xafs_scan plan, energies are not sorted

[Cathyhjj]

Describe the bug

in this xafs scan, the first point of the second region is smaller than the last energy point of the first region, see the scatter plot of the energy as a function of the point number; This will raise an error when processing with larch, because it verifies that energies should be increasing in autobk function

The data uid: 6fc4ee04-ba36-4642-b518-48a685b2eb1d

[Cathyhjj]

With discussion with @canismarko, this occurs because that was how it was set in the plan scan args

[canismarko]

After discussion with the group, we think it belongs in the haven.energy_ranges.merge_ranges function.

[canismarko]

One option for this is for merge_ranges to round to the precision of the positioner, which will make removing duplicates much easier.