Deprecate Array::nelem() and Range::nelem() in favour of ::numel(). * liboctave/array/Array.h (Array::nelem) deprecate in favour of numel(). (Array::capacity, Array:: length): change to call numel() directly. These methods will be deprecated soon. * liboctave/array/PermMatrix.h (PermMatrix::nelem): deprecate in favour of numel(). * liboctave/array/Range.h (Range::numel) new method to replace nelem(). (Range::nelem) deprecate in favour of the new method numel. * liboctave/array/Sparse.h (Sparse::nelem) deprecate in favour of nzmax(). This one is secially bad because unlike the other classes, it is different from numel(). * libinterp/corefcn/debug.cc, libinterp/corefcn/jit-typeinfo.cc, libinterp/corefcn/ls-mat4.cc, libinterp/corefcn/lu.cc, libinterp/corefcn/luinc.cc, libinterp/corefcn/max.cc, libinterp/corefcn/pr-output.cc, libinterp/corefcn/rand.cc, libinterp/corefcn/xpow.cc, libinterp/dldfcn/__magick_read__.cc, libinterp/dldfcn/audioread.cc, libinterp/octave-value/ov-base-int.cc, libinterp/octave-value/ov-bool-mat.cc, libinterp/octave-value/ov-flt-re-mat.cc, libinterp/octave-value/ov-perm.cc, libinterp/octave-value/ov-range.cc, libinterp/octave-value/ov-range.h, libinterp/octave-value/ov-re-mat.cc, libinterp/parse-tree/pt-eval.cc, liboctave/array/Array.cc, liboctave/array/CNDArray.cc, liboctave/array/Range.cc, liboctave/array/dNDArray.cc, liboctave/array/fCNDArray.cc, liboctave/array/fNDArray.cc, liboctave/array/idx-vector.cc, liboctave/array/intNDArray.cc, liboctave/numeric/SparseCmplxLU.cc, liboctave/numeric/SparsedbleLU.cc: replace use of nelem() with numel().

--- a/libinterp/corefcn/pr-output.cc	Mon May 25 16:48:47 2015 -0700
+++ b/libinterp/corefcn/pr-output.cc	Sun May 24 02:41:37 2015 +0100
@@ -2664,7 +2664,7 @@
   double base = r.base ();
   double increment = r.inc ();
   double limit = r.limit ();
-  octave_idx_type num_elem = r.nelem ();
+  octave_idx_type num_elem = r.numel ();
 
   if (plus_format && ! pr_as_read_syntax)
     {