ENH: dont hard-code nanoseconds in overflow exception message (pandas-dev#55983)

Author: jbrockmendel

pandas/_libs/tslibs/conversion.pyx

@@ -30,6 +30,7 @@ from pandas._libs.tslibs.base cimport ABCTimestamp
 from pandas._libs.tslibs.dtypes cimport (
     abbrev_to_npy_unit,
     get_supported_reso,
+    npy_unit_to_attrname,
     periods_per_second,
 )
 from pandas._libs.tslibs.np_datetime cimport (
@@ -38,6 +39,7 @@ from pandas._libs.tslibs.np_datetime cimport (
     NPY_FR_us,
     check_dts_bounds,
     convert_reso,
+    dts_to_iso_string,
     get_conversion_factor,
     get_datetime64_unit,
     get_implementation_bounds,
@@ -214,8 +216,9 @@ cdef int64_t get_datetime64_nanos(object val, NPY_DATETIMEUNIT reso) except? -1:
         try:
             ival = npy_datetimestruct_to_datetime(reso, &dts)
         except OverflowError as err:
+            attrname = npy_unit_to_attrname[reso]
             raise OutOfBoundsDatetime(
-                "Out of bounds nanosecond timestamp: {val}"
+                f"Out of bounds {attrname} timestamp: {val}"
             ) from err
 
     return ival
@@ -248,8 +251,9 @@ cdef class _TSObject:
                 )
             except OverflowError as err:
                 if val is not None:
+                    attrname = npy_unit_to_attrname[creso]
                     raise OutOfBoundsDatetime(
-                        f"Out of bounds nanosecond timestamp: {val}"
+                        f"Out of bounds {attrname} timestamp: {val}"
                     ) from err
                 raise OutOfBoundsDatetime from err
 
@@ -419,7 +423,8 @@ cdef _TSObject convert_datetime_to_tsobject(
     try:
         obj.value = npy_datetimestruct_to_datetime(reso, &obj.dts)
     except OverflowError as err:
-        raise OutOfBoundsDatetime("Out of bounds nanosecond timestamp") from err
+        attrname = npy_unit_to_attrname[reso]
+        raise OutOfBoundsDatetime(f"Out of bounds {attrname} timestamp") from err
 
     if obj.tzinfo is not None and not is_utc(obj.tzinfo):
         offset = get_utcoffset(obj.tzinfo, ts)
@@ -589,18 +594,18 @@ cdef check_overflows(_TSObject obj, NPY_DATETIMEUNIT reso=NPY_FR_ns):
     if obj.dts.year == lb.year:
         if not (obj.value < 0):
             from pandas._libs.tslibs.timestamps import Timestamp
-            fmt = (f"{obj.dts.year}-{obj.dts.month:02d}-{obj.dts.day:02d} "
-                   f"{obj.dts.hour:02d}:{obj.dts.min:02d}:{obj.dts.sec:02d}")
+            fmt = dts_to_iso_string(&obj.dts)
+            min_ts = (<_Timestamp>Timestamp(0))._as_creso(reso).min
             raise OutOfBoundsDatetime(
-                f"Converting {fmt} underflows past {Timestamp.min}"
+                f"Converting {fmt} underflows past {min_ts}"
             )
     elif obj.dts.year == ub.year:
         if not (obj.value > 0):
             from pandas._libs.tslibs.timestamps import Timestamp
-            fmt = (f"{obj.dts.year}-{obj.dts.month:02d}-{obj.dts.day:02d} "
-                   f"{obj.dts.hour:02d}:{obj.dts.min:02d}:{obj.dts.sec:02d}")
+            fmt = dts_to_iso_string(&obj.dts)
+            max_ts = (<_Timestamp>Timestamp(0))._as_creso(reso).max
             raise OutOfBoundsDatetime(
-                f"Converting {fmt} overflows past {Timestamp.max}"
+                f"Converting {fmt} overflows past {max_ts}"
             )
 
 # ----------------------------------------------------------------------

pandas/_libs/tslibs/np_datetime.pyx

@@ -22,6 +22,8 @@ from libc.stdint cimport INT64_MAX
 import_datetime()
 PandasDateTime_IMPORT
 
+import operator
+
 import numpy as np
 
 cimport numpy as cnp
@@ -35,6 +37,10 @@ from numpy cimport (
     uint8_t,
 )
 
+from pandas._libs.tslibs.dtypes cimport (
+    npy_unit_to_abbrev,
+    npy_unit_to_attrname,
+)
 from pandas._libs.tslibs.util cimport get_c_string_buf_and_size
 
 
@@ -217,8 +223,8 @@ cdef check_dts_bounds(npy_datetimestruct *dts, NPY_DATETIMEUNIT unit=NPY_FR_ns):
 
     if error:
         fmt = dts_to_iso_string(dts)
-        # TODO: "nanosecond" in the message assumes NPY_FR_ns
-        raise OutOfBoundsDatetime(f"Out of bounds nanosecond timestamp: {fmt}")
+        attrname = npy_unit_to_attrname[unit]
+        raise OutOfBoundsDatetime(f"Out of bounds {attrname} timestamp: {fmt}")
 
 
 # ----------------------------------------------------------------------
@@ -261,8 +267,9 @@ cdef int64_t pydatetime_to_dt64(datetime val,
     try:
         result = npy_datetimestruct_to_datetime(reso, dts)
     except OverflowError as err:
+        attrname = npy_unit_to_attrname[reso]
         raise OutOfBoundsDatetime(
-            f"Out of bounds nanosecond timestamp: {val}"
+            f"Out of bounds {attrname} timestamp: {val}"
         ) from err
 
     return result
@@ -286,7 +293,8 @@ cdef int64_t pydate_to_dt64(
     try:
         result = npy_datetimestruct_to_datetime(reso, dts)
     except OverflowError as err:
-        raise OutOfBoundsDatetime(f"Out of bounds nanosecond timestamp: {val}") from err
+        attrname = npy_unit_to_attrname[reso]
+        raise OutOfBoundsDatetime(f"Out of bounds {attrname} timestamp: {val}") from err
 
     return result
 
@@ -425,7 +433,7 @@ cpdef ndarray astype_overflowsafe(
     return iresult.view(dtype)
 
 
-# TODO: try to upstream this fix to numpy
+# TODO(numpy#16352): try to upstream this fix to numpy
 def compare_mismatched_resolutions(ndarray left, ndarray right, op):
     """
     Overflow-safe comparison of timedelta64/datetime64 with mismatched resolutions.
@@ -483,9 +491,6 @@ def compare_mismatched_resolutions(ndarray left, ndarray right, op):
     return result
 
 
-import operator
-
-
 cdef int op_to_op_code(op):
     if op is operator.eq:
         return Py_EQ
@@ -538,8 +543,6 @@ cdef ndarray _astype_overflowsafe_to_smaller_unit(
         else:
             new_value, mod = divmod(value, mult)
             if not round_ok and mod != 0:
-                # TODO: avoid runtime import
-                from pandas._libs.tslibs.dtypes import npy_unit_to_abbrev
                 from_abbrev = npy_unit_to_abbrev(from_unit)
                 to_abbrev = npy_unit_to_abbrev(to_unit)
                 raise ValueError(

pandas/_libs/tslibs/strptime.pyx

@@ -51,6 +51,7 @@ from pandas._libs.tslibs.conversion cimport get_datetime64_nanos
 from pandas._libs.tslibs.dtypes cimport (
     get_supported_reso,
     npy_unit_to_abbrev,
+    npy_unit_to_attrname,
 )
 from pandas._libs.tslibs.nattype cimport (
     NPY_NAT,
@@ -412,8 +413,9 @@ def array_strptime(
                 try:
                     value = npy_datetimestruct_to_datetime(creso, &dts)
                 except OverflowError as err:
+                    attrname = npy_unit_to_attrname[creso]
                     raise OutOfBoundsDatetime(
-                        f"Out of bounds nanosecond timestamp: {val}"
+                        f"Out of bounds {attrname} timestamp: {val}"
                     ) from err
                 if out_local == 1:
                     # Store the out_tzoffset in seconds
@@ -451,8 +453,9 @@ def array_strptime(
             try:
                 iresult[i] = npy_datetimestruct_to_datetime(creso, &dts)
             except OverflowError as err:
+                attrname = npy_unit_to_attrname[creso]
                 raise OutOfBoundsDatetime(
-                    f"Out of bounds nanosecond timestamp: {val}"
+                    f"Out of bounds {attrname} timestamp: {val}"
                 ) from err
             result_timezone[i] = tz
 

pandas/_libs/tslibs/timestamps.pyx

@@ -60,6 +60,7 @@ from pandas._libs.tslibs.conversion cimport (
 )
 from pandas._libs.tslibs.dtypes cimport (
     npy_unit_to_abbrev,
+    npy_unit_to_attrname,
     periods_per_day,
     periods_per_second,
 )
@@ -447,15 +448,15 @@ cdef class _Timestamp(ABCTimestamp):
             nanos = other._value
 
             try:
-                new_value = self._value+ nanos
+                new_value = self._value + nanos
                 result = type(self)._from_value_and_reso(
                     new_value, reso=self._creso, tz=self.tzinfo
                 )
             except OverflowError as err:
-                # TODO: don't hard-code nanosecond here
                 new_value = int(self._value) + int(nanos)
+                attrname = npy_unit_to_attrname[self._creso]
                 raise OutOfBoundsDatetime(
-                    f"Out of bounds nanosecond timestamp: {new_value}"
+                    f"Out of bounds {attrname} timestamp: {new_value}"
                 ) from err
 
             return result

pandas/tests/scalar/timestamp/test_constructors.py

@@ -814,7 +814,7 @@ def test_bounds_with_different_units(self):
 
         # With more extreme cases, we can't even fit inside second resolution
         info = np.iinfo(np.int64)
-        msg = "Out of bounds nanosecond timestamp:"
+        msg = "Out of bounds second timestamp:"
         for value in [info.min + 1, info.max]:
             for unit in ["D", "h", "m"]:
                 dt64 = np.datetime64(value, unit)

pandas/tests/tools/test_to_datetime.py

@@ -1143,7 +1143,7 @@ def test_to_datetime_dt64s_out_of_ns_bounds(self, cache, dt, errors):
     def test_to_datetime_dt64d_out_of_bounds(self, cache):
         dt64 = np.datetime64(np.iinfo(np.int64).max, "D")
 
-        msg = "Out of bounds nanosecond timestamp"
+        msg = "Out of bounds second timestamp: 25252734927768524-07-27"
         with pytest.raises(OutOfBoundsDatetime, match=msg):
             Timestamp(dt64)
         with pytest.raises(OutOfBoundsDatetime, match=msg):