ruby--ruby/NEWS

# -*- rdoc -*-

= NEWS for Ruby 2.6.0

This document is a list of user visible feature changes made between
releases except for bug fixes.

Note that each entry is kept so brief that no reason behind or reference
information is supplied with.  For a full list of changes with all
sufficient information, see the ChangeLog file or Redmine
(e.g. <tt>https://bugs.ruby-lang.org/issues/$FEATURE_OR_BUG_NUMBER</tt>)

== Changes since the 2.5.0 release

=== Language changes

* <code>$SAFE</code> now is a process global state and can be set to 0 again. [Feature #14250]

* Refinements take place at block passing.  [Feature #14223]

* Refinements take place at Kernel#public_send.  [Feature #15326]

* Refinements take place at Kernel#respond_to?.  [Feature #15327]

* +else+ without +rescue+ now causes a syntax error.  [EXPERIMENTAL] [Feature #14606]

* Constant names may start with a non-ASCII capital letter. [Feature #13770]

* Endless ranges are introduced.  You can use a Range that has no end,
  like <code>(0..)</code> (or similarly <code>(0...)</code>).  [Feature #12912]

  The following shows typical use cases:

    ary[1..]                              # identical to ary[1..-1]
    (1...).each {|index| block }          # infinite loop from index 1
    ary.zip(1..) {|elem, index| block }   # ary.each.with_index(1) { }

* Non-Symbol keys in a keyword arguments hash cause an exception.

* The "shadowing outer local variable" warning is removed.  [Feature #12490]

  You can now write the following without warning:

    user = users.find {|user| cond(user) }

* Print +cause+ of the exception if the exception is not caught and printed
  its backtraces and error message. [Feature #8257]

* The flip-flop syntax is deprecated. [Feature #5400]

=== Core classes updates (outstanding ones only)

[Array]

  [New methods]

    * Added Array#union and Array#difference instance methods.  [Feature #14097]

  [Modified methods]

    * Array#to_h now accepts a block that maps elements to new key/value pairs.  [Feature #15143]

  [Aliased methods]

    * Array#filter is a new alias for Array#select. [Feature #13784]
    * Array#filter! is a new alias for Array#select!. [Feature #13784]

[Binding]

  [New methods]

    * Added Binding#source_location.  [Feature #14230]

      This method returns the source location of the binding, a 2-element
      array of <code>__FILE__</code> and <code>__LINE__</code>.
      Traditionally, the same information could be retrieved by
      <code>eval("[__FILE__, __LINE__]", binding)</code>, but we are
      planning to change this behavior so that Kernel#eval ignores
      binding's source location [Bug #4352].  So, users should use this
      newly-introduced method instead of Kernel#eval.

[Dir]

  [New methods]

    * Added Dir#each_child and Dir#children instance methods. [Feature #13969]

[Enumerable]

  [New methods]

    * Enumerable#chain returns an enumerator object that iterates over the
      elements of the receiver and then those of each argument
      in sequence.  [Feature #15144]

  [Modified methods]

    * Enumerable#to_h now accepts a block that maps elements to new key/value pairs.  [Feature #15143]

  [Aliased methods]

    * Enumerable#filter is a new alias for Enumerable#select. [Feature #13784]

[Enumerator::ArithmeticSequence]

  * This is a new class to represent a generator of an arithmetic sequence,
    that is a number sequence defined by a common difference. It can be used
    for representing what is similar to Python's slice. You can get an
    instance of this class from Numeric#step and Range#step.

[Enumerator::Chain]

  * This is a new class to represent a chain of enumerables that works as a
    single enumerator, generated by such methods as Enumerable#chain and
    Enumerator#+.

[Enumerator::Lazy]

  [Aliased methods]

    * Enumerator::Lazy#filter is a new alias for
      Enumerator::Lazy#select.  [Feature #13784]

[Enumerator]

  [New methods]

    * Enumerator#+ returns an enumerator object that iterates over the
      elements of the receiver and then those of the other operand.  [Feature #15144]

[ENV]

  [Modified methods]

    * ENV.to_h now accepts a block that maps names and values to new keys and values.  [Feature #15143]

[Exception]

  [New options]

    * Exception#full_message takes +:highlight+ and +:order+
      options. [Bug #14324]

[Hash]

  [Modified methods]

    * Hash#merge, Hash#merge!, and Hash#update now accept multiple
      arguments.  [Feature #15111]

    * Hash#to_h now accepts a block that maps keys and values to new keys and values.  [Feature #15143]

  [Aliased methods]

    * Hash#filter is a new alias for Hash#select.  [Feature #13784]

    * Hash#filter! is a new alias for Hash#select!. [Feature #13784]

[IO]

  [New options]

    * Added new mode character <code>'x'</code> to open files for exclusive
      access. [Feature #11258]

[Kernel]

  [Aliased methods]

    * Kernel#then is a new alias for Kernel#yield_self. [Feature #14594]

  [New options]

    * Kernel#Complex, Kernel#Float, Kernel#Integer, and
      Kernel#Rational take an +:exception+ option to specify the way of
      error handling.  [Feature #12732]

    * Kernel#system takes an +:exception+ option to raise an exception
      on failure.  [Feature #14386]

  [Incompatible changes]

    * Kernel#system and Kernel#exec do not close non-standard file descriptors
      (the default of the +:close_others+ option is changed to +false+,
      but we still set the +FD_CLOEXEC+ flag on descriptors we
      create).  [Misc #14907]

[KeyError]

  [New options]

    * KeyError.new accepts +:receiver+ and +:key+ options to set receiver and
      key in Ruby code.  [Feature #14313]

[Method]

  [New methods]

    * Added Method#<< and Method#>> for Proc composition.  [Feature #6284]

[Module]

  [Modified methods]

    * Module#method_defined?, Module#private_method_defined?, and
      Module#protected_method_defined? now accept the second
      parameter as optional. If it is +true+ (the default value), it checks
      ancestor modules/classes, or checks only the class itself. [Feature #14944]

[NameError]

  [New options]

    * NameError.new accepts a +:receiver+ option to set receiver in Ruby
      code.  [Feature #14313]

[NilClass]

  [New methods]

    * NilClass#=~ is added for compatibility.  [Feature #15231]

[NoMethodError]

  [New options]

    * NoMethodError.new accepts a +:receiver+ option to set receiver in Ruby
      code.  [Feature #14313]

[Numeric]

  [Incompatible changes]

    * Numeric#step now returns an instance of the Enumerator::ArithmeticSequence
      class rather than one of the Enumerator class.

[OpenStruct]

  [Modified methods]

    * OpenStruct#to_h now accepts a block that maps keys and values to new keys and values.  [Feature #15143]

[Proc]

  [New methods]

    * Added Proc#<< and Proc#>> for Proc composition.  [Feature #6284]

  [Incompatible changes]

    * Proc#call doesn't change <code>$SAFE</code> any more.  [Feature #14250]

[Random]

  [New methods]

    * Added Random.bytes.  [Feature #4938]

[Range]

  [New methods]

    * Added Range#% instance method.  [Feature #14697]

  [Incompatible changes]

    * Range#=== now uses the +#cover?+ instead of the +#include?+ method. [Feature #14575]
    * Range#cover? now accepts a Range object. [Feature #14473]
    * Range#step now returns an instance of the Enumerator::ArithmeticSequence
      class rather than one of the Enumerator class.

[Regexp/String]

    * Update Unicode version from 10.0.0 to 11.0.0. [Feature #14802]

      This includes a rewrite of the grapheme cluster (/\X/) algorithm
      and special-casing for Georgian MTAVRULI on String#downcase.

    * Update Emoji version from 5.0 to 11.0.0 [Feature #14802]

[RubyVM::AbstractSyntaxTree]

  [New methods]

    * RubyVM::AbstractSyntaxTree.parse parses a given string and returns AST
      nodes. [experimental]

    * RubyVM::AbstractSyntaxTree.parse_file parses a given file and returns AST
      nodes.  [experimental]

    * RubyVM::AbstractSyntaxTree.of returns AST nodes of the given proc or method.
      [experimental]

[RubyVM]

  [New methods]

    * RubyVM.resolve_feature_path identifies the file that will be loaded by
      "require(feature)". [experimental] [Feature #15230]

[String]

  * String#crypt is now deprecated. [Feature #14915]

  [New features]

    * String#split yields each substring to the block if given. [Feature #4780]

[Struct]

  [Modified methods]

    * Struct#to_h now accepts a block that maps keys and values to new keys and values.  [Feature #15143]

  [Aliased method]

    * Struct#filter is a new alias for Struct#select. [Feature #13784]

[Time]

  [New features]

    * Time.new and Time#getlocal accept a timezone object as well as
      a UTC offset string. Time#+, Time#-, and Time#succ also preserve
      the timezone.  [Feature #14850]

[TracePoint]

  [New features]

    * "script_compiled" event is supported. [Feature #15287]

  [New methods]

    * TracePoint#parameters [Feature #14694]

    * TracePoint#instruction_sequence [Feature #15287]

    * TracePoint#eval_script [Feature #15287]

  [Modified methods]

    * TracePoint#enable accepts new keywords "target:" and "target_line:".
      [Feature #15289]

=== Stdlib updates (outstanding ones only)

[BigDecimal]

  Update to version 1.4.0.  This version includes several compatibility
  issues, see Compatibility issues section below for details.

  [Modified methods]

  * BigDecimal() accepts the new keyword "exception:" similar to Float().

  [Note for the differences among recent versions]

  You should want to know the differences among recent versions of bigdecimal.
  Please select the suitable version of bigdecimal according to the following
  information.

  * 1.3.5 has BigDecimal.new without "exception:" keyword.  You can see the
    deprecation warning of BigDecimal.new when you specify "-w" option.
    BigDecimal(), BigDecimal.new, and Object#to_d methods are the same.

  * 1.4.0 has BigDecimal.new with "exception:" keyword.  You always see the
    deprecation warning of BigDecimal.new.  Object#to_d method is different
    from BigDecimal() and BigDecimal.new.

  * 2.0.0 will be released soon after releasing Ruby 2.6.0.  This version
    will not have the BigDecimal.new method.

[Bundler]

  * Add Bundler to Standard Library. [Feature #12733]

  * Use 1.17.2, the latest stable version.

[Coverage]

  A oneshot_lines mode is added.  [Feature #15022]

  This mode checks "whether each line was executed at least once or not",
  instead of "how many times each line was executed".
  A hook for each line is fired at most once, and after it is fired
  the hook flag is removed, i.e., it runs with zero overhead.

  [New options]

    * Add +:oneshot_lines+ keyword argument to Coverage.start.

    * Add +:stop+ and +:clear+ keyword arguments to Coverage.result.
      If +clear+ is true, it clears the counters to zero.
      If +stop+ is true, it disables coverage measurement.

  [New methods]

    * Coverage.line_stub, which is a simple helper function that
      creates the "stub" of line coverage from a given source code.

[CSV]

  * Upgrade to 3.0.2. This includes performance improvements especially
    for writing. Writing is about 2 times faster.
    See https://github.com/ruby/csv/blob/master/NEWS.md.

[ERB]

  [New options]

    * Add +:trim_mode+ and +:eoutvar+ keyword arguments to ERB.new.
      Now non-keyword arguments other than the first one are softly deprecated
      and will be removed when Ruby 2.5 becomes EOL. [Feature #14256]

    * erb command's <tt>-S</tt> option is deprecated, and will be removed
      in the next version.

[FileUtils]

  [New methods]

    * FileUtils#cp_lr.  [Feature #4189]

[Matrix]

  [New methods]

    * Matrix#antisymmetric?, Matrix#skew_symmetric?

    * Matrix#map!, Matrix#collect! [Feature #14151]

    * Matrix#[]=

    * Vector#map!, Vector#collect!

    * Vector#[]=

[Net]

  [New options]

    * Add +:write_timeout+ keyword argument to Net::HTTP.new. [Feature #13396]

  [New methods]

    * Add Net::HTTP#write_timeout and Net::HTTP#write_timeout=.  [Feature #13396]

  [New constant]

    * Add Net::HTTPClientException to deprecate Net::HTTPServerException,
      whose name is misleading.  [Bug #14688]

[NKF]

  * Upgrade to nkf v2.1.5

[Psych]

  * Upgrade to Psych 3.1.0

[RDoc]

  * Become about 2 times faster.

  * Use SOURCE_DATE_EPOCH to generate files.

  * Fix method line number that slipped off.

  * Enable <code>--width</code>, <code>--exclude</code>,
    and <code>--line-numbers</code> that were ignored.

  * Add support for blockquote by ">>>" in default markup notation.

  * Add support for "Raises" lines in TomDoc notation.

  * Fix syntax error output.

  * Fix many parsing bugs.

[REXML]

  * Upgrade to REXML 3.1.9.
    See https://github.com/ruby/rexml/blob/master/NEWS.md.

  [Improved some XPath implementations]

    * <code>concat()</code> function: Stringify all arguments before concatenating.

    * <code>string()</code> function: Support context node.

    * <code>string()</code> function: Support processing instruction node.

    * Support <code>"*:#{ELEMENT_NAME}"</code> syntax in XPath 2.0.

  [Fixed some XPath implementations]

    * <code>"//#{ELEMENT_NAME}[#{POSITION}]"</code> case

    * <code>string()</code> function: Fix <code>function(document)</code>
      returns nodes that are out of root elements.

    * <code>"/ #{ELEMENT_NAME} "</code> case

    * <code>"/ #{ELEMENT_NAME} [ #{PREDICATE} ]"</code> case

    * <code>"/ #{AXIS}::#{ELEMENT_NAME}"</code> case

    * <code>"#{N}-#{M}"</code> case: One or more white spaces were required
      before <code>"-"</code>

    * <code>"/child::node()"</code> case

    * <code>"#{FUNCTION}()/#{PATH}"</code> case

    * <code>"@#{ATTRIBUTE}/parent::"</code> case

    * <code>"name(#{NODE_SET})"</code> case

[RSS]

  [New options]

    * RSS::Parser.parse now accepts options as Hash. +:validate+ ,
      +:ignore_unknown_element+ , +:parser_class+ options are available.

[RubyGems]

  * Upgrade to RubyGems 3.0.1

  * https://blog.rubygems.org/2018/12/19/3.0.0-released.html

  * https://blog.rubygems.org/2018/12/23/3.0.1-released.html

[Set]

  [Aliased methods]

    * Set#filter! is a new alias for Set#select!.  [Feature #13784]

[URI]

  [New constant]

    * Add URI::File to handle the file URI scheme.  [Feature #14035]

=== Compatibility issues (excluding feature bug fixes)

[Dir]

  * Dir.glob with <code>'\0'</code>-separated pattern list will be deprecated,
    and is now warned.  [Feature #14643]

[File]

  * File.read, File.binread, File.write, File.binwrite, File.foreach, and
    File.readlines do not invoke external commands even if the path starts
    with the pipe character <code>'|'</code>. [Feature #14245]

[Object]

  * Object#=~ is deprecated.  [Feature #15231]

=== Stdlib compatibility issues (excluding feature bug fixes)

* These standard libraries have been promoted to default gems.

  * e2mmap
  * forwardable
  * irb
  * logger
  * matrix
  * mutex_m
  * ostruct
  * prime
  * rexml
  * rss
  * shell
  * sync
  * thwait
  * tracer

[BigDecimal]

  * The following methods are removed.

    * BigDecimal.allocate
    * BigDecimal.ver

  * Every BigDecimal object is frozen. [Feature #13984]

  * BigDecimal() parses the given string similar to Float().

  * String#to_d parses the receiver string similar to String#to_f.

  * BigDecimal.new will be removed in version 2.0.

[Pathname]

  * Pathname#read, Pathname#binread, Pathname#write, Pathname#binwrite,
    Pathname#each_line and Pathname#readlines do not invoke external
    commands even if the path starts with the pipe character <code>'|'</code>.
    This follows [Feature #14245].

=== C API updates

=== Implementation improvements

* Speedup Proc#call because we don't need to care about <code>$SAFE</code>
  any more. [Feature #14318]

  With +lc_fizzbuzz+ benchmark which uses Proc#call many times we can
  measure x1.4 improvements.  [Bug #10212]

* Speedup block.call where +block+ is passed block parameter. [Feature #14330]

  Ruby 2.5 improves block passing performance. [Feature #14045]

  Additionally, Ruby 2.6 improves the performance of passed block calling.

* Introduce an initial implementation of a JIT (Just-in-time) compiler. [Feature #14235] [experimental]

  * <tt>--jit</tt> command line option is added to enable JIT. <tt>--jit-verbose=1</tt>
    is good for inspection.  See <tt>ruby --help</tt> for others.
  * To generate machine code, this JIT compiler uses the C compiler used for building
    the interpreter. Currently GCC, Clang, and Microsoft Visual C++ are supported for it.
  * <tt>--disable-mjit-support</tt> option is added to configure. This is added for JIT debugging,
    but if you get an error on building a header file for JIT, you can use this option to skip
    building it as a workaround.
  * rb_waitpid reimplemented on Unix-like platforms to maintain
    compatibility with processes created for JIT [Bug #14867]

* VM generator script renewal; makes the generated VM more optimized. [GH-1779]

* Thread cache enabled for pthreads platforms (for Thread.new and
  Thread.start).  [Feature #14757]

* timer thread is eliminated for platforms with POSIX timers. [Misc #14937]

* Transient Heap (theap) is supported. [Bug #14858] [Feature #14989]

  theap is a managed heap for short-living memory objects. For example,
  making a small and short-living Hash object is x2 faster. With rdoc benchmark,
  we measured 6-7% performance improvement.

* Native implementations (arm32, arm64, ppc64le, win32, win64, x86, amd64) of
  coroutines to improve performance of Fiber significantly. [Feature #14739]

=== Miscellaneous changes

* On macOS, shared libraries no longer include a full version number of Ruby
  in their names.  This eliminates the burden of each teeny upgrade on the
  platform that users need to rebuild every extension library.

  [Before]
    * libruby.2.6.0.dylib
    * libruby.2.6.dylib -> libruby.2.6.0.dylib
    * libruby.dylib -> libruby.2.6.0.dylib

  [After]
    * libruby.2.6.dylib
    * libruby.dylib -> libruby.2.6.dylib

* Extracted misc/*.el files to https://github.com/ruby/elisp