Merge pull request #1092 from nicoddemus/param-ids-fix

Param ids fix

CHANGELOG

@@ -1,6 +1,14 @@
 2.8.2.dev
 ---------
 
+- fix #1085: proper handling of encoding errors when passing encoded byte
+  strings to pytest.parametrize in Python 2.
+  Thanks Themanwithoutaplan for the report and Bruno Oliveira for the PR.
+
+- fix #1087: handling SystemError when passing empty byte strings to
+  pytest.parametrize in Python 3.
+  Thanks Paul Kehrer for the report and Bruno Oliveira for the PR.
+
 - fix #995: fixed internal error when filtering tracebacks where one entry
   was generated by an exec() statement.
   Thanks Daniel Hahler, Ashley C Straw, Philippe Gauthier and Pavel Savchenko

_pytest/python.py

@@ -1048,6 +1048,8 @@ class Metafunc(FuncargnamesCompatAttr):
 
 
 if _PY3:
+    import codecs
+
     def _escape_bytes(val):
         """
         If val is pure ascii, returns it as a str(), otherwise escapes
@@ -1060,18 +1062,21 @@ if _PY3:
            want to return escaped bytes for any byte, even if they match
            a utf-8 string.
         """
+        if val:
             # source: http://goo.gl/bGsnwC
-        import codecs
             encoded_bytes, _ = codecs.escape_encode(val)
             return encoded_bytes.decode('ascii')
+        else:
+            # empty bytes crashes codecs.escape_encode (#1087)
+            return ''
 else:
     def _escape_bytes(val):
         """
-        In py2 bytes and str are the same, so return it unchanged if it
+        In py2 bytes and str are the same type, so return it unchanged if it
         is a full ascii string, otherwise escape it into its binary form.
         """
         try:
-            return val.encode('ascii')
+            return val.decode('ascii')
         except UnicodeDecodeError:
             return val.encode('string-escape')
 
@@ -1100,7 +1105,7 @@ def _idval(val, argname, idx, idfn):
         # convertible to ascii, return it as an str() object instead
         try:
             return str(val)
-        except UnicodeDecodeError:
+        except UnicodeError:
             # fallthrough
             pass
     return str(argname)+str(idx)

testing/python/metafunc.py

@@ -1,3 +1,4 @@
+# -*- coding: utf-8 -*-
 import re
 
 import pytest, py
@@ -118,6 +119,41 @@ class TestMetafunc:
         assert metafunc._calls[2].id == "x1-a"
         assert metafunc._calls[3].id == "x1-b"
 
+    @pytest.mark.skipif('sys.version_info[0] >= 3')
+    def test_unicode_idval_python2(self):
+        """unittest for the expected behavior to obtain ids for parametrized
+        unicode values in Python 2: if convertible to ascii, they should appear
+        as ascii values, otherwise fallback to hide the value behind the name
+        of the parametrized variable name. #1086
+        """
+        from _pytest.python import _idval
+        values = [
+            (u'', ''),
+            (u'ascii', 'ascii'),
+            (u'ação', 'a6'),
+            (u'josé@blah.com', 'a6'),
+            (u'δοκ.ιμή@παράδειγμα.δοκιμή', 'a6'),
+        ]
+        for val, expected in values:
+            assert _idval(val, 'a', 6, None) == expected
+
+    def test_bytes_idval(self):
+        """unittest for the expected behavior to obtain ids for parametrized
+        bytes values:
+        - python2: non-ascii strings are considered bytes and formatted using
+        "binary escape", where any byte < 127 is escaped into its hex form.
+        - python3: bytes objects are always escaped using "binary escape".
+        """
+        from _pytest.python import _idval
+        values = [
+            (b'', ''),
+            (b'\xc3\xb4\xff\xe4', '\\xc3\\xb4\\xff\\xe4'),
+            (b'ascii', 'ascii'),
+            (u'αρά'.encode('utf-8'), '\\xce\\xb1\\xcf\\x81\\xce\\xac'),
+        ]
+        for val, expected in values:
+            assert _idval(val, 'a', 6, None) == expected
+
     @pytest.mark.issue250
     def test_idmaker_autoname(self):
         from _pytest.python import idmaker