BUG: to_datetime incorrect OverflowError

doc/source/whatsnew/v2.0.0.rst

@@ -800,6 +800,7 @@ Categorical
 Datetimelike
 ^^^^^^^^^^^^
 - Bug in :func:`pandas.infer_freq`, raising ``TypeError`` when inferred on :class:`RangeIndex` (:issue:`47084`)
+- Bug in :func:`to_datetime` incorrectly raising ``OverflowError`` with string arguments corresponding to large integers (:issue:`50533`)
 - Bug in :func:`to_datetime` was raising on invalid offsets with ``errors='coerce'`` and ``infer_datetime_format=True`` (:issue:`48633`)
 - Bug in :class:`DatetimeIndex` constructor failing to raise when ``tz=None`` is explicitly specified in conjunction with timezone-aware ``dtype`` or data (:issue:`48659`)
 - Bug in subtracting a ``datetime`` scalar from :class:`DatetimeIndex` failing to retain the original ``freq`` attribute (:issue:`48818`)

pandas/_libs/tslibs/parsing.pyx

@@ -53,6 +53,7 @@ from pandas._libs.tslibs.nattype cimport (
     c_NaT as NaT,
     c_nat_strings as nat_strings,
 )
+from pandas._libs.tslibs.np_datetime import OutOfBoundsDatetime
 from pandas._libs.tslibs.np_datetime cimport (
     NPY_DATETIMEUNIT,
     npy_datetimestruct,
@@ -298,6 +299,12 @@ def parse_datetime_string(
         # following may be raised from dateutil
         # TypeError: 'NoneType' object is not iterable
         raise ValueError(f'Given date string "{date_string}" not likely a datetime')
+    except OverflowError as err:
+        # with e.g. "08335394550" dateutil raises when trying to pass
+        #  year=8335394550 to datetime.replace
+        raise OutOfBoundsDatetime(
+            f'Parsing "{date_string}" to datetime overflows'
+            ) from err
 
     return dt
 

pandas/core/arrays/datetimes.py

@@ -50,10 +50,7 @@
     TimeNonexistent,
     npt,
 )
-from pandas.errors import (
-    OutOfBoundsDatetime,
-    PerformanceWarning,
-)
+from pandas.errors import PerformanceWarning
 from pandas.util._exceptions import find_stack_level
 from pandas.util._validators import validate_inclusive
 
@@ -2154,18 +2151,14 @@ def objects_to_datetime64ns(
 
     flags = data.flags
     order: Literal["F", "C"] = "F" if flags.f_contiguous else "C"
-    try:
-        result, tz_parsed = tslib.array_to_datetime(
-            data.ravel("K"),
-            errors=errors,
-            utc=utc,
-            dayfirst=dayfirst,
-            yearfirst=yearfirst,
-        )
-        result = result.reshape(data.shape, order=order)
-    except OverflowError as err:
-        # Exception is raised when a part of date is greater than 32 bit signed int
-        raise OutOfBoundsDatetime("Out of bounds nanosecond timestamp") from err
+    result, tz_parsed = tslib.array_to_datetime(
+        data.ravel("K"),
+        errors=errors,
+        utc=utc,
+        dayfirst=dayfirst,
+        yearfirst=yearfirst,
+    )
+    result = result.reshape(data.shape, order=order)
 
     if tz_parsed is not None:
         # We can take a shortcut since the datetime64 numpy array

pandas/tests/tools/test_to_datetime.py

@@ -542,6 +542,29 @@ def test_to_datetime_parse_timezone_keeps_name(self):
 
 
 class TestToDatetime:
+    @pytest.mark.filterwarnings("ignore:Could not infer format")
+    def test_to_datetime_overflow(self):
+        # we should get an OutOfBoundsDatetime, NOT OverflowError
+        # TODO: Timestamp raises VaueError("could not convert string to Timestamp")
+        #  can we make these more consistent?
+        arg = "08335394550"
+        msg = 'Parsing "08335394550" to datetime overflows, at position 0'
+        with pytest.raises(OutOfBoundsDatetime, match=msg):
+            to_datetime(arg)
+
+        with pytest.raises(OutOfBoundsDatetime, match=msg):
+            to_datetime([arg])
+
+        res = to_datetime(arg, errors="coerce")
+        assert res is NaT
+        res = to_datetime([arg], errors="coerce")
+        tm.assert_index_equal(res, Index([NaT]))
+
+        res = to_datetime(arg, errors="ignore")
+        assert isinstance(res, str) and res == arg
+        res = to_datetime([arg], errors="ignore")
+        tm.assert_index_equal(res, Index([arg], dtype=object))
+
     def test_to_datetime_mixed_datetime_and_string(self):
         # GH#47018 adapted old doctest with new behavior
         d1 = datetime(2020, 1, 1, 17, tzinfo=timezone(-timedelta(hours=1)))