Added Pulse.get_sink_by_name and Pulse.get_source_by_name

[natesymer]

I am writing a tool for implementing the function row in Python. To do so, I needed some extra functionality from libpulse, the ability to get a sink given a name.

Yes, I know one could get a list of all the sinks and search them, but, what happens if you want to do this multiple times per second? What happens if you want to use things like @DEFAULT_SINK@? You need to use pa_get_sink_info_by_name(). Using a patched version of pulsectl with this function wrapped gave me a 33% reduction in average memory use for my program.

I urge you to merge this.

[mk-fg]

Thanks!

Indeed, it should be much less code and much more efficient and robust (esp. with large number of sinks/sounrces) than iteration/filtering.
Only reason these funcs were not wrapped is because things get added here on as-needed basis (as libpulse has a ton of stuff), and no one mentioned them yet :)