Re: dmult extension

[John W. Eaton]

On 26-Mar-2008, Jaroslav Hajek wrote:

| On Wed, Mar 26, 2008 at 2:01 AM, John W. Eaton <address@hidden> wrote:
| > On 10-Mar-2008, Jaroslav Hajek wrote:
| >
| >  | please consider the attached patch. It rewrites dmult to allow scaling
| >  | columns of matrix (like A*diag(B)), and also along arbitrary dimension
| >  | (like dmult(ones(3,3,3,3), [1 2 3], 3)).
| >
| >  Will you please edit this to conform more to the Octave coding
| >  conventions (lower case variable names and whitespace) and add
| >  some tests?
| >
| >  Thanks,
| >
| >  jwe
| >
| 
| done. See attached changeset.

Thanks, I applied it.

I think this

| ## @deftypefn {Function File} address@hidden = dmult (@var{a}, @var{b})

should be

  @deftypefn {Function File} address@hidden =} dmult (@var{a}, @var{b})

The arguments for the @deftypefn macro are

  CATEGORY DATA-TYPE NAME ARGUMENTS ...

We are abusing the DATA-TYPE argument for the return value, but the
way you had it, "=" would be picked up as the NAME argument.  This
probably isn't visible in the Info output, but I think you would
notice it in output from TeX since the function name is set in a
normal typewriter font but the DATA-TYPE and ARGUMENTS are set in a
slanted font.

jwe