Bug #11488: String#index and #include? broken for wide encodings

• Status: Closed
• Assignee: -
• Target version: -
• ruby -v: ruby 2.2.3p173 (2015-08-18 revision 51636) [x86_64-darwin14]
• Backport: 2.0.0: UNKNOWN, 2.1: DONE, 2.2: DONE

[tomstuart (Tom Stuart)]

In Ruby 2.2.2, String#index and #include? work as expected for all encodings:

>> Encoding.list.reject(&:dummy?).
     reject { |e| 'abcdef'.encode(e).include?('def'.encode(e)) }.
     map(&:name)
=> []

But in Ruby 2.2.3 and trunk, they have become broken for wide encodings:

>> Encoding.list.reject(&:dummy?).
     reject { |e| 'abcdef'.encode(e).include?('def'.encode(e)) }.
     map(&:name)
=> ["UTF-16BE", "UTF-16LE", "UTF-32BE", "UTF-32LE"]

This broke as a result of revision 51470 (trunk) / 51590 (2.2.3). I attach a test.

[nobu (Nobuyoshi Nakada)]

• Related to Bug #11413: String#split with wchar string added

[nobu (Nobuyoshi Nakada)]

• Status changed from Open to Closed

Applied in changeset r51685.

---

re.c: fix wchar match at EOS

• re.c (rb_memsearch_wchar, rb_memsearch_qchar): test matching
  till the end of string. [ruby-core:70592] [Bug #11488]
• test/ruby/test_m17n.rb (test_include?, tet_index): add tests by
  Tom Stuart.

[nagachika (Tomoyuki Chikanaga)]

• Backport changed from 2.0.0: UNKNOWN, 2.1: UNKNOWN, 2.2: UNKNOWN to 2.0.0: UNKNOWN, 2.1: REQUIRED, 2.2: REQUIRED

This is introduced at r51470 for #11413 and it's backported into ruby_2_1 and ruby_2_2.

[usa (Usaku NAKAMURA)]

• Backport changed from 2.0.0: UNKNOWN, 2.1: REQUIRED, 2.2: REQUIRED to 2.0.0: UNKNOWN, 2.1: DONE, 2.2: REQUIRED

ruby_2_1 r51976 merged revision(s) 51685.

[nagachika (Tomoyuki Chikanaga)]

• Backport changed from 2.0.0: UNKNOWN, 2.1: DONE, 2.2: REQUIRED to 2.0.0: UNKNOWN, 2.1: DONE, 2.2: DONE

Backported into ruby_2_2 branch at r51986.