From: naruse@...
Date: 2018-06-07T10:32:01+00:00
Subject: [ruby-core:87442] [Ruby trunk Feature#14832] Net::ReadTimeout and Net::WriteTimeout should tell the cause socket

Issue #14832 has been updated by naruse (Yui NARUSE).


TCPSocket#inspect shows its address and port from Ruby 2.5 like `#<TCPSocket:fd 10, AF_INET6, ::1, 56090>`.
But if the socket is closed, it becomes `#<TCPSocket:(closed)>`; it's useless.
At least Net::HTTP.start automatically closes the socket when it escapes given block.
It means its message will be like follows:

```
%  ./ruby -Ilib test.rb
Sending 5120000 bytes
Traceback (most recent call last):
    13: from test.rb:15:in `<main>'
    12: from /Users/naruse/work/ruby/lib/net/http.rb:1478:in `request'
    11: from /Users/naruse/work/ruby/lib/net/http.rb:1505:in `transport_request'
    10: from /Users/naruse/work/ruby/lib/net/http.rb:1505:in `catch'
     9: from /Users/naruse/work/ruby/lib/net/http.rb:1506:in `block in transport_request'
     8: from /Users/naruse/work/ruby/lib/net/http/generic_request.rb:123:in `exec'
     7: from /Users/naruse/work/ruby/lib/net/http/generic_request.rb:189:in `send_request_with_body'
     6: from /Users/naruse/work/ruby/lib/net/protocol.rb:247:in `write'
     5: from /Users/naruse/work/ruby/lib/net/protocol.rb:265:in `writing'
     4: from /Users/naruse/work/ruby/lib/net/protocol.rb:248:in `block in write'
     3: from /Users/naruse/work/ruby/lib/net/protocol.rb:275:in `write0'
     2: from /Users/naruse/work/ruby/lib/net/protocol.rb:275:in `each_with_index'
     1: from /Users/naruse/work/ruby/lib/net/protocol.rb:275:in `each'
/Users/naruse/work/ruby/lib/net/protocol.rb:296:in `block in write0': Net::WriteTimeout with #<TCPSocket:(closed)> (Net::WriteTimeout)
```

Maybe WriteTimeout#initialize should get inspect or address.

----------------------------------------
Feature #14832: Net::ReadTimeout and Net::WriteTimeout should tell the cause socket
https://bugs.ruby-lang.org/issues/14832#change-72434

* Author: usa (Usaku NAKAMURA)
* Status: Assigned
* Priority: Normal
* Assignee: naruse (Yui NARUSE)
* Target version: 2.6
----------------------------------------
When `Net::ReadTimeout` occurs, we can't get any information from the exception.
It makes the trouble shooting too difficult.

I propose that `io` member to `Net::ReadTimeout` instance.
Then, when `Net::ReadTimeout` is rescued, we can know the cause IO instance.


```diff
Index: lib/net/protocol.rb
===================================================================
--- lib/net/protocol.rb	(revision 63597)
+++ lib/net/protocol.rb	(working copy)
@@ -75,15 +75,41 @@
   # ReadTimeout, a subclass of Timeout::Error, is raised if a chunk of the
   # response cannot be read within the read_timeout.
 
-  class ReadTimeout            < Timeout::Error; end
+  class ReadTimeout < Timeout::Error
+    def initialize(io = nil)
+      @io = io
+    end
+    attr_reader :io
 
+    def message
+      msg = super
+      if @io
+        msg = "#{msg} with #{@io.inspect}"
+      end
+      msg
+    end
+  end
+
   ##
   # WriteTimeout, a subclass of Timeout::Error, is raised if a chunk of the
   # response cannot be read within the read_timeout.
 
-  class WriteTimeout            < Timeout::Error; end
+  class WriteTimeout < Timeout::Error
+    def initialize(io = nil)
+      @io = io
+    end
+    attr_reader :io
 
+    def message
+      msg = super
+      if @io
+        msg = "#{msg} with #{@io.inspect}"
+      end
+      msg
+    end
+  end
 
+
   class BufferedIO   #:nodoc: internal use only
     def initialize(io, read_timeout: 60, write_timeout: 60, continue_timeout: nil, debug_output: nil)
       @io = io
@@ -188,12 +214,12 @@
         rv.clear
         return
       when :wait_readable
-        @io.to_io.wait_readable(@read_timeout) or raise Net::ReadTimeout
+        (io = @io.to_io).wait_readable(@read_timeout) or raise Net::ReadTimeout.new(io)
         # continue looping
       when :wait_writable
         # OpenSSL::Buffering#read_nonblock may fail with IO::WaitWritable.
         # http://www.openssl.org/support/faq.html#PROG10
-        @io.to_io.wait_writable(@read_timeout) or raise Net::ReadTimeout
+        (io = @io.to_io).wait_writable(@read_timeout) or raise Net::ReadTimeout.new(io)
         # continue looping
       when nil
         raise EOFError, 'end of file reached'
@@ -267,7 +293,7 @@
           end
           # continue looping
         when :wait_writable
-          @io.to_io.wait_writable(@write_timeout) or raise Net::WriteTimeout
+          (io = @io.to_io).wait_writable(@write_timeout) or raise Net::WriteTimeout.new(io)
           # continue looping
         end while need_retry
       end
```



-- 
https://bugs.ruby-lang.org/

Unsubscribe: <mailto:ruby-core-request@ruby-lang.org?subject=unsubscribe>
<http://lists.ruby-lang.org/cgi-bin/mailman/options/ruby-core>