Handle construction of string ExtensionArray from lists

Author: xhochy

pandas/core/construction.py

@@ -470,7 +470,9 @@ def sanitize_array(data, index, dtype=None, copy=False, raise_cast_failure=False
 
     # This is to prevent mixed-type Series getting all casted to
     # NumPy string type, e.g. NaN --> '-1#IND'.
-    if issubclass(subarr.dtype.type, str):
+    if not (
+        is_extension_array_dtype(subarr.dtype) or is_extension_array_dtype(dtype)
+    ) and issubclass(subarr.dtype.type, str):
         # GH#16605
         # If not empty convert the data to dtype
         # GH#19853: If data is a scalar, subarr has already the result

pandas/tests/extension/arrow/bool.py renamed to pandas/tests/extension/arrow/arrays.py

@@ -43,18 +43,27 @@ def _is_boolean(self):
         return True
 
 
-class ArrowBoolArray(ExtensionArray):
-    def __init__(self, values):
-        if not isinstance(values, pa.ChunkedArray):
-            raise ValueError
+@register_extension_dtype
+class ArrowStringDtype(ExtensionDtype):
 
-        assert values.type == pa.bool_()
-        self._data = values
-        self._dtype = ArrowBoolDtype()
+    type = str
+    kind = "U"
+    name = "arrow_string"
+    na_value = pa.NULL
 
-    def __repr__(self):
-        return "ArrowBoolArray({})".format(repr(self._data))
+    @classmethod
+    def construct_from_string(cls, string):
+        if string == cls.name:
+            return cls()
+        else:
+            raise TypeError("Cannot construct a '{}' from '{}'".format(cls, string))
 
+    @classmethod
+    def construct_array_type(cls):
+        return ArrowStringArray
+
+
+class ArrowExtensionArray(ExtensionArray):
     @classmethod
     def from_scalars(cls, values):
         arr = pa.chunked_array([pa.array(np.asarray(values))])
@@ -142,3 +151,29 @@ def any(self, axis=0, out=None):
 
     def all(self, axis=0, out=None):
         return self._data.to_pandas().all()
+
+
+class ArrowBoolArray(ArrowExtensionArray):
+    def __init__(self, values):
+        if not isinstance(values, pa.ChunkedArray):
+            raise ValueError
+
+        assert values.type == pa.bool_()
+        self._data = values
+        self._dtype = ArrowBoolDtype()
+
+    def __repr__(self):
+        return "ArrowBoolArray({})".format(repr(self._data))
+
+
+class ArrowStringArray(ArrowExtensionArray):
+    def __init__(self, values):
+        if not isinstance(values, pa.ChunkedArray):
+            raise ValueError
+
+        assert values.type == pa.string()
+        self._data = values
+        self._dtype = ArrowStringDtype()
+
+    def __repr__(self):
+        return "ArrowStringArray({})".format(repr(self._data))

pandas/tests/extension/arrow/test_bool.py

@@ -7,7 +7,7 @@
 
 pytest.importorskip("pyarrow", minversion="0.10.0")
 
-from .bool import ArrowBoolArray, ArrowBoolDtype  # isort:skip
+from .arrays import ArrowBoolArray, ArrowBoolDtype  # isort:skip
 
 
 @pytest.fixture

pandas/tests/extension/arrow/test_string.py

@@ -0,0 +1,13 @@
+import pytest
+
+import pandas as pd
+
+pytest.importorskip("pyarrow", minversion="0.10.0")
+
+from .arrays import ArrowStringDtype  # isort:skip
+
+
+def test_constructor_from_list():
+    # GH 27673
+    result = pd.Series(["E"], dtype=ArrowStringDtype())
+    assert isinstance(result.dtype, ArrowStringDtype)