- the output write scaling should be applied *after* undoing the
  effects of the specified rotation centre. Fixes #2566

ENH: update option names for transformPoints and surfaceTransformPoints

- prefer  '-auto-centre' and '-centre', but also accept the previous
  options '-auto-origin' and '-origin' as aliases.

  Changing to '-centre' avoids possible confusion with
  coordinate system origin().

COMP: include <limits>

- add overflow() method to the SHA1 streambuf. Previously could rely
  on xsputn for adding to sha1 content, but streams now check pptr()
  first to test for the buffering range and thus overflow() is needed.

- when used for example with wallShearStress, the stress field is
  initially created as incompressible but later updated with the
  correct compressible/incompressible dimensions.

  If this field is sampled as a surface and stored on the registry
  the dimensions should be reset() and not '=' assigned, since that
  causes a dimension check which will obviously fail.

- add "point data" or "face data" note to the boundaryData output
  for easier diagnositics in the future.

- the generic constructor could be called twice in pointPatchField::New,
  which results in an attempt to transfer tokens twice.

- setup writer outside the data loop to ensure that the number of
  output fields is correct (VTK format).

- ignore 'interpolate' on sampled surfaces to ensure proper
  face sampling, never allow point sampling

BUG: incorrect debug-switch for sampledIsoSurface

- number of fields was not set.
- interpolated surfaces incorrectly written for all formats

Should not call map() since it triggers building the
(one-to-one) mapping which conflicts with the self-contained
mapping of cyclicAMI