[SPARK-50667][PYTHON][TESTS] Make jinja2 optional in PySpark Tests

### What changes were proposed in this pull request?
Make `jinja2` optional in PySpark Tests

### Why are the changes needed?
`jinja2` is an optional dependency of `pandas`

https://pypi.org/pypi/pandas/2.2.0/json

```
'jinja2>=3.1.2; extra == "output-formatting"'
```

It is not a mandatory requirement of pyspark, so PySpark tests should succeed even it is not installed

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

### How was this patch tested?
manually test after uninstalling it

### Was this patch authored or co-authored using generative AI tooling?
no

Closes apache#49288 from zhengruifeng/optional_jinja2.

Authored-by: Ruifeng Zheng <ruifengz@apache.org>
Signed-off-by: Hyukjin Kwon <gurwls223@apache.org>

python/pyspark/pandas/frame.py

@@ -2632,7 +2632,7 @@ def to_latex(
         ...                    'mask': ['red', 'purple'],
         ...                    'weapon': ['sai', 'bo staff']},
         ...                   columns=['name', 'mask', 'weapon'])
-        >>> print(df.to_latex(index=False)) # doctest: +NORMALIZE_WHITESPACE
+        >>> print(df.to_latex(index=False))  # doctest: +SKIP
         \begin{tabular}{lll}
         \toprule
               name &    mask &    weapon \\

python/pyspark/pandas/tests/io/test_dataframe_conversion.py

@@ -26,6 +26,7 @@
 from pyspark import pandas as ps
 from pyspark.testing.pandasutils import PandasOnSparkTestCase, TestUtils
 from pyspark.testing.sqlutils import SQLTestUtils
+from pyspark.testing.utils import have_jinja2, jinja2_requirement_message
 
 
 class DataFrameConversionMixin:
@@ -199,6 +200,7 @@ def test_to_clipboard(self):
             psdf.to_clipboard(sep=";", index=False), pdf.to_clipboard(sep=";", index=False)
         )
 
+    @unittest.skipIf(not have_jinja2, jinja2_requirement_message)
     def test_to_latex(self):
         pdf = self.pdf
         psdf = self.psdf

python/pyspark/testing/utils.py

@@ -94,6 +94,9 @@ def have_package(name: str) -> bool:
 have_flameprof = have_package("flameprof")
 flameprof_requirement_message = None if have_flameprof else "No module named 'flameprof'"
 
+have_jinja2 = have_package("jinja2")
+jinja2_requirement_message = None if have_jinja2 else "No module named 'jinja2'"
+
 pandas_requirement_message = None
 try:
     from pyspark.sql.pandas.utils import require_minimum_pandas_version