Merge pull request #1470 from ceridwen/features

Escape both bytes and unicode strings for "ids" in Metafunc.parametrize

Author: nicoddemus

.gitignore

@@ -32,3 +32,4 @@ env/
 .coverage
 .ropeproject
 .idea
+.hypothesis

CHANGELOG.rst

@@ -27,6 +27,10 @@
 
 **Changes**
 
+* Fix (`#1351`_):
+  explicitly passed parametrize ids do not get escaped to ascii.
+  Thanks `@ceridwen`_ for the PR.
+
 * parametrize ids can accept None as specific test id. The
   automatically generated id for that argument will be used.
   Thanks `@palaviv`_ for the complete PR (`#1468`_).
@@ -41,19 +45,20 @@
 .. _@novas0x2a: https://github.com/novas0x2a
 .. _@kalekundert: https://github.com/kalekundert
 .. _@tareqalayan: https://github.com/tareqalayan
+.. _@ceridwen: https://github.com/ceridwen
 .. _@palaviv: https://github.com/palaviv
 .. _@omarkohl: https://github.com/omarkohl
 
 .. _#1428: https://github.com/pytest-dev/pytest/pull/1428
 .. _#1444: https://github.com/pytest-dev/pytest/pull/1444
 .. _#1441: https://github.com/pytest-dev/pytest/pull/1441
 .. _#1454: https://github.com/pytest-dev/pytest/pull/1454
+.. _#1351: https://github.com/pytest-dev/pytest/issues/1351
 .. _#1468: https://github.com/pytest-dev/pytest/pull/1468
 .. _#1474: https://github.com/pytest-dev/pytest/pull/1474
 .. _#1502: https://github.com/pytest-dev/pytest/pull/1502
 .. _#372: https://github.com/pytest-dev/pytest/issues/372
 
-
 2.9.2.dev1
 ==========
 

_pytest/python.py

@@ -1079,67 +1079,76 @@ def addcall(self, funcargs=None, id=_notexists, param=_notexists):
         self._calls.append(cs)
 
 
+
 if _PY3:
     import codecs
 
-    def _escape_bytes(val):
-        """
-        If val is pure ascii, returns it as a str(), otherwise escapes
-        into a sequence of escaped bytes:
+    def _escape_strings(val):
+        """If val is pure ascii, returns it as a str().  Otherwise, escapes
+        bytes objects into a sequence of escaped bytes:
+
         b'\xc3\xb4\xc5\xd6' -> u'\\xc3\\xb4\\xc5\\xd6'
 
+        and escapes unicode objects into a sequence of escaped unicode
+        ids, e.g.:
+
+        '4\\nV\\U00043efa\\x0eMXWB\\x1e\\u3028\\u15fd\\xcd\\U0007d944'
+
         note:
            the obvious "v.decode('unicode-escape')" will return
-           valid utf-8 unicode if it finds them in the string, but we
+           valid utf-8 unicode if it finds them in bytes, but we
            want to return escaped bytes for any byte, even if they match
            a utf-8 string.
+
         """
-        if val:
-            # source: http://goo.gl/bGsnwC
-            encoded_bytes, _ = codecs.escape_encode(val)
-            return encoded_bytes.decode('ascii')
+        if isinstance(val, bytes):
+            if val:
+                # source: http://goo.gl/bGsnwC
+                encoded_bytes, _ = codecs.escape_encode(val)
+                return encoded_bytes.decode('ascii')
+            else:
+                # empty bytes crashes codecs.escape_encode (#1087)
+                return ''
         else:
-            # empty bytes crashes codecs.escape_encode (#1087)
-            return ''
+            return val.encode('unicode_escape').decode('ascii')
 else:
-    def _escape_bytes(val):
-        """
-        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.
+    def _escape_strings(val):
+        """In py2 bytes and str are the same type, so return if it's a bytes
+        object, return it unchanged if it is a full ascii string,
+        otherwise escape it into its binary form.
+
+        If it's a unicode string, change the unicode characters into
+        unicode escapes.
+
         """
-        try:
-            return val.decode('ascii')
-        except UnicodeDecodeError:
-            return val.encode('string-escape')
+        if isinstance(val, bytes):
+            try:
+                return val.encode('ascii')
+            except UnicodeDecodeError:
+                return val.encode('string-escape')
+        else:
+            return val.encode('unicode-escape')
 
 
 def _idval(val, argname, idx, idfn):
     if idfn:
         try:
             s = idfn(val)
             if s:
-                return s
+                return _escape_strings(s)
         except Exception:
             pass
 
-    if isinstance(val, bytes):
-        return _escape_bytes(val)
-    elif isinstance(val, (float, int, str, bool, NoneType)):
+    if isinstance(val, (bytes, str)) or (_PY2 and isinstance(val, unicode)):
+        return _escape_strings(val)
+    elif isinstance(val, (float, int, bool, NoneType)):
         return str(val)
     elif isinstance(val, REGEX_TYPE):
-        return _escape_bytes(val.pattern) if isinstance(val.pattern, bytes) else val.pattern
+        return _escape_strings(val.pattern)
     elif enum is not None and isinstance(val, enum.Enum):
         return str(val)
     elif isclass(val) and hasattr(val, '__name__'):
         return val.__name__
-    elif _PY2 and isinstance(val, unicode):
-        # special case for python 2: if a unicode string is
-        # convertible to ascii, return it as an str() object instead
-        try:
-            return str(val)
-        except UnicodeError:
-            # fallthrough
-            pass
     return str(argname)+str(idx)
 
 def _idvalset(idx, valset, argnames, idfn, ids):
@@ -1148,7 +1157,7 @@ def _idvalset(idx, valset, argnames, idfn, ids):
                    for val, argname in zip(valset, argnames)]
         return "-".join(this_id)
     else:
-        return ids[idx]
+        return _escape_strings(ids[idx])
 
 def idmaker(argnames, argvalues, idfn=None, ids=None):
     ids = [_idvalset(valindex, valset, argnames, idfn, ids)

testing/python/metafunc.py

@@ -1,11 +1,18 @@
 # -*- coding: utf-8 -*-
 import re
+import sys
 
 import _pytest._code
 import py
 import pytest
 from _pytest import python as funcargs
 
+import hypothesis
+from hypothesis import strategies
+
+PY3 = sys.version_info >= (3, 0)
+
+
 class TestMetafunc:
     def Metafunc(self, func):
         # the unit tests of this class check if things work correctly
@@ -121,20 +128,29 @@ class A:
         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
+    @hypothesis.given(strategies.text() | strategies.binary())
+    def test_idval_hypothesis(self, value):
+        from _pytest.python import _idval
+        escaped = _idval(value, 'a', 6, None)
+        assert isinstance(escaped, str)
+        if PY3:
+            escaped.encode('ascii')
+        else:
+            escaped.decode('ascii')
+
+    def test_unicode_idval(self):
+        """This tests that Unicode strings outside the ASCII character set get
+        escaped, using byte escapes if they're in that range or unicode
+        escapes if they're not.
+
         """
         from _pytest.python import _idval
         values = [
             (u'', ''),
             (u'ascii', 'ascii'),
-            (u'ação', 'a6'),
-            (u'josé@blah.com', 'a6'),
-            (u'δοκ.ιμή@παράδειγμα.δοκιμή', 'a6'),
+            (u'ação', 'a\\xe7\\xe3o'),
+            (u'josé@blah.com', 'jos\\[email protected]'),
+            (u'δοκ.ιμή@παράδειγμα.δοκιμή', '\\u03b4\\u03bf\\u03ba.\\u03b9\\u03bc\\u03ae@\\u03c0\\u03b1\\u03c1\\u03ac\\u03b4\\u03b5\\u03b9\\u03b3\\u03bc\\u03b1.\\u03b4\\u03bf\\u03ba\\u03b9\\u03bc\\u03ae'),
         ]
         for val, expected in values:
             assert _idval(val, 'a', 6, None) == expected

testing/test_junitxml.py

@@ -610,14 +610,14 @@ def test_pass():
 def test_escaped_parametrized_names_xml(testdir):
     testdir.makepyfile("""
         import pytest
-        @pytest.mark.parametrize('char', ["\\x00"])
+        @pytest.mark.parametrize('char', [u"\\x00"])
         def test_func(char):
             assert char
     """)
     result, dom = runandparse(testdir)
     assert result.ret == 0
     node = dom.find_first_by_tag("testcase")
-    node.assert_attr(name="test_func[#x00]")
+    node.assert_attr(name="test_func[\\x00]")
 
 
 def test_double_colon_split_function_issue469(testdir):

tox.ini

@@ -10,13 +10,15 @@ envlist=
 commands= py.test --lsof -rfsxX {posargs:testing}
 passenv = USER USERNAME
 deps=
+    hypothesis
     nose
     mock
     requests
 
 [testenv:py26]
 commands= py.test --lsof -rfsxX {posargs:testing}
 deps=
+    hypothesis<3.0
     nose
     mock<1.1  # last supported version for py26
 
@@ -43,6 +45,7 @@ commands = flake8 pytest.py _pytest testing
 deps=pytest-xdist>=1.13
     mock
     nose
+    hypothesis
 commands=
   py.test -n1 -rfsxX {posargs:testing}
 
@@ -67,6 +70,7 @@ commands=
 
 [testenv:py27-nobyte]
 deps=pytest-xdist>=1.13
+    hypothesis
 distribute=true
 setenv=
     PYTHONDONTWRITEBYTECODE=1