numpy
diff --git a/‎numpy/core/_internal.py
Lines changed: 20 additions & 27 deletions b/‎numpy/core/_internal.py
Lines changed: 20 additions & 27 deletions
diff --git a/‎numpy/core/tests/test_multiarray.py
Lines changed: 7 additions & 0 deletions b/‎numpy/core/tests/test_multiarray.py
Lines changed: 7 additions & 0 deletions
@@ -470,25 +470,9 @@ def __dtype_from_pep3118(stream, is_subdtype):
         itemsize=0
     )
     offset = 0
-    explicit_name = False
-    this_explicit_name = False
     common_alignment = 1
     is_padding = False
 
-    dummy_name_index = [0]
-
-
-    def next_dummy_name():
-        dummy_name_index[0] += 1
-
-    def get_dummy_name():
-        while True:
-            name = 'f%d' % dummy_name_index[0]
-            if name not in field_spec['names']:
-                return name
-            next_dummy_name()
-
-
     # Parse spec
     while stream:
         value = None
@@ -583,25 +567,19 @@ def get_dummy_name():
             value = dtype((value, shape))
 
         # Field name
-        this_explicit_name = False
         if stream.consume(':'):
             name = stream.consume_until(':')
-            explicit_name = True
-            this_explicit_name = True
         else:
-            name = get_dummy_name()
+            name = None
 
-        if not is_padding or this_explicit_name:
-            if name in field_spec['names']:
+        if not (is_padding and name is None):
+            if name is not None and name in field_spec['names']:
                 raise RuntimeError("Duplicate field name '%s' in PEP3118 format"
                                    % name)
             field_spec['names'].append(name)
             field_spec['formats'].append(value)
             field_spec['offsets'].append(offset)
 
-            if not this_explicit_name:
-                next_dummy_name()
-
         offset += value.itemsize
         offset += extra_offset
 
@@ -612,18 +590,33 @@ def get_dummy_name():
         field_spec['itemsize'] += (-offset) % common_alignment
 
     # Check if this was a simple 1-item type, and unwrap it
-    if (len(field_spec['names']) == 1
+    if (field_spec['names'] == [None]
             and field_spec['offsets'][0] == 0
             and field_spec['itemsize'] == field_spec['formats'][0].itemsize
-            and not explicit_name
             and not is_subdtype):
         ret = field_spec['formats'][0]
     else:
+        _fix_names(field_spec)
         ret = dtype(field_spec)
 
     # Finished
     return ret, common_alignment
 
+def _fix_names(field_spec):
+    """ Replace names which are None with the next unused f%d name """
+    names = field_spec['names']
+    for i, name in enumerate(names):
+        if name is not None:
+            continue
+
+        j = 0
+        while True:
+            name = 'f{}'.format(j)
+            if name not in names:
+                break
+            j = j + 1
+        names[i] = name
+
 def _add_trailing_padding(value, padding):
     """Inject the specified number of padding bytes at the end of a dtype"""
     if value.fields is None:
 
@@ -5927,6 +5927,13 @@ def test_field_order(self):
         self._check("(0)I:a:f:b:", [('a', 'I', (0,)), ('b', 'f')])
         self._check("(0)I:b:f:a:", [('b', 'I', (0,)), ('a', 'f')])
 
+    def test_unnamed_fields(self):
+        self._check('ii',     [('f0', 'i'), ('f1', 'i')])
+        self._check('ii:f0:', [('f1', 'i'), ('f0', 'i')])
+
+        self._check('i', 'i')
+        self._check('i:f0:', [('f0', 'i')])
+
 class TestNewBufferProtocol(object):
     def _check_roundtrip(self, obj):
         obj = np.asarray(obj)