Bug #5836: #slice results inconsistent after #force_encoding to a valid encoding

• Status: Closed
• Assignee: -
• Target version: -
• ruby -v: ruby 1.9.3p0 (2011-10-30 revision 33570) [x86_64-darwin10.8.0]

[brixen (Brian Shirai)]

I was not able to reduce this further. After calling #force_encoding, #slice returns inconsistent results:

ruby-1.9.3-p0 :001 > RUBY_DESCRIPTION
=> "ruby 1.9.3p0 (2011-10-30 revision 33570) [x86_64-darwin10.8.0]"
ruby-1.9.3-p0 :002 > s = "abc"
=> "abc"
ruby-1.9.3-p0 :003 > s.size
=> 3
ruby-1.9.3-p0 :004 > s.encoding
=> #Encoding:UTF-8
ruby-1.9.3-p0 :005 > s.valid_encoding?
=> true
ruby-1.9.3-p0 :006 > s.slice 3
=> nil
ruby-1.9.3-p0 :007 > s.slice 3, 1
=> ""
ruby-1.9.3-p0 :008 > s
=> "abc"
ruby-1.9.3-p0 :009 > s.force_encoding 'ascii-8bit'
=> "abc"
ruby-1.9.3-p0 :010 > s.slice 3
=> nil
ruby-1.9.3-p0 :011 > s.slice 3, 1
=> ""
ruby-1.9.3-p0 :012 > s
=> "abc"
ruby-1.9.3-p0 :013 > s.valid_encoding?
=> true
ruby-1.9.3-p0 :014 > s.force_encoding 'euc-jp'
=> "abc"
ruby-1.9.3-p0 :015 > s.slice 3
=> nil
ruby-1.9.3-p0 :016 > s.slice 3, 1
=> nil
ruby-1.9.3-p0 :017 > s.valid_encoding?
=> true

Thanks,
Brian

[manveru (Michael Fellinger)]

I've reduced the code a bit, calling force_encoding seems to cause it, and calling valid_encoding? repairs the damage.

s = 'abc'
p [s.slice(3), s.slice(3, 1)]
[nil, ""]

s.force_encoding 'euc-jp'
p [s.slice(3), s.slice(3, 1)]
[nil, nil]

s.valid_encoding?
p [s.slice(3), s.slice(3, 1)]
[nil, ""]

s.force_encoding 'euc-jp'
p [s.slice(3), s.slice(3, 1)]
[nil, nil]

[nobu (Nobuyoshi Nakada)]

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

This issue was solved with changeset r34208.
Brian, thank you for reporting this issue.
Your contribution to Ruby is greatly appreciated.
May Ruby be with you.

---

• string.c (str_nth_len): count ascii-only run at the end. this
  bug appears only when single-byte-optimization is disabled due
  to unknown coderange. [ruby-core:41896] [Bug #5836]