Fix build regression against Spark 3.2.x [databricks]

[andygrove]

Closes #10174

The reason this code was failing against 3.2 and not against 3.3 is that 3.2 will populate timestampFormat with a default if none is supplied, whereas 3.3 does not.

Integration tests passed locally for me with this change against both 3.2.3 and 3.3.2

[andygrove]

build

[andygrove]

build

[andygrove]

CI failed due to unrelated issue (on databricks):

2024-01-10T23:39:01.0960149Z [2024-01-10T23:34:21.452Z] [2024-01-10T23:33:07.595Z] ^[[31mFAILED^[[0m ../../src/main/python/dpp_test.py::^[[1mtest_dpp_bypass[true-5-parquet][DATAGEN_SEED=1704919540, INJECT_OOM, IGNORE_ORDER]^[[0m - py4j.protocol.Py4JJavaError: An error occurred while calling z:org.apache.s...
2024-01-10T23:39:01.0961556Z [2024-01-10T23:34:21.452Z] [2024-01-10T23:33:07.595Z] ^[[31mFAILED^[[0m ../../src/main/python/dpp_test.py::^[[1mtest_dpp_bypass[true-5-orc][DATAGEN_SEED=1704919540, IGNORE_ORDER]^[[0m - py4j.protocol.Py4JJavaError: An error occurred while calling z:org.apache.s...
2024-01-10T23:39:01.0963105Z [2024-01-10T23:34:21.452Z] [2024-01-10T23:33:07.595Z] ^[[31mFAILED^[[0m ../../src/main/python/dpp_test.py::^[[1mtest_dpp_via_aggregate_subquery[true-5-parquet][DATAGEN_SEED=1704919540, INJECT_OOM, IGNORE_ORDER]^[[0m - py4j.protocol.Py4JJavaError: An error occurred while calling z:org.apache.s...
2024-01-10T23:39:01.0964581Z [2024-01-10T23:34:21.452Z] [2024-01-10T23:33:07.595Z] ^[[31mFAILED^[[0m ../../src/main/python/dpp_test.py::^[[1mtest_dpp_via_aggregate_subquery[true-5-orc][DATAGEN_SEED=1704919540, IGNORE_ORDER]^[[0m - py4j.protocol.Py4JJavaError: An error occurred while calling z:org.apache.s...

assertion failed: Could not find DynamicPruningExpression in the **** plan

This seems to be related to #10147

[andygrove]

I filed #10182 for the dpp issue

[andygrove]

build