[ruby-core:66840] [ruby-trunk - Bug #10598] [Closed] Cannot make two symbols with same bytes and different encodings

Issue #10598 has been updated by Nobuyoshi Nakada.

Status changed from Open to Closed
% Done changed from 0 to 100

Applied in changeset r48845.

----------
string.c: fix coderange for non-endianness string

* string.c (rb_enc_str_coderange): dummy wchar, non-endianness
  encoding string cannot be ascii only.
  [ruby-core:66835] [Bug #10598]

----------------------------------------
Bug #10598: Cannot make two symbols with same bytes and different encodings
https://bugs.ruby-lang.org/issues/10598#change-50403

* Author: David Grayson
* Status: Closed
* Priority: Normal
* Assignee: 
* Category: core
* Target version: 
* ruby -v: ruby 2.2.0preview2 (2014-11-28 trunk 48628) [x86_64-linux]
* Backport: 2.0.0: UNKNOWN, 2.1: UNKNOWN
----------------------------------------
It looks like Ruby 2.1.1 introduced a bug where it is impossible create two different symbols with the same bytes but different encodings.  Here is a simple script that reproduces the bug:

```{ruby}
sym1 = "ab".force_encoding("UTF-16").to_sym
sym2 = "ab".to_sym
puts sym2.encoding

sym3 = "cd".to_sym
sym4 = "cd".force_encoding("UTF-16").to_sym
puts sym4.encoding
```

I would expect the output of this script to be:

```
US-ASCII
UTF-16
```

The script behaves as expected in Ruby 2.1.0, but in Ruby 2.1.1 and every later version that I tested, it gives incorrect results.  Here is a shell session showing the output of the script when I run it in Ruby 2.1.0, 2.1.1, and 2.2.0-preview2:

```
$ chruby 2.1.0 && ruby -v && ruby symbol_encoding_bug.rb
ruby 2.1.0p0 (2013-12-25 revision 44422) [x86_64-linux]
US-ASCII
UTF-16

$ chruby 2.1.1 && ruby -v && ruby symbol_encoding_bug.rb
ruby 2.1.1p76 (2014-02-24 revision 45161) [x86_64-linux]
UTF-16
US-ASCII

$ chruby 2.2.0-preview2 && ruby -v && ruby symbol_encoding_bug.rb
ruby 2.2.0preview2 (2014-11-28 trunk 48628) [x86_64-linux]
UTF-16
US-ASCII
```

It looks like `String#to_sym` is not properly accounting for the encoding of the string when it searches the symbol table.

This is definitely a bug; the value of `"ab".to_sym.encoding` should be predictable; it should not depend on the state of the symbol table.

By the way, JRuby has a similar bug:  https://github.com/jruby/jruby/issues/1348



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