Issue 3739: The unicode-internal encoder now reports the number of *characters*
consumed like any other encoder (instead of the number of bytes).
diff --git a/Lib/test/test_codecs.py b/Lib/test/test_codecs.py
index cee819c..5df6fe5 100644
--- a/Lib/test/test_codecs.py
+++ b/Lib/test/test_codecs.py
@@ -802,6 +802,12 @@
"UnicodeInternalTest")
self.assertEquals((u"ab", 12), ignored)
+ def test_encode_length(self):
+ # Issue 3739
+ encoder = codecs.getencoder("unicode_internal")
+ self.assertEquals(encoder(u"a")[1], 1)
+ self.assertEquals(encoder(u"\xe9\u0142")[1], 2)
+
# From http://www.gnu.org/software/libidn/draft-josefsson-idn-test-vectors.html
nameprep_tests = [
# 3.1 Map to nothing.
@@ -1292,8 +1298,7 @@
name = "latin_1"
self.assertEqual(encoding.replace("_", "-"), name.replace("_", "-"))
(bytes, size) = codecs.getencoder(encoding)(s)
- if encoding != "unicode_internal":
- self.assertEqual(size, len(s), "%r != %r (encoding=%r)" % (size, len(s), encoding))
+ self.assertEqual(size, len(s), "%r != %r (encoding=%r)" % (size, len(s), encoding))
(chars, size) = codecs.getdecoder(encoding)(bytes)
self.assertEqual(chars, s, "%r != %r (encoding=%r)" % (chars, s, encoding))