Activity
From 01/08/2015 to 01/14/2015
01/14/2015
-
10:15 PM Bug #10741: const_defined? triggers autoload where it did not before
- Boo...my fix was bad because it stopped searching superclasses. Since I did not have an appropriate method to search superclasses, split up :: elements, and check private consts all in one, I reverted JRuby to the autoload-inducing state...
-
- To be more specific, this assertion fails on MRI:
>assert_equal("#{File.dirname(__FILE__)}/autoloaded.rb", Object.autoload?("Autoloaded"))
Because the previous const_defined? call has already triggered the autoload. -
- Test in JRuby that caught this change: https://github.com/jruby/jruby/blob/master/test/jruby/test_autoload.rb#L4-L24
Fix in JRuby: https://github.com/jruby/jruby/commit/496c638d2419e089e93eec9759d9775de0390901 -
- This should not cause b.rb to load, but it does in 2.2:
~/projects/jruby $ cat a.rb
module Foo
autoload :Bar, 'b.rb'
end
p Foo.const_defined?('Bar')
~/projects/jruby $ cat b.rb
puts "in b"
module Foo
Bar = 1
end
~/... -
07:56 PM Feature #10740: Base64 urlsafe methods are not urlsafe
- I suspect the reason the spec is that way is that it's easier to calculate what the decoded length will be if the encoding is always divisible by 4, since it's just `(encoded_len / 4) * 3`. It makes more since in the context of wire pro...
-
07:52 PM Feature #10740: Base64 urlsafe methods are not urlsafe
- Hi Yusuke,
RFC6920 is just an example of an RFC which refers to RFC4648 and stipulates that something encoded in base64url MUST NOT be padded. According to RFC4648 this is allowed.
Specifically in the case of RFC6920, the data leng... -
05:43 AM Feature #10740 (Feedback): Base64 urlsafe methods are not urlsafe
- Hello, I'm a maintainer of lib/base64.
I don't think that this is a bug. RFC 4648 is still the latest standard of Base64. (Note that RFC 6920 does not obsolete RFC 4648.) Because lib/base64 is an implementation of Base64, it should... -
12:44 AM Feature #10740: Base64 urlsafe methods are not urlsafe
- I like this feature.
(I think this issue is a feature, not a bug.)
However I think the current behavior should be choosable for compatibility.
I searched Base64.urlsafe_encode64 in gems: base64-urlsafe-encode64-search-result.txt
... -
06:01 PM Bug #10725: Segfault with ObjectSpace::trace_object_allocations_start
- Nobuyoshi Nakada wrote:
> We don't have the same environment equals to yours, we can't address where it happens without the debugging information.
> ...
I understand the difficulties in handling such kind of issues.
I am not able to... -
09:06 AM Bug #10738 (Closed): Parsing invalid URI results in undefined method `split' for nil:NilClass (NoMethodError)
- Applied in changeset r49254.
----------
* lib/uri/mailto.rb: raising URI::InvalidComponentError instead
of failing with undefined method `split' for nil:NilClass for
mailto: URIs without opaque part. [Bug #10738]
* test/uri/testuri.... -
- of failing with undefined method `split' for nil:NilClass for
mailto: URIs without opaque part. [Bug #10738]
* test/uri/testuri.rb: Test for above
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49254 b2dd03c8-39d4-4d8f-98ff-823fe69... -
- Thank you, matz.
Please commit, nobu. -
08:59 AM Feature #10714: Array#reject! nonlinear performance problem
- I agree having linear performance is crucial, and the proposed new behavior is acceptable.
Go ahead, and make the change.
Matz.
-
08:02 AM Bug #10711: Incorrect error message in coerce failed
- memo: see also [ruby-core:67555], [ruby-core:67557], r49234 and r49235.
-
- Backported into `ruby_2_0_0` at r49253.
- 07:53 AM Revision 1ac6292d (git): merge revision(s) 49034: [Backport #10662]
- * ext/tk/lib/tkextlib/tcllib/plotchart.rb: fix to invoke correct function
of tcllib. Patch by @zalt50 [fix GH-787]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_0_0@49253 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Backported r49701, r49702, r49703 and r49708 into `ruby_2_0_0` at r49252.
- 07:49 AM Revision 568a5d52 (git): merge revision(s) 49071-49073,49078: [Backport #10550]
- * lib/resolv.rb (Resolv::DNS::Name#==): DNS is case-insensitive, so the
comparison should be case-insensitive as well.
[ruby-core:66498] [Bug #10550]
* lib/resolv.rb (Resolv::DNS::Name): names with different ... -
07:45 AM Revision a9e033b1 (git): downloader.rb: verify gems
- * tool/downloader.rb (RubyGems.download): verify downloaded gem
packages. LowSecurity to allow untrusted certificates now.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49251 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Backported into `ruby_2_0_0` at r49250.
- 07:32 AM Revision f9d88c60 (git): merge revision(s) 48948: [Backport #10591]
- * lib/net/http/response.rb (Net::HTTPResponse): require one or more
spaces [Bug #10591].
by leriksen <leif.eriksen.au@gmail.com>
https://github.com/ruby/ruby/pull/782 fix GH-782
NOTE: graph.faceboo... -
- Backported into `ruby_2_0_0` at r49249.
- 07:25 AM Revision 06383012 (git): merge revision(s) 48923: [Backport #10633]
- * ext/openssl/ossl_cipher.c (ossl_cipher_update_long): update huge
data gradually not to exceed INT_MAX. workaround of OpenSSL API
limitation. [ruby-core:67043] [Bug #10633]
git-svn-id: svn+ssh://ci.ruby-lang.org/... -
- Backported into `ruby_2_0_0` at r49248.
- 07:15 AM Revision 1c65dd9b (git): merge revision(s) 48886: [Backport #10210]
- * configure.in (rb_cv_binary_elf): get rid of -e option of cat
which is not available on BusyBox, use tr instead.
[ruby-core:64824] [Bug #10210]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_0_0@49248 ... -
- Backported r48884 and r48885 into `ruby_2_0_0` at r49247.
- 07:08 AM Revision ce65dfa3 (git): merge revision(s) 48884,48885: [Backport #10615]
- * signal.c (ruby_signal): since SIGKILL is not supported by MSVCRT,
should be treated before calling signal(2).
[Bug #10615]
should be treated before calling signal(3).
git-svn-id: svn+ssh://ci.ruby-lang.o... -
- Backported into `ruby_2_0_0` at r49246.
note: thread.c is a little different from trunk. - 07:04 AM Revision 9120d051 (git): merge revision(s) 48744,48752: [Backport #10579]
- * eval.c (rb_frame_last_func): return the most recent frame method
name.
* thread.c (recursive_list_access): use the last method name,
instead of the current method name which can be unset in some
c... -
- Backported into `ruby_2_0_0` at r49245.
- 06:55 AM Revision e6a5f817 (git): merge revision(s) 48534: [Backport #10412]
- * lib/resolv.rb: fall back if canonicalization fails.
Thanks Vit Ondruch for the patch! [ruby-core:65836]
* test/resolv/test_dns.rb: test for patch
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_0_0@492... -
- Applied in changeset r49244.
----------
readline.c: initialize before rl_refresh_line
* ext/readline/readline.c (readline_s_refresh_line): initialize
before rl_refresh_line(), as some function make the internal
state non-clean but ... -
02:12 AM Bug #6232: Ruby : Segmentation fault on Readline.refresh_line after screen resize.
- It seems like GNU Readline behavior.
`Readline.refresh_line` calls `rl_refresh_line(0, 0)` -> `rl_clear_to_eol(0)` -> `tputs (_rl_term_clreol, 1, _rl_output_character_function)` -> `_rl_output_character_function(c)` -> `putc (c, _rl_out... -
- * ext/readline/readline.c (readline_s_refresh_line): initialize
before rl_refresh_line(), as some function make the internal
state non-clean but rl_refresh_line() does not re-initialize it.
[ruby-core:43957] [Bug #6232]
git-svn-id...
01/13/2015
-
- * test/socket/test_basicsocket.rb (socks): use dynamically chosen
port number, and remove never used argument.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49243 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49242 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- I ran into this problem trying to implement RFC6920 in this program:
https://github.com/cryptosphere/cryptor/blob/master/lib/cryptor/encoding.rb#L20
RFC6920 says:
~~~
Digest Value: The digest value MUST be encoded using the ba... -
- https://github.com/ruby/ruby/pull/815
-
- Note that SecureRandom.urlsafe_base64 does the right thing by default, with the note "By default, padding is not generated because "=" may be used as a URL delimiter."
-
- Base64.urlsafe_decode64 is not to spec, because it currently REQUIRES appropriate trailing '=' characters.
Base64.urlsafe_encode64 produces trailing '=' characters.
'=' is not web safe, and is not recommended for base64url. Some spe... - 11:22 PM Revision 2705eeea (git): * remove trailing spaces.
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49241 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 11:22 PM Revision d3f7e5d3 (git): * 2015-01-14
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49240 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- and close_write using TCPServer/Socket as a target.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49239 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
10:01 PM Feature #10739 (Open): merge upstream of json-1.8.2
-
05:14 PM Feature #10739 (Feedback): merge upstream of json-1.8.2
- I think we need to fix the regression in gem before we merge to trunk.
We can take advantage of gem's CI for fixing regression. -
- `GET_STATE_TO` macro seems no longer necessary.
-
- I create patch of json-1.8.2 changes. see attached file.
json-1.8.2 implement different way of Typed_Data macro.
we need to apply upstream changes into ruby-core, or backport request to json gem.
nobu, nalsh: how do you think this... -
- * tool/rbinstall.rb (gem): fix changing permissions of installed
bundled gems. [Fix GH-812]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49238 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Applied in changeset r49237.
----------
common.mk: more clean
* common.mk (distclean-local): remove autom4te.cache generated by
autoconf.
* common.mk (realclean-local): remove id sources and dummy header
for dtrace. [ruby-core:675... -
10:30 AM Bug #10737 (Closed): probes.dmyh (and others) not removed on make clean / make distclean
- Some files that are generated by building ruby are not removed upon make clean or make distclean.
This is problematic when one of these files where generated with wrong content but are not regenerated after make clean.
In my case p... -
- * common.mk (distclean-local): remove autom4te.cache generated by
autoconf.
* common.mk (realclean-local): remove id sources and dummy header
for dtrace. [ruby-core:67562] [Bug #10737]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/tr... -
- separate imprementation of Typed_Data macro.
https://github.com/flori/json/compare/v1.8.1...v1.8.2
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49236 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * numeric.c (coerce_failed): fix the error message on non-flonum
platforms.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49235 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * compar.c (rb_cmperr): show float values in the error message
always, not only flonum.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49234 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- I tried to parse an invalid URI with the URI library. I expected an error such as URI::InvalidComponentError or URI::InvalidComponentError, but I got "undefined method `split' for nil:NilClass (NoMethodError)".
The simplest way to che... -
- `BASERUBY` is not necessary when all files have been generated, e.g., unpacked from tarballs, shared with another machine, etc.
-
- The ./configure script should check if a version of ruby is installed.
Currently not such check is performed.
Current behaviour:
configure succeeds even if the is no version of ruby available.
make fails with: "executable host ru... -
08:59 AM Bug #10470: TracePoint cannot trace attr_accessor/reader/writer method
- すいません。
元のサンプルコードを流用してしまったので、tracepointの性能測定としては余り適切じゃない感じのコードになってたことに気付きました。
~~~ruby
trace = TracePoint.new(:return, :c_return) do |tp|
[tp.lineno, tp.defined_class, tp.method_id, tp.event, tp.binding, tp.return_value]
end
~~~
... -
- * tool/rbinstall.rb (gem): reuse options hash, which is invariant
for all gems.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49233 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * tool/rbinstall.rb (install): get rid of shadowing outer local
variables.
* tool/rbinstall.rb: remove unused variables.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49232 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
07:27 AM Bug #10689: `unexpected break' occurs when TracePoint#binding is called
- なんのために 0x02 の bit について気にしていたか覚えていないのですが（クラスとか、その辺でしたっけ）、
よろしいと思います！ よろしくお願いします。
-
- Applied in changeset r49231.
----------
test_http.rb: test without DNS access
* test/net/http/test_http.rb: get rid of accessing DNS actually
for some servers returning wrong results.
[ruby-core:67454] [Bug #10721] -
- * test/net/http/test_http.rb: get rid of accessing DNS actually
for some servers returning wrong results.
[ruby-core:67454] [Bug #10721]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49231 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- This issue may be due to Mac bundled OpenSSL like [Bug #6886].
Two crash reports (this and [Bug #6886]'s) are similar in that crashed threads have called '`CRYPTO_get_ex_data`' before '`_sigtramp`' and crashes have been triggered by '`o... -
- * range.c (range_bsearch): trivial optimizations, for Fixnum, and
by keeping the last satisfied element.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49230 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 03:47 AM Revision 75a43dbe (git): * 2015-01-13
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49229 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * range.c (range_bsearch): preserve encoding of class name in an
exception message.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49228 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * range.c (BSEARCH_CHECK): get rid of conversion of the argument
multiple times.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49227 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
01/12/2015
-
09:12 PM Bug #10735: Memory leak in openssl ossl_pkey_sign
- Zachary Scott wrote:
> I know you tried this with 1.9.3, but could you try to repro on trunk (and newer rubies) first?
> ...
Hi Zachary,
Just ran the test case against Ruby 2.3.0dev (2015-01-12 trunk 49226) [x86_64-darwin14] and see... -
- I know you tried this with 1.9.3, but could you try to repro on trunk (and newer rubies) first?
1.9.3 will be EOL soon, and I want to make sure we fix it upstream before attempting any backports -
- Similar to the memory leak fixed in https://bugs.ruby-lang.org/issues/9743 there is an issue with ossl_pkey_sign. The ruby heap usage reported through GC.stat remains very stable, while the process heap grows linearly with the number of ...
-
06:23 PM Feature #3187: Allow dynamic Fiber stack size
- ~~~
Ruby 2.0 already has
RUBY_VM_FIBER_VM_STACK_SIZE
RUBY_FIBER_MACHINE_STACK_SIZE
~~~
Are these environment variables to configure the stack size? Is there documentation on usage? (I mean, what are the units? bytes, kb?) Is th... -
04:40 PM Bug #10733: Time.httpdate raises ArgumentError when DateTime.now.httpdate is provided as input
After some experimentation it looks like this has something do with the fact that `DateTime#httpdate` returns an `US-ASCII` encoded string.
```ruby
require 'time'
from_httpdate = DateTime.now.httpdate
as_string = "Mon, 12 J...-
03:21 PM Bug #10733: Time.httpdate raises ArgumentError when DateTime.now.httpdate is provided as input
- Another example for Time.httpdate failing.
The third call of Time.httpdate fails, which should be the same as the first call of Time.httpdate.
```ruby
require 'time'
puts RUBY_VERSION
datetime_in_httpdate = DateTime.now.httpda... -
- An irb session demonstrating the bug:
```
irb(main):001:0> require 'time'
=> true
irb(main):002:0> Time.httpdate("Mon, 12 Jan 2015 12:04:15 GMT")
=> 2015-01-12 12:04:15 UTC
irb(main):003:0> DateTime.now.httpdate.to_s
=> "Mon, 12... -
02:33 PM Bug #10734 (Closed): Segfault when combining default arguments with keyword arguments
-
02:15 PM Bug #10734: Segfault when combining default arguments with keyword arguments
- if i use ruby2.2.0 it does happen for me too ...
and it seems it some kind of stack loop error because it seems to repeat itself
i tested it against ruby-head and there it seems the problem is fixed but not backported yet
-
01:19 PM Bug #10734 (Closed): Segfault when combining default arguments with keyword arguments
- Combining default arguments with a keyword argument produces a segfault when the function is called with hash that contains a string key.
See the following transcript:
~~~
natano@ketzer:~$ irb
2.2.0 :001 > def foo(data=nil, foo: :b... -
- > Let me then tell you about my use case.
Thank you, looks good to me.
Also, I just noticed that Array#bsearch is weaker than C's BSEARCH(3). BSEARCH(3) returns the pointer to the array, so it effectively returns not only the valu... -
01:56 PM Feature #10730: Implement Array#bsearch_index
- >
> ...
Let me then tell you about my use case. There is a sparse array of dates and I want to slice out a part of it that falls within minimum and maximum date. It is then later used to retrieve same values associated with those dates... -
- > Why do we have `Array#bsearch`?
Just because matz wanted it: https://redmine.ruby-lang.org/issues/4766#note-2
Personally, I don't think `Array#bsearch` is necessarily required. `Range#bsearch` is more general and powerful. Howe... -
- Yusuke Endoh wrote:
> You may want to use `Range#bsearch` for the case.
> ...
Yusuke, thanks, that is a very clever approach.
But if we do that then we are just one small step away from having `Array#bsearch` based on `Range#bsearc... -
- You may want to use `Range#bsearch` for the case.
~~~ruby
i = (0...ary.size).bsearch {|i| predicate(ary[i]) }
~~~
--
Yusuke Endoh <mame@ruby-lang.org> - 02:22 PM Revision 3e818c61 (git): * gems/bundled_gems: update test-unit to 3.0.9.
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49226 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 02:09 PM Revision 14804d6b (git): * test/ruby/test_numeric.rb (TestNumeric#test_coerce): fixed wrong message.
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49225 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
-
-
12:44 PM Bug #10708: In a function call, double splat of an empty hash still calls the function with an argument
- @Kolja: I wasn't aware of your post on stackoverflow when I posted this bug report, but this is indeed a nice coincidence!
For context I sometime want to apply some methods from a module without including the module, so I have a functio... -
11:02 AM Bug #10711 (Closed): Incorrect error message in coerce failed - Applied in changeset r49224.
----------
numeric.c: correct error message when coerce fails
* numeric.c (bit_coerce): use original value for error message
[ruby-core:67405] [Bug #10711]
* test/ruby/test_numeric.rb (test_coerce): check... -
- Applied in changeset r49223.
----------
rdoc/text.rb: fix infinite loop
* lib/rdoc/text.rb (expand_tabs): get rid of infinite loop with
CR. should check if substitution occurred too.
[ruby-dev:48813] [Bug #10732] -
07:10 AM Bug #10732 (Closed): rdoc:ファイル中に単独の\rがあるとrdocがハングアップする
- 以下で再現できます。
~~~
echo -e '\ra\t' > dummy.txt
rdoc dummy.txt
~~~
原因は、rdoc/text.rbのexpand_tabsで、each_line が上記のダミーデータを1行扱いするのに、gsub!の正規表現にマッチせず、until line !~ /\t/ が無限ループになるためです。
(libyajl2 のインストールでri/rdocの処理が終わらないのはこれが原因) -
- You can use `assert_raise_with_message`.
~~~diff
diff --git a/test/net/http/test_http.rb b/test/net/http/test_http.rb
index 4d82fd7..4f5093d 100644
--- a/test/net/http/test_http.rb
+++ b/test/net/http/test_http.rb
@@ -1,5 +1,6 @@... -
09:28 AM Misc #10721: Failing test because of DNS server
- e@zzak.io wrote:
> Were you working on a similar patch?
Not really, r49175 was because I was offline completely.
Anthony's ISP is online and doing bad things, but his patch
looks good to me.
I can commit _v2 of his patch in a d... -
- Were you working on a similar patch?
- 09:56 AM Revision 34289fff (git): numeric.c: correct error message when coerce fails
- * numeric.c (bit_coerce): use original value for error message
[ruby-core:67405] [Bug #10711]
* test/ruby/test_numeric.rb (test_coerce): check error message
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49224 b2dd03c8-39d4-4d8f-98... -
- Could you show ext/curses/mkmf.log and ext/curses/extconf.h files?
-
- seems an issue by therubyracer.
-
- * lib/rdoc/text.rb (expand_tabs): get rid of infinite loop with
CR. should check if substitution occurred too.
[ruby-dev:48813] [Bug #10732]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49223 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * method.h (UNDEFINED_REFINED_METHOD_P): macro to tell if refined
original method is defined.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49222 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Applied in changeset r49221.
----------
vm_method.c: NameError at refined method alias
* vm_method.c (rb_alias): raise a NameError when creating alias to
a refined method if the original method of the refined method is
not defined.... -
05:09 AM Bug #10731: Segmentation fault when create alias to refined method
- the patch for this.
-
- following code cause segmentation fault, both `foo` and `bar` is not defined in refined class, in trunk and 2.2.0, 2.1.5, 2.0.0
``` ruby
class C
end
module RefinementBug
refine C do
def foo
end
def bar
en... -
- * vm_method.c (rb_alias): raise a NameError when creating alias to
a refined method if the original method of the refined method is
not defined. [ruby-core:67523] [Bug #10731]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49221... -
- Two different libruby are used.
> ~~~
> ...
-
-
- Applied in changeset r49220.
----------
* math.c (math_atan2): improve documentation.
[Feature #10323][ruby-core:65400][ci skip] -
- [Feature #10323][ruby-core:65400][ci skip]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49220 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Applied in changeset r49219.
----------
* ext/bigdecimal/bigdecimal.c: fixes documentation like labeled lists,
code examples etc. [ruby-core:66730][Bug #10576][ci skip] -
- code examples etc. [ruby-core:66730][Bug #10576][ci skip]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49219 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- [misc #10608][ruby-core:66901][ci skip]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49218 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- [misc #10469][ruby-core:66063]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49217 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
-
- * array.c (rb_ary_bsearch): trivial optimizations, for Fixnum, and
by keeping the last satisfied element.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49216 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
01/11/2015
-
- * array.c (rb_ary_bsearch): preserve encoding of class name in an
exception message.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49215 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
We currently have Array#bsearch but no Array#bsearch_index and to me it seems that violates the principle of least surprise, especially when we consider the other combinations of existing methods that find either an element or it’s ind...-
- RbConfig::SIZEOF['long'] is a drop-in replacement but Fixnum::MAX is not.
Also, I feel Fixnum is a implementation detail.
- 03:58 PM Revision cb95d38a (git): * 2015-01-12
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49214 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49213 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- I negate the part "others don't requested", ... somebody requested it, ... perhaps.
-
- I think nobu has a right to accept this feature because the documented bahavior which will be changed was implemented and committed by nobu and others don't requested.
If nobu don't think so, we need approval of matz or naruse. -
- Here are the results for `reject!` using this benchmark:
```ruby
require 'derailed_benchmarks'
require 'derailed_benchmarks/tasks'
namespace :perf do
desc "Array#reject!"
task :array_reject => [:setup] do
Benchmark.ips... -
-
- Ooops, there was a bug in my test, and after fixing it and updating the results above it seems that nobu's patch improves things slightly.
I'm going to post the results for both `select!` and `reject!` soon.
@Yura, you can see tha... -
02:29 PM Feature #10729: Array method to subtract in place
- The block form can be achieved with Enumerable#partition.
I agree in general splitting an array, according to some property using the order of the elements (no take_drop_while) or to some other array (this request) is more difficult t... -
- vm_throw_start内にあるflagが立っている場合の処理は
以下のように整理できるんじゃないかと思うのですがどうでしょう。
~~~diff
diff --git a/vm.c b/vm.c
index 45734e1..80b3bbe 100644
--- a/vm.c
+++ b/vm.c
@@ -1523,7 +1523,7 @@ vm_exec(rb_thread_t *th)
}
}
}
- else i... -
- * ext/extmk.rb: do not exclude readline by default on win32. it
would not be completely impossible.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49212 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * test/webrick/test_utils.rb (test_create_listeners): use
dynamically chosen port number, not hardcoded port number.
[ruby-core:67508]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49211 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * ext/extmk.rb: move the default execluded extensions from
configure.in so that it works on win32 too.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49210 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 03:54 AM Revision d201761d (git): * 2015-01-11
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49209 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * ext/fiddle/extconf.rb: requires assembler for src/win{32,64}.S.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49208 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * ext/fiddle/win32/libffi-3.2.1-mswin.patch (include/ffi.h.in):
old version VCs did not support long long, use __int64 instead.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49207 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- I think we could do better to warn users when doing things out of order.
In this case we have a stateful thing, which should know whether key is set. In the case it's not yet, we should raise. This could be considered behavior change,...
01/10/2015
-
10:36 PM Feature #10729 (Open): Array method to subtract in place
- I request a method on array that takes another array, subtract that from self in place (= destructively), and return the subtracted elements:
a = [1, 2, 3, 4, 5]
a.some_method([2, 4, 6]) #=> [2, 4]
a #=> [1, 3, 5]
Ide... -
06:56 PM Feature #10714: Array#reject! nonlinear performance problem
- Zachary, patch fixes `reject!` and you test `select!`, ie patch fixes case when most items were deleted, and you test case when no item is deleted.
Based on numbers you present, patched version is so close to unpatched so I could not ... -
- Using the following benchmark I compared nobu's patch vs. 2.2:
```ruby
require 'derailed_benchmarks'
require 'derailed_benchmarks/tasks'
namespace :perf do
desc "Array#select! and friends"
task :array_select => [:setup] do
... -
- https://github.com/nobu/ruby/compare/Feature%2310714-Array-linear-performance
-
- Now, I think this issue is a feature instead of a bug because it changes a documented behavior.
-
- Akira Tanaka wrote:
> So the problem is which is important between the linear performance and the observability.
> ...
I fully agree. Nonlinear performance is a killer. Observability and ability to break are just nice-to-have. I thin... -
- It seems that several people found this problem.
- http://qiita.com/Nabetani/items/623df6f738864b5ed005
- https://twitter.com/grafi_tt/status/263097832158924800
- http://d.hatena.ne.jp/plonk123/20140603/1401832299
-
- > Ok, then try if something dies?
I think we should try.
We can explain the reason of this change.
It may be a good idea to describe this issue in NEWS because a documented behavior changed, though. -
- Forgot a patch of the tests
~~~diff
diff --git a/test/ruby/test_array.rb b/test/ruby/test_array.rb
index 31f33dd..33fc5d6 100644
--- a/test/ruby/test_array.rb
+++ b/test/ruby/test_array.rb
@@ -661,7 +661,7 @@ class TestArray < Te... -
- Ok, then try if something dies?
~~~diff
diff --git a/array.c b/array.c
index 0de7231..f2f7352 100644
--- a/array.c
+++ b/array.c
@@ -2824,6 +2824,48 @@ rb_ary_select(VALUE ary)
return result;
}
+struct select_bang_arg... -
- Hm.
The failed test is introduced by you after [Bug #5752].
The bug report describes as "If this is indeed the intended behaviour, ...".
It means that the reporter is not sure that the current behavior is intended or not.
Why cur... -
- I know, and `make test-all` failed.
-
- I don't think the modification must be observable from the given block.
[Bug #2545] doesn't discuss the observability in the block.
It discusses the receiver after "break".
If the modification is not required to be observable from... -
- The target of `Array#reject!` is the receiver itself, so the modification can be observed from the given block, or the block can exit by `break`.
Therefore the compaction is necessary each times.
I can't think of the way to avoid it,... -
- Apart from the performance problem, I feel following exmaple should show [7,8].
```
a = [5,6,7,8,9,10]; a.keep_if { |x| break if x > 8; x >= 7 }; p a
```
Because the method name is "keep_if", the method should keep only elements... -
- r49196 causes nonlinear performance problem.
```
% ./ruby -v -e '
20.times {|i|
a = [nil]*i*10000;
t1 = Time.now
a.keep_if { false }
t2 = Time.now
t = t2 - t1
p ["*" * (t * 20).to_i , t]
}
'
ruby 2.3.0dev (2015-... -
- Applied in changeset r49196.
----------
array.c: keep consistency
* array.c (rb_ary_select_bang): keep the array consistent by
removing unselected values soon. [ruby-dev:48805] [Bug #10722] -
01:03 PM Feature #10728: Warning for Fixnum#size to use RbConfig::SIZEOF['long']
- I took a look at `fixnum-size-search.txt` and using `RbConfig::SIZEOF['long']` instead usually isn't the correct replacement. In many lines the size of `long` in bytes isn't of interest, the min- and maximum value of a fixnum is (see fea...
-
- How about add a warning for Fixnum#size ?
```
% ./ruby -e 'p 0.size'
-e:1: warning: Use RbConfig::SIZEOF['long'] instead of Fixnum#size
8
```
Currently it returns sizeof(long) which is 4 or 8.
However it is implementation de... -
- * test/lib/test/unit.rb (ExcludesOption): allow directory list by
PATH_SEPARATOR to --excludes-dir option.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49206 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49205 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- 添付したパッチと同じ様にEXEC_EVENT_HOOKを追加して性能測定をしてみました。
測定にはbenchmark-ipsというgemを使って、100msでの実行命令数を計測しています。
測定に使ったソースコードは以下の通りです。
~~~ruby
require 'benchmark/ips'
class Foo
attr_accessor :hoge
def bar
@bar
end
def initial... -
- * test/etc/test_etc.rb (test_getgrgid, test_getgrnam): relax
comparisons. getgrent() does not return mem properly in some
circumstances (possibly, involved in Open Directory on OSX).
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trun... -
04:33 AM Bug #10727: Segfault with newlines + multibyte characters in exception message
- Thanks!
-
- Applied in changeset r49201.
----------
* eval_error.c (error_print): pos and len parameters of rb_str_substr()
are counted by characters, not bytes. use rb_str_subseq() instead.
[Bug #10727] [ruby-core:67473] -
- I can reproduce this.
```shell
% ruby -v issue10727.rb
issue10727.rb:7:in `<main>': にほんご (E)
issue10727.rb: [BUG] Segmentation fault at 0x00000000000008
ruby 2.3.0dev (2015-01-09 trunk 49195) [x86_64-darwin13]
``` -
- In Ruby 2.2.0, this simple Ruby script results in segfault.
```ruby
class E < StandardError
def initialize
super("にほんご\n改行")
end
end
raise E
```
It's caused by r48637. -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49203 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Applied in changeset r49202.
----------
* ext/zlib/zlib.c: fix document of method signatures.
[Bug #10668][ruby-core:67186][ci skip] -
- [Bug #10668][ruby-core:67186][ci skip]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49202 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 03:35 AM Revision 27c679b8 (git): * eval_error.c (error_print): pos and len parameters of rb_str_substr()
- are counted by characters, not bytes. use rb_str_subseq() instead.
[Bug #10727] [ruby-core:67473]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49201 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49200 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49199 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Applied in changeset r49198.
----------
* rational.c: removed commented-out code.
[Feature #10376][ruby-core:65643] -
- [Feature #10376][ruby-core:65643]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49198 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
-
- DL has been removed from Ruby 2.2.0 or later.
but I think we need to backport this issue into Ruby 2.1 and 2.0.0
https://github.com/freebsd/freebsd-ports/blob/master/lang/ruby19/files/patch-ext__dl__extconf.rb -
01:22 AM Feature #10726 (Open): [PATCH 4/4] * New methods: Set#power
- Hi, there.
Here are 4 patches.
```
implement_set#power.patch ... Implement Set#power
update_test_set.rb.patch ... Add tests for Set#power
update_NEWS.patch ... Update NEWS for Set#power
fix_indent_NEWS.patch ... Ad... -
- We don't have the same environment equals to yours, we can't address where it happens without the debugging information.
And, as you use some extension libraries, especially ffi, it might be impossible to fix by us. -
While trying to pin down a memory leak issue I enabled trace_object_allocations_start but after a while my application (a daemon, not rails) crashes. I attached the output.
- 01:12 AM Revision 9f7179c5 (git): * 2015-01-10
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49197 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * array.c (rb_ary_select_bang): keep the array consistent by
removing unselected values soon. [ruby-dev:48805] [Bug #10722]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49196 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
12:33 AM Misc #10721: Failing test because of DNS server
- Improved the original solution by adding assert_raises.
01/09/2015
-
- * test/rubygems: ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49195 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
10:43 AM Bug #10724 (Closed): [TracePointAPI] Missing return event from initialize method when using `domain_name` gem
- The domain_name gem defines a DomainName object which has a pretty complex `initialize` method. The TracePoint API misses the return event from this method when instantiating a DomainName object.
I attach a failing test case.
Thanks!! -
09:41 AM Bug #10710: can't compile ruby 2.2.0 with gcc versions older than 4.4.0
- here's the build output:
~~~
~/src/ruby-2.2.0 $ make
CC = gcc
LD = ld
LDSHARED = gcc -shared
CFLAGS = -O3 -fno-fast-math -ggdb3 -Wall -Wextra -Wno-unused-parameter -Wno-parentheses -Wno-long-long -Wno-missing-field-initialize... -
- the problem is that the compiler accepts the warning flag but fails to compile a piece of code as soon as it tries to emit some warning.
~~~
configure:7495: checking whether -Wno-packed-bitfield-compat is accepted as CFLAGS
configur... -
- Could you show your config.log file, around checking `-Wno-packed-bitfield-compat`?
-
09:40 AM Bug #10723: [PERF] bm_tread_create_join 20% slower
- Opps it should bm_vm_thread_create_join.rb
-
- Relevant commits: https://github.com/ruby/ruby/compare/39fd4a8...5697b2f
Chart showing the regression: http://rubybench.org/ruby/ruby/commits?result_type=vm_thread_create_join
For those unable to view the chart, the benchmark resul... -
- ref. [Bug #2545]
```
$ ruby -e 'a = [5,6,7,8,9,10]; a.keep_if { |x| break if x > 8; x >= 7 }; p a'
[7, 8, 7, 8, 9, 10]
$ ruby -e 'a = [5,6,7,8,9,10]; a.delete_if { |x| break if x > 8; x < 7 }; p a'
[7, 8, 9, 10]
```
I was expe... -
- Although `*args` includes and passes keywords too, but seems you want to add/remove/change some of keyword arguments.
It sounds reasonable to me. -
12:57 AM Bug #10708: In a function call, double splat of an empty hash still calls the function with an argument
- If I am not mistaken, even latest Ruby 2.2 selects keyword arguments as the last method's argument and of Hash type.
Let's imagine an example where both simple and keyword optional arguments are used:
~~~ruby
def call_multiargs(me... -
07:51 AM Bug #10716: Erroneous semicolon after AC_CASE
- Nobuyoshi Nakada wrote:
> Vit Ondruch wrote:
> ...
Ok, that makes sense. So I'll happily change the AC_PREREQ back to 2.63 for RHEL 6/CentOS 6 packages. Thanks for applying the patch. -
- Applied in changeset r49192.
----------
configure.in: Remove superfluous semicolon
* configure.in (RUBY_SETJMP_TYPE): Remove superfluous semicolon
which causes a syntax error with autoconf 2.63.
[ruby-core:67429] [Bug #10716] -
- Vit Ondruch wrote:
> BTW I am wondering why Ruby actually enforces autoconf 2.67 and makes plenty of RHEL 6/CentOS 6 users [1] unnecessary issues, when it builds with 2.63 just fine. If there was some justification at least [2].
Beca... -
- The following test is failing:
[ 5/52] TestNetHTTP#test_failure_message_includes_failed_domain_and_port = 0.32 s
1) Failure:
TestNetHTTP#test_failure_message_includes_failed_domain_and_port [/vagrant/test/net/http/test_http.rb:196... -
04:03 AM Revision 30253e6a (git): fix a typo [ci skip]
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49194 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- I don't expect immediate stdlib update.
-
- I think the API is improved but I'm not sure we need to find & replace
every occurrence in the stdlib
-
- Thank you for proposing this. I think this will simplify working with
Ruby IO and make me happier since I work with a lot of IO-related code.
Outside of test cases, I don't forsee compatibility problems either. -
- I'd like to change IO#close.
It should not raise IOError on closed IO objects.
We sometimes invoke IO#close only when the IO object is not closed as:
```
f.close if !f.closed?
```
If this issue is accepted, we can write it si... -
02:23 AM Feature #10720 (Rejected): A proposal for something like: attr_reader :foo? - with the trailing '?' question mark
- Hi guys,
Sorry to burden you with another suggestion, no problem
if it is not accepted, I wanted to just make it. :-)
We have a way to define reader methods easily, such
as in:
attr_reader :foo
This should be equivalen... -
- Applied in changeset r49193.
----------
parse.y: eliminate empty hashes
* parse.y (assocs, assoc): eliminate splatting empty literal
hashes. [ruby-core:67446] [Bug #10719]
* compile.c (compile_array_): supprt splatted hash in hash t... -
- ~~~
$ ruby -e 'foo(a: 1, **{})'
ruby: [BUG] Segmentation fault at 0x00000000000018
ruby 2.3.0dev (2015-01-09 trunk 49192) [universal.x86_64-darwin14]
~~~
-
- * parse.y (assocs, assoc): eliminate splatting empty literal
hashes. [ruby-core:67446] [Bug #10719]
* compile.c (compile_array_): supprt splatted hash in hash type.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49193 b2dd03c8-39d... -
- * configure.in (RUBY_SETJMP_TYPE): Remove superfluous semicolon
which causes a syntax error with autoconf 2.63.
[ruby-core:67429] [Bug #10716]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49192 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49191 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
01/08/2015
-
10:25 PM Revision 344b8bbd (git): * ext/psych/lib/psych/visitors/yaml_tree.rb: correctly quote non-ascii
- letters. Thanks @jirutka for the patch.
* test/psych/test_string.rb: test for change
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49190 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- subclasses. Fixes github.com/tenderlove/psych/issues/196
* test/psych/test_hash.rb: test for change
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49189 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * ext/psych/lib/psych/visitors/yaml_tree.rb: dump hashes with ivars.
Fixes github.com/psych/issues/43
* test/psych/test_hash.rb: test for change
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49188 b2dd03c8-39d4-4d8f-98ff-823fe69b... -
09:11 PM Bug #10708: In a function call, double splat of an empty hash still calls the function with an argument
- Just to inform everyone, this issue stems from [this post on Stack Overflow](http://stackoverflow.com/questions/27821422/how-can-i-collapse-double-splat-arguments-into-nothing).
Also, I have now idea how the ruby parser works, so if i... -
- By my subjective opinion I don't find this a bug but a feature.
> Consider this:
> ...
Here you define a method without a (keyword) argument placeholder so it does not expect a Hash argument, which is effectively used to pass keyword... -
- no. it aborts the build.
-
- ありがとうございます！
頂いたパッチを元に、cfp を使うように整理してみました。
http://www.atdot.net/sp/view/e5ivhn
* vm_throw() を簡素化
* マクロっぽい名前を関数っぽく
一応、test-all/test-rubyspec は通っていますが、良さそうでしたらコミットしてもらえないでしょうか。
-
03:58 PM Bug #10697: WIN32OLE: WIN32OLE_RECORD を使用したスクリプト終了時にruby.exe がクラッシュすることがある
- ありがとうございます。
現象自体は確認しており、メモリーの二重開放まではわかっていたのですが、
COMサーバ側で開放しているとは思いませんでした。
ちょっと時間が取れてなくてパッチの方は詳しく見ていないのですが、別のアプローチの
VT_BYREF|VT_RECORDで渡したら、こちらでも現象は起こらなくなったみたいです。
わざわざ、VT_VARIANT | VT_BYREF で渡しているのは参照渡しにしたいからなので、
参照渡しをするのなら、VT_RECOR... - 03:38 PM Revision 2b2f9b75 (git): * 2015-01-09
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49187 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49186 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- ~~~
$ rpm -q cmake
cmake-2.6.4-5.el6.x86_64
$ make test-all TESTS=-n\ TestGemExtCmakeBuilder#test_self_build
<snip>
./miniruby -I./lib -I. -I.ext/common ./tool/runruby.rb --extout=.ext -- --disable-gems "./test/runner.rb" --... -
02:02 PM Bug #10699: m(*a, **b) doesn't recognize integer options.
- Thank you akr san for your consideration about backports.
I'll fill Backport field with 2.1: REQUIRED only for r49173. -
- Forgot to mention the actual error:
~~~
$ autoconf
$ ./configure
./configure: line 63225: syntax error near unexpected token `;'
./configure: line 63225: `; }; then'
~~~ -
- This is issue when building with autoconf 2.63 available in RHEL6/CentOS6
BTW I am wondering why Ruby actually enforces autoconf 2.67 and makes plenty of RHEL 6/CentOS 6 users [1] unnecessary issues, when it builds with 2.63 just fine... -
- DL has been removed at Ruby 2.2.0
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49185 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
09:37 AM Bug #10715 (Closed): WEBrick::HTTPResponse#to_s calls send_response(), and make debugging quite difficult.
- WEBrick::HTTPResponse#to_s calls send_response(), and this makes debugging very difficult.
If I run webrick on debugger or just do "print" debug, and show status of the object in some point,
then debugger or "print" may calls "to_s" an... -
- According to `git bisect`, it caused by r32373 (related to [Bug #2545]) .
-
- I found Array#reject! is too slow.
I measured it and it seems the performance is nonlinear.
```
% ./ruby -v -e '
20.times {|i|
a = [nil]*i*10000;
t1 = Time.now
a.reject! { true }
t2 = Time.now
t = t2 - t1
p ["*"... -
- Applied in changeset r49184.
----------
vm_method.c: no super klass, no original method entry
* vm_method.c (rb_method_entry): if no super class, no original
method entry. [ruby-core:67389] [Bug #10707] -
- * vm_method.c (rb_method_entry): if no super class, no original
method entry. [ruby-core:67389] [Bug #10707]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49184 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Applied in changeset r49183.
----------
ref of r49182 [Bug #10706] -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49183 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49182 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- i also would like if there are block variants of them like that
data = 0..10
data.prev_value {|o| o == 5} #=> 4
data.next_value {|o| o == 5} #=> 6
and i would use prev_value and next value instead of prev and next because it coll... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49181 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 06:31 AM Revision a8565ad2 (git): * test/ruby/test_symbol.rb (TestSymbol#test_symbol_fstr_leak): get rid of a
- false positive on mswin CI.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49180 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 06:26 AM Revision 93425341 (git): * test/test_open3.rb (TestOpen3#test_numeric_file_descriptors): passing FDs
- bigger than 2 is not supported on Windows.
fixed test failure introcuded at r49173.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49179 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
04:42 AM Bug #10713: Assigning default value for a Hash as an empty Array creating unpredictable results
- Arvinder Singh wrote:
> ~~~
> ...
The `letters[:a]` part of the second line returns the "default value" of the Hash because the `:a` key does not exist in the Hash. The `<< 1` part pushes a `1` onto the end of the default value, but i... -
01:38 AM Bug #10713: Assigning default value for a Hash as an empty Array creating unpredictable results
- It's quite expected because the default array is created exactly once:
letters = Hash.new([])
letters.default.object_id # => 70310393550400
letters[:a].object_id # => 70310393550400
letters[:b].object_id # => 70310393550400
The... -
01:33 AM Bug #10713: Assigning default value for a Hash as an empty Array creating unpredictable results
- Martin Dürst wrote:
> Hiroshi SHIBATA wrote:
> ...
I would agree that this is surprising behaviour. It would appear that in this case, the append operator is not re-assigning the value, the way it does any other time it is used. And it... -
- Hiroshi SHIBATA wrote:
> It's expected behavior
Hiroshi, can you tell us why it's expected behavior? It looks quite surprising. -
- It's expected behavior
-
- Applied in changeset r49178.
----------
dir.c: OSX case-folding
* dir.c (glob_helper): match in case-folding only if the directory
resides on a case-insensitve file system, on OSX.
[ruby-core:67364] [Bug #10700] -
03:10 AM Bug #10700: On case-sensitive filesystem on OS X, Dir.glob("*.TXT") matches case-insensitively
- Nobuyoshi Nakada wrote:
> Can you try https://github.com/nobu/ruby/compare/Bug%2310700-OSX-case-folding ?
Confirmed that it works.
~~~
$ touch foo.txt foo.TXT FOO.txt FOO.TXT
$ ls -l foo* FOO* ... -
- * dir.c (glob_helper): match in case-folding only if the directory
resides on a case-insensitve file system, on OSX.
[ruby-core:67364] [Bug #10700]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49178 b2dd03c8-39d4-4d8f-98ff-823f... -
- * lib/open3.rb (popen_run): use Hash.try_convert for duck typing.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49177 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- [fix GH-809] Patch by @gxworld
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49176 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
02:31 AM Feature #10119 (Feedback): [PATCH] [net/imap] Add CHANGEDSINCE FETCH modifier + complementary add MODSEQ Message Data Item in FETCH Command
- It seems to be a bug that store_internal is changed not to use flags.
What was your intention?
Could you write tests for this change?