Re: Don't report a programming error when trying to align grob with an empty extent (issue 7533046)

[janek . lilypond]

On 2013/03/24 12:45:38, dak wrote:
> On 2013/03/24 12:37:14, Trevor Daniels wrote:
> > I was a little concerned that problems might
> > result when a non-empty stencil was given an
> > empty extent, but as this passes all tests it
> > looks like this fear was unfounded.

> I don't think your fear is unfounded: there just does not seem to be
any
> sensible behavior when we have a non-empty stencil with an empty
extent: you
> can't align on an empty extent.

I think there may be use-cases when someone could want to keep the
stencil but make its extent empty for some spacing purposes.  However, i
didn't found any example yet.

> If we can detect this case reliably (non-empty stencil more or less
being
> defined as anything putting ink to paper) it may well be worth to warn
about
> this _particular_ variant.

I'll look into this.


https://codereview.appspot.com/7533046/diff/15001/lily/self-alignment-interface.cc
File lily/self-alignment-interface.cc (right):

https://codereview.appspot.com/7533046/diff/15001/lily/self-alignment-interface.cc#newcode153
lily/self-alignment-interface.cc:153: // empty extent isn't a problem -
we simply don't align such grobs.
On 2013/03/24 12:37:14, Trevor Daniels wrote:
> I think this comment would be better associated
> with the if condition so it covers the entire
> conditional clause.

Done.

https://codereview.appspot.com/7533046/