously implemented by overloading `columns` and `columns_hash`, to make them return column objects which were modified with the attribute information. This meant that those methods LIED! We didn't change the database schema. We changed the attribute information on the class. That is wrong! It should be the other way around, where schema loading just calls the attributes API for you. And now it does! Yes, this means that there is nothing that happens in automatic schema loading that you couldn't manually do yourself. (There's still some funky cases where we hit the connection adapter that I need to handle, before we can turn off automatic schema detection entirely.) There were a few weird test failures caused by this that had to be fixed. The main source came from the fact that the attribute methods are now defined in terms of `attribute_names`, which has a clause like `return [] unless table_exists?`. I don't *think* this is an issue, since the only place this caused failures were in a fake adapter which didn't override `table_exists?`. Additionally, there were a few cases where tests were failing because a migration was run, but the model was not reloaded. I'm not sure why these started failing from this change, I might need to clear an additional cache in `reload_schema_from_cache`. Again, since this is not normal usage, and it's expected that `reset_column_information` will be called after the table is modified, I don't think it's a problem. Still, test failures that were unrelated to the change are worrying, and I need to dig into them further. Finally, I spent a lot of time debugging issues with the mutex used in `define_attribute_methods`. I think we can just remove that method entirely, and define the attribute methods *manually* in the call to `define_attribute`, which would simplify the code *tremendously*. Ok. now to make this damn thing public, and work on moving it up to Active Model.
2015-01-30 16:03:36 -05:00
end
test "the given default value is cast from user" do
custom_type = Class.new(Type::Value) do
2015-02-17 15:39:42 -05:00
def cast(*)
Attribute assignment and type casting has nothing to do with columns It's finally finished!!!!!!! The reason the Attributes API was kept private in 4.2 was due to some publicly visible implementation details. It was previously implemented by overloading `columns` and `columns_hash`, to make them return column objects which were modified with the attribute information. This meant that those methods LIED! We didn't change the database schema. We changed the attribute information on the class. That is wrong! It should be the other way around, where schema loading just calls the attributes API for you. And now it does! Yes, this means that there is nothing that happens in automatic schema loading that you couldn't manually do yourself. (There's still some funky cases where we hit the connection adapter that I need to handle, before we can turn off automatic schema detection entirely.) There were a few weird test failures caused by this that had to be fixed. The main source came from the fact that the attribute methods are now defined in terms of `attribute_names`, which has a clause like `return [] unless table_exists?`. I don't *think* this is an issue, since the only place this caused failures were in a fake adapter which didn't override `table_exists?`. Additionally, there were a few cases where tests were failing because a migration was run, but the model was not reloaded. I'm not sure why these started failing from this change, I might need to clear an additional cache in `reload_schema_from_cache`. Again, since this is not normal usage, and it's expected that `reset_column_information` will be called after the table is modified, I don't think it's a problem. Still, test failures that were unrelated to the change are worrying, and I need to dig into them further. Finally, I spent a lot of time debugging issues with the mutex used in `define_attribute_methods`. I think we can just remove that method entirely, and define the attribute methods *manually* in the call to `define_attribute`, which would simplify the code *tremendously*. Ok. now to make this damn thing public, and work on moving it up to Active Model.
2015-01-30 16:03:36 -05:00
"from user"
end
def deserialize(*)
Attribute assignment and type casting has nothing to do with columns It's finally finished!!!!!!! The reason the Attributes API was kept private in 4.2 was due to some publicly visible implementation details. It was previously implemented by overloading `columns` and `columns_hash`, to make them return column objects which were modified with the attribute information. This meant that those methods LIED! We didn't change the database schema. We changed the attribute information on the class. That is wrong! It should be the other way around, where schema loading just calls the attributes API for you. And now it does! Yes, this means that there is nothing that happens in automatic schema loading that you couldn't manually do yourself. (There's still some funky cases where we hit the connection adapter that I need to handle, before we can turn off automatic schema detection entirely.) There were a few weird test failures caused by this that had to be fixed. The main source came from the fact that the attribute methods are now defined in terms of `attribute_names`, which has a clause like `return [] unless table_exists?`. I don't *think* this is an issue, since the only place this caused failures were in a fake adapter which didn't override `table_exists?`. Additionally, there were a few cases where tests were failing because a migration was run, but the model was not reloaded. I'm not sure why these started failing from this change, I might need to clear an additional cache in `reload_schema_from_cache`. Again, since this is not normal usage, and it's expected that `reset_column_information` will be called after the table is modified, I don't think it's a problem. Still, test failures that were unrelated to the change are worrying, and I need to dig into them further. Finally, I spent a lot of time debugging issues with the mutex used in `define_attribute_methods`. I think we can just remove that method entirely, and define the attribute methods *manually* in the call to `define_attribute`, which would simplify the code *tremendously*. Ok. now to make this damn thing public, and work on moving it up to Active Model.
2015-01-30 16:03:36 -05:00
"from database"
end
end
klass = Class.new(OverloadedType) do
attribute :wibble, custom_type.new, default: "default"
end
model = klass.new
assert_equal "from user", model.wibble
end
test "procs for default values" do
klass = Class.new(OverloadedType) do
@@counter = 0
attribute :counter, :integer, default: -> { @@counter += 1 }
end
assert_equal 1, klass.new.counter
assert_equal 2, klass.new.counter
end
test "procs are memoized before type casting" do
klass = Class.new(OverloadedType) do
@@counter = 0
attribute :counter, :integer, default: -> { @@counter += 1 }
end
model = klass.new
assert_equal 1, model.counter_before_type_cast
assert_equal 1, model.counter_before_type_cast
end
test "user provided defaults are persisted even if unchanged" do
model = OverloadedType.create!
assert_equal "the overloaded default", model.reload.string_with_default
end
if current_adapter?(:PostgreSQLAdapter)
2015-07-20 11:22:30 -04:00
test "array types can be specified" do
klass = Class.new(OverloadedType) do
attribute :my_array, :string, limit: 50, array: true
attribute :my_int_array, :integer, array: true
end
string_array = ConnectionAdapters::PostgreSQL::OID::Array.new(
Type::String.new(limit: 50))
int_array = ConnectionAdapters::PostgreSQL::OID::Array.new(
Type::Integer.new)
assert_not_equal string_array, int_array
assert_equal string_array, klass.type_for_attribute("my_array")
assert_equal int_array, klass.type_for_attribute("my_int_array")
end
test "range types can be specified" do
klass = Class.new(OverloadedType) do
attribute :my_range, :string, limit: 50, range: true
attribute :my_int_range, :integer, range: true
end
string_range = ConnectionAdapters::PostgreSQL::OID::Range.new(
Type::String.new(limit: 50))
int_range = ConnectionAdapters::PostgreSQL::OID::Range.new(
Type::Integer.new)
assert_not_equal string_range, int_range
assert_equal string_range, klass.type_for_attribute("my_range")
assert_equal int_range, klass.type_for_attribute("my_int_range")
end
end
test "attributes added after subclasses load are inherited" do
parent = Class.new(ActiveRecord::Base) do
self.table_name = "topics"
end
child = Class.new(parent)
child.new # => force a schema load
parent.attribute(:foo, Type::Value.new)
assert_equal(:bar, child.new(foo: :bar).foo)
end
test "attributes not backed by database columns are not dirty when unchanged" do
refute OverloadedType.new.non_existent_decimal_changed?
end
test "attributes not backed by database columns are always initialized" do
OverloadedType.create!
model = OverloadedType.first
assert_nil model.non_existent_decimal
model.non_existent_decimal = "123"
assert_equal 123, model.non_existent_decimal
end
test "attributes not backed by database columns return the default on models loaded from database" do
child = Class.new(OverloadedType) do
attribute :non_existent_decimal, :decimal, default: 123
end
child.create!
model = child.first
assert_equal 123, model.non_existent_decimal
end
test "attributes not backed by database columns properly interact with mutation and dirty" do
child = Class.new(ActiveRecord::Base) do
self.table_name = "topics"
attribute :foo, :string, default: "lol"
end
child.create!
model = child.first
assert_equal "lol", model.foo
model.foo << "asdf"
assert_equal "lolasdf", model.foo
assert model.foo_changed?
model.reload
assert_equal "lol", model.foo
model.foo = "lol"
refute model.changed?
end
end
end