From: Simon M. <sim...@sc...> - 2013-11-22 09:40:48
|
First, I never tried SWIG Scilab on OSX. For now, SWIG Scilab is supposed to run only on Linux, it is mentioned now in the documentation page. We'll try to support other platforms in the (near ?) future. But, let's have a look at your problem. On OSX, it is clear that your configure failed when finding scilab. I suspect the tilde in (in ~/Desktop) to be the source of problem. Could you try again with an absolute path /home/.... ? And if this errors occurs again, please send us the log of configure (config.log). For Xubuntu, maybe we have the same issue, I am not sure. Could you test also a configure with an absolute path ? Simon Le 21/11/2013 23:49, Ian Bell a écrit : > No good on linux or OSX. New errors now... > > Trying on xubuntu, I build swig with > > git clone https://github.com/swig/swig -b gsoc2012-scilab swig-scilab > cd swig-scilab > sh autogen.sh > ./configure --with-scilab-inc=~/Desktop/scilab-5.4.1/include > --with-scilab=~/Desktop/scilab-5.4.1/bin/scilab-cli > --prefix=/home/xubuntu/Desktop/swig-bin --disable-ccache > make > > then cd swig-scilab/Examples/scilab/simple > > when I run make, I get > > Scilab 5.4.1 (Mar 29 2013, 16:52:15) > !--error 10000 > ilib_gen_Make: An error occurred during the detection of the > compiler(s). Set ilib_verbose(2) for more information. > at line 24 of function generateConfigure called by : > at line 56 of function ilib_gen_Make_unix called by : > at line 49 of function ilib_gen_Make called by : > at line 99 of function ilib_build called by : > ilib_build(ilib_name, table, files, libs, [], ldflags, cflags); > at line 24 of exec file called by : > exec('SCI/etc/scilab.start',-1);;exec('builder.sce',-1) > > Argh! Am I doing something wrong? Or is it not working properly? > > OSX, same problem: > > Ians-Mac-mini:Desktop Ian$ cd swig-scilab/Examples/scilab/simple/ > Ians-Mac-mini:simple Ian$ make > make -f ../../Makefile SRCS='example.c' SWIG='../../../preinst-swig' \ > TARGET='example_wrap.c' INTERFACE='example.i' scilab > ../../../preinst-swig -scilab -addsrc "example.c" example.i > if [ -f builder.sce ]; then \ > env LD_LIBRARY_PATH=.:$LD_LIBRARY_PATH MAKEFLAGS="-j1" -f > builder.sce; \ > fi > env: -f: No such file or directory > make[1]: *** [scilab] Error 127 > make: *** [build] Error 2 > > > Ian > > > On Sun, Nov 17, 2013 at 12:17 AM, Ian Bell <ian...@gm... > <mailto:ian...@gm...>> wrote: > > I'm away from my development computers right now, but I will let > you know as soon as I get home whether I can get the examples to > build properly now. *fingers crossed*. > > And then on to my own code. > > Ian > > > On Fri, Nov 15, 2013 at 12:25 PM, Simon Marchetto > <sim...@sc... > <mailto:sim...@sc...>> wrote: > > Le 13/11/2013 19:43, Ian Bell a écrit : >> Simon, >> >> That macro definition seems to be part of the problem, but >> there do seem to be quite a lot of usability issues with >> 5.4.1 and swig. I would be very appreciative if you could >> take a look. Ideally it would be great if you could run the >> SWIG test suite before you release a new version to be sure >> it isn't going to break things. >> >> Ian >> >> > I had the same compilation error as you: > > ilib_compile: An error occurred during the compilation: > In file included from example_wrap.c:759:0: > /home/ian/Desktop/scilab-5.4.1/share/scilab/../../include/scilab/stack-c.h:19:6: > error: #error Using stack-c.h is deprecated. Please use > api_scilab instead (try 'help api_scilab'). Note the stack-c.h > API will be removed after Scilab 6.0. You can define > __USE_DEPRECATED_STACK_FUNCTIONS__ > > I committed the fix that does the define of > __USE_DEPRECATED_STACK_FUNCTIONS__. > > With that fix, on an installed Scilab 5.4.1 on a Debian 7.10 > x64, I successfully ran all the examples. > I ran also the test suite. The test suite has still yet some > problems, essentially about typemaps, but these problems are > known and are planned to be to fixed (I hope soon....). > For me 5.4.1 should run as same as 5.3.3, and master 5.5. > > Could you give me some details about your issues ? > > Simon > > PS: the example in documentation (which comes from the simple > example you are testing) was wrong, it did not run. > I fixed this, among many other things in the documentation, > which was outdated (but there is still some work to do). > > > |