Issue #10285 has been updated by Tomoyuki Chikanaga.

Backport changed from 2.0.0: REQUIRED, 2.1: REQUIRED to 2.0.0: REQUIRED, 2.1: DONE

Backported into `ruby_2_1` at r47982.

----------------------------------------
Bug #10285: StringIO with encodings broken due to #9769
https://bugs.ruby-lang.org/issues/10285#change-49498

* Author: Dirkjan Bussink
* Status: Closed
* Priority: Normal
* Assignee: 
* Category: 
* Target version: 
* ruby -v: 2.1.3
* Backport: 2.0.0: REQUIRED, 2.1: DONE
----------------------------------------
It looks like the change in https://bugs.ruby-lang.org/issues/9769 resulted in a behavior change with how StringIO works with different encodings. 

The following snippet is broken and now raises:

~~~
test_string_io_encoding.rb:8:in `write': incompatible character encodings: ASCII-8BIT and Windows-1252 (Encoding::CompatibilityError)
	from test_string_io_encoding.rb:8:in `<main>'
~~~

~~~
require 'stringio'

io = StringIO.new
io.set_encoding(Encoding::ASCII_8BIT)
io.write("quz \x83 mat".force_encoding(Encoding::ASCII_8BIT))

str = "foo \x97 bar".force_encoding(Encoding::WINDOWS_1252)
io.write(str)

p io.string
~~~

What is the intended behavior here? If I change the code to not set the encoding on the StringIO object, it does work somehow:

~~~
require 'stringio'

io = StringIO.new
io.write("quz \x83 mat".force_encoding(Encoding::ASCII_8BIT))

str = "foo \x97 bar".force_encoding(Encoding::WINDOWS_1252)
io.write(str)

p io.string
~~~ 

In this case it sees io.string as UTF-8 encoded, but this is invalid. It does allow the second StringIO#write here though.



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