[PYTHON] Avoid shadowing python built-ins in python function variable naming

Given that spark 4.0.0 is upcoming I wonder if we should at least consider renaming certain function variable naming in python. Otherwise, we may need to wait another 4 years to do so.

Example

https://github.com/apache/spark/blob/e6b7950f553cff5adc02b8b5195e79cffff3c97c/python/pyspark/sql/functions/builtin.py#L12768

There are 8 uses of `len` and 35 `str` as variable names, both of which are python built-ins. Shadowing `str` is somewhat dangerous in that the following would be nonsensical:

def foo(str: "ColumnOrName", bar: "ColumnOrName"):
    # str is variable now, cannot be used as type
    bar = if lit(bar) if isinstance(bar, str) else bar

Now obviously this would be breaking change for user code if the function is called with kwargs style. If we rename `str` to `src` or `col`, certain old code using kwargs would break:

# breaks:
foo(str="x", bar="y")

# okay:
foo("x", bar="y")

Is this change a possibility for 4.0? Or are we thinking that the kwargs breaking change is too big to make compared to the benefit?