ENH: Fix Python 3.13 test failures & enable CI (pandas-dev#59065)

* ENH: Fix Python 3.13 test failures & enable CI

x-ref pandas-dev#58734

Co-authored-by: Thomas Li <[email protected]>

* Cast npy_intp to int to fix Windows CI

---------

Co-authored-by: Thomas Li <[email protected]>

Author: 2 people

.github/workflows/unit-tests.yml

@@ -329,7 +329,7 @@ jobs:
     #    To freeze this file, uncomment out the ``if: false`` condition, and migrate the jobs
     #    to the corresponding posix/windows-macos/sdist etc. workflows.
     # Feel free to modify this comment as necessary.
-    if: false # Uncomment this to freeze the workflow, comment it to unfreeze
+    # if: false # Uncomment this to freeze the workflow, comment it to unfreeze
     defaults:
       run:
         shell: bash -eou pipefail {0}
@@ -361,7 +361,7 @@ jobs:
       - name: Set up Python Dev Version
         uses: actions/setup-python@v5
         with:
-          python-version: '3.12-dev'
+          python-version: '3.13-dev'
 
       - name: Build Environment
         run: |

pandas/_libs/src/vendored/ujson/python/objToJSON.c

@@ -410,8 +410,8 @@ static void NpyArr_iterBegin(JSOBJ _obj, JSONTypeContext *tc) {
   npyarr->type_num = PyArray_DESCR(obj)->type_num;
 
   if (GET_TC(tc)->transpose) {
-    npyarr->dim = PyArray_DIM(obj, npyarr->ndim);
-    npyarr->stride = PyArray_STRIDE(obj, npyarr->ndim);
+    npyarr->dim = PyArray_DIM(obj, (int)npyarr->ndim);
+    npyarr->stride = PyArray_STRIDE(obj, (int)npyarr->ndim);
     npyarr->stridedim = npyarr->ndim;
     npyarr->index[npyarr->ndim] = 0;
     npyarr->inc = -1;
@@ -452,8 +452,8 @@ static void NpyArrPassThru_iterEnd(JSOBJ obj, JSONTypeContext *tc) {
     return;
   }
   const PyArrayObject *arrayobj = (const PyArrayObject *)npyarr->array;
-  npyarr->dim = PyArray_DIM(arrayobj, npyarr->stridedim);
-  npyarr->stride = PyArray_STRIDE(arrayobj, npyarr->stridedim);
+  npyarr->dim = PyArray_DIM(arrayobj, (int)npyarr->stridedim);
+  npyarr->stride = PyArray_STRIDE(arrayobj, (int)npyarr->stridedim);
   npyarr->dataptr += npyarr->stride;
 
   NpyArr_freeItemValue(obj, tc);
@@ -524,8 +524,8 @@ static int NpyArr_iterNext(JSOBJ _obj, JSONTypeContext *tc) {
   }
   const PyArrayObject *arrayobj = (const PyArrayObject *)npyarr->array;
 
-  npyarr->dim = PyArray_DIM(arrayobj, npyarr->stridedim);
-  npyarr->stride = PyArray_STRIDE(arrayobj, npyarr->stridedim);
+  npyarr->dim = PyArray_DIM(arrayobj, (int)npyarr->stridedim);
+  npyarr->stride = PyArray_STRIDE(arrayobj, (int)npyarr->stridedim);
   npyarr->index[npyarr->stridedim] = 0;
 
   ((PyObjectEncoder *)tc->encoder)->npyCtxtPassthru = npyarr;

pandas/_libs/tslibs/offsets.pyx

@@ -4960,7 +4960,12 @@ cpdef to_offset(freq, bint is_period=False):
     if result is None:
         raise ValueError(INVALID_FREQ_ERR_MSG.format(freq))
 
-    if is_period and not hasattr(result, "_period_dtype_code"):
+    try:
+        has_period_dtype_code = hasattr(result, "_period_dtype_code")
+    except ValueError:
+        has_period_dtype_code = False
+
+    if is_period and not has_period_dtype_code:
         if isinstance(freq, str):
             raise ValueError(f"{result.name} is not supported as period frequency")
         else:

pandas/tests/groupby/test_groupby.py

@@ -2816,7 +2816,9 @@ def test_rolling_wrong_param_min_period():
     test_df = DataFrame([name_l, val_l]).T
     test_df.columns = ["name", "val"]
 
-    result_error_msg = r"__init__\(\) got an unexpected keyword argument 'min_period'"
+    result_error_msg = (
+        r"^[a-zA-Z._]*\(\) got an unexpected keyword argument 'min_period'"
+    )
     with pytest.raises(TypeError, match=result_error_msg):
         test_df.groupby("name")["val"].rolling(window=2, min_period=1).sum()
 

pandas/tests/io/parser/test_dialect.py

@@ -26,7 +26,7 @@ def custom_dialect():
         "escapechar": "~",
         "delimiter": ":",
         "skipinitialspace": False,
-        "quotechar": "~",
+        "quotechar": "`",
         "quoting": 3,
     }
     return dialect_name, dialect_kwargs

pandas/tests/io/test_common.py

@@ -485,7 +485,10 @@ def test_warning_missing_utf_bom(self, encoding, compression_):
                 df.to_csv(path, compression=compression_, encoding=encoding)
 
             # reading should fail (otherwise we wouldn't need the warning)
-            msg = r"UTF-\d+ stream does not start with BOM"
+            msg = (
+                r"UTF-\d+ stream does not start with BOM|"
+                r"'utf-\d+' codec can't decode byte"
+            )
             with pytest.raises(UnicodeError, match=msg):
                 pd.read_csv(path, compression=compression_, encoding=encoding)
 

pandas/tests/io/xml/test_xml.py

@@ -1044,7 +1044,7 @@ def test_utf16_encoding(xml_baby_names, parser):
         UnicodeError,
         match=(
             "UTF-16 stream does not start with BOM|"
-            "'utf-16-le' codec can't decode byte"
+            "'utf-16(-le)?' codec can't decode byte"
         ),
     ):
         read_xml(xml_baby_names, encoding="UTF-16", parser=parser)

pandas/tests/scalar/timedelta/test_arithmetic.py

@@ -622,6 +622,7 @@ def test_td_floordiv_invalid_scalar(self):
             [
                 r"Invalid dtype datetime64\[D\] for __floordiv__",
                 "'dtype' is an invalid keyword argument for this function",
+                "this function got an unexpected keyword argument 'dtype'",
                 r"ufunc '?floor_divide'? cannot use operands with types",
             ]
         )