ENH: Update DataFrame.to_stata to handle pd.NA and None values in strL columns (#61286)

* ENH: Update DataFrame.to_stata to handle pd.NA and None values in strL columns

* Update pandas/io/stata.py

Co-authored-by: Matthew Roeschke <[email protected]>

* Moved changelog msg to 3.0.0 and adapted phrasing

---------

Co-authored-by: Jesse <[email protected]>
Co-authored-by: Matthew Roeschke <[email protected]>

Author: 3 people

Diff for: doc/source/whatsnew/v3.0.0.rst

@@ -767,6 +767,7 @@ I/O
 - Bug in :meth:`DataFrame.to_dict` raises unnecessary ``UserWarning`` when columns are not unique and ``orient='tight'``. (:issue:`58281`)
 - Bug in :meth:`DataFrame.to_excel` when writing empty :class:`DataFrame` with :class:`MultiIndex` on both axes (:issue:`57696`)
 - Bug in :meth:`DataFrame.to_excel` where the :class:`MultiIndex` index with a period level was not a date (:issue:`60099`)
+- Bug in :meth:`DataFrame.to_stata` when exporting a column containing both long strings (Stata strL) and :class:`pd.NA` values (:issue:`23633`)
 - Bug in :meth:`DataFrame.to_stata` when writing :class:`DataFrame` and ``byteorder=`big```. (:issue:`58969`)
 - Bug in :meth:`DataFrame.to_stata` when writing more than 32,000 value labels. (:issue:`60107`)
 - Bug in :meth:`DataFrame.to_string` that raised ``StopIteration`` with nested DataFrames. (:issue:`16098`)

Diff for: pandas/io/stata.py

@@ -3196,8 +3196,8 @@ def generate_table(self) -> tuple[dict[str, tuple[int, int]], DataFrame]:
         for o, (idx, row) in enumerate(selected.iterrows()):
             for j, (col, v) in enumerate(col_index):
                 val = row[col]
-                # Allow columns with mixed str and None (GH 23633)
-                val = "" if val is None else val
+                # Allow columns with mixed str and None or pd.NA (GH 23633)
+                val = "" if isna(val) else val
                 key = gso_table.get(val, None)
                 if key is None:
                     # Stata prefers human numbers

Diff for: pandas/tests/io/test_stata.py

@@ -2587,3 +2587,17 @@ def test_many_strl(temp_file, version):
     lbls = ["".join(v) for v in itertools.product(*([string.ascii_letters] * 3))]
     value_labels = {"col": {i: lbls[i] for i in range(n)}}
     df.to_stata(temp_file, value_labels=value_labels, version=version)
+
+
+@pytest.mark.parametrize("version", [117, 118, 119, None])
+def test_strl_missings(temp_file, version):
+    # GH 23633
+    # Check that strl supports None and pd.NA
+    df = DataFrame(
+        [
+            {"str1": "string" * 500, "number": 0},
+            {"str1": None, "number": 1},
+            {"str1": pd.NA, "number": 1},
+        ]
+    )
+    df.to_stata(temp_file, version=version)