BUG: raise on non-hashable in __contains__

Author: jbrockmendel

pandas/_libs/index.pyx

@@ -72,9 +72,10 @@ cdef class IndexEngine:
         self.over_size_threshold = n >= _SIZE_CUTOFF
         self.clear_mapping()
 
-    def __contains__(self, object val):
+    def __contains__(self, val: object) -> bool:
+        # We assume before we get here:
+        #  - val is hashable
         self._ensure_mapping_populated()
-        hash(val)
         return val in self.mapping
 
     cpdef get_value(self, ndarray arr, object key, object tz=None):
@@ -85,7 +86,6 @@ cdef class IndexEngine:
         """
         cdef:
             object loc
-            void* data_ptr
 
         loc = self.get_loc(key)
         if isinstance(loc, slice) or util.is_array(loc):
@@ -101,7 +101,6 @@ cdef class IndexEngine:
         """
         cdef:
             object loc
-            void* data_ptr
 
         loc = self.get_loc(key)
         value = convert_scalar(arr, value)
@@ -409,7 +408,9 @@ cdef class DatetimeEngine(Int64Engine):
     cdef _get_box_dtype(self):
         return 'M8[ns]'
 
-    def __contains__(self, object val):
+    def __contains__(self, val: object) -> bool:
+        # We assume before we get here:
+        #  - val is hashable
         cdef:
             int64_t loc
 
@@ -717,7 +718,9 @@ cdef class BaseMultiIndexCodesEngine:
 
         return indexer
 
-    def __contains__(self, object val):
+    def __contains__(self, val: object) -> bool:
+        # We assume before we get here:
+        #  - val is hashable
         # Default __contains__ looks in the underlying mapping, which in this
         # case only contains integer representations.
         try:

pandas/core/indexes/category.py

@@ -390,6 +390,7 @@ def __contains__(self, key) -> bool:
         if is_scalar(key) and isna(key):
             return self.hasnans
 
+        hash(key)
         return contains(self, key, container=self._engine)
 
     def __array__(self, dtype=None) -> np.ndarray:

pandas/core/indexes/datetimelike.py

@@ -154,6 +154,7 @@ def equals(self, other) -> bool:
 
     @Appender(_index_shared_docs["contains"] % _index_doc_kwargs)
     def __contains__(self, key):
+        hash(key)
         try:
             res = self.get_loc(key)
             return (

pandas/core/indexes/interval.py

@@ -385,6 +385,7 @@ def __contains__(self, key) -> bool:
         -------
         bool
         """
+        hash(key)
         if not isinstance(key, Interval):
             return False
 

pandas/core/indexes/numeric.py

@@ -474,6 +474,7 @@ def equals(self, other) -> bool:
             return False
 
     def __contains__(self, other) -> bool:
+        hash(other)
         if super().__contains__(other):
             return True
 

pandas/core/indexes/period.py

@@ -377,11 +377,11 @@ def __contains__(self, key) -> bool:
             else:
                 return key.ordinal in self._engine
         else:
+            hash(key)
             try:
                 self.get_loc(key)
                 return True
-            except (TypeError, KeyError):
-                # TypeError can be reached if we pass a tuple that is not hashable
+            except KeyError:
                 return False
 
     @cache_readonly

pandas/tests/indexes/common.py

@@ -19,6 +19,7 @@
     PeriodIndex,
     RangeIndex,
     Series,
+    Float64Index,
     TimedeltaIndex,
     UInt64Index,
     isna,
@@ -883,3 +884,11 @@ def test_getitem_2d_deprecated(self):
             res = idx[:, None]
 
         assert isinstance(res, np.ndarray), type(res)
+
+    def test_contains_requires_hashable(self):
+        idx = self.create_index()
+        with pytest.raises(TypeError, match="unhashable type"):
+            [] in idx
+
+        with pytest.raises(TypeError):
+            {} in idx._engine