1 files changed, 115 insertions, 0 deletions
diff --git a/spec/ruby/core/string/unicode_normalize_spec.rb b/spec/ruby/core/string/unicode_normalize_spec.rb
new file mode 100644
index 0000000000..138c0455fd
--- /dev/null
+++ b/spec/ruby/core/string/unicode_normalize_spec.rb
@@ -0,0 +1,115 @@
+# -*- encoding: utf-8 -*-
+require File.expand_path('../../../spec_helper', __FILE__)
+
+# Examples taken from http://www.unicode.org/reports/tr15/#Norm_Forms
+
+describe "String#unicode_normalize" do
+  before :each do
+    @accented_f = "\u1e9b\u0323"
+    @angstrom = "\u212b"
+    @ohm = "\u2126"
+  end
+
+  it "normalizes code points in the string according to the form that is specified" do
+    @accented_f.unicode_normalize(:nfc).should == "\u1e9b\u0323"
+    @accented_f.unicode_normalize(:nfd).should == "\u017f\u0323\u0307"
+    @accented_f.unicode_normalize(:nfkc).should == "\u1e69"
+    @accented_f.unicode_normalize(:nfkd).should == "\u0073\u0323\u0307"
+  end
+
+  it "defaults to the nfc normalization form if no forms are specified" do
+    @accented_f.unicode_normalize.should == "\u1e9b\u0323"
+    @angstrom.unicode_normalize.should == "\u00c5"
+    @ohm.unicode_normalize.should == "\u03a9"
+  end
+
+  # http://unicode.org/faq/normalization.html#6
+  context "returns normalized form of string by default" do
+    it "03D3 (ϓ) GREEK UPSILON WITH ACUTE AND HOOK SYMBOL" do
+      "\u03D3".unicode_normalize(:nfc).should == "\u03D3"
+      "\u03D3".unicode_normalize(:nfd).should == "\u03D2\u0301"
+      "\u03D3".unicode_normalize(:nfkc).should == "\u038E"
+      "\u03D3".unicode_normalize(:nfkd).should == "\u03A5\u0301"
+    end
+
+    it "03D4 (ϔ) GREEK UPSILON WITH DIAERESIS AND HOOK SYMBOL" do
+      "\u03D4".unicode_normalize(:nfc).should == "\u03D4"
+      "\u03D4".unicode_normalize(:nfd).should == "\u03D2\u0308"
+      "\u03D4".unicode_normalize(:nfkc).should == "\u03AB"
+      "\u03D4".unicode_normalize(:nfkd).should == "\u03A5\u0308"
+    end
+
+    it "1E9B (ẛ) LATIN SMALL LETTER LONG S WITH DOT ABOVE" do
+      "\u1E9B".unicode_normalize(:nfc).should == "\u1E9B"
+      "\u1E9B".unicode_normalize(:nfd).should == "\u017F\u0307"
+      "\u1E9B".unicode_normalize(:nfkc).should == "\u1E61"
+      "\u1E9B".unicode_normalize(:nfkd).should == "\u0073\u0307"
+    end
+  end
+
+  it "raises an Encoding::CompatibilityError if string is not in an unicode encoding" do
+    lambda do
+      [0xE0].pack('C').force_encoding("ISO-8859-1").unicode_normalize(:nfd)
+    end.should raise_error(Encoding::CompatibilityError)
+  end
+
+  it "raises an ArgumentError if the specified form is invalid" do
+    lambda {
+      @angstrom.unicode_normalize(:invalid_form)
+    }.should raise_error(ArgumentError)
+  end
+end
+
+describe "String#unicode_normalize!" do
+  it "normalizes code points and modifies the receiving string" do
+    angstrom = "\u212b"
+    angstrom.unicode_normalize!
+    angstrom.should == "\u00c5"
+    angstrom.should_not == "\u212b"
+  end
+
+  it "modifies original string (nfc)" do
+    str = "a\u0300"
+    str.unicode_normalize!(:nfc)
+
+    str.should_not == "a\u0300"
+    str.should == "à"
+  end
+
+  it "modifies self in place (nfd)" do
+    str = "\u00E0"
+    str.unicode_normalize!(:nfd)
+
+    str.should_not == "\u00E0"
+    str.should == "a\u0300"
+  end
+
+  it "modifies self in place (nfkc)" do
+    str = "\u1E9B\u0323"
+    str.unicode_normalize!(:nfkc)
+
+    str.should_not == "\u1E9B\u0323"
+    str.should == "\u1E69"
+  end
+
+  it "modifies self in place (nfkd)" do
+    str = "\u1E9B\u0323"
+    str.unicode_normalize!(:nfkd)
+
+    str.should_not == "\u1E9B\u0323"
+    str.should == "s\u0323\u0307"
+  end
+
+  it "raises an Encoding::CompatibilityError if the string is not in an unicode encoding" do
+    lambda {
+      [0xE0].pack('C').force_encoding("ISO-8859-1").unicode_normalize!
+    }.should raise_error(Encoding::CompatibilityError)
+  end
+
+  it "raises an ArgumentError if the specified form is invalid" do
+    ohm = "\u2126"
+    lambda {
+      ohm.unicode_normalize!(:invalid_form)
+    }.should raise_error(ArgumentError)
+  end
+end