All conversions using `latin-1` have been converted to
`raw_unicode_escape` which is like `latin-1`, but handles unicode
characters appropriately (with a `\u` prefix).
Since this is like `latin-1` it should have no impact on things that ran
previously, but those that would fail with a unicode error now will
present an encoded unicode string. There may be situations where the
binary representation of unicode would be better (timeliner file
output?), but those can be changed when/if it's determined necessary.
Fixes#274.
This should have no impact functionality-wise.
The statistics plugin was left out a) as an example and b) because it
was committed by mistake in the first place and was never meant to be a
real plugin.
Mike Auty