1 files changed, 94 insertions, 0 deletions
diff --git a/spec/ruby/core/array/pack/shared/unicode.rb b/spec/ruby/core/array/pack/shared/unicode.rb
new file mode 100644
index 0000000000..e16110c491
--- /dev/null
+++ b/spec/ruby/core/array/pack/shared/unicode.rb
@@ -0,0 +1,94 @@
+# -*- encoding: utf-8 -*-
+
+describe :array_pack_unicode, shared: true do
+  it "encodes ASCII values as a Unicode codepoint" do
+    [ [[0],   "\x00"],
+      [[1],   "\x01"],
+      [[8],   "\x08"],
+      [[15],  "\x0f"],
+      [[24],  "\x18"],
+      [[31],  "\x1f"],
+      [[127], "\x7f"],
+      [[128], "\xc2\x80"],
+      [[129], "\xc2\x81"],
+      [[255], "\xc3\xbf"]
+    ].should be_computed_by(:pack, "U")
+  end
+
+  it "encodes UTF-8 BMP codepoints" do
+    [ [[0x80],    "\xc2\x80"],
+      [[0x7ff],   "\xdf\xbf"],
+      [[0x800],   "\xe0\xa0\x80"],
+      [[0xffff],  "\xef\xbf\xbf"]
+    ].should be_computed_by(:pack, "U")
+  end
+
+  it "constructs strings with valid encodings" do
+    str = [0x85].pack("U*")
+    str.should == "\xc2\x85"
+    str.valid_encoding?.should be_true
+  end
+
+  it "encodes values larger than UTF-8 max codepoints" do
+    [
+      [[0x00110000], [244, 144, 128, 128].pack('C*').force_encoding('utf-8')],
+      [[0x04000000], [252, 132, 128, 128, 128, 128].pack('C*').force_encoding('utf-8')],
+      [[0x7FFFFFFF], [253, 191, 191, 191, 191, 191].pack('C*').force_encoding('utf-8')]
+    ].should be_computed_by(:pack, "U")
+  end
+
+  it "encodes UTF-8 max codepoints" do
+    [ [[0x10000],   "\xf0\x90\x80\x80"],
+      [[0xfffff],   "\xf3\xbf\xbf\xbf"],
+      [[0x100000],  "\xf4\x80\x80\x80"],
+      [[0x10ffff],  "\xf4\x8f\xbf\xbf"]
+    ].should be_computed_by(:pack, "U")
+  end
+
+  it "encodes the number of array elements specified by the count modifier" do
+    [ [[0x41, 0x42, 0x43, 0x44], "U2",  "\x41\x42"],
+      [[0x41, 0x42, 0x43, 0x44], "U2U", "\x41\x42\x43"]
+    ].should be_computed_by(:pack)
+  end
+
+  it "encodes all remaining elements when passed the '*' modifier" do
+    [0x41, 0x42, 0x43, 0x44].pack("U*").should == "\x41\x42\x43\x44"
+  end
+
+  it "calls #to_int to convert the pack argument to an Integer" do
+    obj = mock('to_int')
+    obj.should_receive(:to_int).and_return(5)
+    [obj].pack("U").should == "\x05"
+  end
+
+  it "raises a TypeError if #to_int does not return an Integer" do
+    obj = mock('to_int')
+    obj.should_receive(:to_int).and_return("5")
+    lambda { [obj].pack("U") }.should raise_error(TypeError)
+  end
+
+  it "ignores NULL bytes between directives" do
+    [1, 2, 3].pack("U\x00U").should == "\x01\x02"
+  end
+
+  it "ignores spaces between directives" do
+    [1, 2, 3].pack("U U").should == "\x01\x02"
+  end
+
+  it "raises a RangeError if passed a negative number" do
+    lambda { [-1].pack("U") }.should raise_error(RangeError)
+  end
+
+  it "raises a RangeError if passed a number larger than an unsigned 32-bit integer" do
+    lambda { [2**32].pack("U") }.should raise_error(RangeError)
+  end
+
+  it "sets the output string to UTF-8 encoding" do
+    [ [[0x00].pack("U"),     Encoding::UTF_8],
+      [[0x41].pack("U"),     Encoding::UTF_8],
+      [[0x7F].pack("U"),     Encoding::UTF_8],
+      [[0x80].pack("U"),     Encoding::UTF_8],
+      [[0x10FFFF].pack("U"), Encoding::UTF_8]
+    ].should be_computed_by(:encoding)
+  end
+end