Re: R: EXEC_PATH

[John W. Eaton]

On 25-Nov-2010, Marco Atzeri wrote:

| --- Gio 25/11/10, Ben Abbott <address@hidden> ha scritto:
| 
| > > 
| > > print("prova.jpg", "-djpg")
| > > 
| > > it complains that ghostscript and other programs are
| > not in 
| > > EXEC_PATH.
| > > 
| > > gs is on /usr/bin that is missing from EXEC_PATH
| > > 
| > 
| > Does 'system ("which gs")'  return "/usr/bin/gs"?
| > 
| > Ben
| > 
| 
| 
| just build from scratch on a x86_64 fedora 14
| 
| all the documentation break at: 
| 
| ../../run-octave -f -q -H -p ../../../octave/doc/interpreter --eval 
"geometryimages ('voronoi', 'png');"
| warning: print.m: ghostscript not found in EXEC_PATH.
| warning: print.m: Ghostscript binary is not available.
| warning: print.m: epstool binary is not available.
| warning: print.m: fig2dev binary is not available.
| warning: print.m: pstoedit binary is not available.

I checked in the following change to adapt the Octave sources to this
change in meaning for EXEC_PATH:

  http://hg.savannah.gnu.org/hgweb/octave/rev/4ecc7bc5bc83

It looks like there are a few files in Octave Forge packages that will
need similar changes.

It might actually be best if instead of doing something like

  if (file_in_path (getenv ("PATH"), program))

to see if a program exists, we just tried to execute the program in
some way (say asking for a version number) to see if it is found and
works.  That seems better than just looking in the path since that
doesn't tell us that the program will actually run.

jwe