BUG: Fix string addition promoter to work with dtype=

Also makes it reject forced unicode selection, since that doens't
work.

Author: seberg

numpy/_core/src/umath/stringdtype_ufuncs.cpp

@@ -1028,6 +1028,25 @@ all_strings_promoter(PyObject *NPY_UNUSED(ufunc),
                      PyArray_DTypeMeta *const signature[],
                      PyArray_DTypeMeta *new_op_dtypes[])
 {
+    if (op_dtypes[0] != &PyArray_StringDType &&
+            op_dtypes[1] != &PyArray_StringDType &&
+            op_dtypes[2] != &PyArray_StringDType) {
+        /*
+         * This promoter was triggered with only unicode arguments, so use
+         * unicode.  This can happen due to `dtype=` support which sets the
+         * output DType/signature.
+         */
+        new_op_dtypes[0] = NPY_DT_NewRef(&PyArray_UnicodeDType);
+        new_op_dtypes[1] = NPY_DT_NewRef(&PyArray_UnicodeDType);
+        new_op_dtypes[2] = NPY_DT_NewRef(&PyArray_UnicodeDType);
+        return 0;
+    }
+    if (signature[0] == &PyArray_UnicodeDType &&
+            signature[1] == &PyArray_UnicodeDType &&
+            signature[2] == &PyArray_UnicodeDType) {
+        /* Unicode forced, but didn't override a string input: invalid */
+        return -1;
+    }
     new_op_dtypes[0] = NPY_DT_NewRef(&PyArray_StringDType);
     new_op_dtypes[1] = NPY_DT_NewRef(&PyArray_StringDType);
     new_op_dtypes[2] = NPY_DT_NewRef(&PyArray_StringDType);
@@ -2532,6 +2551,17 @@ init_stringdtype_ufuncs(PyObject *umath)
         return -1;
     }
 
+    PyArray_DTypeMeta *out_strings_promoter_dtypes[] = {
+        &PyArray_UnicodeDType,
+        &PyArray_UnicodeDType,
+        &PyArray_StringDType,
+    };
+
+    if (add_promoter(umath, "add", out_strings_promoter_dtypes, 3,
+                     all_strings_promoter) < 0) {
+        return -1;
+    }
+
     INIT_MULTIPLY(Int64, int64);
     INIT_MULTIPLY(UInt64, uint64);
 

numpy/_core/tests/test_stringdtype.py

@@ -828,6 +828,22 @@ def test_add_promoter(string_list):
         assert_array_equal(op + arr, lresult)
         assert_array_equal(arr + op, rresult)
 
+    # The promoter should be able to handle things if users pass `dtype=`
+    res = np.add("hello", string_list, dtype=StringDType)
+    assert res.dtype == StringDType()
+
+    # The promoter should not kick in if users override the input,
+    # which means arr is cast, this fails because of the unknown length.
+    with pytest.raises(TypeError, match="cannot cast dtype"):
+        np.add(arr, "add", signature=("U", "U", None), casting="unsafe")
+
+    # But it must simply reject the following:
+    with pytest.raises(TypeError, match=".*did not contain a loop"):
+        np.add(arr, "add", signature=(None, "U", None))
+
+    with pytest.raises(TypeError, match=".*did not contain a loop"):
+        np.add("a", "b", signature=("U", "U", StringDType))
+
 
 def test_add_promoter_reduce():
     # Exact TypeError could change, but ensure StringDtype doesn't match