#11554: reactivate test_email_codecs, and make it pass.

The fix is to charset.py, which was not doing the encoding to the
correct output character set when doing a body_encode for either
the shift-jis or euc-jp charsets.  There's also a fix for handling
a bytes input in encoders.py.

Patch by Michael Henry, comment changes by me.

4 files changed, 46 insertions, 22 deletions

diff --git a/Lib/email/charset.py b/Lib/email/charset.py
index 24d5545..f22be2c 100644
--- a/Lib/email/charset.py
+++ b/Lib/email/charset.py
@@ -263,7 +263,7 @@ class Charset:
 
         Returns "quoted-printable" if self.body_encoding is QP.
         Returns "base64" if self.body_encoding is BASE64.
-        Returns "7bit" otherwise.
+        Returns conversion function otherwise.
         """
         assert self.body_encoding != SHORTEST
         if self.body_encoding == QP:
@@ -381,7 +381,10 @@ class Charset:
         """Body-encode a string by converting it first to bytes.
 
         The type of encoding (base64 or quoted-printable) will be based on
-        self.body_encoding.
+        self.body_encoding.  If body_encoding is None, we assume the
+        output charset is a 7bit encoding, so re-encoding the decoded
+        string using the ascii codec produces the correct string version
+        of the content.
         """
         # 7bit/8bit encodings return the string unchanged (module conversions)
         if self.body_encoding is BASE64:
@@ -391,4 +394,6 @@ class Charset:
         elif self.body_encoding is QP:
             return email.quoprimime.body_encode(string)
         else:
+            if isinstance(string, str):
+                string = string.encode(self.output_charset).decode('ascii')
             return string
diff --git a/Lib/email/encoders.py b/Lib/email/encoders.py
index 0ea441d..c66f4cc 100644
--- a/Lib/email/encoders.py
+++ b/Lib/email/encoders.py
@@ -54,10 +54,13 @@ def encode_7or8bit(msg):
         # There's no payload.  For backwards compatibility we use 7bit
         msg['Content-Transfer-Encoding'] = '7bit'
         return
-    # We play a trick to make this go fast.  If encoding to ASCII succeeds, we
-    # know the data must be 7bit, otherwise treat it as 8bit.
+    # We play a trick to make this go fast.  If encoding/decode to ASCII
+    # succeeds, we know the data must be 7bit, otherwise treat it as 8bit.
     try:
-        orig.encode('ascii')
+        if isinstance(orig, str):
+            orig.encode('ascii')
+        else:
+            orig.decode('ascii')
     except UnicodeError:
         # iso-2022-* is non-ASCII but still 7-bit
         charset = msg.get_charset()
diff --git a/Lib/email/test/test_email.py b/Lib/email/test/test_email.py
index f8bdaa2..a6ea1d1 100644
--- a/Lib/email/test/test_email.py
+++ b/Lib/email/test/test_email.py
@@ -3365,9 +3365,9 @@ class TestCharset(unittest.TestCase):
         # built-in encodings where the header encoding is QP but the body
         # encoding is not.
         from email import charset as CharsetModule
-        CharsetModule.add_charset('fake', CharsetModule.QP, None)
+        CharsetModule.add_charset('fake', CharsetModule.QP, None, 'utf-8')
         c = Charset('fake')
-        eq('hello w\xf6rld', c.body_encode('hello w\xf6rld'))
+        eq('hello world', c.body_encode('hello world'))
 
     def test_unicode_charset_name(self):
         charset = Charset('us-ascii')
diff --git a/Lib/email/test/test_email_codecs.py b/Lib/email/test/test_email_codecs.py
index acc19c3..ca85f57 100644
--- a/Lib/email/test/test_email_codecs.py
+++ b/Lib/email/test/test_email_codecs.py
@@ -13,7 +13,7 @@ from email.message import Message
 # We're compatible with Python 2.3, but it doesn't have the built-in Asian
 # codecs, so we have to skip all these tests.
 try:
-    str('foo', 'euc-jp')
+    str(b'foo', 'euc-jp')
 except LookupError:
     raise unittest.SkipTest
 
@@ -22,11 +22,14 @@ except LookupError:
 class TestEmailAsianCodecs(TestEmailBase):
     def test_japanese_codecs(self):
         eq = self.ndiffAssertEqual
-        j = Charset("euc-jp")
-        g = Charset("iso-8859-1")
+        jcode = "euc-jp"
+        gcode = "iso-8859-1"
+        j = Charset(jcode)
+        g = Charset(gcode)
         h = Header("Hello World!")
-        jhello = '\xa5\xcf\xa5\xed\xa1\xbc\xa5\xef\xa1\xbc\xa5\xeb\xa5\xc9\xa1\xaa'
-        ghello = 'Gr\xfc\xdf Gott!'
+        jhello = str(b'\xa5\xcf\xa5\xed\xa1\xbc\xa5\xef\xa1\xbc'
+                     b'\xa5\xeb\xa5\xc9\xa1\xaa', jcode)
+        ghello = str(b'Gr\xfc\xdf Gott!', gcode)
         h.append(jhello, j)
         h.append(ghello, g)
         # BAW: This used to -- and maybe should -- fold the two iso-8859-1
@@ -36,13 +39,17 @@ class TestEmailAsianCodecs(TestEmailBase):
         # encoded word.
         eq(h.encode(), """\
 Hello World! =?iso-2022-jp?b?GyRCJU8lbSE8JW8hPCVrJUkhKhsoQg==?=
- =?iso-8859-1?q?Gr=FC=DF?= =?iso-8859-1?q?_Gott!?=""")
+ =?iso-8859-1?q?Gr=FC=DF_Gott!?=""")
         eq(decode_header(h.encode()),
-           [('Hello World!', None),
-            ('\x1b$B%O%m!<%o!<%k%I!*\x1b(B', 'iso-2022-jp'),
-            ('Gr\xfc\xdf Gott!', 'iso-8859-1')])
-        int = 'test-ja \xa4\xd8\xc5\xea\xb9\xc6\xa4\xb5\xa4\xec\xa4\xbf\xa5\xe1\xa1\xbc\xa5\xeb\xa4\xcf\xbb\xca\xb2\xf1\xbc\xd4\xa4\xce\xbe\xb5\xc7\xa7\xa4\xf2\xc2\xd4\xa4\xc3\xa4\xc6\xa4\xa4\xa4\xde\xa4\xb9'
-        h = Header(int, j, header_name="Subject")
+           [(b'Hello World!', None),
+            (b'\x1b$B%O%m!<%o!<%k%I!*\x1b(B', 'iso-2022-jp'),
+            (b'Gr\xfc\xdf Gott!', gcode)])
+        subject_bytes = (b'test-ja \xa4\xd8\xc5\xea\xb9\xc6\xa4\xb5'
+            b'\xa4\xec\xa4\xbf\xa5\xe1\xa1\xbc\xa5\xeb\xa4\xcf\xbb\xca\xb2'
+            b'\xf1\xbc\xd4\xa4\xce\xbe\xb5\xc7\xa7\xa4\xf2\xc2\xd4\xa4\xc3'
+            b'\xa4\xc6\xa4\xa4\xa4\xde\xa4\xb9')
+        subject = str(subject_bytes, jcode)
+        h = Header(subject, j, header_name="Subject")
         # test a very long header
         enc = h.encode()
         # TK: splitting point may differ by codec design and/or Header encoding
@@ -50,15 +57,24 @@ Hello World! =?iso-2022-jp?b?GyRCJU8lbSE8JW8hPCVrJUkhKhsoQg==?=
 =?iso-2022-jp?b?dGVzdC1qYSAbJEIkWEVqOUYkNSRsJD8lYSE8JWskTztKGyhC?=
  =?iso-2022-jp?b?GyRCMnE8VCROPjVHJyRyQlQkQyRGJCQkXiQ5GyhC?=""")
         # TK: full decode comparison
-        eq(h.__unicode__().encode('euc-jp'), int)
+        eq(str(h).encode(jcode), subject_bytes)
+
+    def test_payload_encoding_utf8(self):
+        jhello = str(b'\xa5\xcf\xa5\xed\xa1\xbc\xa5\xef\xa1\xbc'
+                     b'\xa5\xeb\xa5\xc9\xa1\xaa', 'euc-jp')
+        msg = Message()
+        msg.set_payload(jhello, 'utf-8')
+        ustr = msg.get_payload(decode=True).decode(msg.get_content_charset())
+        self.assertEqual(jhello, ustr)
 
     def test_payload_encoding(self):
-        jhello = '\xa5\xcf\xa5\xed\xa1\xbc\xa5\xef\xa1\xbc\xa5\xeb\xa5\xc9\xa1\xaa'
         jcode  = 'euc-jp'
+        jhello = str(b'\xa5\xcf\xa5\xed\xa1\xbc\xa5\xef\xa1\xbc'
+                     b'\xa5\xeb\xa5\xc9\xa1\xaa', jcode)
         msg = Message()
         msg.set_payload(jhello, jcode)
-        ustr = str(msg.get_payload(), msg.get_content_charset())
-        self.assertEqual(jhello, ustr.encode(jcode))
+        ustr = msg.get_payload(decode=True).decode(msg.get_content_charset())
+        self.assertEqual(jhello, ustr)