- persisted? is the API defined in ActiveModel
- makes it easier for extension libraries to conform to ActiveModel APIs
  without concern for whether the extended object is specifically
  ActiveRecord

[#5927 state:committed]
Signed-off-by: NSantiago Pastorino <santiago@wyeworks.com>

Convert :primary_key in association to a string before comparing to column names, so that for example :primary_key => :another_pk works as well [#5605 state:resolved]

Signed-off-by: NSantiago Pastorino <santiago@wyeworks.com>

code gardening: we have assert_(nil|blank|present), more concise, with better default failure messages - let's use them

[#2786 state:committed]
Signed-off-by: NSantiago Pastorino <santiago@wyeworks.com>

[#5153 state:resolved]
Signed-off-by: NJosé Valim <jose.valim@gmail.com>

[#4645 state:resolved]
Signed-off-by: NJosé Valim <jose.valim@gmail.com>

Signed-off-by: NJosé Valim <jose.valim@gmail.com>

Signed-off-by: NJosé Valim <jose.valim@gmail.com>

[#765 state:committed]
Signed-off-by: NJeremy Kemper <jeremy@bitsweat.net>

Signed-off-by: NJosé Valim <jose.valim@gmail.com>

Signed-off-by: NPratik Naik <pratiknaik@gmail.com>

Oracle stores '' string as NULL
Oracle cannot have identifiers larger than 30 characters
added missing fixtures to test setup method

[#765 state:committed]
Signed-off-by: NJeremy Kemper <jeremy@bitsweat.net>

Signed-off-by: NPratik Naik <pratiknaik@gmail.com>

[#1617 state:resolved]

Ensure belongs_to association with a counter cache in name spaced model works [#1678 state:resolved]
Signed-off-by: NPratik Naik <pratiknaik@gmail.com>

Ensure methods called on association proxies respect access control. [#1083 state:resolved] [Adam Milligan, Pratik]

There are several situations it doesn't cater for, and the inconsistency isn't worth blocking 2.2.

Support for updating a belongs to association from the foreign key (without saving and reloading the record)
Signed-off-by: NMichael Koziarski <michael@koziarski.com>
[#142 state:committed]

Signed-off-by: NJoshua Peek <josh@joshpeek.com>

Signed-off-by: NPratik Naik <pratiknaik@gmail.com>

belongs_to polymorphic association assignments update the foreign_id and foreign_type fields regardless of whether the record being assigned is new or not.

fixes the following scenarios:
* I have validates_inclusion_of on the type field for a polymorphic belongs_to association.  I assign a new record to the model's polymorphic relationship of the proper type.  validation fails because the type field has not been updated.

* I replace the value for a ppolymorphic association to a new record of another class.  The type field still says its the previous class, and the id field points to the previous record as well.

[#191 state:closed]

belongs_to polymorphic association assignments update the foreign_id and foreign_type fields regardless of whether the record being assigned is new or not.

fixes the following scenarios:
* I have validates_inclusion_of on the type field for a polymorphic belongs_to association.  I assign a new record to the model's polymorphic relationship of the proper type.  validation fails because the type field has not been updated.

* I replace the value for a ppolymorphic association to a new record of another class.  The type field still says its the previous class, and the id field points to the previous record as well.

Signed-off-by: NMichael Koziarski <michael@koziarski.com>
[#241 state:resolved]

Revert "Ensure that save on child object fails for invalid belongs_to association. Closes #11555. [rubyruy]"

This reverts commit c67e9859.

git-svn-id: http://svn-commit.rubyonrails.org/rails/trunk@9247 5ecf4fe2-1ee6-0310-87b1-e25e094e27de

git-svn-id: http://svn-commit.rubyonrails.org/rails/trunk@9233 5ecf4fe2-1ee6-0310-87b1-e25e094e27de