Add the turn events emptier.

Create the Great Lakes branch.

This involves making a few things configurable, reading pop figures from a
different table, and a few other tweaks that should probably be more
configurable.

Nerf cities that don't have any popdata (the pre-existing ones in central/eastern Europe).  This prevents them from dominating economically/pop-wise if they grow at high 1900 rates for several decades.

Don't return null for pop figures, but a dummy figure, if there aren't any
available.

Tweak lat/lon adjustment.  For EEIII I should really make a formula based on
the curvature of the Earth and the map, as these tweaks are hacky, but this is
good enough for now.

Write to turn events so I don't have to copy it.

Set resources to average for everything b/c it's a 1.0 = average thing, not a 0-100 thing.

Update TurnEvents.xml based on country-level conflict data.

Tested with the Great War and it's working well.

Set the max XML ID based on the database.  This allows things to not break when setting permanent XML IDs in the DB.

Set flag based on year for Germany, which AFAIK is the only country we have whose flag changes without changing names.

Still imperfect, e.g. Russian cities won't change name or flag under the USSR yet.

Zero-out infrastructure/manufacturing for non-updated cities, which were often higher than non-updated ones.

Set countries/flags of XML cities across all files.  So far they'll stick with the 1900 city/flag the whole time, except where already updated by the base turn events.

Calculate the figures for existing XML-and-DB cities from scratch.  This brings their stats in line with the rest of the DB cities.

1. Interpolate pop figures when needed.  Fixes the "no English manufacturing growth in 1900" problem.
2. Tweak the urbanization calculation.  This makes city size a direct component, and also means regions are no longer uniform in this stat.
3. Always use region 2.  The car pops aren't moddable by region, so this must always be 2.

Also remove a few sysouts that should really be debugging level.

Tweak the lat/longitude offsets for England and Portugal.

Set resources/infrastructure/skill/wage/manufacturing/manufacturing growth, based on initial formulas.

This basically works, we just need better data especially for urbanization levels in earlier years in areas such as Switzerland and Italy.

Calculate the next XML ID automatically based on the source XML file.

All decades are now enabled.

Progress towards interpolating exchange rates; for now it just uses the "before" figure.  Still, that allows calculation of approximate values.  Which does much good for Spain's recovering economy from 1940 to 1950.

1940 is now enabled.

Add some city growth stats for fun.  Birmingham takes the cake for fastest decade of growth so far, at 5.75% between 1910 and 1920.

Spain is now working due to finding an exchange rate from 1940, but the closest to 1950 is 1948 so that'll need interpolation.

Add a general-purpose interpolater, and use it to interpolate GDP values for countries when we don't have exact data for specific years, e.g. Italy has 1936 and 1950, but not 1940, so now it's interpolated.

Refactor years into a static variable, enabled 1920, which is now working.

1930 isn't working yet.

Generate the 1910 XML file.

This involved a bit of re-arranging so we try to create population maps up front, and then create the XML cities just prior to exporting to XML.  It does mean there's some re-processing of the "source" EE1 1900 XML file to re-import things, and those cities don't get the nice progression that newer ones do.  This also highlights our lack of interpolation in intermediary years.  Still, it is a step forward.

Automatically run the desktop copier after rebuilding the files.

This avoids potentially forgetting a step and having things not work while testing.