premisrw: always return unicode in text getter

In metsrw 0.2.4 or older, premisrw was returning:

  - In Py2: binary (bytes)
  - In Py3: text (unicode)

In metsrw 0.3.0, the Py2 code path was updated to explicitly use the `utf-8`
codec since the default is `ascii` which is usually a problem during encoding.

This commit updates the same code path again so in Py2 we also return unicode
making the API more uniform.

metsrw/plugins/premisrw/premis.py

@@ -699,10 +699,13 @@ def data_find_text(data, path):
              if not isinstance(child, (tuple, list, dict))]
     if not texts:
         return None
-    if six.PY2:
-        return ' '.join(
-            [x.encode('utf-8', errors='ignore') for x in texts])
-    return ' '.join([str(x) for x in texts])
+    return ' '.join([
+        # How should we deal with decoding errors when `x` is binary?
+        # For now, we're using the ``strict`` mode. Other options here:
+        # https://docs.python.org/3/library/functions.html#open.
+        six.ensure_text(x, encoding='utf-8', errors='strict')
+        for x in texts
+    ])
 
 
 def data_find_text_or_all(data, path, dyn_cls=False):

tests/plugins/premisrw/test_premis.py

@@ -3,6 +3,7 @@
 
 import pytest
 from lxml import etree
+import six
 
 import metsrw
 import metsrw.plugins.premisrw as premisrw
@@ -36,6 +37,38 @@ def test_roundtrip_unicode(self):
         assert data[2][1][1] == u'𝕡𝕣𝕖𝕤𝕖𝕣𝕧𝕒𝕥𝕚𝕠𝕟 𝕤𝕪𝕤𝕥𝕖𝕞'
         assert data[2][2][1] == u'𝓊𝓃𝒾𝒸𝑜𝒹𝑒'
 
+    @pytest.mark.skipif(six.PY3, reason='lxml in py3 does not accept binary')
+    def test_roundtrip_unicode_from_binary(self):
+        """Test that premisrw returns unicode values in all cases."""
+        lxml_el = premisrw.data_to_premis((
+            'agent',
+            premisrw.PREMIS_META,
+            (
+                'agent_identifier',
+                ('agent_identifier_type', b'foo'),
+                ('agent_identifier_value', b'bar'),
+            )
+        ))
+        data = premisrw.premis_to_data(lxml_el)
+        assert data[2][1][1] == u'foo'
+        assert data[2][2][1] == u'bar'
+
+    @pytest.mark.skipif(six.PY3, reason='lxml in py3 does not accept binary')
+    def test_with_invalid_binary(self):
+        """Test lxml's ``ValueError`` with invalid byte sequences."""
+        invalid_sequence = b'\x78\x9a\xbc\xde\xf0'
+        data = ((
+            'agent',
+            premisrw.PREMIS_META,
+            (
+                'agent_identifier',
+                ('agent_identifier_type', b'type'),
+                ('agent_identifier_value', invalid_sequence),
+            )
+        ))
+        with pytest.raises(ValueError):
+            premisrw.data_to_premis(data)
+
     def test_premis_event_cls_data(self):
         """Tests that you can pass a Python tuple as the ``data`` argument to
         ``PREMISEvent`` to construct an instance.