gperf

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

gperf

Lippert, Ross A.

I think the problem is that oct-gperf.h isn't being generated properly.
I think I need to install gperf or something.  ./configure did not complain
about this to me at all, and I certainly didn't get any sort of notice that
it had failed other than oct-gperf.h ended up being an empty file.


-r


Reply | Threaded
Open this post in threaded view
|

gperf

John W. Eaton-6
On 30-Oct-2000, Lippert, Ross A. <[hidden email]> wrote:

| I think the problem is that oct-gperf.h isn't being generated properly.
| I think I need to install gperf or something.  ./configure did not complain
| about this to me at all, and I certainly didn't get any sort of notice that
| it had failed other than oct-gperf.h ended up being an empty file.

Octave's configure script does test for gperf, and complains if it
doesn't find it.

Was gperf missing on your system, or was it an old version that didn't
do the right thing?

jwe


Reply | Threaded
Open this post in threaded view
|

RE: gperf

Lippert, Ross A.
In reply to this post by Lippert, Ross A.
gperf does not exist on my system.  @$GPERF in Makeconf was
blank, and the command that got executed when
"$GPERF -a octave.gperf" or whatever got executed was
"-a octave.gperf > oct-gperf.h" with error "-a not found", but
that not being enough an error to stop the make.  Further
downstream, the make derails because oct-gperf.h is an empty file.

I have installed flex, bison, and gperf (the programs which I
think are complained inadequately about) now, and have run from
a fresh CVS this morning and it built just fine.


-r

-----Original Message-----
From: John W. Eaton [mailto:[hidden email]]
Sent: Monday, October 30, 2000 2:44 PM
To: Lippert, Ross A.
Cc: '[hidden email]'
Subject: gperf


On 30-Oct-2000, Lippert, Ross A. <[hidden email]> wrote:

| I think the problem is that oct-gperf.h isn't being generated properly.
| I think I need to install gperf or something.  ./configure did not
complain
| about this to me at all, and I certainly didn't get any sort of notice
that
| it had failed other than oct-gperf.h ended up being an empty file.

Octave's configure script does test for gperf, and complains if it
doesn't find it.

Was gperf missing on your system, or was it an old version that didn't
do the right thing?

jwe


Reply | Threaded
Open this post in threaded view
|

RE: gperf

John W. Eaton-6
On 30-Oct-2000, Lippert, Ross A. <[hidden email]> wrote:

| gperf does not exist on my system.  @$GPERF in Makeconf was
| blank, and the command that got executed when
| "$GPERF -a octave.gperf" or whatever got executed was
| "-a octave.gperf > oct-gperf.h" with error "-a not found", but
| that not being enough an error to stop the make.  Further
| downstream, the make derails because oct-gperf.h is an empty file.
|
| I have installed flex, bison, and gperf (the programs which I
| think are complained inadequately about) now, and have run from
| a fresh CVS this morning and it built just fine.

OK, I added a `missing' script, similar to the one that is distributed
with automake, except that it never tries to create any output files
(it will exit with a warning and failure status).

Now if bison is missing, Octave's configure script will set the YACC
to '$(top_srcdir)/missing bison' so when you try to use $YACC, you
should see a warning message and make should quit.

jwe