From 2cf12ca386108a48f89271eff638fe6d39076a65 Mon Sep 17 00:00:00 2001
From: Hiroshi Shirosaki <h.shirosaki@gmail.com>
Date: Fri, 29 Mar 2013 16:03:22 +0900
Subject: [PATCH] load.c: fix require with non ascii path

---
 internal.h                |    1 +
 load.c                    |    6 +++---
 string.c                  |   13 +++++++++++--
 test/ruby/test_require.rb |   20 ++++++++++++++++++++
 4 files changed, 35 insertions(+), 5 deletions(-)

diff --git a/internal.h b/internal.h
index fccbeb6..02d6a39 100644
--- a/internal.h
+++ b/internal.h
@@ -295,6 +295,7 @@ int rb_str_buf_cat_escaped_char(VALUE result, unsigned int c, int unicode_p);
 int rb_str_symname_p(VALUE);
 VALUE rb_str_quote_unprintable(VALUE);
 VALUE rb_id_quote_unprintable(ID);
+VALUE rb_str_subseq_without_enc(VALUE, long, long);
 #define QUOTE(str) rb_str_quote_unprintable(str)
 #define QUOTE_ID(id) rb_id_quote_unprintable(id)
 
diff --git a/load.c b/load.c
index 7c69461..d04c475 100644
--- a/load.c
+++ b/load.c
@@ -247,16 +247,16 @@ features_index_add(VALUE feature, VALUE offset)
 	if (p < feature_str)
 	    break;
 	/* Now *p == '/'.  We reach this point for every '/' in `feature`. */
-	short_feature = rb_str_substr(feature, p + 1 - feature_str, feature_end - p - 1);
+	short_feature = rb_str_subseq_without_enc(feature, p + 1 - feature_str, feature_end - p - 1);
 	features_index_add_single(short_feature, offset);
 	if (ext) {
-	    short_feature = rb_str_substr(feature, p + 1 - feature_str, ext - p - 1);
+	    short_feature = rb_str_subseq_without_enc(feature, p + 1 - feature_str, ext - p - 1);
 	    features_index_add_single(short_feature, offset);
 	}
     }
     features_index_add_single(feature, offset);
     if (ext) {
-	short_feature = rb_str_substr(feature, 0, ext - feature_str);
+	short_feature = rb_str_subseq_without_enc(feature, 0, ext - feature_str);
 	features_index_add_single(short_feature, offset);
     }
 }
diff --git a/string.c b/string.c
index 8bbd8a4..924152e 100644
--- a/string.c
+++ b/string.c
@@ -1667,7 +1667,7 @@ rb_str_sublen(VALUE str, long pos)
 }
 
 VALUE
-rb_str_subseq(VALUE str, long beg, long len)
+rb_str_subseq_without_enc(VALUE str, long beg, long len)
 {
     VALUE str2;
 
@@ -1681,12 +1681,21 @@ rb_str_subseq(VALUE str, long beg, long len)
 	RB_GC_GUARD(str);
     }
 
-    rb_enc_cr_str_copy_for_substr(str2, str);
     OBJ_INFECT(str2, str);
 
     return str2;
 }
 
+VALUE
+rb_str_subseq(VALUE str, long beg, long len)
+{
+    VALUE str2 = rb_str_subseq_without_enc(str, beg, len);
+
+    rb_enc_cr_str_copy_for_substr(str2, str);
+
+    return str2;
+}
+
 static char *
 rb_str_subpos(VALUE str, long beg, long *lenp)
 {
diff --git a/test/ruby/test_require.rb b/test/ruby/test_require.rb
index cb17b91..56041c1 100644
--- a/test/ruby/test_require.rb
+++ b/test/ruby/test_require.rb
@@ -639,4 +639,24 @@ class TestRequire < Test::Unit::TestCase
   ensure
     script.close(true) if script
   end
+
+  def test_require_with_non_ascii_path
+    bug8165 = '[ruby-core:53733] [Bug #8165]'
+    Dir.mktmpdir {|tmp|
+      Dir.chdir(tmp) {
+        dir = "\u3042" * 5
+        Dir.mkdir(dir)
+        path = File.join(tmp, dir, 'foo.rb').force_encoding('UTF-8')
+        open(path, "w") {|f|
+          f.puts "p :ok"
+        }
+        assert_in_out_err([], <<-INPUT, %w(:ok), [], bug8165)
+          # coding: UTF-8
+          $:.replace([IO::NULL])
+          require '#{path}'
+          p :ng if require '#{path}'
+        INPUT
+      }
+    }
+  end
 end
-- 
1.7.10.4