Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Merge pull request #6368 from Antiarchitect/store-indifferent-access | Jeremy Kemper | 2012-05-22 | 1 | -0/+34 |
|\ | | | | | Support migrating from Hash to HashWithIndifferentAccess for accessing attributes. | ||||
| * | Convert Hash to HashWithIndifferentAccess in ActiveRecord::Store. | Andrey Voronkov | 2012-05-22 | 1 | -0/+34 |
| | | | | | | | | | | | | | | In order to make migration from 3.x apps easier, we should try to convert Hash instances to HashWithIndifferentAccess, to allow accessing values with both symbol and a string. This is follow up to changes in 3c0bf043. | ||||
* | | Restore the frozen state on rollback. Fixes #6417. | Godfrey Chan | 2012-05-21 | 1 | -0/+10 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently, when saving a frozen record, an exception would be thrown which causes a rollback. However, there is a bug in active record that "defrost" the record as a side effect: >> t = Topic.new => #<Topic id: nil, ...> >> t.freeze => #<Topic id: nil, ...> >> t.save RuntimeError: can't modify a frozen Hash >> t.frozen? => false >> t.save => true This patch fixes the bug by explictly restoring the frozen state on the attributes Hash after every rollback. | ||||
* | | Remove not used variables from connection poll test | Carlos Antonio da Silva | 2012-05-21 | 1 | -2/+2 |
| | | | | | | | | These variables were issuing some "not used" warnings. | ||||
* | | Merge pull request #6416 from pmahoney/threadsafe-connection-pool | Aaron Patterson | 2012-05-21 | 2 | -1/+116 |
|\ \ | | | | | | | Make connection pool fair with respect to waiting threads. | ||||
| * | | Make connection pool fair with respect to waiting threads. | Patrick Mahoney | 2012-05-20 | 2 | -1/+116 |
| | | | |||||
* | | | Merge pull request #6143 from senny/composed_of_converter_returns_nil | Aaron Patterson | 2012-05-21 | 1 | -0/+18 |
|\ \ \ | |/ / |/| | | allow the :converter Proc form composed_of to return nil | ||||
| * | | allow the :converter Proc form composed_of to return nil | Yves Senn | 2012-05-03 | 1 | -0/+18 |
| | | | | | | | | | | | | | | | | | | | | | This makes it possible to filter invalid input values before they are passed into the value-object (like empty strings). This behaviour is only relevant if the :allow_nil options is set to true. Otherwise you will get the resulting NoMethodError. | ||||
* | | | fixed typo in word finiding | Ivan Kukobko | 2012-05-20 | 1 | -1/+1 |
| | | | |||||
* | | | Ensure that CollectionAssociation#replace returns proper target | Piotr Sarnacki | 2012-05-19 | 1 | -2/+2 |
| | | | | | | | | | | | | | | | | | | | | | | | | The fix commited in e2a070c was returning the `new_target`, as a try to return whatever user replaced association with. The problem is, the resulting association target may be ordered differently. In such case we want to return the target that will be later used for that association. | ||||
* | | | Merge pull request #6376 from jgaskins/timestamp-microseconds | Jeremy Kemper | 2012-05-18 | 1 | -1/+1 |
|\ \ \ | | | | | | | | | Increase numeric-timestamp precision to nanoseconds | ||||
| * | | | Increase `AR#cache_key` precision to nanoseconds | Jamie Gaskins | 2012-05-19 | 1 | -1/+1 |
| | | | | |||||
* | | | | fix #delete_all with habtm with :delete_sql | Jon Leighton | 2012-05-18 | 1 | -0/+6 |
| | | | | |||||
* | | | | Perf: Don't load the association for #delete_all. | Jon Leighton | 2012-05-18 | 1 | -0/+12 |
| | | | | | | | | | | | | | | | | Bug #6289 | ||||
* | | | | port some mocha to minitest/mock | Aaron Patterson | 2012-05-18 | 3 | -222/+222 |
| |_|/ |/| | | |||||
* | | | quarantine more deprecated stuff | Jon Leighton | 2012-05-18 | 4 | -36/+36 |
| | | | |||||
* | | | quarantine deprecated tests | Jon Leighton | 2012-05-18 | 10 | -578/+605 |
|/ / | |||||
* | | Merge pull request #6250 from iGEL/dont_destroy_readonly_models | Aaron Patterson | 2012-05-17 | 1 | -0/+1 |
|\ \ | | | | | | | Don't allow to destroy readonly models | ||||
| * | | Don't destroy readonly models | Johannes Barre | 2012-05-17 | 1 | -0/+1 |
| | | | |||||
* | | | Fix warning: ambiguous first argument. | kennyj | 2012-05-18 | 1 | -1/+1 |
|/ / | |||||
* | | Merge pull request #6058 from ↵ | Jon Leighton | 2012-05-17 | 2 | -1/+13 |
|\ \ | | | | | | | | | | | | | RStankov/relation-from-to-accept-other-relation-objects Relation#from to accept other Relation objects | ||||
| * | | Relation#from to accept other Relation objects | Radoslav Stankov | 2012-05-17 | 2 | -1/+13 |
| | | | | | | | | | | | | Record.from("(#{sub_query.to_sql})") -> Record.from(sub_query) Record.from("(#{sub_query.to_sql}) a") -> Record.from(sub_query, :a) | ||||
* | | | Merge pull request #5453 from ↵ | Aaron Patterson | 2012-05-16 | 1 | -0/+6 |
|\ \ \ | | | | | | | | | | | | | | | | | JonRowe/patch_uniq_has_and_belongs_to_many_when_already_loaded When Active Record has already loaded a unique association `.size` returns the wrong number. | ||||
| * | | | when using a preloaded array and the uniq flag is set then return the size ↵ | Jon Rowe | 2012-03-15 | 1 | -0/+6 |
| | | | | | | | | | | | | | | | | of the uniq array | ||||
* | | | | Merge pull request #6358 from jfirebaugh/schema_define | Aaron Patterson | 2012-05-16 | 1 | -0/+7 |
|\ \ \ \ | | | | | | | | | | | Call methods on the correct instance in AR::Schema.define | ||||
| * | | | | Call methods on the correct instance in AR::Schema.define | John Firebaugh | 2012-05-16 | 1 | -0/+7 |
| | |/ / | |/| | | | | | | | | | | | | | | Now that migrations support instance methods, we should use the same instance rather than relying on delegation to a global instance. This allows subclassing AR::Schema. | ||||
* | | | | Fix CollectionAssociation#replace to return new target (closes #6231) | Piotr Sarnacki | 2012-05-16 | 1 | -0/+12 |
| | | | | |||||
* | | | | Merge pull request #6192 from ↵ | Aaron Patterson | 2012-05-16 | 2 | -8/+32 |
|\ \ \ \ | | | | | | | | | | | | | | | | | | | | | danmcclain/add_inet_and_cidr_types_to_postgresql_adapter Add support for macaddr, inet, and cidr types to PostgreSQL adapter | ||||
| * | | | | Removes NetAddr dependency | Dan McClain | 2012-05-08 | 1 | -3/+4 |
| | | | | | |||||
| * | | | | Converts inet and cidr columns to NetAddr::CIDR | Dan Seaver | 2012-05-07 | 1 | -4/+6 |
| | | | | | |||||
| * | | | | Adds migration and schema dump support for INET, CIDR, and MACADDR | Dan Seaver | 2012-05-05 | 2 | -3/+24 |
| | | | | | |||||
* | | | | | Merge pull request #5535 from markmcspadden/issue_5527_rollbacks | Aaron Patterson | 2012-05-16 | 1 | -0/+37 |
|\ \ \ \ \ | |_|/ / / |/| | | | | Allow manual rollbacks in after_save to reset object correctly | ||||
| * | | | | Allow manual rollbacks in after_save to reset object correctly | Mark McSpadden | 2012-03-21 | 1 | -0/+37 |
| | | | | | |||||
* | | | | | MySQL returns "SUM(DISTINCT(credit_limit))" as the column name unless | Aaron Patterson | 2012-05-16 | 1 | -1/+5 |
| | | | | | | | | | | | | | | | | | | | | | | | | | an alias is provided. Without the alias, the column cannot be found and properly typecast. | ||||
* | | | | | Merge pull request #6136 from mhfs/sqlbypass_fixes | Aaron Patterson | 2012-05-16 | 1 | -0/+5 |
|\ \ \ \ \ | | | | | | | | | | | | | Bring SqlByPassTest to light and fix broken tests | ||||
| * | | | | | Rename test file so that the test suite matches it, implement `persisted?` ↵ | Marcelo Silveira | 2012-05-16 | 1 | -0/+5 |
| | | | | | | | | | | | | | | | | | | | | | | | | for SqlBypass as expected by tests and convert session_id to string before using on queries to get correct quotes on postgresql (avoid casting error). | ||||
* | | | | | | Merge pull request #6349 from erichmenge/patch-raise-type-errors | Aaron Patterson | 2012-05-16 | 3 | -6/+33 |
|\ \ \ \ \ \ | | | | | | | | | | | | | | | Integer limit out of range should be allowed to raise. Closes #6272 | ||||
| * | | | | | | Integer limit out of range should be allowed to raise. Closes #6272 | Erich Menge | 2012-05-16 | 3 | -6/+33 |
| |/ / / / / | |||||
* | | | | | | Merge pull request #6336 from erichmenge/patch-pluck-fragment | Carlos Antonio da Silva | 2012-05-16 | 1 | -0/+3 |
|\ \ \ \ \ \ | | | | | | | | | | | | | | | Fixes issue where SQL fragments prevented type casting based on column | ||||
| * | | | | | | Fixes issue where SQL fragments prevented type casting based on column type. | Erich Menge | 2012-05-16 | 1 | -0/+3 |
| |/ / / / / | |||||
* | | | | | | Remove unneeded tests | Piotr Sarnacki | 2012-05-16 | 1 | -12/+0 |
| | | | | | | |||||
* | | | | | | SQLite3 doesn't actually support the 'time' type. | Erich Menge | 2012-05-15 | 2 | -3/+14 |
|/ / / / / | |||||
* | | | | | Merge pull request #6325 from kennyj/fix_5832-3 | Carlos Antonio da Silva | 2012-05-15 | 1 | -5/+1 |
|\ \ \ \ \ | | | | | | | | | | | | | Remove unnecessary assertions. | ||||
| * | | | | | Remove unnecessary assertions. We should only test the before_seq isn't same ↵ | kennyj | 2012-05-16 | 1 | -5/+1 |
| | | | | | | | | | | | | | | | | | | | | | | | | to the after_seq when using sequencer. | ||||
* | | | | | | Restore support for Model.pluck('sql fragment') | Jeremy Kemper | 2012-05-15 | 1 | -0/+8 |
|/ / / / / | |||||
* | | | | | Set the primary key during #copy_table if necessary. Fixes [#2312] | Sean Scally | 2012-05-14 | 1 | -0/+8 |
| | | | | | |||||
* | | | | | Merge pull request #5807 from Antiarchitect/store-improvement | Rafael Mendonça França | 2012-05-13 | 1 | -1/+35 |
|\ \ \ \ \ | | | | | | | | | | | | | Custom coders support for ActiveRecord::Store. | ||||
| * | | | | | Custom coders support for ActiveRecord::Store. JSON, YAML, Marshal can be ↵ | Andrey Voronkov | 2012-05-09 | 1 | -1/+35 |
| | | | | | | | | | | | | | | | | | | | | | | | | used out of the box. | ||||
* | | | | | | fix todo: use minitest skip instead of return nil | Francesco Rodriguez | 2012-05-13 | 1 | -9/+8 |
| | | | | | | |||||
* | | | | | | Merge pull request #6193 from mhfs/fix_oracle_spec | Rafael Mendonça França | 2012-05-12 | 1 | -1/+1 |
|\ \ \ \ \ \ | | | | | | | | | | | | | | | Fix failing AR test on Oracle |