Activity
From 08/19/2016 to 08/25/2016
08/25/2016
-
08:55 PM Bug #12599: For CLang, increase inline-threshold to get 7%-10% speedup of optcarrot
- Including ALWAYS_INLINE in the header and then defining the method in a .c file doesn't seem to successfully inline the function invocations -- they still show up in a GPerfTools profiling listing, for instance. So I think that using ALW...
-
- Based on diffs of long profiling runs of optcarrot, I think the following functions aren't inlined by default, and (I suspect) should be. Working on code changes for that now.
rb_get_alloc_func, rb_ary_rotate, rb_ary_modify.
I'm al... -
06:44 PM Feature #12697: Why shouldn't Module meta programming methods be public?
- Hi,
As stated by bughit, a typical case where we have to resort to `send` for these is in meta programming, say:
def self.included(base)
base.send(:define_method, :foo) { ... } unless base.column_names.include?(:foo)
... -
06:43 PM Bug #12706 (Closed): Hash#each yields inconsistent number of args
- ```ruby
def foo(a, b)
p [a, b]
end
def bar(a, b = 2)
p [a, b]
end
foo_lambda = method(:foo).to_proc
bar_lambda = method(:bar).to_proc
{a: 1}.each(&foo_lambda)
{a: 1}.each(&bar_lambda)
```
From #12705, yielding t... -
- a related point is that lambda procs with the same arity (normal_lambda, method_lambda) should behave the same
-
- ```ruby
def yield_test
yield 1, 2
yield [1, 2]
end
def foo(a, b)
p a, b
end
method_lambda = method(:foo).to_proc
normal_lambda = ->a, b{p a, b}
yield_test(&normal_lambda)
yield_test(&method_lambda)
```
the yi... -
05:40 PM Feature #12700: regexg heredoc support
- Added pull request https://github.com/ruby/ruby/pull/1418
-
05:27 PM Feature #12676: Significant performance increase, and code conciseness, for prime_division method in prime.rb
- This is neater.
replace
```ruby
raise ZeroDivisionError if self == 0
```
with
```ruby
return [] if self | 1 == 1
``` -
- Question:
Why do you raise an error for the value '0'?
```ruby
1.prime_division => []
```
Why not also:
```ruby
0.prime_division => []
```
This is more mathematically consistent because neither
are prime so neither have p... -
01:57 AM Feature #12676 (Assigned): Significant performance increase, and code conciseness, for prime_division method in prime.rb
-
02:28 PM Bug #12704 (Closed): Fix typo in ossl_ocspreq_sign
- The code of ossl_ocspreq_sign reads:
~~~ c
static VALUE
ossl_ocspreq_sign(int argc, VALUE *argv, VALUE self)
{
VALUE signer_cert, signer_key, certs, flags, digest;
OCSP_REQUEST *req;
X509 *signer;
EVP_PKEY *key;... -
12:53 PM Bug #12681: Gemfied tk issue
- Actually this line is suspicious:
https://github.com/ruby/ruby/blob/trunk/ext/extmk.rb#L568
It expands to `$(topdir)/.ext/gems` and creates the `$(topdir)` directory indeed. But that doesn't look to be the original intention here. ... -
- Moreover, I still see the `tcltklib.so` and `tkutil.so` installed into `/usr/lib64/ruby/2.4.0/x86_64-linux/` but they should not be there at all.
-
- The r55950 does not resolve the issue completely. The make install does not work (neither the gem itself probably):
```
# make install DESTDIR=/builddir/build/BUILDROOT/ruby-2.4.0-0.1.r56008.fc26.x86_64
... snip ...
./miniruby ... -
- Hiroshi SHIBATA wrote:
> >Also, the "gem install tk" does not build the extension, so that is another source of possible issues.
> ...
Yes, the 0.1.1 looks good. Thx. -
10:43 AM Bug #12353: Regression with Marshal.dump on ruby 2.3.1p112 (2016-04-26 revision 54768) [x86_64-darwin15]
- So, can we revert the backport?
This issue spotted a feature bug and it got fixed in trunk, but the original problem (`Marshall.dump(1.day)` does not work with ruby 2.3+ and activesupport 4.1.x) was not solved by the fix, causing a mo... -
05:52 AM Bug #12703 (Rejected): String.gsub returns unexpected results with numeric match characters
- What you want is `"z'z".gsub(/(')/, '\\\\\1')`.
You need to escape twice, for a literal, and for `gsub`. -
02:23 AM Bug #12703 (Rejected): String.gsub returns unexpected results with numeric match characters
- I'm unable to prepend a backslash to numeric group matches using gsub
~~~
irb(main):053:0> "z'z".gsub(/(')/, '\\1')
=> "z'z"
~~~
Expected Result
~~~
=> "z\'z"
~~~
Adding an extra backslash gives me this:
~~~
irb(main... -
03:03 AM Bug #11846 (Closed): Block passed by &block is GC-ed
- wanabe san, Thank you for your investigation!
So this issue is already fixed on both 2.3/trunk and not exist on 2.2 and before.
I'll close this ticket and fill Backport field. -
- * win32/file.c (home_dir, replace_to_long_name): use ALLOC_N
instead of casted xmalloc with multiplication.
win32/file.c (rb_file_expand_path_internal): ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56008 b2dd03c8-39d4-4d8... -
- * win32/file.h (rb_w32_filecp): add declaration.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56007 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
12:50 AM Bug #12699: Crash in the VM - maybe garbage collector bug
- Adding a Dockerfile that makes it easy to reproduce the problem, under Valgrind. The output of this Dockerfile gives the following report:
`
==1== Memcheck, a memory error detector
==1== Copyright (C) 2002-2015, and GNU GPL'd, by Juli...
08/24/2016
-
10:52 PM Feature #12690: Improve GC with external library that may use large memory
- Nobu's implementation: https://github.com/ruby/ruby/compare/trunk...nobu:gc_trigger
-
- Koichi agreed:
https://twitter.com/_ko1/status/768278836882526208
> I don't see this ticket details yet but implementing the concept is OK to me.
(Translated to English from Japanese by me.)
-
- Why doesn't your service launcher set `HOME`?
-
02:34 AM Feature #12695: File.expand_path should resolve ~/ using /etc/passwd when HOME is not set
- +1 it makes sense to me.
-
- One last simple tweek to increase overall peformance, in `prime_division5`.
Instead of selecting the optimum pg based on the number's size, first
suck out any factors of some base primes, then determine the optimum
`pg` based on the `... -
- * win32/file.c (append_wstr): remove a codepage argument, and use
INVALID_CODE_PAGE for conversion by econv.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56006 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 03:39 PM Revision 2025f057 (git): * 2016-08-25
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56005 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * win32/file.c (append_wstr): exclude the terminator from the
result length when input len == -1.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56004 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- ruby_2_3 r56003 merged revision(s) 53707,53708,53709,53712,53777,53781.
-
- I overlooked this ticket because it's "Open". The stable branch maintainers watch closed ticket with Backport field filed with REQUIRED.
-
- * lib/rubygems.rb, lib/rubygems/*, test/rubygems/*: Update rubygems-2.5.2.
It supports to enable frozen string literal and add `--norc` option for
disable to `.gemrc` configuration.
See 2.5.2 release notes f... -
- ruby_2_3 r56002 merged revision(s) 53586,53587,54030.
-
- Thank you for request and comments. I overlook this ticket because it's "Open". I'll close this ticket. (We the stable branch maintainers watch the closed ticket whose Backport field filled with REQUIRED).
-
- increase timeout for ARMv7
http://rubyci.s3.amazonaws.com/scw-9d6766/ruby-trunk/log/20160113T091704Z.diff.html.gz
* test/ruby/test_array.rb: split out the test for no stack error
on large input for test_permuta... -
- Yukihiro Matsumoto wrote:
> I still believe
> ...
In many meta programming scenarios you are not including or defining literals at top level but expressions and in methods, so it's not
```ruby
class Class1
include Module1
... -
01:18 PM Feature #12697 (Feedback): Why shouldn't Module meta programming methods be public?
- I still believe
```
class String
include Term::ANSIColor
end
```
is far better than `String.include Term::ANSIColor`. It is clearer and has more space to optimize.
Besides that the fact that `include` etc may have huge perfo... -
- I made the same proposal years ago (#6539), and Matz stated that he ["thinks class/module operations should be done in the scope"](https://bugs.ruby-lang.org/issues/6539#note-3).
I still strongly believe that `include` should be publi... - 01:42 PM Revision 8ddc191d (git): * gc.c (gc_reset_malloc_info): Remove too much ";".
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56001 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
01:28 PM Feature #12698: Method to delete a substring by regex match
- Active Support already has `String#remove`.
```
$ irb -r irb/completion --simple-prompt
>> require 'active_support/all'
=> true
> ...
=> "abab"
>> "abcabc".remove(/\zc/)
=> "abcabc"
> ...
=> "abcab"
>> "abcabc".remove(/\Aabc/)... -
04:42 AM Feature #12698: Method to delete a substring by regex match
- Perhaps my examples were not clear enough. My point is that, unlike `String#delete`, the given string argument should not be interpreted as character classes, but rather as a substring. So, the following should also hold:
```ruby
"ab... -
- There is frequent need to delete a substring from a string. There already are methods `String#delete` and `String#delete!`, but their feature is a little bit different from the use cases that I am mentioning here.
I request methods th... -
- * win32/file.c (home_dir): use enum instead of magic numbers.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56000 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
11:46 AM Bug #12702 (Rejected): DRb client fails when calling out-of-process remote method with a block
- See attached code for test case.
applies to ruby-2.3.1 and ruby-2.4.0dev (2016-08-24 trunk 55999)
The problem starts in DRbObject#method_missing where
the code following DRb.here? is used for the in-process case which works.
Bu... -
11:13 AM Bug #12701 (Closed): compile error about ALWAYS_INLINE
- Applied in changeset r55998 (and r55999)
-
- 遅くとも r55884 以降、Solaris 10上の古い Fujitsu C Compiler (fcc) にて、以下のコンパイルエラーが発生します。
```
fcc -O2 -KV9 -KVIS2 -DRUBY_EXPORT -I/usr/local/64/lib/libffi-3.0.10/include -I/usr/local/64/include -D_XOPEN_SOURCE=500 -I. -I.ext/include/sparc64-s... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55999 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Fix compile error with compilers that do not have force inline
attribute, including old version of fcc on Solaris 10.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55998 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- There is support for ', ", and ` heredocs, but there is no support for /.
Example code with new feature:
~~~ ruby
first = 'first'
orig = /#{first}
match\s # match
this # match this
/x
right = <</REGEXP/x
#{first}
mat... -
07:56 AM Revision fe5f47a6 (git): * .gdbinit: follow r55766's VM change.
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55997 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Basically, we were investigating this: https://github.com/grpc/grpc/issues/7661
Our investigation led to realize that this assert in the protobuf code is being triggered, but only if the garbage collector has been exercised enough: ht... -
05:50 AM Feature #6842: Add Optional Arguments to String#strip
- I am working on implementing this at https://github.com/ruby/ruby/compare/trunk...sonots:lstrip_arg?expand=1
```ruby
"hello".lstrip!("hell") #=> "o"
"hello".lstrip!("ello") #=> nil
"hello".lstrip!(/hell/) #=> "o"
"hello".lstrip!(/... -
- I will continue this issue on https://bugs.ruby-lang.org/issues/6842, let me close.
=> https://bugs.ruby-lang.org/issues/6842 is too broad. Let us reopen this ticket, and talk about a way removing heading substr at this thread. -
- * common.mk (PRELUDE_C): remove unnecessary dependents, which are
not included directly.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55996 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Applied in changeset r55994.
----------
object.c: fix {Module,Class}.new rdoc [ci skip]
* object.c (rb_mod_initialize, rb_class_initialize): [DOC] these
methods do not invoke module_eval/class_eval, just eval the
given block under ... - 03:57 AM Revision fbd49d38 (git): * 2016-08-24
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55995 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * object.c (rb_mod_initialize, rb_class_initialize): [DOC] these
methods do not invoke module_eval/class_eval, just eval the
given block under the new module/class but sharing the context
with the surrounding scope like those metho...
08/23/2016
-
11:23 PM Feature #12697: Why shouldn't Module meta programming methods be public?
- This is somewhat an interesting comment made here, epecially in regards to .send().
Take code snippets such as here:
https://gist.github.com/melborne/665406/4a2a732b1c35b290ddb5a08661453704b8d3c047#file-irc-rb
Quote:
... -
- Methods like alias_method, attr_accessor, define_method, and similar
I don't think Ruby discourages this kind of meta programming, so why make it less convenient, by necessitating `send` or `module_eval`? -
- define_method used to have the same (or similar) problem
https://bugs.ruby-lang.org/issues/9005
which was fixed, so why shouldn't this be? -
- why should top level visibility (which applies to methods defined in the Object class) have any effect on other classes?
this also applies to any other module which you might be in
```ruby
class Class1
end
module SomeUnrelated... -
10:55 PM Bug #12372: Resolv::DNS does not fetch authority section resources when calling getresource(s)
- Ping again.
-
08:18 PM Bug #12696 (Closed): Defining anonymous classes via `Class.new {}` does not run `class_eval` on the block
- When defining anonymous classes, the documentation comment (https://github.com/ruby/ruby/blob/b9a82eaa13a5438318d78aee0afb907a1d143a51/object.c#L1754) explains that
~~~
If a block is given, it is passed the class object, and the block ... -
12:20 PM Bug #12688: Thread unsafety in autoload
- I have the same expectations as you, Charles, and as so I'd also expect this to be a bug.
-
- Thanks, I found an issue which is 3 years old https://bugs.ruby-lang.org/issues/6842
-
- IIRC, there were proposals of `lstrip` and `rstrip` enhancement.
I'd expect them to accept char class(es), than a sub-string.
Note that `rb_str_index_m` accepts a `Regexp` too.
Calling `str_strlen` on it will segfault.
-
- I often write codes like:
```ruby
str = 'abcdef'
substr = 'abc'
str[substr.size..-1] if str.start_with?(substr) #=> 'def'
# or
str.sub(/\A#{Regexp.escape(substr)}/, '') #=> 'def'
```
I want a short hand which is something lik... -
11:36 AM Feature #12695 (Closed): File.expand_path should resolve ~/ using /etc/passwd when HOME is not set
- I would expect a call like File.expand_path('~/.ssh/known_hosts') to work even though HOME is not set.
A reason this is important to fix is that libraries (in my use case Net::SSH) do calls like File.expand_path('~/.ssh/known_hosts'),... -
- * goruby.c (init_golf): call Init_golf, because ruby_init_ext only
registers the function since r43514.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55993 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
-
- It worked like a charm :)
test.rb
```
puts Process.pid
trap(:CONT) do
raise 'CONT'
end
_pid = fork do
sleep 10
end
begin
Process.waitpid(-1)
rescue => e
puts e.backtrace.join("\n")
end
```
Sending sign... -
- You can raise exception on main threadh by Thread.main.raise in trap.
-
- If adding an option to raise EINTR is not acceptable, adding timeout option is acceptable for **my** case.
-
- I am currently using Timeout hack for when I wanted to get EINTR like https://github.com/sonots/ruby-server-starter/blob/7fbc8db2548c215aae02afef385caa683040bd8a/lib/server/starter.rb#L215-L223.
I was glad if I did not need to write s... -
- Naotoshi Seo wrote:
> However, ruby implementation of Process.waitpid does not raise EINTR
No you don't. It is an extremely bad idea to use signal to interrupt execution.
Is it you must use signal for some reason or you just want... -
- http://man7.org/linux/man-pages/man2/waitpid.2.html tells EINTR error if WNOHANG flag was not set and an unblocked signal or a SIGCHLD was caught.
However, ruby implementation of Process.waitpid does not raise EINTR https://github.com... -
- * test/webrick/test_ssl_server.rb (assert_self_signed_cert): close
underlying TCP socket to fix FD leak.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55992 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * test/psych/test_psych.rb (test_load_file_with_fallback): fix
Tempfile leak. https://github.com/tenderlove/psych/pull/288
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55991 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * string.c (rb_fs_setter): check and convert $; value at
assignment.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55990 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
08/22/2016
-
- ruby_2_3 r55989 merged revision(s) 55554.
-
- * test/fiddle/test_pointer.rb (test_to_str, test_to_s, test_aref_aset):
Attempt to use independent strings for destructive tests that
directly modify values on memory by using Fiddle::Pointer.
[Bug #12537] [... -
- We can make it faster across ranges, with just a little bit more code,
by adaptively selecting the Prime Generator to use based on the number size.
The method select_pg has a profile of the best PG to use for given number ranges.
It's... -
- ruby_2_3 r55988 merged revision(s) 55547,55551,55552,55555,55557,55559,55575,55691,55568.
-
- * string.c: Fix memory corruptions when using UTF-16/32 strings.
[Bug #12536] [ruby-dev:49699]
* string.c (TERM_LEN_MAX): Macro for the longest TERM_FILL length,
the same as largest value of rb_enc_mbminlen(e... -
- Yui Naruse: you suggest I find out which functions are inlined in this way and add ALWAYS_INLINE for them? I'll see if there's a way I can do that.
-
- The change inlines many functions by 3MB, but I think this speed up is come from few functions which are too complex to inline.
If so such functions should be refactored to ease compiles optimize them, or simply add ALWAYS_INLINE.
Fo... - 05:10 PM Revision 2e2f7df7 (git): * 2016-08-23
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55987 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * string.c (rb_str_split_m): show $; name in error message when it
is a wrong object.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55986 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Okay, so sounds like just a documentation fix for ReXML. That makes sense. Thanks!
-
- Noah Gibbs wrote:
> Sounds good to me. Do we need to fix ReXML, if it depends on the current rounding mode?
The rdoc of `REXML::Functions.number` states:
> ...
sign followed by a Number followed by whitespace is converted to the IEE... -
- Noah Gibbs wrote:
> Another reason not to add an argument: `Float#round` already takes an argument for the number of digits to round to.
It's easy to distinguish a `Fixnum` and a `Symbol`.
Or keyword arguments will work well. -
- Another reason not to add an argument: `Float#round` already takes an argument for the number of digits to round to.
-
12:58 PM Bug #12421: Please backport r53816, r53817
- backporting those fixes would make ruby compatible with a syntax handled by other tools. otherwise it means people need to take special care for ruby and things like that are often used against ruby later on.
also do you see any poten... -
- That's why i would deprecate the $ variables and make people use match objects all the time.
I mean the stdlib even has code that reads
```
matchdata = $~
```
That feels just wrong.
Maybe 2.4 could start issue warnings abou... -
- Applied in changeset r55985.
----------
csv.rb: performance with very long quoted lines
* lib/csv.rb (CSV#shift): store partial quoted strings in an array
and join at last, to improve performance with very long quoted
lines. [ruby... -
- * lib/csv.rb (CSV#shift): store partial quoted strings in an array
and join at last, to improve performance with very long quoted
lines. [ruby-core:76987] [Bug #12691]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55985 b2dd03c... -
07:23 AM Bug #12688: Thread unsafety in autoload
- Charles Nutter wrote:
> I need clarification here. I expected, based on Ruby's assertion that autoloads are thread-safe, that the following code would never error. Instead, it gets a couple iterations in and raises NameError:
> ...
`$"... -
- Applied in changeset r55984.
----------
irb.1: useless option [ci skip]
* man/irb.1: remove useless -width option.
[ruby-dev:49767] [Bug #12692] -
- * man/irb.1: remove useless -width option.
[ruby-dev:49767] [Bug #12692]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55984 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * array.c (rb_ary_splice): consider elements in middle of self.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55983 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 12:03 AM Revision 7ac136c9 (git): * 2016-08-22
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55982 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * iseq.c (Init_ISeq): undefine allocator of InstructionSequence,
to get rid of segfaults at method call on uninitialized object.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55981 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
08/21/2016
-
- Guyren Howe wrote:
> I believe this behavior is wrong and should be fixed.
> ...
There is a simple fix for your use-case, if you just want to fowrard arguments, don't use ** at all:
(it's not like in Python, keyword arguments are less... -
06:09 AM Bug #10856: Splat with empty keyword args gives unexpected results
- I believe this behavior is wrong and should be fixed.
This gets in the way of simple functional programming idioms. eg "Call each of these functions with these args until one doesn't fail"
~~~ ruby
class FnSeries
def initialize... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55980 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Applied in changeset ruby_2_3|r55979.
----------
merge revision(s) 53433: [Backport #12668]
test_ostruct.rb: sort method names
* test/ostruct/test_ostruct.rb (test_accessor_defines_method):
should not depend on the order of methods. -
- ruby_2_3 r55979 merged revision(s) 53433.
-
-
- test_ostruct.rb: sort method names
* test/ostruct/test_ostruct.rb (test_accessor_defines_method):
should not depend on the order of methods.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_3@55979 b2dd03c8-39d4-4d... -
05:14 AM Revision 009a2956 (git): * ChangeLog: fix typo.
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55978 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
05:13 AM Feature #12347: [PATCH] Make CSV::Row#each etc. return Enumerator when no block given
- I will commit it in a month.
Let me know if you have any concerns.
-
- * array.c (rb_ary_splice): use pointer and length pair instead of
an array object to replace.
* array.c (rb_ary_insert): get rid of creating temporary array.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55977 b2dd03c8-39d4-4d8f-... -
- r55500 breaks `1.days`, while it fixed Marshal.dump with object which respond to marshal_dump with method_missing.
Even though it's obviously a bug-fix, introduce more critical issue as for activesupport-4.1.x.
I think I should revert ...
08/20/2016
- 08:51 PM Revision 48e19b66 (git): * 2016-08-21
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55976 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 08:51 PM Revision d661b277 (git): enum.c (enum_sort): prevent wasteful array duplicaion
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55975 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55974 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
05:30 AM Bug #12692 (Closed): remove RUBYLIB_PREFIX from man/irb.1
- remove RUBYLIB_PREFIX like r45599?
* https://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/man/irb.1?view=markup&pathrev=53867#l122 -
- I think it's intentional, and don't see why it should contain a new line.
-
- * rubystub.c: generalize win32/stub.c.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55973 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 02:19 AM Revision e79b23c1 (git): * 2016-08-20
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55972 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * win32/win32.c (StartSockets): rb_fatal does not need a newline
at the end.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55971 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
08/19/2016
-
10:43 PM Feature #11815: Proposal for method `Array#difference`
- Implemented the method in a clearer and more efficient manner.
-
10:09 PM Bug #12691 (Closed): CSV performance problem on large files that are misformatted (unclosed quoted field)
- If you have a large file which has an unclosed quoted field in it, the amount of time it takes for the CSV parser to determine that error increases worse-than-exponentially. My example tests -
60k records - takes 50 seconds to determ... -
- An additional change to the implementation of `prime_division2` provides an
another 3x increase in speed over using the generators implemented as classes.
Using class based generators, and enumerable methods, incur a metaprogramming ... -
07:21 PM Feature #8365: Make variables objects
- I deleted my account by mistake so my earlier posts show as 'Anonymous'. I'm back and curious how this feature request resolves in the end.
-
- Marcus Rückert wrote:
> I wonder, if moving away from those special $ variables to explicit match objects wouldn't be a possible solution to this.
If you always use the returned MatchData then you can avoid these problems. This only ... -
- I wonder, if moving away from those special $ variables to explicit match objects wouldn't be a possible solution to this.
-
- Maybe $~ is always set in the surrounding method frame, but never in a block frame?
There is still a lot of weird cases to explain though. -
- To clarify the one-liners' behavior: when the thread's top-level frame is the same as a proc's frame that it calls, it will see thread-local values. When the proc's frame is not the top-level frame for the thread, the memory location for...
-
- We are debating what is correct behavior now, and what should be correct behavior in the future, for the thread-visibility of the special variables `%~` and `$_`
We have several examples from https://github.com/jruby/jruby/issues/3031... -
- LGTM. Iff Koichi agrees, it should be merged.
Matz.
-
- GC cares memory allocated by only Ruby. GC doesn't care memory allocated by external library. GC isn't run frequently when external library allocates large memory but Ruby allocates only small memory. In the case, process uses large memo...
-
- * ruby.c (ruby_cmdline_options_t): typedef.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55970 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- win32/stub.c only works with malloced argv as win32/win32.c, to
realloc it.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55969 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * cygwin/GNUmakefile.in (scriptbin): enable on cygwin.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55968 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * cygwin/GNUmakefile.in (STUBPROGRAM): strip.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55967 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Oh, sorry...the source of bar.rb is trivial:
```ruby
class Foo
Bar = 1
end
``` -
- I need clarification here. I expected, based on Ruby's assertion that autoloads are thread-safe, that the following code would never error. Instead, it gets a couple iterations in and raises NameError:
```ruby
loop do
class Foo
... -
- * ext/extmk.rb (extmake): now yield makefile content at creation.
substitute makefiles for gems in the block, not to rewrite when
nothing changed.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55966 b2dd03c8-39d4-4d8f-98ff-823fe... -
- Applied in changeset r55965.
----------
parse.y: lparen_arg statement
* parse.y (primary): allow parenthesised statement as a method
argument. [Feature #12686] -
- * parse.y (primary): allow parenthesised statement as a method
argument. [Feature #12686]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55965 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * vm.c (vm_set_main_stack): TOPLEVEL_BINDING must be built.
http://www.viva64.com/en/b/0414/#ID0EQ1CI [ruby-core:76973]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55964 b2dd03c8-39d4-4d8f-98ff-823fe69b080e