1 files changed, 113 insertions, 23 deletions
diff --git a/activesupport/CHANGELOG.md b/activesupport/CHANGELOG.md
index 7514faabb2..74d57180fe 100644
--- a/activesupport/CHANGELOG.md
+++ b/activesupport/CHANGELOG.md
@@ -1,3 +1,93 @@
+*   Corrected Inflector#underscore handling of multiple successive acroynms.
+
+    *James Le Cuirot*
+
+*   Delegation now works with ruby reserved words passed to `:to` option.
+
+    Fixes #16956.
+
+    *Agis Anastasopoulos*
+
+*   Added method `#eql?` to `ActiveSupport::Duration`, in addition to `#==`.
+
+    Currently, the following returns `false`, contrary to expectation:
+
+        1.minute.eql?(1.minute)
+
+    Adding method `#eql?` will make this behave like expected. Method `#eql?` is
+    just a bit stricter than `#==`, as it checks whether the argument is also a duration. Their
+    parts may be different though.
+
+        1.minute.eql?(60.seconds)  # => true
+        1.minute.eql?(60)          # => false
+
+    *Joost Lubach*
+
+*   Time#change can now change nanoseconds (:nsec) as a higher-precision
+    alternative to microseconds (:usec).
+
+    *Agis Anastasooulos*
+
+*   `MessageVerifier.new` raises an appropriate exception if the secret is `nil`.
+    This prevents `MessageVerifier#generate` from raising a cryptic error later on.
+
+    *Kostiantyn Kahanskyi*
+
+*   Introduced new configuration option `active_support.test_order` for
+    specifying the order in which test cases are executed. This option currently defaults
+    to `:sorted` but will be changed to `:random` in Rails 5.0.
+
+    *Akira Matsuda*, *Godfrey Chan*
+
+*   Fixed a bug in Inflector#underscore where acroynms in nested constant names
+    are incorrectly parsed as camelCase.
+
+    Fixes #8015.
+
+    *Fred Wu*, *Matthew Draper*
+
+*   Make Time#change throw an exception if the :usec option is out of range and
+    the time has an offset other than UTC or local.
+
+    *Agis Anastasopoulos*
+
+*   `Method` objects now report themselves as not `duplicable?`. This allows
+    hashes and arrays containing `Method` objects to be `deep_dup`ed.
+
+    *Peter Jaros*
+
+*   `determine_constant_from_test_name` does no longer shadow `NameError`s
+    which happens during constant autoloading.
+
+    Fixes #9933.
+
+    *Guo Xiang Tan*
+
+*   Added instance_eval version to Object#try, so you can do this:
+
+      person.try { name.first }
+
+    instead of:
+
+      person.try { |person| person.name.first }
+
+    *DHH*
+
+*   Fix the `ActiveSupport::Duration#instance_of?` method to return the right
+    value with the class itself since it was previously delegated to the
+    internal value.
+
+    *Robin Dupret*
+
+*   Fix rounding errors with #travel_to by resetting the usec on any passed time to zero, so we only travel
+    with per-second precision, not anything deeper than that.
+
+    *DHH*
+
+*   Fix DateTime comparison with DateTime::Infinity object.
+
+    *Rafael Mendonça França*
+
 *   Added Object#itself which returns the object itself. Useful when dealing with a chaining scenario, like Active Record scopes:
 
         Event.public_send(state.presence_in([ :trashed, :drafted ]) || :itself).order(:created_at)
@@ -45,8 +135,8 @@
 
 *   Always instrument `ActiveSupport::Cache`.
 
-    Since `ActiveSupport::Notifications` only instrument items when there
-    are subscriber we don't need to disable instrumentation.
+    Since `ActiveSupport::Notifications` only instruments items when there
+    are attached subscribers, we don't need to disable instrumentation.
 
     *Peter Wagenet*
 
@@ -69,17 +159,17 @@
     `ActiveSupport::TimeWithZone#-` should return the same result as if we were
     using `Time#-`:
 
-        Time.now.end_of_day - Time.now.beginning_of_day #=> 86399.999999999
+        Time.now.end_of_day - Time.now.beginning_of_day # => 86399.999999999
 
     Before:
 
-        Time.zone.now.end_of_day.nsec #=> 999999999
-        Time.zone.now.end_of_day - Time.zone.now.beginning_of_day #=> 86400.0
+        Time.zone.now.end_of_day.nsec # => 999999999
+        Time.zone.now.end_of_day - Time.zone.now.beginning_of_day # => 86400.0
 
     After:
 
         Time.zone.now.end_of_day - Time.zone.now.beginning_of_day
-        #=> 86399.999999999
+        # => 86399.999999999
 
     *Gordon Chan*
 
@@ -88,12 +178,12 @@
     Before:
 
         ActiveSupport::NumberHelper.number_to_rounded Rational(1000, 3), precision: 2
-        #=> "330.00"
+        # => "330.00"
 
     After:
 
         ActiveSupport::NumberHelper.number_to_rounded Rational(1000, 3), precision: 2
-        #=> "333.33"
+        # => "333.33"
 
     See #15379.
 
@@ -175,9 +265,9 @@
 
     *Xavier Noria*
 
-*   Fixed backward compatibility isues introduced in 326e652.
+*   Fixed backward compatibility issues introduced in 326e652.
 
-    Empty Hash or Array should not present in serialization result.
+    Empty Hash or Array should not be present in serialization result.
 
         {a: []}.to_query # => ""
         {a: {}}.to_query # => ""
@@ -196,20 +286,20 @@
 
     This fixes the current situation of:
 
-        1.second.eql?(1.second) #=> false
+        1.second.eql?(1.second) # => false
 
     `eql?` also requires that the other object is an `ActiveSupport::Duration`.
     This requirement makes `ActiveSupport::Duration`'s behavior consistent with
     the behavior of Ruby's numeric types:
 
-        1.eql?(1.0) #=> false
-        1.0.eql?(1) #=> false
+        1.eql?(1.0) # => false
+        1.0.eql?(1) # => false
 
-        1.second.eql?(1) #=> false (was true)
-        1.eql?(1.second) #=> false
+        1.second.eql?(1) # => false (was true)
+        1.eql?(1.second) # => false
 
         { 1 => "foo", 1.0 => "bar" }
-        #=> { 1 => "foo", 1.0 => "bar" }
+        # => { 1 => "foo", 1.0 => "bar" }
 
         { 1 => "foo", 1.second => "bar" }
         # now => { 1 => "foo", 1.second => "bar" }
@@ -217,11 +307,11 @@
 
     And though the behavior of these hasn't changed, for reference:
 
-        1 == 1.0 #=> true
-        1.0 == 1 #=> true
+        1 == 1.0 # => true
+        1.0 == 1 # => true
 
-        1 == 1.second #=> true
-        1.second == 1 #=> true
+        1 == 1.second # => true
+        1.second == 1 # => true
 
     *Emily Dobervich*
 
@@ -231,9 +321,9 @@
 
     *Pavel Pravosud*
 
-*   `HashWithIndifferentAccess` better respects `#to_hash` on objects it's
-    given. In particular, `.new`, `#update`, `#merge`, `#replace` all accept
-    objects which respond to `#to_hash`, even if those objects are not Hashes
+*   `HashWithIndifferentAccess` better respects `#to_hash` on objects it
+    receives. In particular, `.new`, `#update`, `#merge`, and `#replace` accept
+    objects which respond to `#to_hash`, even if those objects are not hashes
     directly.
 
     *Peter Jaros*