merge revision(s) 5b1bf8dd2d08ae7371ecf025967376bb794ed651: [Backport #16099]

	UTF LE is fixed at least the first 2 bytes

	* io.c (io_strip_bom): if the first 2 bytes are 0xFF0xFE, it
	should be a little-endian UTF, 16 or 32.  [Bug #16099]

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_6@67746 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

4 files changed, 9 insertions, 12 deletions

diff --git a/io.c b/io.c
index e581145564..94e425a277 100644
--- a/io.c
+++ b/io.c
@@ -6129,12 +6129,9 @@ io_strip_bom(VALUE io)
 		    return ENCINDEX_UTF_32LE;
 		}
 		rb_io_ungetbyte(io, b4);
-		rb_io_ungetbyte(io, b3);
-	    }
-	    else {
-		rb_io_ungetbyte(io, b3);
-		return ENCINDEX_UTF_16LE;
 	    }
+            rb_io_ungetbyte(io, b3);
+            return ENCINDEX_UTF_16LE;
 	}
 	rb_io_ungetbyte(io, b2);
 	break;
diff --git a/test/ruby/test_file.rb b/test/ruby/test_file.rb
index 36c154d36c..3deab76e93 100644
--- a/test/ruby/test_file.rb
+++ b/test/ruby/test_file.rb
@@ -87,7 +87,7 @@ class TestFile < Test::Unit::TestCase
   end
 
   def test_bom_32le
-    assert_bom(["\xFF\xFE\0", "\0"], __method__)
+    assert_bom(["\xFF", "\xFE\0\0"], __method__)
   end
 
   def test_truncate_wbuf
diff --git a/test/ruby/test_io_m17n.rb b/test/ruby/test_io_m17n.rb
index 8dfa5d5500..416a24ba6b 100644
--- a/test/ruby/test_io_m17n.rb
+++ b/test/ruby/test_io_m17n.rb
@@ -2082,14 +2082,14 @@ EOT
 
   def test_strip_bom
     with_tmpdir {
-      text = "\uFEFFa"
-      stripped = "a"
+      text = "\uFEFF\u0100a"
+      stripped = "\u0100a"
       %w/UTF-8 UTF-16BE UTF-16LE UTF-32BE UTF-32LE/.each do |name|
         path = '%s-bom.txt' % name
         content = text.encode(name)
         generate_file(path, content)
         result = File.read(path, mode: 'rb:BOM|UTF-8')
-        assert_equal(content[1].force_encoding("ascii-8bit"),
+        assert_equal(content[1..-1].force_encoding("ascii-8bit"),
                      result.force_encoding("ascii-8bit"))
         result = File.read(path, mode: 'rb:BOM|UTF-8:UTF-8')
         assert_equal(Encoding::UTF_8, result.encoding)
@@ -2099,10 +2099,10 @@ EOT
       bug3407 = '[ruby-core:30641]'
       path = 'UTF-8-bom.txt'
       result = File.read(path, encoding: 'BOM|UTF-8')
-      assert_equal("a", result.force_encoding("ascii-8bit"), bug3407)
+      assert_equal(stripped.b, result.force_encoding("ascii-8bit"), bug3407)
 
       bug8323 = '[ruby-core:54563] [Bug #8323]'
-      expected = "a\xff".force_encoding("utf-8")
+      expected = (stripped.b + "\xff").force_encoding("utf-8")
       open(path, 'ab') {|f| f.write("\xff")}
       result = File.read(path, encoding: 'BOM|UTF-8')
       assert_not_predicate(result, :valid_encoding?, bug8323)
diff --git a/version.h b/version.h
index 14a158f826..5ea21f14f4 100644
--- a/version.h
+++ b/version.h
@@ -1,6 +1,6 @@
 #define RUBY_VERSION "2.6.3"
 #define RUBY_RELEASE_DATE RUBY_RELEASE_YEAR_STR"-"RUBY_RELEASE_MONTH_STR"-"RUBY_RELEASE_DAY_STR
-#define RUBY_PATCHLEVEL 96
+#define RUBY_PATCHLEVEL 97
 
 #define RUBY_RELEASE_YEAR 2019
 #define RUBY_RELEASE_MONTH 8