Bug #1929

str.dup.force_encodingが元のstrに影響を与えることがある

Added by kazuhiko (Kazuhiko Shiozaki) almost 3 years ago. Updated about 1 year ago.

[ruby-dev:39068]
Status:Closed Start date:08/12/2009
Priority:Normal Due date:
Assignee:- % Done:

100%

Category:core
Target version:1.9.2
ruby -v:ruby 1.9.2dev (2009-08-09 trunk 24484) [x86_64-linux]

Description

かずひこです。

tDiary trunk 3514をruby-trunkで動かすとカテゴリプラグインで日本語カテゴリのリンクをたどると
incompatible character encodings: ASCII-8BIT and UTF-8 (Encoding::CompatibilityError)
になります。
http://www.cozmixng.org/retro/projects/tdiary/tickets/171

つきつめていくと、どうやら ERB::Util.url_encodeの
    def url_encode(s)
      s.to_s.dup.force_encoding("ASCII-8BIT").gsub(/[^a-zA-Z0-9_\-.]/n) {
        sprintf("%%%02X", $&.unpack("C")[0])
      }
    end
で、元のsが影響をうけているようなのです。
http://vvvvvv.sakura.ne.jp/ds14050/diary/20090223.html#p01

例えば、url_encodeの呼び出し側を次のように修正するとエラーが再現しません。
-    a = @category.map {|c| "category=#{u c}"}.join(';')
+    a = @category.map {|c| "category=#{u ''+c}"}.join(';')

もしくは、url_encodeを次のように修正するとエラーが再現しません。
    def url_encode(s)
      "#{s}".force_encoding("ASCII-8BIT").gsub(/[^a-zA-Z0-9_\-.]/n) {
        sprintf("%%%02X", $&.unpack("C")[0])
      }
    end

短い再現コードが書けなかったのですが、どうぞよろしくお願いします。
かずひこ

Associated revisions

Revision 24509
Added by nobu (Nobuyoshi Nakada) almost 3 years ago

* string.c (rb_str_new_frozen): must not change encoding of frozen shared string. [ruby-dev:39068]

History

Updated by znz (Kazuhiro NISHIYAMA) almost 3 years ago

まだ長いように思いますが、単体で再現できるものができたので送っておきます。

% cat c.rb
#!/usr/bin/env ruby-trunk
# -*- coding: utf-8 -*-
require 'pstore'
require 'cgi'
require 'erb'

def cache_file(c)
  'c.db'
end

params = CGI::parse("category=%E3%81%A6%E3%81%99%E3%81%A8")
categories = params['category']
categorized = {}
$dump_categorized = proc do |file, line|
  $stderr.puts [file, line, categories.map{|e|[e[0].object_id, e[0].encoding]}].inspect
  $stderr.puts [file, line, categorized.map{|e|[e[0].object_id, e[0].encoding]}].inspect
end
$dump_categorized[__FILE__, __LINE__]
categories.each do |c|
  PStore.new(cache_file(c)).transaction do |db|
    categorized[c] = db['category']
    db.abort
  end
end

$dump_categorized[__FILE__, __LINE__]
categorized.keys.each do |c|
  PStore.new(cache_file(c)).transaction do |db|
    db['category'] = {} unless db.root?('category')
    db['category'].update(categorized[c] || {})
  end
end
$dump_categorized[__FILE__, __LINE__]
categories.map{|e|ERB::Util.u(e)}
$dump_categorized[__FILE__, __LINE__]
% ruby-trunk c.rb -v
["c.rb", 18, [[10938200, #<Encoding:UTF-8>]]]
["c.rb", 18, []]
["c.rb", 26, [[10937000, #<Encoding:UTF-8>]]]
["c.rb", 26, [[10937100, #<Encoding:UTF-8>]]]
["c.rb", 33, [[10935120, #<Encoding:UTF-8>]]]
["c.rb", 33, [[10937100, #<Encoding:UTF-8>]]]
["c.rb", 35, [[10932440, #<Encoding:UTF-8>]]]
["c.rb", 35, [[10937100, #<Encoding:ASCII-8BIT>]]]
% ruby-trunk -v
ruby 1.9.2dev (2009-08-07 trunk 24439) [x86_64-linux]
%

Updated by nobu (Nobuyoshi Nakada) almost 3 years ago

  • Status changed from Open to Closed
  • % Done changed from 0 to 100
Applied in changeset r24509.

Also available in: Atom PDF