Replacing lru_cache with cache, post-3.8

[HalfWhitt]

Now that Toga's dropping 3.8 support, I remembered a spot in images where I used functools.lru_cache because cache wasn't added until 3.9.

From the docs on cache:

Returns the same as lru_cache(maxsize=None), creating a thin wrapper around a dictionary lookup for the function arguments. Because it never needs to evict old values, this is smaller and faster than lru_cache() with a size limit.

So in that case it's a one-to-one swap. But I noticed one other place too, in platform. There it's maxsize=1 instead of None, which could behave differently — only keeping the most recent call instead of all calls — except that get_platform_factory() accepts no arguments, so it can only ever cache one value anyway. So setting it to None makes the result potentially smaller and faster without changing behavior, and that in turn is the same as cache.

PR Checklist:

• All new features have been tested
• All new features have been documented
• I have read the CONTRIBUTING.md file
• I will abide by the code of conduct

[freakboy3742]

A small cleanup, but a nice one - thanks!