[petsc-dev] implementation header incluing ../src

[Barry Smith]

  Is it intended for endusers -> petschash.h (requires decent documentation and error checking) or for use by PETSc -> petsc-private/hashimpl.h 


> On Feb 19, 2015, at 3:47 PM, Matthew Knepley <knepley at gmail.com> wrote:
> 
> On Thu, Feb 19, 2015 at 2:39 PM, Jed Brown <jed at jedbrown.org> wrote:
> Matthew Knepley <knepley at gmail.com> writes:
> 
> > On Thu, Feb 19, 2015 at 1:42 AM, Lisandro Dalcin <dalcinl at gmail.com> wrote:
> >
> >> Please take look to the last line of the following output:
> >>
> >> $ grep include $PETSC_DIR/include/petsc-private/dmpleximpl.h
> >> #include <petscmat.h>       /*I      "petscmat.h"          I*/
> >> #include <petscdmplex.h> /*I      "petscdmplex.h"    I*/
> >> #include <petscbt.h>
> >> #include <petscsf.h>
> >> #include <petsc-private/dmimpl.h>
> >> #include <../src/sys/utils/hash.h>
> >>
> >
> > It does not look like anything else uses hash.h. Can we just move it to
> > include/petschash.h?
> 
> Why does it need to be included by a header?
> 
> I thought something else was using this, but it looks like I am the only one. We
> should put it in a header and clear out the cruft no one uses anymore:
> 
> knepley/feature-harmonic-advection *$:/PETSc3/petsc/petsc-dev$ find ./src -type f | xargs grep hash.h
> find ./src -type f | xargs grep hash.h
> ./src/dm/impls/plex/plexinterpolate.c:#include <../src/sys/utils/hash.h>
> ./src/dm/impls/plex/plexinterpolate.c~:#include <../src/sys/utils/hash.h>
> ./src/dm/impls/plex/plextree.c:#include <../src/sys/utils/hash.h>
> ./src/sys/examples/tests/ex26.c:#include <../src/sys/utils/hash.h>
> ./src/sys/utils/hash.h:#include <khash.h>
>  1knepley/feature-harmonic-advection *$:/PETSc3/petsc/petsc-dev$ find ./include -type f | xargs grep hash.h
> find ./include -type f | xargs grep hash.h
> ./include/petsc-private/dmpleximpl.h:#include <../src/sys/utils/hash.h>
> 
>   Matt
> 
> -- 
> What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.
> -- Norbert Wiener