[pypy-commit] pypy default: let host ord handle non-surrogate cases so we get proper exceptions, test

[bdkearns]

Author: Brian Kearns <bdkearns at gmail.com>
Branch: 
Changeset: r60902:093fe9b62735
Date: 2013-02-06 07:06 -0500
http://bitbucket.org/pypy/pypy/changeset/093fe9b62735/

Log:	let host ord handle non-surrogate cases so we get proper exceptions,
	test

diff --git a/rpython/rlib/runicode.py b/rpython/rlib/runicode.py
--- a/rpython/rlib/runicode.py
+++ b/rpython/rlib/runicode.py
@@ -31,9 +31,7 @@
         ch2 = ord(u[1])
         if 0xD800 <= ch1 <= 0xDBFF and 0xDC00 <= ch2 <= 0xDFFF:
             return (((ch1 - 0xD800) << 10) | (ch2 - 0xDC00)) + 0x10000
-    if len(u) != 1:
-        raise ValueError('not a surrogate')
-    return ord(u[0])
+    return ord(u)
 
 if MAXUNICODE > sys.maxunicode:
     # A version of unichr which allows codes outside the BMP
diff --git a/rpython/rlib/test/test_runicode.py b/rpython/rlib/test/test_runicode.py
--- a/rpython/rlib/test/test_runicode.py
+++ b/rpython/rlib/test/test_runicode.py
@@ -5,16 +5,26 @@
 from rpython.rlib import runicode
 
 def test_unichr():
-    a = runicode.UNICHR(0xffff)
-    assert a == u'\uffff'
+    assert runicode.UNICHR(0xffff) == u'\uffff'
     if runicode.MAXUNICODE > 0xffff:
-        a = runicode.UNICHR(0x10000)
         if sys.maxunicode < 0x10000:
-            assert len(a) == 2      # surrogates
+            assert runicode.UNICHR(0x10000) == u'\ud800\udc00'
         else:
-            assert len(a) == 1
+            assert runicode.UNICHR(0x10000) == u'\U00010000'
     else:
         py.test.raises(ValueError, runicode.UNICHR, 0x10000)
+    py.test.raises(TypeError, runicode.UNICHR, 'abc')
+
+def test_ord():
+    assert runicode.ORD(u'\uffff') == 0xffff
+    if runicode.MAXUNICODE > 0xffff:
+        if sys.maxunicode < 0x10000:
+            assert runicode.ORD(u'\ud800\udc00') == 0x10000
+        else:
+            assert runicode.ORD(u'\U00010000') == 0x10000
+    else:
+        py.test.raises(TypeError, runicode.ORD, u'\ud800\udc00')
+    py.test.raises(TypeError, runicode.ORD, 'abc')
 
 
 class UnicodeTests(object):