TYP: internals.construction

pandas/core/dtypes/cast.py

@@ -1620,6 +1620,7 @@ def maybe_cast_to_datetime(
                                 dta = dta.tz_localize(None)
                             value = dta
                         elif is_datetime64tz:
+                            dtype = cast(DatetimeTZDtype, dtype)
                             # The string check can be removed once issue #13712
                             # is solved. String data that is passed with a
                             # datetime64tz is assumed to be naive which should
@@ -1700,6 +1701,8 @@ def ensure_nanosecond_dtype(dtype: DtypeObj) -> DtypeObj:
     dtype('<M8[ns]')
 
     >>> ensure_nanosecond_dtype(np.dtype("m8[ps]"))
+    Traceback (most recent call last):
+        ...
     TypeError: cannot convert timedeltalike to dtype [timedelta64[ps]]
     """
     msg = (

pandas/core/frame.py

@@ -1939,12 +1939,11 @@ def from_records(
                         arr_columns_list.append(k)
                         arrays.append(v)
 
-                arrays, arr_columns = reorder_arrays(arrays, arr_columns_list, columns)
+                arr_columns = Index(arr_columns_list)
+                arrays, arr_columns = reorder_arrays(arrays, arr_columns, columns)
 
         elif isinstance(data, (np.ndarray, DataFrame)):
             arrays, columns = to_arrays(data, columns)
-            if columns is not None:
-                columns = ensure_index(columns)
             arr_columns = columns
         else:
             arrays, arr_columns = to_arrays(data, columns)
@@ -1954,9 +1953,7 @@ def from_records(
                         arrays[i] = lib.maybe_convert_objects(arr, try_float=True)
 
             arr_columns = ensure_index(arr_columns)
-            if columns is not None:
-                columns = ensure_index(columns)
-            else:
+            if columns is None:
                 columns = arr_columns
 
         if exclude is None:

pandas/core/internals/construction.py

@@ -79,6 +79,7 @@
 if TYPE_CHECKING:
     from numpy.ma.mrecords import MaskedRecords
 
+
 # ---------------------------------------------------------------------
 # BlockManager Interface
 
@@ -91,7 +92,7 @@ def arrays_to_mgr(
     dtype: Optional[DtypeObj] = None,
     verify_integrity: bool = True,
     typ: Optional[str] = None,
-):
+) -> Manager:
     """
     Segregate Series based on type and coerce into matrices.
 
@@ -109,11 +110,11 @@ def arrays_to_mgr(
         # don't force copy because getting jammed in an ndarray anyway
         arrays = _homogenize(arrays, index, dtype)
 
-        columns = ensure_index(columns)
     else:
-        columns = ensure_index(columns)
         index = ensure_index(index)
 
+    columns = ensure_index(columns)
+
     # from BlockManager perspective
     axes = [columns, index]
 
@@ -140,9 +141,8 @@ def rec_array_to_mgr(
     fdata = ma.getdata(data)
     if index is None:
         index = _get_names_from_index(fdata)
-        if index is None:
-            index = ibase.default_index(len(data))
-    index = ensure_index(index)
+    else:
+        index = ensure_index(index)
 
     if columns is not None:
         columns = ensure_index(columns)
@@ -215,14 +215,14 @@ def mgr_to_mgr(mgr, typ: str):
 
 def ndarray_to_mgr(
     values, index, columns, dtype: Optional[DtypeObj], copy: bool, typ: str
-):
+) -> Manager:
     # used in DataFrame.__init__
-    # input must be a ndarray, list, Series, index
+    # input must be a ndarray, list, Series, Index, ExtensionArray
 
     if isinstance(values, ABCSeries):
         if columns is None:
             if values.name is not None:
-                columns = [values.name]
+                columns = Index([values.name])
         if index is None:
             index = values.index
         else:
@@ -309,7 +309,9 @@ def ndarray_to_mgr(
     return create_block_manager_from_blocks(block_values, [columns, index])
 
 
-def dict_to_mgr(data: Dict, index, columns, dtype: Optional[DtypeObj], typ: str):
+def dict_to_mgr(
+    data: Dict, index, columns, dtype: Optional[DtypeObj], typ: str
+) -> Manager:
     """
     Segregate Series based on type and coerce into matrices.
     Needs to handle a lot of exceptional cases.
@@ -531,21 +533,18 @@ def extract_index(data) -> Index:
     return ensure_index(index)
 
 
-def reorder_arrays(arrays, arr_columns, columns):
+def reorder_arrays(
+    arrays: List[ArrayLike], arr_columns: Index, columns: Optional[Index]
+) -> Tuple[List[ArrayLike], Index]:
     # reorder according to the columns
-    if (
-        columns is not None
-        and len(columns)
-        and arr_columns is not None
-        and len(arr_columns)
-    ):
+    if columns is not None and len(columns) and len(arr_columns):
         indexer = ensure_index(arr_columns).get_indexer(columns)
         arr_columns = ensure_index([arr_columns[i] for i in indexer])
         arrays = [arrays[i] for i in indexer]
     return arrays, arr_columns
 
 
-def _get_names_from_index(data):
+def _get_names_from_index(data) -> Index:
     has_some_name = any(getattr(s, "name", None) is not None for s in data)
     if not has_some_name:
         return ibase.default_index(len(data))
@@ -560,7 +559,7 @@ def _get_names_from_index(data):
             index[i] = f"Unnamed {count}"
             count += 1
 
-    return index
+    return Index(index)
 
 
 def _get_axes(