aboutsummaryrefslogtreecommitdiffstats
path: root/actionpack/lib
Commit message (Collapse)AuthorAgeFilesLines
* cache host on the stack so we only look it up onceAaron Patterson2014-06-111-8/+9
|
* only pull :protocol from the options hash onceAaron Patterson2014-06-111-2/+3
|
* cache protocol on the stack to reduce options hash lookupsAaron Patterson2014-06-111-7/+7
|
* eliminate nil checks in normalize_portAaron Patterson2014-06-111-7/+7
|
* reduce hash lookups and disconnect normalize_port from the options hashAaron Patterson2014-06-111-9/+9
|
* Handle client disconnect during live streamingMatthew Draper2014-06-082-13/+95
| | | | .. even when the producer is blocked for a write.
* adds some details to the rationale of converted_arrays [ci skip]Xavier Noria2014-06-071-0/+4
|
* Revert "Convert StrongParameters cache to a hash. This fixes an unbounded"Xavier Noria2014-06-071-6/+6
| | | | | | | | | | | | | | | | | | | | We cannot cache keys because arrays are mutable. We rather want to cache the arrays. This behaviour is tailor-made for the usage pattern strongs params is designed for. In a forthcoming commit I am going to add a test that covers why we need to cache by value. Every strong params instance has a live span of a request, the cache goes away with the object. Since strong params have such a concrete intention, it would be interesting to see if there are actually any real-world use cases that are an actual leak, one that practically may matter. I am not convinced that the theoretical leak has any practical consequences, but if it can be shown there are, then I believe we should either get rid of the cache (which is an optimization), or else wipe it in the mutating API. This reverts commit e63be2769c039e4e9ada523a8497ce3206cc8a9b.
* [ci skip] Fix capitalizationAkshay Vishnoi2014-06-072-3/+3
|
* remove another wasteful AS::SafeBuffer allocationAaron Patterson2014-06-061-1/+1
|
* eliminate wasteful AS::SafeBuffer allocationAaron Patterson2014-06-061-1/+1
|
* Use a frozen Set instance for CGI_VARIABLES.Nathaniel Bibler2014-06-051-8/+21
| | | Also expand the CGI_VARIABLE name listing to multiple lines for cleaner diffs and legibility.
* Avoid misuse of underscore argumentCorey Ward2014-06-051-2/+2
| | | Per convention, underscore-only argument names should be used for unused parameters.
* Merge branch 'master' of github.com:rails/railsAaron Patterson2014-06-044-36/+14
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * 'master' of github.com:rails/rails: For our build, stick with mail 2.5.x for now Correct result, previously showing wrong result Collapse PG default extractoin of most types to single regex Change wording of explanation about precision & scale of decimal numbers [ci skip] Cleaned up duplicated CHANGELOG entry [ci skip] reuse available belongs_to? method Convert StrongParameters cache to a hash. This fixes an unbounded memory leak demonstrated on @tenderlove's latest blog post: Partially revert deprecation of *_filter Pluralize params Add default_i18n_subject to the guides Fix regression on eager loading association based on SQL query rather than existing column. Relax mail gem constraint from ~> 2.5.4 to ~> 2.5, >= 2.5.4 Keep column defaults in type cast form Return a null column when no column exists for an attribute Refactor XML serializer to not depend on `serialized_attributes` Test the serialized types of virtual columns in XML implement ActiveRecord::Base#pretty_print + changelog Remove duplicated HashWithIndifferentAccess#with_indifferent_access.
| * Merge pull request #15498 from zenspider/fix_memory_leakGuillermo Iguaran2014-06-031-6/+6
| |\ | | | | | | Convert StrongParameters cache to a hash. This fixes an unbounded memory leak
| | * Convert StrongParameters cache to a hash. This fixes an unboundedRyan Davis2014-06-031-6/+6
| | | | | | | | | | | | | | | | | | memory leak demonstrated on @tenderlove's latest blog post: http://tenderlovemaking.com/2014/06/02/yagni-methods-are-killing-me.html
| * | Partially revert deprecation of *_filterRafael Mendonça França2014-06-031-27/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We are going to deprecate only on Rails 5 to make easier plugin maintainers support different Rails versions. Right now we are only discouraging their usage. This reverts commit 6c5f43bab8206747a8591435b2aa0ff7051ad3de. Conflicts: actionpack/CHANGELOG.md
| * | Merge pull request #15349 from tgxworld/remove_duplicated_method_callRafael Mendonça França2014-06-032-3/+2
| |\ \ | | | | | | | | Remove duplicated HashWithIndifferentAccess#with_indifferent_access.
| | * | Remove duplicated HashWithIndifferentAccess#with_indifferent_access.Guo Xiang Tan2014-05-262-3/+2
| | | |
* | | | push splitting "to" up the callstackAaron Patterson2014-06-031-15/+20
|/ / /
* | | Routes specifying 'to:' must be a string that contains a "#" or a rackAaron Patterson2014-06-031-3/+21
| | | | | | | | | | | | | | | application. Use of a symbol should be replaced with `action: symbol`. Use of a string without a "#" should be replaced with `controller: string`.
* | | push some options munging to an alternate constructorAaron Patterson2014-06-031-13/+19
| | |
* | | move options_constraints tests next to each otherAaron Patterson2014-06-031-2/+2
| | |
* | | only do is_a checks on `formatted` onceAaron Patterson2014-06-031-9/+5
| | |
* | | push options_constraints processing upAaron Patterson2014-06-031-22/+24
| | | | | | | | | | | | | | | this lets us reduce is_a checks on the options_constraints and push "callable constraints" verification to the right place.
* | | remove options as an ivarAaron Patterson2014-06-031-6/+8
| | |
* | | only do Regexp === option onceAaron Patterson2014-06-031-3/+4
| | |
* | | return early from add_request_methodAaron Patterson2014-06-031-11/+11
| | |
* | | only loop over `options` once (hopefully)Aaron Patterson2014-06-031-13/+11
| | |
* | | only loop through constraints onceAaron Patterson2014-06-031-19/+15
| | |
* | | change defaults allocation to a one-linerAaron Patterson2014-06-031-3/+2
| | |
* | | no longer need the scope ivarAaron Patterson2014-06-031-2/+1
| | |
* | | only look up scope[:module] onceAaron Patterson2014-06-031-6/+6
| | |
* | | remove IGNORE_OPTIONSAaron Patterson2014-06-031-4/+8
| | |
* | | :only is never used in Mapping, so rm the keyAaron Patterson2014-06-031-1/+2
| | |
* | | remove :defaults from the IGNORE_OPTIONS listAaron Patterson2014-06-031-4/+4
| | |
* | | remove :constraints from IGNORE_OPTIONSAaron Patterson2014-06-031-4/+5
| | |
* | | disconnect normalize_defaults! from options[:constraints]Aaron Patterson2014-06-031-6/+6
| | |
* | | disconnect options and scope from the `blocks` methodAaron Patterson2014-06-031-7/+8
| | |
* | | disconnect the constraints method from the options and scope hashesAaron Patterson2014-06-031-12/+15
| | |
* | | :on is removed from options before the Mapping is instantiated, so remove it ↵Aaron Patterson2014-06-031-1/+1
| | | | | | | | | | | | from IGNORE_OPTIONS
* | | pull up via extraction and remove it from options / IGNORE_OPTIONSAaron Patterson2014-06-031-8/+6
| | |
* | | always make :via a listAaron Patterson2014-06-031-3/+2
| | |
* | | always pull out a via variable and simplify logicAaron Patterson2014-06-031-12/+15
| | |
* | | add a test for missing "via" parameterAaron Patterson2014-06-031-1/+1
| | |
* | | remove :as and :anchor from IGNORE_OPTIONSAaron Patterson2014-06-031-3/+5
| | |
* | | remove another value from IGNORE_OPTIONSAaron Patterson2014-06-031-2/+2
| | |
* | | shorten up IGNORE_OPTIONSAaron Patterson2014-06-031-2/+2
| | | | | | | | | | | | | | | | | | since we are now passing the format value around, we can remove it from the options hash, which means we don't need to consult as many values from IGNORE_OPTIONS
* | | only look up the format option from the hash onceAaron Patterson2014-06-031-16/+18
| | |
* | | Mapping never actually uses @set, so rmAaron Patterson2014-05-291-3/+3
| | |
generated by cgit v1.2.3 (git 2.25.1) at 2024-07-02 14:17:39 +0000