gpg_get_key_addresses: fix handling non-UTF-8 uid

The Python 3 transition introduced a regression in
2f95fde0 which assumed all of
gpg's output is UTF-8.

Closes: #980182

daklib/utils.py

@@ -714,11 +714,11 @@ def gpg_keyring_args(keyrings=None):
 def _gpg_get_addresses_from_listing(output: bytes):
     addresses = []
 
-    for l in six.ensure_str(output).split('\n'):
-        parts = l.split(':')
-        if parts[0] not in ("uid", "pub"):
+    for line in output.split(b'\n'):
+        parts = line.split(b':')
+        if parts[0] not in (b"uid", b"pub"):
             continue
-        if parts[1] in ("i", "d", "r"):
+        if parts[1] in (b"i", b"d", b"r"):
             # Skip uid that is invalid, disabled or revoked
             continue
         try:
@@ -726,9 +726,11 @@ def _gpg_get_addresses_from_listing(output: bytes):
         except IndexError:
             continue
         try:
-            uid = six.ensure_text(uid)
+            uid = uid.decode(encoding='utf-8')
         except UnicodeDecodeError:
-            uid = uid.decode("latin1") # does not fail
+            # If the uid is not valid UTF-8, we assume it is an old uid
+            # still encoding in Latin-1.
+            uid = uid.decode(encoding='latin1')
         m = re_parse_maintainer.match(uid)
         if not m:
             continue

tests/test_utils.py

@@ -23,6 +23,7 @@ from base_test import DakTestCase
 from daklib.utils import (is_in_debug_section,
                           parse_built_using,
                           extract_component_from_section,
+                          _gpg_get_addresses_from_listing,
                           ArchKey)
 
 apt_pkg.init()
@@ -120,6 +121,22 @@ class UtilsTest(DakTestCase):
         for v, r in data:
             self.assertEqual(extract_component_from_section(v), r)
 
+    def test_gpg_get_addresses_from_listing(self):
+        # Test with output containing a uid encoded in Latin-1 and UTF-8 each
+        data = (
+            b"tru::1:1610798976:1673870895:3:1:5\n"
+            b"pub:u:3072:1:4847924E3DEDDF1E:1610798895:1673870895::u:::scESC::::::23::0:\n"
+            b"fpr:::::::::AC6DFC4B78223BFCC1C064004847924E3DEDDF1E:\n"
+            b"uid:u::::1610798964::1D4AB6BD4AB8A446C196A1801F367B83B6141CC0::Markus Mustermann-S\xf6der <mustermann-soeder@example.org>::::::::::0:\n"
+            b"uid:u::::1610798895::F60CF610557A38E2943DECEEA5B02EE57B1ECC24::Markus Mustermann-S\xc3\xb6der <mustermann-soeder@example.com>::::::::::0:\n"
+            b"sub:u:3072:1:C3BCC6A42035DDB3:1610798895:1673870895:::::e::::::23:\n"
+            b"fpr:::::::::C2CC5D5744746B5E7F5F7286C3BCC6A42035DDB3:\n"
+        )
+
+        addresses = _gpg_get_addresses_from_listing(data)
+        expected = ['mustermann-soeder@example.org', 'mustermann-soeder@example.com']
+        self.assertEqual(addresses, expected)
+
 
 if __name__ == '__main__':
     unittest.main()