[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[aspell-devel] Ported to Solaris / Sun WorkShop 6 compiler
From: |
Flemming Frandsen |
Subject: |
[aspell-devel] Ported to Solaris / Sun WorkShop 6 compiler |
Date: |
Tue, 20 Jan 2004 10:05:16 +0100 |
Kevin Atkinson wrote:
> If that does not work (it should) than use the get method as I said.
Sorry, I got confused there, I've changed the code to use mumble.get() and it
works ok.
> There should not be a C compiler involved you are linking one C++
> library with another, both of which will compiled with the same
> compiler.
Uhm, well, the problem is that the makefile that the example program (source is
example/example-c.c) is using the C compiler for linking (it's a C program so
who could blame it), but that means that libaspell.so doesn't get linked to the
stl lib.
That's why the build script changed CC=cc to CC=CC in the examples/Makefile
> Thats fine but please remove your "There is no good reason" comment
> and reverse the if (ie use ifndef). if you don't do I will so its not
> a major deal.
Ok, done.
> There is one more thing I noticed. What is going on with
> string_buffer?
For some reason the symbols defined in string_buffer.cpp do not get linked into
libaspell.so, the error I get is:
CC -g -o .libs/example-c example-c.o ../lib/.libs/libaspell.so
-R/home/ffr/projects/spell/test/lib
ild: (undefined symbol) acommon::StringBuffer::StringBuffer() -- referenced in
../lib/.libs/libaspell.so
ild: (undefined symbol) char*acommon::StringBuffer::alloc(unsigned) --
referenced in ../lib/.libs/libaspell.so
I think there is some problem with a makefile somewhere, but I can't really
change it, making the code inline in stead fixes the problem, I'd be happy to
hear how to fix the problem correctly.
The strange thing is that it works fine with gcc, so it may be a similar
problem to the one with iostreams.hpp in prog/aspell.cpp, but I really can't
make heads or tails of the make files.
I could ifdef the change, but as the implementation is rather trivial I think
it's valid to keep it inline in general.
> Ok I will include that. Can I direct people too you if they have
> question?
Sure.
--
Flemming Frandsen / Systems Designer
solaris-v4-aspell-0.50.4.1.diff.tar.gz
Description: Binary data