[SPARK-43710][PS][FOLLOWUP] Fix date_part invocations

### What changes were proposed in this pull request?
fix bc20e85:
1, should use `F.lit("HOUR")` as the parameter;
2, should also change `def microseconds` since the old function was removed;

### Why are the changes needed?
to fix wrong usage

### Does this PR introduce _any_ user-facing change?
no

### How was this patch tested?
existing UT

Closes #41911 from zhengruifeng/spark_43710_followup.

Authored-by: Ruifeng Zheng <[email protected]>
Signed-off-by: Ruifeng Zheng <[email protected]>

Author: zhengruifeng

python/pyspark/pandas/indexes/timedelta.py

@@ -25,7 +25,6 @@
 from pyspark.pandas.indexes.base import Index
 from pyspark.pandas.missing.indexes import MissingPandasLikeTimedeltaIndex
 from pyspark.pandas.series import Series
-from pyspark.pandas.spark import functions as SF
 from pyspark.sql import functions as F
 
 
@@ -150,9 +149,9 @@ def seconds(self) -> Index:
 
         @no_type_check
         def get_seconds(scol):
-            hour_scol = F.date_part("HOUR", scol)
-            minute_scol = F.date_part("MINUTE", scol)
-            second_scol = F.date_part("SECOND", scol)
+            hour_scol = F.date_part(F.lit("HOUR"), scol)
+            minute_scol = F.date_part(F.lit("MINUTE"), scol)
+            second_scol = F.date_part(F.lit("SECOND"), scol)
             return (
                 F.when(
                     hour_scol < 0,
@@ -178,7 +177,7 @@ def microseconds(self) -> Index:
 
         @no_type_check
         def get_microseconds(scol):
-            second_scol = SF.date_part("SECOND", scol)
+            second_scol = F.date_part(F.lit("SECOND"), scol)
             return (
                 (
                     F.when(