Implement \u notation for strings and regexes.

This change adds the \u notation for double quoted strings and regular
expressions.  It does not implement the \u notation for character literals.
Both the \uNNNN and \u{NNNN} notations are supported.

\uNNNN is implemented by emitting equivalent UTF-8; that is, "\u4000" is
equivalent to "\xE4\x80\x80".

Unlike CRuby, the \u{NNNN} notation allows only one character per pair of
braces; I see no way to lift this restriction without remodeling the parser.

1 files changed, 60 insertions, 0 deletions

diff --git a/test/t/unicode.rb b/test/t/unicode.rb
new file mode 100644
index 000000000..8f94421e6
--- /dev/null
+++ b/test/t/unicode.rb
@@ -0,0 +1,60 @@
+# Test of the \u notation
+
+assert('bare \u notation test') do
+  # Mininum and maximum one byte characters
+  assert_equal("\u0000", "\x00")
+  assert_equal("\u007F", "\x7F")
+
+  # Mininum and maximum two byte characters
+  assert_equal("\u0080", "\xC2\x80")
+  assert_equal("\u07FF", "\xDF\xBF")
+
+  # Mininum and maximum three byte characters
+  assert_equal("\u0800", "\xE0\xA0\x80")
+  assert_equal("\uFFFF", "\xEF\xBF\xBF")
+
+  # Four byte characters require the \U notation
+end
+
+assert('braced \u notation test') do
+  # Mininum and maximum one byte characters
+  assert_equal("\u{0000}", "\x00")
+  assert_equal("\u{007F}", "\x7F")
+
+  # Mininum and maximum two byte characters
+  assert_equal("\u{0080}", "\xC2\x80")
+  assert_equal("\u{07FF}", "\xDF\xBF")
+
+  # Mininum and maximum three byte characters
+  assert_equal("\u{0800}", "\xE0\xA0\x80")
+  assert_equal("\u{FFFF}", "\xEF\xBF\xBF")
+
+  # Mininum and maximum four byte characters
+  assert_equal("\u{10000}",  "\xF0\x90\x80\x80")
+  assert_equal("\u{10FFFF}", "\xF4\x8F\xBF\xBF")
+end
+
+# Test regular expressions only if implemented
+begin
+  Regexp
+  have_regexp = true
+rescue NameError
+  have_regexp = false
+end
+if have_regexp then
+  assert('Testing in regular expressions') do
+    # The regular expression uses the unbraced notation where the string uses
+    # the braced notation, and vice versa, so these tests will fail if the \u
+    # modification is not applied
+
+    # Test of unbraced \u notation in a regular expression
+    assert_false(/\u0300/ =~ "\u{02FF}")
+    assert_true( /\u0300/ =~ "\u{0300}")
+    assert_false(/\u0300/ =~ "\u{0301}")
+
+    # Test of braced \u notation in a regular expression
+    assert_false(/\u{0300}/ =~ "\u02FF")
+    assert_true( /\u{0300}/ =~ "\u0300")
+    assert_false(/\u{0300}/ =~ "\u0301")
+  end
+end