Activity
From 11/18/2015 to 11/24/2015
11/24/2015
-
11:40 PM Bug #11199: IO.copy_stream degrade that can't call pathname to pathname
- Hi!
We run into this here: https://github.com/janko-m/shrine/commit/399adcecad64c66ca4ffb1b1619777ff8551e761. That test passes in 2.2.3 but fails in 2.3.0.preview1, apparently due to this issue. Is fixing this planned for 2.3.0?
Th... -
09:03 PM Feature #11737: Pass in expression to then block in `case expression`
- Please fix the when scenario in the last example
~~~ruby
case Enumerator.new do |y| y << 1; y << 2; y << 3; end
when ->e{ 2.times{e.next}; true}
then _.peek
end == 3
~~~ -
- Ruby's `case <expression>` scenario allows the expression to be tested in the `when` test with `:===`, but in the then block the expression is not accessible. Only if the expression has been assigned to a variable beforehand can it be c...
-
08:48 PM Feature #11735: Porting String#squish and String#squish! from Ruby on Rails' Active Support
- > ~~~diff
> +static VALUE
> +rb_str_squish_bang(VALUE str)
> +{
> + static const char before_regex_source[] = "\\A[[:space:]]+";
> + static const char after_regex_source[] = "[[:space:]]+\\z";
> + static const char ... -
07:38 PM Feature #11735: Porting String#squish and String#squish! from Ruby on Rails' Active Support
- I also just created a GitHub PR in case that will be easier to give feedback to the code: https://github.com/ruby/ruby/pull/1113
Thank you,
Prem -
- Hi,
I have been using this `String#squish` method so many time when I'm using Rails, and I think it should be a useful addition to core.
Here's the method on Rails' documentation: http://api.rubyonrails.org/v4.2.5/classes/String.ht... -
- `&` is associated more with proc. I think `_` would be closer to the kind of thing you're looking for.
~~~ruby
some_long_expression = :baz
_ ? _.to_s : 'foobar'
# => "baz"
~~~
Although I tried in-lining it with a semicolon and... -
03:10 PM Feature #11734: Improved ternary operator
- why not:
```ruby
(x = some_long_expression) ? x.to_s : 'foobar'
``` -
02:33 PM Feature #11734 (Closed): Improved ternary operator - In ternary operator it would be nice to be able to pass expression result from *condition* into *value_if_true/value_if_false* in such way:
`some_long_expression ? &.to_s : 'foobar'`
where `&` refers to `some_long_expression`
Ins... -
08:10 PM Bug #11736 (Closed): Float(arg) not consistent with documentation
- Float(arg) → float
Returns arg converted to a float. Numeric types are converted directly, the rest are converted using arg.to_f. Converting nil generates a TypeError.
Strings are NOT converted using arg.to_f
~~~
'123_INVALID'.to... -
04:10 PM Revision 1e662300 (git): bump patchlevel for previous commit.
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@52744 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * ext/digest/sha1/sha1ossl.c: fix defs.h path to catch up changes in
r52739.[Bug #3231]
* ext/digest/rmd160/rmd160ossl.c: ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@52743 b2dd03c8-39d4-4d8f-9... -
- Backported into `ruby_2_2` branch at r52742.
-
- * parse.y (kwd_append): fix segv after invalid keyword argument,
preceding keyword list is NULL when syntax error is there.
[ruby-core:71356] [Bug #11663]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_... -
- r52244 and (partially) r50727 were backported into `ruby_2_2` branch at r52741.
-
- * lib/ipaddr.rb, test/test_ipaddr.rb: Reject invalid address contained
EOL string. Patch by @kachick [fix GH-942][Bug #11513]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@52741 b2dd03c8-39d4-4d8f-98ff-823fe69... -
- Backported into `ruby_2_2` branch at r52740.
-
- * compile.c (iseq_build_from_ary_body): register cdhash to the
iseq constant body instead of compile time mark array, not to
get GCed. [ruby-core:70708] [Feature #8543]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/b... -
- Backported into `ruby_2_2` branch at r52739.
Add similar patch to ext/digest/sha2_sha2ossl.c. -
- * ext/digest/*/*.[ch]: include ruby.h before digest.h to avoid
includeing ext/digest/extconf.h. [Bug #3231]
https://msdn.microsoft.com/library/36k2cdd4.aspx
* ext/digest/*/extconf.rb: remove ext/digest from i... -
- Backported into `ruby_2_2` branch at r52738.
-
- Backported into `ruby_2_2` branch at r52738.
-
- Applied in changeset backport22:r52738.
----------
merge revision(s) 52683,52684: [Backport #11719] [Backport #11722]
* lib/net/http.rb: Fixed regression for Net::HTTP::PUT with "Expect-100"
header.
* test/net/http/test_http.rb: ... -
- * lib/net/http.rb: Fixed regression for Net::HTTP::PUT with "Expect-100"
header.
* test/net/http/test_http.rb: added test.
* Added missing reference of GitHub
header. [fix GH-949]
git-svn-id: svn+s... - 03:03 PM Revision ff838167 (git): * 2015-11-25
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52737 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
03:00 PM Revision 06c04b38 (git): * configure.in: On Solaris, it is safe to define _LARGEFILE_SOURCE
- when _FILE_OFFSET_BITS=64 is defined (= when 32-bit compile).
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52736 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
01:12 PM Revision 17f2a12f (git): fix a typo [ci skip]
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52735 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
09:23 AM Bug #11718: Constant access on `nil`
- This is a (intentional) bug. Because I expect nobody reveal such code :p
-
07:12 AM Revision b0fb3ff1 (git): parse.y: ripper for warnings
- * parse.y (parser_yylex): deal with magic comment warnings also in
ripper.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52734 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * parse.y (dispatch_scan_event): fix wrong macro at r52547.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52733 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
06:20 AM Feature #11181: Add a line directive to Ruby
- I have changed the assignee from ruby-core (all committers) to Matz to make sure this doesn't show up in everybody's todo list.
-
01:00 AM Revision ba516bc3 (git): * lib/rubygems/installer.rb: Fix two double-word typos.
- [ci skip][fix GH-1108] Patch by @jwworth
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52732 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
12:58 AM Bug #11733: Compile of ruby 2.2.3 fails on AIX 6.1 TL07 SP03
- Another set of errors (why I gave up on modifying fs.c were these:
compiling /usr/work/src/ruby-2.2.3/ext/-test-/file/fs.c
/usr/work/src/ruby-2.2.3/ext/-test-/file/fs.c: In function 'get_fsname':
/usr/work/src/ruby-2.2.3/ext/-test-/... -
- The build on AIX fails in the ext/-test-/file directory with errors such as
compiling /usr/work/src/ruby-2.2.3/ext/-test-/file/fs.c
/usr/work/src/ruby-2.2.3/ext/-test-/file/fs.c: In function 'get_fsname':
/usr/work/src/ruby-2.2.3/ex... -
- * beos: Drop support for BeOS now that Haiku is stable.
[Fix GH-1112]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52731 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * internal.h (rb_gc_for_fd): move to export, as referred by
ext/socket.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52730 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * file.c: [DOC] add a missing period to File docs, to terminate
the sentence and separate from the next sentence. [Fix GH-1111]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52729 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
11/23/2015
-
- * NEWS: Fix the issue number of `Struct#dig`, which should be
[Feature #11688]. [Fix GH-1110]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52728 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 10:57 PM Revision d51743af (git): ext/socket/init.c (rsock_accept): handle ENOMEM
- accept(2) documents ENOMEM as a possible error, handle it
consistent with all of our other FD-allocating wrappers.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52727 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 10:50 PM Revision 61e5fe06 (git): use rb_gc_for_fd for more callers
- * dir.c (dir_initialize): use rb_gc_for_fd for ENOMEM
* ext/socket/init.c (rsock_socket): ditto
* ext/socket/socket.c (rsock_socketpair): ditto
* internal.h (rb_gc_for_fd): prototype
* io.c (rb_gc_for_fd): remove static
[ruby-core:7162... - 09:57 PM Feature #11727 (Closed): [PATCH] io.c (rb_gc_for_fd): wrapper for retrying FD creation
- Applied in changeset r52725.
----------
io.c (rb_gc_for_fd): wrapper for retrying FD creation
This simplifies callers and makes error handling more consistent
between our pipe, open, fdopen, and dup wrappers.
This adds missing ENOMEM ... - 09:57 PM Revision a936bd5c (git): io.c (rb_gc_for_fd): wrapper for retrying FD creation
- This simplifies callers and makes error handling more consistent
between our pipe, open, fdopen, and dup wrappers.
This adds missing ENOMEM handling as documented in the open(2),
pipe(2freebsd), and fdopen(3posix) manpages on my system.... -
09:37 PM Feature #11599: Dump entries of hash in ObjectSpace
- Aaron Patterson wrote:
> We should probably get Aman's opinion on this since he wrote the initial heap dumping code. Personally, I would like to see this feature enabled with a flag. It seems like it will significantly increase the si... -
09:34 PM Feature #11599: Dump entries of hash in ObjectSpace
- We should probably get Aman's opinion on this since he wrote the initial heap dumping code. Personally, I would like to see this feature enabled with a flag. It seems like it will significantly increase the size of the dump file, and a...
- 09:24 PM Revision 1946607b (git): * 2015-11-24
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52724 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 09:21 PM Feature #11607 (Closed): [PATCH] fiddle: release GVL for ffi_call
- Applied in changeset r52723.
----------
fiddle: release GVL for ffi_call
Some external functions I wish to call may take a long time
and unnecessarily block other threads. This may lead to performance
regressions for fast functions as... -
- On Fri, Nov 13, 2015 at 05:08:36AM +0000, Eric Wong wrote:
> So, should I commit the patch as-is, or perhaps add a new
> option/method for releasing the GVL?
>
> Naming new options/methods hard for me :<
Yes, please commit it! :... - 09:20 PM Revision 15476c69 (git): fiddle: release GVL for ffi_call
- Some external functions I wish to call may take a long time
and unnecessarily block other threads. This may lead to performance
regressions for fast functions as releasing/acquiring the GVL is not
cheap, but can improve performance for ... -
04:44 PM Feature #11731: Make ruby gem install to user-install by default
- https://github.com/rubygems/rubygems/issues/1394 opened to track. Sorry for putting it in the wrong place.
Haiku is a little special. /boot represents the os "boot" drive (and not where only the kernel lives)
https://www.haiku-os.o... -
- Rubygems upstream is https://github.com/rubygems/rubygems.
Maybe to fallback to user-install when no permission to the default place.
BTW, is ruby installed under `/boot` filesystem, on Haiku? -
02:40 PM Bug #11718: Constant access on `nil`
- I don't think it's intended, that commit should not introduce new behavior except on singleton class constants.
@ko1: Could you confirm this is a bug? -
02:25 PM Bug #11718: Constant access on `nil`
- looks like it's somehow intended behavior
https://github.com/ruby/ruby/blob/trunk/insns.def#L179-L190
/**
@c variable
@e
Get constant variable id. If klass is Qnil, constants
are searched... -
01:34 PM Feature #11630: possibility to serialize Proc or Lambda
- In my mind, there is no perfect behavior when it comes to serializing closures with bind variables. This being said, it should not prevent us from serializing pure functions.
Since the behavior of a serialized proc with closure can ea... -
11:17 AM Bug #11729: String#split(nil) incorrectly uses $;
- I think this is intended.
[Comment of MRI](https://github.com/ruby/ruby/blob/6965964df65c1c3679b92b92101225e4a709ba73/string.c#L6583)
and
[rubyspec](https://github.com/ruby/rubyspec/blob/d439436f9c73e491e6b137eb41bcd1e9b9de9b4c/... -
- Coverageのrestart後に新しいファイルをrequireしたときの挙動がテストに記述されていなかったので、`test_restarting_coverage`に追記しました。
-
- in $ansi_options is often also needed in CPPFLAGS,
because some feature definitions vary depending on such
standards options.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52722 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Applied in changeset r52721.
----------
Haiku now best effort support
* configure.in: remove obsolete workarounds for Haiku.
* dln.c, file.c, io.c: remove obsolete Haiku workarounds.
* thread_pthread.c: add stack bounds detection for H... -
- * configure.in: remove obsolete workarounds for Haiku.
* dln.c, file.c, io.c: remove obsolete Haiku workarounds.
* thread_pthread.c: add stack bounds detection for Haiku.
* signal.c: get stack pointer from signal context on Haiku.
[rub... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52720 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
11/22/2015
- 11:55 PM Revision 165cc0d9 (git): * 2015-11-23
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52719 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52718 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Shouldn't site-wide installations use packaged gem files as they install to /usr?
Installing non-packaged files to /usr violates the LSB.
A few linux distros seem to also set --user-install as default:
https://projects.archlinux.o... -
11:18 PM Feature #11731: Make ruby gem install to user-install by default
- On 23 November 2015 at 08:56, <kallisti5@unixzen.com> wrote:
> Issue #11731 has been updated by Alexander von Gluck.
>
>
> yeah, the large number of workarounds show this is a pretty default
> workflow for people. Thus raises the... -
- yeah, the large number of workarounds show this is a pretty default workflow for people. Thus raises the question on why it isn't the default behavior.
-
09:40 PM Feature #11731: Make ruby gem install to user-install by default
- Aren't most people using rvm or rbenv/bundler to address this need?
-
- There would be a few minor logic issues around gem looking for --install-dir + --user-directory and erroring out... that check would need to be changed to --no-user-directory + --install-dir
This definitely would need some discussion... -
- Installing gem's system-wide doesn't make sense 99% of the time...
* Linux -- Need root access via sudo
* Windows -- Need administrator
* Max OS X -- Need root access via sudo
* Haiku -- /boot/system read-only
gem currently inst... -
- Even shorter `Kernel#with`
~~~ruby
my_hash = {}
my_hash.with.update, b, c = ->{ {a: 1}, 5, 6 }.call
~~~ -
- This issue is resolved via the pull-request here:
https://github.com/ruby/ruby/pull/1109 -
- Backported into `ruby_2_2` branch at r52717.
-
- * vm_insnhelper.c (vm_invoke_block): we should not expect ci->argc is
stable after invoking a block. [Bug #11451]
* test/ruby/test_yield.rb: add a test. This test script is given by
Alex Dowad.
git-svn-id: ... - 01:41 PM Revision 88376e63 (git): * gc.c (rb_raw_obj_info): fix compile errors when USE_RGENGC
- is 0.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52716 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Applied in changeset r52715.
----------
* lib/cmath.rb: methods which has suffix '!' are now deprecated.
Re-apply r52469 made by Kazuki Tanaka, with fixing bug about
mathn.rb compatibility. [ruby-core:68528] [Feature #10974] -
- Re-apply r52469 made by Kazuki Tanaka, with fixing bug about
mathn.rb compatibility. [ruby-core:68528] [Feature #10974]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52715 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- [ci skip][fix GH-1007] Patch by @arthurnn
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52714 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- AC_TRY_CPP and AC_TRY_COMPILE pass, because some options
(e.g. -std=iso9899:1999) are not set when running C preprocessor
or building ext.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52713 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- `IO#write_nonblock` call-seq has `[, options]`, but `IO#read_nonblock` does not.
In prelude.rb:
~~~
# call-seq:
# ios.read_nonblock(maxlen) -> string
# ios.read_nonblock(maxlen, outbuf) -> outbuf
~... -
- * compile.c (iseq_compile_each): vals is NULL here, show the
position of the parent node.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52712 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * compile.c (iseq_peephole_optimize): remove unreachable code
chunk after jump/leave.
* parse.y: move dead code elimination of logical operation to
compile.c. not to warn logical operation of literal constants.
git-svn-id: svn+ssh:... -
- * compile.c (iseq_peephole_optimize): eliminate always/never
branches after a literal object and when the value is used after
the branch.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52710 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * template/fake.rb.in: move shell code inside a block comment.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52709 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * compile.c (validate_labels): free labels table before raising an
exception.
* compile.c (rb_iseq_build_from_ary): wrap labels table to ensure
freeing.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52708 b2dd03c8-39d4-4d8f-98f... -
- * compile.c: use rb_compile_bug instead of rb_bug to show error
position precisely.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52707 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * compile.c: use COMPILE_ERROR to save an exception to be raised
at cleanup_iseq_build.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52706 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * compile.c (dump_disasm_list): flush disassembled list.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52705 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * compile.c (iseq_setup): check return values of each steps.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52704 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
02:14 AM Bug #11729 (Closed): String#split(nil) incorrectly uses $;
- Hi,
Is this intentional?
```
$; = "@"
"a@b".split(nil) # => ["a", "b"]
```
I expect `["a@b"]` because I found the following sentence in the Japanese documentation.
```
1 バイトの空白文字 ' ' か nil
先頭と末尾の空白を除いたうえで、空白文字列で分割する。
``...
11/21/2015
-
10:22 PM Bug #11602: Compile error on CentOS: Segmentation fault after "linking miniruby"
- Ruben Ambarcumyanc wrote:
> Some additional information. Version 2.2.3 didn't compiled with the same error message. But Ruby from the stable snapshot (https://ftp.ruby-lang.org/pub/ruby/stable-snapshot.tar.gz) compiled successfully:
> ... -
- if available, mainly for enabling some features in sockets.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52703 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Applied in changeset r52701.
----------
* test/socket/test_socket.rb (test/socket/test_socket.rb): skip the test when Socket::SO_TIMESTAMP is not defined.
Fix error on Solaris 10. [Bug #11728] [ruby-dev:49377] -
- Solaris 10 にて、CFLAGSに -D_XOPEN_SOURCE=500 を追加して、SUSv2準拠にてコンパイルした場合、make test-all にて以下のエラーが発生します。
~~~
Finished tests in 1518.357144s, 10.4659 tests/s, 1475.1002 assertions/s.
1) Error:
TestSocket#test_timestamp:
SocketError: un... - 04:19 PM Revision c3bc3b16 (git): * 2015-11-22
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52702 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Fix error on Solaris 10. [Bug #11728] [ruby-dev:49377]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52701 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52700 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52699 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- **Can I get a fair explanation of why this issue abandoned?**
-
- io.c (rb_gc_for_fd): wrapper for retrying FD creation
This simplifies callers and makes error handling more consistent
between our pipe, open, fdopen, and dup wrappers.
This adds missing ENOMEM handling as documented in the open(2... -
- * ruby.c (need_argument): move frozen-string-literal-debug option
from --enable to --debug. [Feature #11725]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52698 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * ruby.c (proc_options): fix pointer overrun. do not advance argv
until it is valid.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52697 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
06:36 AM Revision 8fdebec0 (git): fix r52690
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52696 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Assert `Coverage.peek_result` raises Error if `Coverage.start` was not called before.
-
- SHA_CTX. http://openssl.org/docs/man0.9.8/crypto/SHA1.html
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52695 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
-
- includeing ext/digest/extconf.h. [Bug #3231]
https://msdn.microsoft.com/library/36k2cdd4.aspx
* ext/digest/*/extconf.rb: remove ext/digest from include search path
to avoid confusion of cl.exe.
* ext/digest/*/*.[ch]: explicitly spe... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52693 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52692 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Applied in changeset r52691.
----------
* thread_sync.c: reduce the specification of Queue#close.
* Queue#close accepts no arguments.
* deq'ing on closed queue returns nil, always.
[Feature #10600]
* test/thread/test_queue.rb: ca... -
- I decide to reduce specification of `Queue#close`. For closed queues, `deq` returns `nil`.
No exception is raised for `deq`.
Other tokens are also not supported.
We can introduce them as new feature.
For Ruby 2.3 (or just now),... -
- * Queue#close accepts no arguments.
* deq'ing on closed queue returns nil, always.
[Feature #10600]
* test/thread/test_queue.rb: catch up this fix.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52691 b2dd03c8-39d4-4d8f-98ff-... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52690 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- nobu@ruby-lang.org wrote:
> The new function looks fine to me, but why deprecating `rb_autoload`?
In general, I prefer to make the C API smaller so it is easier to
support and improve Ruby internals. The rb_autoload API is also
t...
11/20/2015
-
- You run `static_test3` twice, not `static_test4`.
-
- Applied in changeset r52688.
----------
* compile.c (iseq_compile_each): add debug information to NODE_STR
strings as default.
[Feature #11725]
* insns.def (freezestring): add new instruction to support adding
debug information f... -
- # Background
Debug frozen string literal is difficult because frozen string can be modified at far from created locations. For example, library X creates one string and modify this string is common situation. Also library X can pass t... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52689 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- strings as default.
[Feature #11725]
* insns.def (freezestring): add new instruction to support adding
debug information for dynamically constracted strings.
* compile.c (iseq_compile_each): support adding debug information
for N... -
- The new function looks fine to me, but why deprecating `rb_autoload`?
We used to append `_str` in many cases, IIRC. -
- normalperson@yhbt.net wrote:
> https://bugs.ruby-lang.org/issues/11664
Comments? It's low-risk, but does deprecate a (probably unused)
public C-API.
Thanks. -
09:42 PM Feature #10674: Net::HTTP retries idempotent requests once after a timeout, but its not configurable
- I can confirm this exact behavior on 2.1.7
-
09:32 PM Bug #11724 (Closed): SNIでsessionが無効だったときにhostnameがサーバに送られない
- これも https://github.com/ruby/ruby/pull/964 を見る限り相当手痛いバグなので要backportと思います。
が、確かにこの修正でnet/httpのSNIできねー問題は直るんでしょうけど、そもそもこれはおかしいのはext/opensslの方だと思いますので、むしろそっちを直すべきだと強く感じます。
あとテストがない。
なので、Closedにはしておきますが、backportする前にどないかしたい(またはしてほしい)ですね。 -
- Applied in changeset r52686.
----------
symbol.c: not freeze the receiver
* symbol.c (rb_str_intern): should not freeze the receiver itself
unexpectedly. [ruby-core:71611] [Bug #11721] -
05:33 PM Bug #11721 (Closed): String used as IO object buffer unexpectedly frozen if #to_sym called on it
- A string with forced encoding used a buffer for reading from a File becomes frozen unexpectedly if #to_sym is called on it
file = File.open('some_file')
str = ""
str.force_encoding(Encoding::US_ASCII)
str.frozen? # => false
file.... - 07:34 PM Revision 89435216 (git): * 2015-11-21
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52687 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- * symbol.c (rb_str_intern): should not freeze the receiver itself
unexpectedly. [ruby-core:71611] [Bug #11721]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52686 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
07:06 PM Feature #11723 (Open): CGI library should give access to raw request body
- Hi everyone,
I was trying to write a little CGI script that simply stencils out e-mail autoconfig as defined by Microsoft’s [autodiscover protocol](https://technet.microsoft.com/en-us/library/cc511507%28v=office.14%29.aspx), however, ... -
- バックポート管理用チケットです。
r52683 の net/http の修正部分は差分のところだけみると 2.0.0 から同様のようなのでバックポートが必要かもしれません。 -
04:18 PM Feature #11717: Object#trap -- pass object to block and return result
- > "trap" already means "trap a signal", it comes from long-standing Unix terminology
Ooops. Completely forgot about this one :(
> ...
Yeah, can see it now.
Thinking further, I wonder if just `Object#yield` could be parsed correc... -
03:34 PM Bug #10984: Hash#contain? to check whether hash contains other hash
- I second Ilya's opinion regarding partially ordered sets. But propose to implement the comparision similiar to classes - `Hash` and `Class` both satisfy reflexive, antisymmetric, and transitive relations. So like the `Class`, `Hash` can ...
-
- The infinite loop observed during RubySpec test in r52469
(http://rubyci.s3.amazonaws.com/ubuntu1510/ruby-trunk/log/20151106T153002Z.fail.html.gz )
is caused by the following code in mathn.rb.
~~~
unless defined?(Math.exp!)
Obje... -
02:16 PM Revision 8594cab1 (git): * lib/logger.rb: expose logger mutex
- [fix GH-541] Patch by @arthurnn
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52685 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
12:07 PM Misc #11580 (Closed): Add documentation for #thread_list_all in lib/debug.rb
- Fixed by r52451.
Thank you for your patch! -
- yeah that should raise an error. A little test shows that this exist a while way back:
```ruby
p nil::String
```
for 1.9.3:
```ruby
String
```
for 1.8.7:
```ruby
# is not a class/module (TypeError)
```
-
- I think this should really raise an error like:
$ ruby -e 'p nil::String'
-e:1:in `<main>': nil is not a class/module (TypeError)
-
- hm it might be that ```nil::Foo``` is parsed as ```::Foo``` which is doing the top-level access.
but i am not deep enough in the parser to fix that. -
02:38 AM Bug #11718 (Closed): Constant access on `nil`
- It is possible to access top-level constants by doing `nil::CONSTANT`, which looks like a bug [according to Matz](https://github.com/opal/opal/issues/1197#issuecomment-158078778). Here are a couple of examples:
```
$ ruby -ve "Foo = ... -
07:37 AM Misc #11681 (Closed): Refinements documentation is based on older implementation
- Merged by r52629
-
06:55 AM Bug #11692: [PATCH] Re-enable GC if stack overflow was caught from signal handler
- Koichi, do you plan to do backport to 2.2 version of this patch?
-
- 余談ですが、
そういう用途だったらopenに最近O_TMPFILEってフラグが追加されているので、最初は名前無しファイルで作って、
終わったらlinkat(2)で永続化するっていうのを提供してもよいかもしれませんねぇ。
http://linuxjm.osdn.jp/html/LDP_man-pages/man2/open.2.html -
01:37 AM Feature #11715: Tempfile.create unlinked file silently fails
- 納得しました。ありがとうございました。
-
- that might not be a bug itself but because from the XMLRPC doesnt support dates earlier than 1970 ...
otherwise i do wonder why such a extra class is needed and why it doesn't use ::DateTime itself -
06:24 AM Feature #11720 (Rejected): XMLRPC::DateTime.to_time does not support times prior to 1970.
- While Time (as of at least version 1.9.2) now supports at least 1823-11-12, XMLRPC::DateTime.to_time still does not allow for dates earlier than 1970.
-
- Ok, I understood. This issue can be closed now, thank you
-
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52684 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Net::HTTP::PUT の挙動が regression している問題の修正である r52683 のバックポートをお願いします。
https://github.com/ruby/ruby/pull/949 -
- header.
* test/net/http/test_http.rb: added test.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52683 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Applied in changeset r52682.
----------
* lib/net/http.rb: set hostname before call ossl_ssl_set_session.
[Bug #11401][ruby-core:70152][fix GH-964] Patch by @mkarnebeek -
- [Bug #11401][ruby-core:70152][fix GH-964] Patch by @mkarnebeek
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52682 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * tool/enc-unicode.rb: add --header option to emit name2ctype.h
directly.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52681 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- [ci skip][fix GH-1028] Patch by @leriksen
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52680 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- It's typically not a support option.
[ci skip][fix GH-990] Patch by @eam
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52679 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- [ci skip][fix GH-1105] Patch by @davydovanton
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52678 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Applied in changeset r52677.
----------
* vm.c (rb_vm_cref_replace_with_duplicated_cref): added.
CREFs should not be shared by methods between `using'.
[Bug #11247]
* vm_insnhelper.c (vm_cref_replace_with_duplicated_cref): ditto.
... -
- CREFs should not be shared by methods between `using'.
[Bug #11247]
* vm_insnhelper.c (vm_cref_replace_with_duplicated_cref): ditto.
* vm.c (vm_cref_dup): should copy refinements correctly.
* eval.c: use rb_vm_cref_replace_with_dupl...
11/19/2015
-
11:41 PM Feature #11717: Object#trap -- pass object to block and return result
- Victor Shepelev wrote:
> Even if `#trap` will be implemented as a simple alias of `#instance_eval`...
If you did in fact know that you were essentially requesting an alias for `#instance_eval`, this was a remarkably roundabout way to... -
- Victor Shepelev wrote:
>
> ...
"trap" already means "trap a signal", it comes from long-standing Unix terminology; see [Signal#trap](http://ruby-doc.org/core-2.2.0/Signal.html#method-c-trap) -
- Some related issues and historical readings:
* #10095 `Object#as`
* from #6373 `Object#self`
* #6721 `Object#yield_seld`
* #6684 `Object#do`
* #7388 `Object#embed`
Clearly this is something the Ruby community wants. Just... -
- Nope.
I'm aware of `#instance_eval` last 10 years or so (I even can recall times when `#instance_exec` were external library method, not part of the core).
Primary goal/usage of `#instance_eval` is to "dig inside". Primary goal/us... -
- You're looking for `#instance_eval`:
```ruby
'foo'.instance_eval { |obj| obj.size } # => 3
``` -
- `Object#trap` can be thought as useful counterpart for `Object#tap`: like tap, it passes object to the block; **unlike** tap, it returns results of the block, not object itself.
**Rationale**
`#trap` could allow to gracefully chain... -
10:55 PM Bug #11716 (Rejected): Truncation error in Time xmlschema with fraction_digits.
- It's because 1447961542.26 is actually 1447961542.2599999904632568359375.
```
% ruby -e 'p "%.1000g" % 1447961542.26'
"1447961542.2599999904632568359375"
```
If 1447961542.26 is exactly given as rational, xmlschema(2) generates ... -
08:02 PM Bug #11716 (Rejected): Truncation error in Time xmlschema with fraction_digits.
- It looks like xmlschema with the fraction_digits parameter truncates the fractional second part.
Sometimes it truncates too much.
ex : 1447961542.26 with 2 fraction digits gives 1447961542.25 instead of 1447961542.26
Could be due to f... - 10:10 PM Revision 895b44dc (git): * properties.
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52676 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 10:10 PM Revision c1a461e5 (git): * 2015-11-20
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52675 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 10:10 PM Feature #11713 (Closed): [PATCH] add .gitattributes
- Applied in changeset r52674.
----------
add .gitattributes
This improves method detection for Ruby source files;
making output of "git diff" and "git log -p -W" more
readable for git users looking at our Ruby code.
git will automatica... -
- This improves method detection for Ruby source files;
making output of "git diff" and "git log -p -W" more
readable for git users looking at our Ruby code.
git will automatically notice this file and start using
it when generating ... - 10:09 PM Revision c2088414 (git): add .gitattributes
- This improves method detection for Ruby source files;
making output of "git diff" and "git log -p -W" more
readable for git users looking at our Ruby code.
git will automatically notice this file and start using
it when generating diff ... -
- /tmp のような複数のユーザが共有するディレクトリでは不確実な操作
(たとえば存在しないかもしれないファイルの削除) はしないほうがいいと思います。
ライブラリ側でファイルを消すのがエラーになって困るというのであれば、
ブロックを使わない形で Tempfile.create を呼び出してください。
そうすればファイルは消されないので、rename するなり消すなり、アプリケーション側で自由に行えます。 -
- ファイルを作成してデータを書き込んでから読み込んで利用するプログラムにおいて、
全て書き込み終わるまでは他のプロセスからアクセスされたくないので、
ファイルをアトミックに作成する方法として、Tempfileに一旦書き込んでからrenameする方法が使われています。
https://github.com/rails/rails/blob/7e62a65e18d1e97debfa1c6215b7541157489115/activesupport/lib/acti... -
- buffer fullness is more robust to detect the message is too big for
the buffer.
AIX 7.1 recvmsg doesn't set MSG_TRUNC for rflags when MSG_PEEK is
given.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52673 b2dd03c8-39d4-4d8f-... -
- Unfortunately, it is known problem (2nd time machine stack overflow we can not capture correctly).
* 1st machine stack overflow
* SEGV
* check machine stack overflow
* raise an error from signal handler (*1) by longjmp.
* 2n... -
08:55 AM Bug #11692: [PATCH] Re-enable GC if stack overflow was caught from signal handler
- Tried with ruby head, this solves the GC problems now, but it doesn't solve the other problem that is visible if you run my example -
https://gist.github.com/ebeigarts/44648eb7b2773e102335
~~In ruby 2.2 it looks like if a stack overf... -
- Thank you for your great survey!
-
- Applied in changeset r52668.
----------
* signal.c: should also clear ruby_disable_gc.
[Bug #11692] -
01:39 PM Bug #11602: Compile error on CentOS: Segmentation fault after "linking miniruby"
- This is **always** reproducible on centos 5.x latest, running XEN kernel domU. No issues on regular centos 5.x (KVM or not).
Exact same stack trace.
Anybody care to fix this? -
- Sorry, last comment was my mistake.
ObjectSpace.dump_all() can output garbage (but not collected yet) objects.
This is current implementation.
```C
/* dump all objects */
rb_objspace_each_objects(heap_i, &dc);
```
T... -
- After running this example (dead_thread.rb) you can grep by printed address of thread and find out that it doesn't have parent
-
- It can be a bug of object traversing API. It is not well tested.
Do you have any script I can try?
@tmm1 do you have any idea?
Thanks,
Koichi -
- Sorry for late response.
I try this on my environment, but not so big difference.
Please reopen if you find it is still slow.
-
- even if it is GC.disable'd.
* test/ruby/test_gc.rb: add a test.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52672 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
12:04 PM Bug #11714 (Rejected): `include Module` affects scope of code loaded by require statements on Ruby 2.2+
- This seems to be a regression with Ruby 2.2 and beyond, I can still reproduce this in Ruby trunk from today. Ruby 2.1 works just fine.
~~~
$ cat /tmp/test.rb
module MyThing
class Date < String
end
end
include MyThing
req... -
- name modified
vm1_gc_short_lived* 1.015
vm1_gc_short_with_complex_long* 1.014
vm1_gc_short_with_long* 1.000
vm1_gc_short_with_symbol* 1.016
vm1_gc_wb_ary* 1... -
- Applied in changeset r52670.
----------
ruby.c: fix ARGV doc [ci skip]
* ruby.c (ruby_prog_init): [DOC] ARGV does not contain the name of
the executable. [ruby-core:71561] [Bug #11711] -
- * ruby.c (ruby_prog_init): [DOC] ARGV does not contain the name of
the executable. [ruby-core:71561] [Bug #11711]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52670 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- mistake.
-
- Please update other `enc/unicode/name2ctype.*` files too.
-
- It's an easy code to show the concept.
Both return `false`. -
01:38 AM Bug #10984: Hash#contain? to check whether hash contains other hash
- > Akira Tanaka wrote:
> ...
This implies that
```ruby
{a: 1, b: 2} <= {a: 1, b: 3}
{a: 1, b: 3} <= {a: 1, b: 2}
```
Is that expected? This breaks the main use-case in testing method results. -
- * ruby.c (feature_option): unify enable_option and disable_option
not to repeat feature names twice.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52669 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- I don't touch this behavior because Ruby 2.3 preview 1 was already released.
I try it on Ruby 2.4.
-
- On 2015/08/21 19:45, Eric Wong wrote:
> Sorry, I am mildly against this patch now because it makes
> the binary bigger and (slightly) slower.
>
> Do you have upcoming plans/changes to compile.c which can
> benefit from this? If s... -
-
- [Bug #11692]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52668 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Applied in changeset r52667.
----------
* compile.c (iseq_compile_each): T_IMEMO/iseq objects should be
wrap with ISeq wrappers. [Bug #11676] -
- wrap with ISeq wrappers. [Bug #11676]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52667 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- this version includes #1114, #1314, #1322, #1375, #1383, #1387
* test/rubygems: ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52666 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * ruby.c (process_options): set instruction compilation options at
once, and set disabled options to false explicitly.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52665 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 05:16 AM Revision 2e865ba8 (git): * win32/win32.c (finish_overlapped_socket): return value of this
- function should be only 0 or SOCKET_ERROR.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52664 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * compile.c (iseq_tailcall_optimize): apply tail call optimization
before conversion to specialized instructions. when looking
back from `leave` instruction, `send` instructions have been
translated already.
git-svn-id: svn+ssh:/... - 04:59 AM Revision 01a54cf4 (git): * win32/win32.c (finish_overlapped_socket): ignore EMSGSIZE when input,
- because POSIX platforms just do so. fixes test errors revealed by
r52647.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52662 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Backport r51445 into `ruby_2_2` branch at r52661 to suppress warning in CI.
-
- Backport r51445 into `ruby_2_2` branch at r52661 to suppress warning in CI.
-
- envutil.rb: -w for assert_separately
* test/lib/envutil.rb (assert_separately): always add -w option
for warnings.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@52661 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
11/18/2015
-
- * gc.c (rb_raw_obj_info): BUILTIN_TYPE works fro non-special_const
objects.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52660 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52659 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Private methods are implementation details which may change in
the future. Avoid cluttering our API documentation with them or
possibly encouraging their use.
[ruby-core:71540]
I'll commit in a day or two and make changes to sys*_... -
07:28 PM Feature #11710: [PATCH] Replace Set#merge with Set#merge! and make Set#merge non-mutating.
- (updating patch with a better change to the | method)
-
- The Set#merge method currently mutates its caller. I propose changing its behavior to non-mutating, and replace its current behavior with a mutating Set#merge! method.
For example, the current behavior:
```
> s = Set.new [1,2,3] ... -
- If we wanted to avoid chomping the equals sign on methods that have the equals already we'd have to write in a check for that. But I'm not sure that would be as performant.
~~~ruby
class Applicable
def initialize(obj)
@targe... -
- I was thinking of what word(s) may best clarify this and not conflict with other potential uses. One issue about the word `assigned` is it seems like `self = ` or `replace`. I at first was thinking of `to_be` ... but that's sounds more...
- 05:53 PM Revision 3006e5ea (git): * ext/socket/ancdata.c (bsock_recvmsg_internal): stretch the buffer size
- only when vmaxdatlen is nil.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52658 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
05:31 PM Bug #11711 (Closed): Documentation for ARGV behavior is incorrect
- http://ruby-doc.org/core-2.2.3/Object.html#ARGV
The above documentation states that ARGV[0] will contain the name of the Ruby program. While this is the behavior of ARGV in other places (C, Python, etc), it is *not* the behavior in Ru... -
- Applied in changeset r52657.
----------
* test/socket/test_socket.rb (test_udp_recvmsg_truncation): rflags is
nil on Solaris 10 which have no HAVE_STRUCT_MSGHDR_MSG_CONTROL.
Reported by Naohisa Goto. [ruby-core:71557] [Bug #11709] -
- On Solaris 10, the following failure occurs.
(observed in r52648)
~~~
1) Failure:
TestSocket#test_udp_recvmsg_truncation [/XXXXX-52648/test/socket/test_socket.rb:709]:
<32> expected but was
<false>.
~~~
The test code ignore... -
- nil on Solaris 10 which have no HAVE_STRUCT_MSGHDR_MSG_CONTROL.
Reported by Naohisa Goto. [ruby-core:71557] [Bug #11709]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52657 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 04:48 PM Revision f74c229c (git): * configure.in: add -static-libgcc for mingw automatically if available.
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52656 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- r51292, r51439 and r51440 were backported into `ruby_2_2` branch at r52655.
-
- r51292, r51439 and r51440 were backported into `ruby_2_2` branch at r52655.
r51292 seems required to r51440 works well. -
- * load.c (rb_load_internal0): do not raise any exceptions but
return the result tag state.
* load.c (rb_load_protect): reduce nested EXEC_TAGs.
* load.c (rb_load_internal0): extra check before returning
... - 03:55 PM Revision 3c075955 (git): * 2015-11-19
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52654 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- 03:55 PM Revision adbf8723 (git): * ext/extmk.rb (--extflags): new option to pass EXTLDFLAGS to children,
- expecially exts.mk.
* common.mk (EXTMK_ARGS): use above option.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52653 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- r52016, r52017, r52019, r52020 and r52021 were backported into `ruby_2_2` branch at r52652.
-
- * enc/euc_jp.c (mbc_case_fold): check given string is valid or not,
and if invalid, return 1. [Bug #11486]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@52652 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Backported into `ruby_2_2` branch at r52651.
-
- ruby_2_1 r52643 merged revision(s) 52556,52557.
-
- * ext/openssl/ossl_pkey.c: Merge ruby/openssl@b9ea8ef [Bug #10735]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@52651 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Backported into `ruby_2_2` branch at r52650.
-
- Backported into `ruby_2_2` branch at r52650.
-
- * io.c (argf_next_argv): check ARGV element type, and try
conversion if necessary. [ruby-core:71140] [Bug #11610]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@52650 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Backported into `ruby_2_2` branch at r52649.
-
- ruby_2_1 r52642 merged revision(s) 51722,51724.
-
- * ext/openssl/ossl_ssl.c (static const struct): Only add SSLv3 support
if the SSL library supports it. Thanks Kurt Roeckx <kurt@roeckx.be>
[Bug #11376]
* ext/openssl/extconf.rb: check for SSLv3 support in th... -
- It's common to create simple data-object with some constraints. One can either implement custom class or use `Struct`. Struct is generally simpler and helps to avoid some mistakes as non-defined `#hash` and `#eql?`. But at the same time ...
-
- Applied in changeset r52648.
----------
* vm_method.c (rb_class_clear_method_cache): should clear all
RCLASS_CALLABLE_M_TBLs of all sub-classes (T_ICLASS).
RCLASS_CALLABLE_M_TBL() caches complemented method entries.
It should be ... -
- RCLASS_CALLABLE_M_TBLs of all sub-classes (T_ICLASS).
RCLASS_CALLABLE_M_TBL() caches complemented method entries.
It should be cleared when the modules are cleared.
On previous version clears only for direct children.
It is enou... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52647 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- RCLASS_CALLABLE_M_TBL() instead of accessing a filed directly.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52646 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- ruby_2_1 r52645 merged revision(s) 50104,50105.
- 11:57 AM Revision 18d73ae8 (git): merge revision(s) 50104,50105: [Backport #10906]
- * ext/date/extconf.rb: try_cflags("-std=iso9899:1999") [Bug #10906]
ruby itself (including numeric.c) is built with strict compile
options including -std=iso9899:1999, but ext/date is not.
By the way -std=is... -
- ruby_2_1 r52638 merged revision(s) 51569,51649.
-
- ruby_2_1 r52644 merged revision(s) 50985.
- 11:48 AM Revision 5d428a7e (git): merge revision(s) 50985: [Backport #11693]
- runruby.rb: don't close other fds
* tool/runruby.rb: rubyspec now requires other FDs not to be
closed since 7b6ce1fee.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_1@52644 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 11:39 AM Revision dacb9169 (git): merge revision(s) 52556,52557: [Backport #10735]
- * ext/openssl/ossl_pkey.c: Merge ruby/openssl@b9ea8ef [Bug #10735]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_1@52643 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 11:34 AM Revision 936a2775 (git): merge revision(s) 51722,51724: [Backport #11376]
- * ext/openssl/ossl_ssl.c (static const struct): Only add SSLv3 support
if the SSL library supports it. Thanks Kurt Roeckx <kurt@roeckx.be>
[Bug #11376]
* ext/openssl/extconf.rb: check for SSLv3 support in th... -
- ruby_2_1 r52641 merged revision(s) 52476,52477.
- 11:26 AM Revision 3233891d (git): merge revision(s) 52476,52477: [Backport #11603]
- * vm_trace.c (rb_threadptr_exec_event_hooks_orig):
maintain trace_running counter on internal events.
This patch is made by Takashi Kokubun <takashikkbn@gmail.com>.
[Bug #11603] https://github.com/ruby/ruby/... -
- ruby_2_1 r52640 merged revision(s) 48609.
- 11:24 AM Revision 9f8d00ff (git): merge revision(s) 48609: [Backport #10449] [Backport #11651]
- * compile.c (iseq_compile_each): remove duplicated line event.
[Bug #10449]
* test/ruby/test_settracefunc.rb: add and fix tests.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_1@52640 b2dd03c8-39d4-4d8f-... -
- ruby_2_1 r52639 merged revision(s) 51713.
- 11:17 AM Revision bc090868 (git): merge revision(s) 51713: [Backport #11492] [Backport #11646]
- * vm.c (hook_before_rewind): prevent kicking :return event while
finishing vm_exec func because invoke_block_from_c() kick a :return
event for bmethods.
[Bug #11492]
* test/ruby/test_settracefunc.rb... -
- ruby_2_1 r52638 merged revision(s) 51569,51649.
- 11:13 AM Revision 69e671aa (git): merge revision(s) 51569,51649: [Backport #11367]
- test_ssl_session.rb: check SSL method
* test/openssl/test_ssl_session.rb (test_ctx_server_session_cb):
ensure the method to be tested is supported.
* test/openssl/test_ssl_session.rb: Fix tests so that they take in to
... -
- ruby_2_1 r52637 merged revision(s) 49099.
- 11:08 AM Revision 7e9ce9f5 (git): merge revision(s) 49099: [Backport #10046]
- * test/openssl/test_ssl_session.rb (OpenSSL#test_ctx_client_session_cb):
fix test failure with OpenSSL disabled SSLv3 protocol.
[ruby-core:63772] [Bug #10046]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/rub... -
- ```
logger = Logger.new(nil)
logger.fatal { 'foo' } # no print
```
Is this not enough? -
-
- * rb_method_definition_t::complemented_count to count shared method
entries because of complemented method entries and separate from
alias_count.
Shared `def' only by complemented method entries should not prevent
method... -
- * support SPECIAL_CONSTs.
* fix IMEMO/ment outputs.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52635 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
07:51 AM Bug #11707 (Rejected): (a * b) / b yields incorrect result when a is BigDecimal & b is Rational with large terms
- When performing arithmetic on BigDecimals and Rationals, (a * b) / b sometimes doesn't yield a result that is equal to a.
Here is an example program:
~~~
require 'bigdecimal'
a = BigDecimal.new('5.0')
b = Rational(889644323052... -
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52634 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- The files name2ctype.{h.blt,kwd,src} in etc/unicode are intermediate products that are not needed in the repository, and haven't been committed consistently. I propose to remove them.
[I'm not sure this is a bug or a feature, but it d... -
- * compile.c (iseq_peephole_optimize): eliminate always/never
branches after a literal object. this sequence typically
appears by defined? operator for a method call on a local
variable.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby... -
- Eric Wong <normalperson@yhbt.net> wrote:
> Will followup with other IO/Socket/SSL methods
Done for normal socket, asked about SSL in [ruby-core:71538]
I might ignore optimizing ARGF.read_nonblock(... exception: false)
for now si... -
- * compile.c (iseq_peephole_optimize): check element type before
checking insn_id which is valid only for INSN elements.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52632 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Edgars Beigarts wrote:
> Do I need to open a new issue for this?
I think it is the same problem. I wrote small explanation what is happening: https://gist.github.com/gazay/54da61919b85eb2e0d42 - 01:37 AM Revision d8e19392 (git): * ext/socket/ancdata.c (bsock_recvmsg_internal): stretch the buffer size
- when EMSGSIZE occurs on non HAVE_STRUCT_MSGHDR_MSG_CONTROL platforms
(such as, Windows). fixes a test error revealed by r52625.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52631 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 01:18 AM Revision ca5f5871 (git): socket: fix typo for default size
- * ext/socket/ancdata.c (bsock_recvmsg_internal): use 4096 as
default size to match pre-r52610, which also maps to a common
page size.
Oops; absolutely no idea why I made this typo...
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trun... -
- recent changes.
[ci skip] [ruby-core:71466] [Misc #11681] Patch by James Adam
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52629 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- I've missed absence of `<=>` first. Yes, you are right. And it can slightly reduce damage.
But anyway it doesn't resolve issues with misinterpretation of comparison negation. My qsort function is just the one problem which lie on surf... -
- Thanks, I will try with vs2013, but anyway committed :)
-
- Applied in changeset r52628.
----------
* test/logger/test_logdevice.rb: Fix tests of logger to make it work on
windows (windows can not remove opened file) [Bug #11702] -
- windows (windows can not remove opened file) [Bug #11702]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52628 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * configure.in (ruby-pc): fix help message for --with-ruby-pc
option.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52627 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- * configure.in: add additional flag variable descriptions, which
are appended to common flags, e.g., CFLAGS, CPPFLAGS, etc.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52626 b2dd03c8-39d4-4d8f-98ff-823fe69b080e