REF: extract_array earlier in block construction (#40026)

Author: jbrockmendel

pandas/_testing/asserters.py

@@ -18,6 +18,7 @@
     is_numeric_dtype,
     needs_i8_conversion,
 )
+from pandas.core.dtypes.dtypes import PandasDtype
 from pandas.core.dtypes.missing import array_equivalent
 
 import pandas as pd
@@ -630,12 +631,12 @@ def raise_assert_detail(obj, message, left, right, diff=None, index_values=None)
 
     if isinstance(left, np.ndarray):
         left = pprint_thing(left)
-    elif is_categorical_dtype(left):
+    elif is_categorical_dtype(left) or isinstance(left, PandasDtype):
         left = repr(left)
 
     if isinstance(right, np.ndarray):
         right = pprint_thing(right)
-    elif is_categorical_dtype(right):
+    elif is_categorical_dtype(right) or isinstance(right, PandasDtype):
         right = repr(right)
 
     msg += f"""

pandas/core/internals/blocks.py

@@ -178,7 +178,7 @@ def _maybe_coerce_values(cls, values):
 
         Parameters
         ----------
-        values : np.ndarray, ExtensionArray, Index
+        values : np.ndarray or ExtensionArray
 
         Returns
         -------
@@ -350,7 +350,7 @@ def __getstate__(self):
     @final
     def __setstate__(self, state):
         self.mgr_locs = libinternals.BlockPlacement(state[0])
-        self.values = state[1]
+        self.values = extract_array(state[1], extract_numpy=True)
         self.ndim = self.values.ndim
 
     def _slice(self, slicer):
@@ -1623,7 +1623,7 @@ def _maybe_coerce_values(cls, values):
 
         Parameters
         ----------
-        values : Index, Series, ExtensionArray
+        values : np.ndarray or ExtensionArray
 
         Returns
         -------
@@ -2105,7 +2105,7 @@ def _maybe_coerce_values(cls, values):
 
         Parameters
         ----------
-        values : array-like
+        values : np.ndarray or ExtensionArray
             Must be convertible to datetime64/timedelta64
 
         Returns

pandas/core/internals/managers.py

@@ -45,7 +45,6 @@
 from pandas.core.dtypes.dtypes import ExtensionDtype
 from pandas.core.dtypes.generic import (
     ABCDataFrame,
-    ABCPandasArray,
     ABCSeries,
 )
 from pandas.core.dtypes.missing import (
@@ -316,6 +315,8 @@ def __getstate__(self):
     def __setstate__(self, state):
         def unpickle_block(values, mgr_locs, ndim: int):
             # TODO(EA2D): ndim would be unnecessary with 2D EAs
+            # older pickles may store e.g. DatetimeIndex instead of DatetimeArray
+            values = extract_array(values, extract_numpy=True)
             return make_block(values, placement=mgr_locs, ndim=ndim)
 
         if isinstance(state, tuple) and len(state) >= 4 and "0.14.1" in state[3]:
@@ -1212,6 +1213,7 @@ def insert(self, loc: int, item: Hashable, value, allow_duplicates: bool = False
             # TODO(EA2D): special case not needed with 2D EAs
             value = ensure_block_shape(value, ndim=2)
 
+        # TODO: type value as ArrayLike
         block = make_block(values=value, ndim=self.ndim, placement=slice(loc, loc + 1))
 
         for blkno, count in _fast_count_smallints(self.blknos[loc:]):
@@ -1673,16 +1675,20 @@ def create_block_manager_from_blocks(blocks, axes: List[Index]) -> BlockManager:
         raise construction_error(tot_items, blocks[0].shape[1:], axes, e)
 
 
+# We define this here so we can override it in tests.extension.test_numpy
+def _extract_array(obj):
+    return extract_array(obj, extract_numpy=True)
+
+
 def create_block_manager_from_arrays(
     arrays, names: Index, axes: List[Index]
 ) -> BlockManager:
     assert isinstance(names, Index)
     assert isinstance(axes, list)
     assert all(isinstance(x, Index) for x in axes)
 
-    # ensure we dont have any PandasArrays when we call get_block_type
-    # Note: just calling extract_array breaks tests that patch PandasArray._typ.
-    arrays = [x if not isinstance(x, ABCPandasArray) else x.to_numpy() for x in arrays]
+    arrays = [_extract_array(x) for x in arrays]
+
     try:
         blocks = _form_blocks(arrays, names, axes)
         mgr = BlockManager(blocks, axes)
@@ -1692,7 +1698,12 @@ def create_block_manager_from_arrays(
         raise construction_error(len(arrays), arrays[0].shape, axes, e)
 
 
-def construction_error(tot_items, block_shape, axes, e=None):
+def construction_error(
+    tot_items: int,
+    block_shape: Shape,
+    axes: List[Index],
+    e: Optional[ValueError] = None,
+):
     """ raise a helpful message about our construction """
     passed = tuple(map(int, [tot_items] + list(block_shape)))
     # Correcting the user facing error message during dataframe construction
@@ -1716,7 +1727,9 @@ def construction_error(tot_items, block_shape, axes, e=None):
 # -----------------------------------------------------------------------
 
 
-def _form_blocks(arrays, names: Index, axes: List[Index]) -> List[Block]:
+def _form_blocks(
+    arrays: List[ArrayLike], names: Index, axes: List[Index]
+) -> List[Block]:
     # put "leftover" items in float bucket, where else?
     # generalize?
     items_dict: DefaultDict[str, List] = defaultdict(list)
@@ -1836,21 +1849,14 @@ def _multi_blockify(tuples, dtype: Optional[Dtype] = None):
 
 def _stack_arrays(tuples, dtype: np.dtype):
 
-    # fml
-    def _asarray_compat(x):
-        if isinstance(x, ABCSeries):
-            return x._values
-        else:
-            return np.asarray(x)
-
     placement, arrays = zip(*tuples)
 
     first = arrays[0]
     shape = (len(arrays),) + first.shape
 
     stacked = np.empty(shape, dtype=dtype)
     for i, arr in enumerate(arrays):
-        stacked[i] = _asarray_compat(arr)
+        stacked[i] = arr
 
     return stacked, placement
 
@@ -1874,7 +1880,7 @@ def _interleaved_dtype(blocks: Sequence[Block]) -> Optional[DtypeObj]:
     return find_common_type([b.dtype for b in blocks])
 
 
-def _consolidate(blocks):
+def _consolidate(blocks: Tuple[Block, ...]) -> List[Block]:
     """
     Merge blocks having same dtype, exclude non-consolidating blocks
     """

pandas/tests/extension/test_numpy.py

@@ -20,13 +20,26 @@
     ExtensionDtype,
     PandasDtype,
 )
+from pandas.core.dtypes.generic import ABCPandasArray
 
 import pandas as pd
 import pandas._testing as tm
 from pandas.core.arrays.numpy_ import PandasArray
+from pandas.core.internals import managers
 from pandas.tests.extension import base
 
 
+def _extract_array_patched(obj):
+    if isinstance(obj, (pd.Index, pd.Series)):
+        obj = obj._values
+    if isinstance(obj, ABCPandasArray):
+        # TODO for reasons unclear, we get here in a couple of tests
+        #  with PandasArray._typ *not* patched
+        obj = obj.to_numpy()
+
+    return obj
+
+
 @pytest.fixture(params=["float", "object"])
 def dtype(request):
     return PandasDtype(np.dtype(request.param))
@@ -51,6 +64,7 @@ def allow_in_pandas(monkeypatch):
     """
     with monkeypatch.context() as m:
         m.setattr(PandasArray, "_typ", "extension")
+        m.setattr(managers, "_extract_array", _extract_array_patched)
         yield
 
 

pandas/tests/internals/test_internals.py

@@ -123,7 +123,7 @@ def create_block(typestr, placement, item_shape=None, num_offset=0):
         assert m is not None, f"incompatible typestr -> {typestr}"
         tz = m.groups()[0]
         assert num_items == 1, "must have only 1 num items for a tz-aware"
-        values = DatetimeIndex(np.arange(N) * 1e9, tz=tz)
+        values = DatetimeIndex(np.arange(N) * 1e9, tz=tz)._data
     elif typestr in ("timedelta", "td", "m8[ns]"):
         values = (mat * 1).astype("m8[ns]")
     elif typestr in ("category",):