Ruby Issue Tracking System: Issueshttps://redmine.ruby-lang.org/https://redmine.ruby-lang.org/favicon.ico?17113305112013-04-12T22:54:05ZRuby Issue Tracking System
Redmine Ruby master - Bug #8261 (Rejected): module_function for methods of same namehttps://redmine.ruby-lang.org/issues/82612013-04-12T22:54:05Zmanveru (Michael Fellinger)m.fellinger@gmail.com
<p>This affects all versions from 1.8 to 2.0 I have tested.</p>
<p>I'd expect all three examples below to have the same output, but not only does the example with module_function end up with a private instance method, it also has the wrong one.</p>
<p>If the name of the method is different, it behaves as one would expect, this only happens if the methods have the same name.</p>
<p>module A<br>
def a<br>
1<br>
end</p>
<p>module_function</p>
<p>def a<br>
2<br>
end<br>
end</p>
<p>class X<br>
include A<br>
end</p>
<p>module B<br>
def b<br>
1<br>
end</p>
<p>def self.b<br>
2<br>
end<br>
end</p>
<p>class Y<br>
include B<br>
end</p>
<p>module C<br>
def c<br>
1<br>
end</p>
<p>class << self<br>
def c<br>
2<br>
end<br>
end<br>
end</p>
<p>class Z<br>
include C<br>
end</p>
<p>p [A.a, X.new.send(:a)]</p>
<a name="gt-2-2"></a>
<h1 >=> [2, 2]<a href="#gt-2-2" class="wiki-anchor">¶</a></h1>
<p>p [B.b, Y.new.b]</p>
<a name="gt-2-1"></a>
<h1 >=> [2, 1]<a href="#gt-2-1" class="wiki-anchor">¶</a></h1>
<p>p [C.c, Z.new.c]</p>
<a name="gt-2-1-2"></a>
<h1 >=> [2, 1]<a href="#gt-2-1-2" class="wiki-anchor">¶</a></h1> Ruby master - Bug #4037 (Rejected): Net/http keep alive bughttps://redmine.ruby-lang.org/issues/40372010-11-08T23:26:12Zmanveru (Michael Fellinger)m.fellinger@gmail.com
<p>=begin<br>
I've found a bug where if the target server sends an eof to the socket, net/http will raise an error. I've attached<br>
a sample script to prove the bug and a patch to fix it. I wrote the patch against 1.8, but I think it should apply<br>
to 1.9.</p>
<p>-Aaron</p>
<p>(This bug is from <a href="https://rubyforge.org/tracker/index.php?func=detail&aid=17778&group_id=426&atid=1700" class="external">https://rubyforge.org/tracker/index.php?func=detail&aid=17778&group_id=426&atid=1700</a> - I simply copied it here to get attention and proper tracking)</p>
<p>Some more information to reproduce the issue: <a href="http://pastie.org/1281626" class="external">http://pastie.org/1281626</a><br>
=end</p> Ruby master - Bug #3844 (Closed): Kernel.methods(false) not documentedhttps://redmine.ruby-lang.org/issues/38442010-09-17T19:02:46Zmanveru (Michael Fellinger)m.fellinger@gmail.com
<p>=begin<br>
The documentation of Kernel.methods doesn't cover the optional argument.<br>
It's covered by RubySpec, so I think the docs should mention it.<br>
=end</p> Ruby master - Feature #2759 (Closed): Regexp /g and /G optionshttps://redmine.ruby-lang.org/issues/27592010-02-19T04:01:51Zmanveru (Michael Fellinger)m.fellinger@gmail.com
<p>=begin<br>
Oniguruma has flags called ONIG_OPTION_CAPTURE_GROUP, ONIG_OPTION_DONT_CAPTURE_GROUP.<br>
There has been a discussion about it in the past, which can be found at <a href="http://blade.nagaokaut.ac.jp/cgi-bin/vframe.rb/ruby/ruby-dev/28446?28233-28923" class="external">http://blade.nagaokaut.ac.jp/cgi-bin/vframe.rb/ruby/ruby-dev/28446?28233-28923</a> (Japanese).<br>
As far as I can tell, the proposed patch hasn't been applied yet due to lack of demand.<br>
With this Issue, I'd like to vote for application of the patch, as I need the functionality in my code.<br>
=end</p> Ruby master - Bug #1892 (Closed): mkmf.rb:1008:in check_sizeof: undefined local variable or metho...https://redmine.ruby-lang.org/issues/18922009-08-06T07:02:04Zmanveru (Michael Fellinger)m.fellinger@gmail.com
<p>=begin<br>
An issue introduced by r23369<br>
It references "opts", but that doesn't exist.<br>
=end</p> Ruby master - Bug #1616 (Closed): Net::BufferedIO tries to raise Timeout::TimeoutErrorhttps://redmine.ruby-lang.org/issues/16162009-06-13T02:04:02Zmanveru (Michael Fellinger)m.fellinger@gmail.com
<p>=begin<br>
in lib/net/protocol.rb in the Net::BufferedIO#rbuf_fill method, in some cases Timeout::TimeoutError is raised, but this class does not exist anymore.<br>
Timeout::TimeoutError has been deprecated? Yet it's being used and mentioned in the stdlib, but only absolutely referenced in lib/net/protocol.rb while it is being set in the main namespace.</p>
<pre><code> # Another name for Timeout::Error, defined for backwards compatibility with
# earlier versions of timeout.rb.
TimeoutError = Timeout::Error
</code></pre>
<p>Other places where it's mentioned:</p>
<p>lib/net/smtp.rb<br>
373: # TimeoutError is raised.<br>
378: # TimeoutError is raised.<br>
458: # * TimeoutError<br>
519: # * TimeoutError<br>
649: # * TimeoutError<br>
703: # * TimeoutError</p>
<p>lib/net/pop.rb<br>
491: # it raises a TimeoutError exception.<br>
496: # it raises a TimeoutError exception.</p>
<p>lib/net/protocol.rb<br>
140: raise Timeout::TimeoutError<br>
148: raise Timeout::TimeoutError</p>
<p>lib/net/telnet.rb<br>
250: # TimeoutError to be raised. The default value is 10 seconds.<br>
355: rescue TimeoutError<br>
356: raise TimeoutError, "timed out while opening a connection to the host"<br>
513: # before raising a TimeoutError. If set to false,<br>
560: raise TimeoutError, "timed out while waiting for more data"</p>
<p>lib/net/http.rb<br>
522: # it raises a TimeoutError exception.<br>
527: # it raises a TimeoutError exception.</p>
<p>lib/timeout.rb<br>
87:TimeoutError = Timeout::Error<br>
93: p timeout(5, TimeoutError) {</p>
<p>lib/resolv.rb<br>
162: class ResolvTimeout < TimeoutError; end</p>
<p>lib/webrick/httprequest.rb<br>
356: rescue TimeoutError</p>
<p>test/drb/drbtest.rb<br>
171: assert_raise(TimeoutError) do<br>
174: assert_raise(TimeoutError) do</p>
<p>test/ruby/test_readpartial.rb<br>
53: assert_raise(TimeoutError) {<br>
67: assert_raise(TimeoutError) {<br>
=end</p> Backport187 - Bug #1246 (Closed): Segfault with Symbol#to_proc and yield without argumenthttps://redmine.ruby-lang.org/issues/12462009-03-05T09:58:10Zmanveru (Michael Fellinger)m.fellinger@gmail.com
<p>=begin<br>
Ruby 1.8.7 should raise an ArgumentError instead of segfaulting:</p>
<p>sigma ~ % ruby -ve 'def a; yield; end; a(&:p)'<br>
ruby 1.8.7 (2008-08-11 patchlevel 72) [x86_64-linux]<br>
-e:1: [BUG] Segmentation fault</p>
<p>sigma ~ % ruby19 -ve 'def a; yield; end; a(&:p)'<br>
ruby 1.9.2dev (2009-02-13 trunk 22277) [x86_64-linux]<br>
-e:1:in <code>a': no receiver given (ArgumentError) from -e:1:in </code>'<br>
=end</p> Ruby master - Bug #1053 (Closed): Obsolete methods in Pathnamehttps://redmine.ruby-lang.org/issues/10532009-01-27T01:19:17Zmanveru (Michael Fellinger)m.fellinger@gmail.com
<p>=begin<br>
There are lot of methods in Pathname that have been marked obsolete and give warnings since 2003, shouldn't they be removed by 1.9?<br>
=end</p> Ruby master - Bug #717 (Closed): Unneccesary statement in lib/irb.rbhttps://redmine.ruby-lang.org/issues/7172008-11-07T10:12:28Zmanveru (Michael Fellinger)m.fellinger@gmail.com
<p>=begin<br>
In lib/irb.rb in the <code>eval_input</code> method around line 110, there is a <code>@context.prompt_i</code> too much, <code>else</code> doesn't care.<br>
This is the case in all versions of Ruby back to 1.8.2 (that's as far as i cared to look).</p>
<p>The context:</p>
<p>if ltype<br>
f = @context.prompt_s<br>
elsif continue<br>
f = @context.prompt_c<br>
elsif indent > 0<br>
f = @context.prompt_n<br>
else @context.prompt_i<br>
f = @context.prompt_i<br>
end<br>
=end</p>