[pypy-commit] pypy py3k: add a new space method to unwrap unicode identifiers into UTF-8 encoded strings

[antocuni]

Author: Antonio Cuni <anto.cuni at gmail.com>
Branch: py3k
Changeset: r57043:259f31766894
Date: 2012-08-31 09:47 +0200
http://bitbucket.org/pypy/pypy/changeset/259f31766894/

Log:	add a new space method to unwrap unicode identifiers into UTF-8
	encoded strings

diff --git a/pypy/interpreter/baseobjspace.py b/pypy/interpreter/baseobjspace.py
--- a/pypy/interpreter/baseobjspace.py
+++ b/pypy/interpreter/baseobjspace.py
@@ -1371,6 +1371,14 @@
                                  self.wrap('argument must be a unicode'))
         return self.unicode_w(w_obj)
 
+    def identifier_w(self, w_obj):
+        """
+        Unwrap an object which is used as an identifier (i.e. names of
+        variables, methdods, functions, classes etc.). In py3k, identifiers
+        are unicode strings and are unwrapped as UTF-8 encoded byte strings.
+        """
+        return self.unicode_w(w_obj).encode('utf-8')
+
     def bool_w(self, w_obj):
         # Unwraps a bool, also accepting an int for compatibility.
         # This is here mostly just for gateway.int_unwrapping_space_method().
diff --git a/pypy/interpreter/test/test_objspace.py b/pypy/interpreter/test/test_objspace.py
--- a/pypy/interpreter/test/test_objspace.py
+++ b/pypy/interpreter/test/test_objspace.py
@@ -1,3 +1,4 @@
+# -*- encoding: utf-8 -*-
 from py.test import raises
 from pypy.interpreter.error import OperationError
 from pypy.interpreter.function import Function
@@ -179,6 +180,12 @@
         assert space.unicode0_w(w(u"123")) == u"123"
         exc = space.raises_w(space.w_TypeError, space.unicode0_w, w(u"123\x004"))
 
+    def test_identifier_w(self):
+        space = self.space
+        x = u'&#224;&#232;&#236;'
+        w_name = space.wrap(x)
+        assert space.identifier_w(w_name) == x.encode('utf-8')
+
     def test_getindex_w(self):
         w_instance1 = self.space.appexec([], """():
             class X(object):