Hi, sorry to come in with so many problems, I am generally quite please with MacPorts. However, it seems that the Python/Science ports I am trying to install these days somehow don't work for me... Anyway, that's what I get:
sudo port install py-gsl ---> Fetching py-gsl ---> Verifying checksum(s) for py-gsl ---> Extracting py-gsl ---> Configuring py-gsl ---> Building py-gsl with target build Error: Target com.apple.build returned: shell command "cd "/opt/local/ var/db/dports/build/ _opt_local_var_db_dports_sources_rsync.rsync.darwinports.org_dpupdate_dp orts_python_py-gsl/work/pygsl-0.2.0" && CFLAGS="-I/opt/local/include - I/opt/local/include/gsl" LDFLAGS=-L/opt/local/lib /opt/local/bin/ python2.4 setup.py build" returned error 1 Command output: swig_src/block_wrap.c: In function 'SWIG_ConvertPtr': swig_src/block_wrap.c:450: warning: suggest explicit braces to avoid ambiguous 'else' swig_src/block_wrap.c: In function '_wrap_gsl_vector_complex_set_all': swig_src/block_wrap.c:16391: warning: unused variable 'tmp2' swig_src/block_wrap.c: In function '_wrap_gsl_matrix_complex_set_all': swig_src/block_wrap.c:17102: warning: unused variable 'tmp2' swig_src/block_wrap.c: At top level: swig_src/block_wrap.c:170: warning: 'SWIG_TypeDynamicCast' defined but not used swig_src/block_wrap.c:182: warning: 'SWIG_TypeName' defined but not used swig_src/block_wrap.c:188: warning: 'SWIG_TypeQuery' defined but not used swig_src/block_wrap.c:370: warning: 'SWIG_addvarlink' defined but not used swig_src/block_wrap.c:423: warning: 'SWIG_ConvertPtr' defined but not used swig_src/block_wrap.c:515: warning: 'SWIG_ConvertPacked' defined but not used /usr/bin/gcc-4.0 -bundle -undefined dynamic_lookup -L/opt/local/lib - I/opt/local/include -I/opt/local/include/gsl build/temp.darwin-8.9.1- i386-2.4/swig_src/block_wrap.o -L/opt/local/lib -Lbuild/ temp.darwin-8.9.1-i386-2.4 -lpygsl -lgsl -lgslcblas -lm -lpygsl -o build/lib.darwin-8.9.1-i386-2.4/pygsl/__block.so building '_siman' extension creating build/temp.darwin-8.9.1-i386-2.4/src /usr/bin/gcc-4.0 -fno-strict-aliasing -Wno-long-double -no-cpp- precomp -fno-common -fno-common -dynamic -DNDEBUG -g -O3 -Wall - Wstrict-prototypes -I/opt/local/include -I/opt/local/include/gsl - DSWIG_COBJECT_TYPES=1 -DNUMERIC=1 -DPYGSL_GSL_MAJOR_VERSION=1 - DPYGSL_GSL_MINOR_VERSION=9 -UNDEBUG -I/opt/local/include -IInclude - I. -I/opt/local/Library/Frameworks/Python.framework/Versions/2.4/ include/python2.4 -I/opt/local/include/python2.4 -c src/simanmodule.c -o build/temp.darwin-8.9.1-i386-2.4/src/simanmodule.o src/simanmodule.c: In function 'PyGSL_siman_efunc': src/simanmodule.c:129: error: array initialized from non-constant array expression src/simanmodule.c: In function 'PyGSL_siman_step': src/simanmodule.c:189: error: array initialized from non-constant array expression src/simanmodule.c: In function 'PyGSL_siman_metric': src/simanmodule.c:243: error: array initialized from non-constant array expression src/simanmodule.c: In function 'PyGSL_siman_print': src/simanmodule.c:302: error: array initialized from non-constant array expression src/simanmodule.c: In function 'PyGSL_siman_copy': src/simanmodule.c:352: error: array initialized from non-constant array expression src/simanmodule.c: In function 'PyGSL_siman_solve': src/simanmodule.c:551: error: array initialized from non-constant array expression error: command '/usr/bin/gcc-4.0' failed with exit status 1
Error: Status 1 encountered during processing. Greetings, Jochen -- Fritz-Haber-Institut der MPG -- Department of Molecular Physics Faradayweg 4-6 (C1.03) D-14195 Berlin, Germany phone: +49-30-84135686 fax: +49-30-84135892
Jochen Küpper <jochen@fhi-berlin.mpg.de> on Saturday, April 14, 2007 at 8:23 AM -0800 wrote:
sorry to come in with so many problems, I am generally quite please with MacPorts. However, it seems that the Python/Science ports I am trying to install these days somehow don't work for me...
I just tried it and it failed for me too (Intel), though I forgot to check to see if the errors were the same. Anyway, the version was pretty old (0.20) and since it has no maintainer I updated it to 0.3.2 and it builds for me now. There is a version 0.9.0 but I couldn't get it to work. So give it a half day or so and do a selfupdate and try again and hopefully 0.3.2 will install ok for you. Mark
Hi Mark,
try again and hopefully 0.3.2 will install ok for you.
Indeed it does work for me;) Thanks a lot! Greetings, Jochen -- Einigkeit und Recht und Freiheit http://www.Jochen- Kuepper.de Liberté, Égalité, Fraternité GnuPG key: CC1B0B4D Sex, drugs and rock-n-roll
Hi Mark, On 14.04.2007, at 22:35, markd@macports.org wrote:
Jochen Küpper <jochen@fhi-berlin.mpg.de> on Saturday, April 14, 2007 at 8:23 AM -0800 wrote:
sorry to come in with so many problems, I am generally quite please with MacPorts. However, it seems that the Python/Science ports I am trying to install these days somehow don't work for me...
I just tried it and it failed for me too (Intel), though I forgot to check to see if the errors were the same. Anyway, the version was pretty old (0.20) and since it has no maintainer I updated it to 0.3.2 and it builds for me now. There is a version 0.9.0 but I couldn't get it to work. So give it a half day or so and do a selfupdate and try again and hopefully 0.3.2 will install ok for you.
As I told you, the 0.3.2-port you prepared seems to work for me as well. However, in repsonse for above message I also got the following response from Pierre Schnizer:
Thanks for your report. Please could you insert the attached file to src/simanmodule.c and send me any remaining errors? Please which compiler version are you using.
Thus I downloaded the pygsl-0.9.0 tarball from sourceforge and run python setup.py build in there -- simply worked for me (This is on a PPC machine). Since 0.3.2 is quite old and also does not support numpy, IIUC, maybe we can get a 0.9.0 port? What was your problem with the 0.9.0 version? - Would the new simanmodule.c help? - Could it be you don't have py-numpy installed, but that y-numpy should be a dependency for py-gsl, instead of py-numeric? Maybe sending the output to Pierre or the pygsl ml would immediately show them the problem? Anyway, thank you all for your help! /* * Original Author: * * author: Jochen K"upper * * created: April 2002 * * author: Pierre Schnizer * created: December 2003 * file: pygsl/src/simanmodule.c * $Id: simanmodule.c,v 1.7 2005/01/13 17:54:53 schnizer Exp $ * * Jochen K"upper wrote the original version of this module. In December 2003 I * rewrote it. Now I only support the variable type as it is more pythonic. * */ #include <Python.h> #include <math.h> #include <stdlib.h> #include <stdio.h> #include <setjmp.h> #include <gsl/gsl_siman.h> #include <gsl/gsl_nan.h> #include <pygsl/error_helpers.h> #include <pygsl/general_helpers.h> #include <pygsl/rng.h> #include <pygsl/rng_helpers.h> /* * Common to all objects of one problem * Currently the individual method pointers are not used. Instead the method * is resolved every time. Do you know which method is faster? Is it advicable * to resolve it once and store it here? Pierre */ typedef struct{ /* PyObject * efunc; PyObject * step; PyObject * metric; PyObject * print; */ PyObject * rng; jmp_buf buffer; }pygsl_siman_func_t; /* * The Linked list keeping the reference to the individual objects. * * Necessary as I have to use longjmp as the current implementation does * not allow error propagation yet! */ struct _pygsl_siman_t{ pygsl_siman_func_t *func; PyObject * x; struct _pygsl_siman_t * prev; struct _pygsl_siman_t * next; }; typedef struct _pygsl_siman_t pygsl_siman_t; static const char module_doc[] = "C Implementation needed for the siman module."; static PyObject *module = NULL; static const char filename[] = __FILE__; /* * Naming of the various methods the object has provide as callbacks. */ static const char EFunc_name[] = "EFunc"; static const char Metric_name[] = "Metric"; static const char Step_name[] = "Step"; static const char Clone_name[] = "Clone"; static const char Print_name[] = "Print"; /* * Get a callable method from a object. * * Basically it is a flat wrapper around PyObject_GetAttrString but * checks if the method is callable and adds a Traceback frame * * Flag usage: * flag == 1 Must exist and must be callable * flag == 2 If it exists it must be callable * * For the traceback frame all arguments following the module are needed. * If you do not know the module, you can pass a NULL pointer. */ static PyObject * PyGSL_get_callable_method(PyObject *o, const char * attr, int flag, PyObject *module, const char * filename, const char * func_name, int lineno) { PyObject * method = NULL; FUNC_MESS_BEGIN(); method = PyObject_GetAttrString(o, (char *) attr); if(method == NULL){ if(flag == 1){ PyGSL_add_traceback(module, filename, func_name, lineno); }else if(flag == 2){ /* Clear the error otherwise it will show up later on! */ PyErr_Clear(); } return NULL; } if(!(PyCallable_Check(method))) { /* I must add the method name! I must change it to an more descriptive exception! */ PyGSL_add_traceback(module, (const char *) filename, func_name, lineno); PyErr_SetString(PyExc_TypeError, "Found a attribute which was not callable!" "XXX must add the method name!"); return NULL; } DEBUG_MESS(2, "Found a method at %p", (void *) method); FUNC_MESS_END(); return method; } /* This function type should return the energy of a configuration XP. */ static double PyGSL_siman_efunc(void *xp) { PyObject *result = NULL, *callback = NULL, *arglist = NULL; PyGSL_error_info info; pygsl_siman_t *x; int flag=GSL_EFAILED; double value; /* static char *functionname = __FUNCTION__; */ FUNC_MESS_BEGIN(); assert(xp); x = (pygsl_siman_t *) xp; DEBUG_MESS(2, "Found a pygsl_siman_t at %p and a pygsl_siman_func_t at %p and x at %p", (void *)x, (void *) x->func, (void *) x->x); assert(x); assert(x->func); callback = PyGSL_get_callable_method(x->x, EFunc_name, 1, module, filename, __FUNCTION__, __LINE__); if(callback == NULL) goto fail; info.callback = callback; info.message = __FUNCTION__; info.error_description = "and the description ???"; info.argnum = 1; arglist = PyTuple_New(0); result = PyEval_CallObject(callback, arglist); Py_DECREF(arglist); if((flag = PyGSL_CHECK_PYTHON_RETURN(result, 1, &info)) != GSL_SUCCESS){ PyGSL_add_traceback(module, filename, __FUNCTION__, __LINE__); goto fail; } if((flag = PyGSL_PYFLOAT_TO_DOUBLE(result, &value, &info)) != GSL_SUCCESS){ PyGSL_add_traceback(module, filename, __FUNCTION__, __LINE__); goto fail; } Py_DECREF(result); FUNC_MESS_END(); return value; fail: FUNC_MESS("In Fail"); Py_XDECREF(result); longjmp(x->func->buffer, flag); return GSL_NAN; } /* * This function type should modify the configuration XP using a random step * taken from the generator R, up to a maximum distance of STEP_SIZE. */ static void PyGSL_siman_step(const gsl_rng *r, void *xp, double step_size) { PyObject *result = NULL, *arglist = NULL, *callback = NULL; PyGSL_error_info info; pygsl_siman_t *x; int flag=GSL_EFAILED; /* static const char * functionname = __FUNCTION__; */ FUNC_MESS_BEGIN(); x = (pygsl_siman_t *) xp; DEBUG_MESS(2, "Found x at %p", xp); callback = PyGSL_get_callable_method(x->x, Step_name, 1, module, filename, __FUNCTION__, __LINE__); if(callback == NULL) goto fail; info.callback = callback; info.message = __FUNCTION__; info.error_description = "???"; info.argnum = 1; assert(PyGSL_RNG_Check(x->func->rng)); assert(((PyGSL_rng *) x->func->rng)->rng == r); /* create argument list */ arglist = PyTuple_New(2); PyTuple_SET_ITEM(arglist, 0, x->func->rng); Py_INCREF(x->func->rng); /* Don't forget tuple is owner! */ PyTuple_SET_ITEM(arglist, 1, PyFloat_FromDouble(step_size)); result = PyEval_CallObject(callback, arglist); Py_DECREF(arglist); if((flag = PyGSL_CHECK_PYTHON_RETURN(result, 0, &info)) != GSL_SUCCESS){ PyGSL_add_traceback(module, filename, __FUNCTION__, __LINE__); goto fail; } Py_DECREF(result); FUNC_MESS_END(); return; fail: FUNC_MESS("In Fail"); Py_XDECREF(result); longjmp(x->func->buffer, flag); return; } /* This function type should return the distance between two configurations XP and YP. */ static double PyGSL_siman_metric(void *xp, void *yp) { PyObject *result = NULL, *arglist = NULL, *callback = NULL; PyGSL_error_info info; pygsl_siman_t *x, *y; int flag=GSL_EFAILED; double value; /* static const char * functionname = __FUNCTION__; */ FUNC_MESS_BEGIN(); x = (pygsl_siman_t *) xp; y = (pygsl_siman_t *) yp; DEBUG_MESS(2, "Found x at (%p,%p) and y at (%p %p)", (void *) x, (void *)x->x, (void *) y, (void *) y->x); assert(x); assert(y); assert(x->x); assert(y->x); callback = PyGSL_get_callable_method(x->x, Metric_name, 1, module, filename, __FUNCTION__, __LINE__); if(callback == NULL) goto fail; info.callback = callback; info.message = __FUNCTION__; info.error_description = "???"; info.argnum = 1; arglist = PyTuple_New(1); PyTuple_SET_ITEM(arglist, 0, y->x); Py_INCREF(y->x); /* Tuple is owner! */ result = PyEval_CallObject(callback, arglist); Py_XDECREF(arglist); if((flag = PyGSL_CHECK_PYTHON_RETURN(result, 0, &info)) != GSL_SUCCESS){ PyGSL_add_traceback(module, filename, __FUNCTION__, __LINE__); goto fail; } if((flag = PyGSL_PYFLOAT_TO_DOUBLE(result, &value, &info)) != GSL_SUCCESS){ PyGSL_add_traceback(module, filename, __FUNCTION__, __LINE__); goto fail; } Py_DECREF(result); FUNC_MESS_END(); return value; fail: FUNC_MESS("In Fail"); Py_XDECREF(result); longjmp(x->func->buffer, flag); return GSL_NAN; } /* This function type should print the contents of the configuration XP. */ static void PyGSL_siman_print(void *xp) { PyObject *result = NULL, *callback=NULL, *arglist=NULL; PyGSL_error_info info; pygsl_siman_t *x; int flag=GSL_EFAILED; /* static const char * functionname = __FUNCTION__; */ FUNC_MESS_BEGIN(); x = (pygsl_siman_t *) xp; callback = PyGSL_get_callable_method(x->x, Print_name, 1, module, filename, __FUNCTION__, __LINE__); if(callback == NULL) goto fail; info.callback = callback; info.message = __FUNCTION__; info.error_description = "what goes here ???"; info.argnum = 1; arglist = PyTuple_New(0); result = PyEval_CallObject(callback, arglist); Py_DECREF(arglist); if((flag = PyGSL_CHECK_PYTHON_RETURN(result, 0, &info)) != GSL_SUCCESS){ PyGSL_add_traceback(module, (const char*)filename, __FUNCTION__, __LINE__); goto fail; } Py_DECREF(result); FUNC_MESS_END(); return; fail: FUNC_MESS("In Fail"); Py_XDECREF(result); longjmp(x->func->buffer, flag); return; } /* * As Python is generating the new objects I can not skip the reference * counting, but must add a reference for each object in construct and * depose it here and in the siman_destroy function. * * siman_solve will call siman_release_x anyway at the end to clear up * all objects. Necessary to non existing error propagation in the siman * module. */ static void PyGSL_siman_copy(void *src, void *dst) { PyObject *callback=NULL, *new = NULL, *arglist = NULL; PyGSL_error_info info; pygsl_siman_t *x, *y; int flag=GSL_EFAILED; /* static const char * functionname = __FUNCTION__; */ FUNC_MESS_BEGIN(); x = (pygsl_siman_t *) src; y = (pygsl_siman_t *) dst; DEBUG_MESS(2, "Got source at %p, Destination at %p", (void *) x, (void *) y); assert(x->x); callback = PyGSL_get_callable_method(x->x, Clone_name, 1, module, filename, __FUNCTION__, __LINE__); if(callback == NULL) goto fail; arglist = PyTuple_New(0); new = PyEval_CallObject(callback, arglist); Py_DECREF(arglist); info.callback = callback; info.message = __FUNCTION__; info.error_description = "???"; info.argnum = 1; if((flag = PyGSL_CHECK_PYTHON_RETURN(new, 1, &info)) != GSL_SUCCESS){ PyGSL_add_traceback(module, (const char*)filename, __FUNCTION__, __LINE__); goto fail; } Py_XDECREF(y->x); /* Py_INCREF(new); Necessary? I don't think so. */ y->x = new; FUNC_MESS_END(); return; fail: FUNC_MESS("Fail"); Py_XDECREF(new); longjmp(x->func->buffer, flag); } static void * PyGSL_siman_copy_construct(void *new) { pygsl_siman_t * ret, * n, *p; int flag = GSL_ENOMEM; FUNC_MESS_BEGIN(); n = (pygsl_siman_t *) new; /* The pointers next and prev are checked against NULL */ ret = (pygsl_siman_t *) calloc(1, sizeof(pygsl_siman_t)); DEBUG_MESS(2, "New was %p, Constructed a new object at %p", new, (void *) ret); if(ret == NULL){ gsl_error("Could not allocate the object for the linked list", filename, __LINE__ - 3, GSL_ENOMEM); goto fail; } /* Put the Link to the old object so that I can clone when I need to copy */ ret->x = n->x; /* Eventually I will dispose the object */ Py_INCREF(ret->x); /* Pointer to the func struct. This information is the same for all objects */ ret->func = n->func; /* Find the first open object in the linked list ... */ p = n; while(p->next != NULL){ p = p->next; } DEBUG_MESS(2, "I found a open object at %p", (void *) p); /* and connect the links */ p->next = ret; ret->prev = p; FUNC_MESS_END(); return ret; fail: FUNC_MESS("Fail"); longjmp(n->func->buffer, flag); return NULL; } static void PyGSL_siman_destroy(void * old) { pygsl_siman_t * o; FUNC_MESS_BEGIN(); o = (pygsl_siman_t *) old; assert(o); /* fprintf(stderr, "Destroying: Previous object = %p, Next Object = % p\n", (void *)o->prev, (void *)o->next); */ /* Reconnect the linked list */ if (o->prev && o->next){ /* Connect the both */ o->prev->next = o->next; o->next->prev = o->prev; } else if (o->prev && o->next == NULL){ /* Prev last element. Terminate the list */ o->prev->next = NULL; } else if (o->prev == NULL && o->next == NULL){ /* Last Element, better to leave it */ DEBUG_MESS(2, "I do not dispose the last element %p!", (void *) o); return; } /* Dispose the object */ Py_XDECREF(o->x); free(o); FUNC_MESS_END(); } /* Clean up of the linked list of objects */ int PyGSL_siman_release_x(pygsl_siman_t * myargs, pygsl_siman_t * x) { pygsl_siman_t *p=NULL; FUNC_MESS_BEGIN(); p = myargs; /* fprintf(stderr, "Releasing list!\n"); */ while(1){ /* fprintf(stderr, "Previous object = %p, Next Object = %p\n", (void *)p->prev, (void *)p->next); */ /* Don't delete the object containing the result! */ if(p != x){ /* fprintf(stderr, "Deleting object at %p\n", (void *) p); */ PyGSL_siman_destroy((void *) p); } if(p->next == NULL) break; p = p->next; } FUNC_MESS_END(); return GSL_SUCCESS; } static const char pygsl_siman_solve_doc[] = "Simulated annealing driver.\n\ \n\ Usage:\n\ result = solve(r, x0, ...)\n\ \n\ Input:\n\ r ... a random generator from pygsl.rng\n\ x0 ... a configuration. It must be an object providing the following\n\ methods:\n\ EFunc()\n\ Metric()\n\ Step()\n\ Clone()\n\ If you want to use the print functionality you must provide\n\ the following method:\n\ Print()\n\ \n\ Output:\n\ result ... a object of type x0 with the final value.\n\ \n\ Keywords:\n\ n_tries = 200 ... how many points to try for each step\n\ iters_fixed_T = 10 ... how many iterations at each temperature? \n\ step_size = 10 ... max step size in the random walk\n\ \n\ parameters for the Boltzmann distribution\n\ k = 1.0 ... Boltzmann constant\n\ t_initial = 0.002 ... initial temperature\n\ mu_t = 1.005 ... damping factor for the temperature\n\ t_min = 2.0e-6\n\ \n\ do_print = 0 ... print the status of the annealing process\n\ (== 0: do not print)\n\ ( > 0: print)\n\ "; /* wrapper functions */ static PyObject * PyGSL_siman_solve(PyObject *self, PyObject *args, PyObject *kw) { PyObject *result = NULL; PyObject *efunc = NULL, *step = NULL, *metric = NULL, *clone = NULL, *print = NULL, *r_o = NULL, *x_o=NULL; gsl_rng *rng = NULL; gsl_siman_print_t a_print= PyGSL_siman_print; gsl_siman_params_t params = {200, 10, 10.0, 1.0, 0.002, 1.005, 2.0e-6}; pygsl_siman_func_t myargs_func = {NULL}; pygsl_siman_t myargs = {NULL, NULL, NULL, NULL}; /* static const char * functionname = __FUNCTION__; */ int flag=GSL_EFAILED, do_print=0; void * x0 = NULL; static const char * kwlist[] = {"rng", "x0", "n_tries", "iters_fixed_T", "step_size", "k", "t_initial", "mu_t", "t_min", "do_print", NULL}; FUNC_MESS_BEGIN(); /* python arguments are (rng, x0, settings) */ if(! PyArg_ParseTupleAndKeywords(args, kw, "OO|iidddddi", (char **) kwlist, &r_o, &x_o, ¶ms.n_tries, ¶ms.iters_fixed_T, ¶ms.step_size, ¶ms.k, ¶ms.t_initial, ¶ms.mu_t, ¶ms.t_min, &do_print)) return NULL; /* The following methods must exist */ efunc = PyGSL_get_callable_method(x_o, EFunc_name, 1, module, filename, __FUNCTION__, __LINE__); step = PyGSL_get_callable_method(x_o, Step_name, 1, module, filename, __FUNCTION__, __LINE__); metric = PyGSL_get_callable_method(x_o, Metric_name, 1, module, filename, __FUNCTION__, __LINE__); clone = PyGSL_get_callable_method(x_o, Clone_name, 1, module, filename, __FUNCTION__, __LINE__); if( efunc == NULL || step == NULL || metric == NULL || clone == NULL){ return NULL; } /* optional print */ if(do_print == 0){ a_print = NULL; } else { print = PyGSL_get_callable_method(x_o, Print_name, 1, module, filename, __FUNCTION__, __LINE__); if(print == NULL){ DEBUG_MESS(2, "Did not get a print method! print = %p", print); a_print = NULL; return NULL; } } rng = PyGSL_gsl_rng_from_pyobject(r_o); if(rng == NULL){ return NULL; } /* initialize/assign functions */ Py_INCREF(x_o); /* myargs_func.efunc = efunc; myargs_func.step = step; myargs_func.metric = metric; myargs_func.print = print; */ myargs_func.rng = r_o; myargs.func = &myargs_func; myargs.x = x_o; myargs.prev = NULL; myargs.next = NULL; x0 = (void *) &myargs; DEBUG_MESS(2, "x0 @ %p; myargs at %p; myargs_func at %p", x0, (void *) &myargs, (void *) &myargs_func); DEBUG_MESS(2, "Found a pygsl_siman_t at %p and a pygsl_siman_func_t at %p", (void *) x0, (void *) (((pygsl_siman_t *) x0)->func)); if((flag = setjmp(myargs_func.buffer)) == 0){ FUNC_MESS("Starting siman"); gsl_siman_solve(rng, x0, PyGSL_siman_efunc, PyGSL_siman_step, PyGSL_siman_metric, a_print, PyGSL_siman_copy, PyGSL_siman_copy_construct, PyGSL_siman_destroy, 0, /* Only variable mode supported by this wrapper. */ params); FUNC_MESS("End siman"); }else{ PyGSL_add_traceback(module, filename, __FUNCTION__, __LINE__); goto fail; } Py_DECREF(x_o); DEBUG_MESS(2, "I found x0 at %p", x0); result = ((pygsl_siman_t *) x0)->x; PyGSL_siman_release_x(&myargs, x0); FUNC_MESS_END(); return result; fail: FUNC_MESS("In Fail"); PyGSL_siman_release_x(&myargs, x0); Py_XDECREF(x_o); PyGSL_error_flag(flag); return NULL; } /* module initialization */ static PyMethodDef simanMethods[] = { {"solve", (PyCFunction) PyGSL_siman_solve, METH_VARARGS | METH_KEYWORDS, (char *) pygsl_siman_solve_doc}, {NULL, NULL} /* Sentinel */ }; DL_EXPORT(void) init_siman(void) { PyObject *m = NULL; FUNC_MESS_BEGIN(); m = Py_InitModule("_siman", simanMethods); module = m; init_pygsl(); import_pygsl_rng(); FUNC_MESS_END(); return; } /* * Local Variables: * mode: C * c-file-style: "python" * End: */ Greetings, Jochen -- Fritz-Haber-Institut der MPG -- Department of Molecular Physics Faradayweg 4-6 (C1.03) D-14195 Berlin, Germany phone: +49-30-84135686 fax: +49-30-84135892
Jochen Küpper <jochen@fhi-berlin.mpg.de> on Monday, April 16, 2007 at 8:26 AM -0800 wrote:
As I told you, the 0.3.2-port you prepared seems to work for me as well.
However, in repsonse for above message I also got the following response from Pierre Schnizer:
Thanks for your report. Please could you insert the attached file to src/simanmodule.c and send me any remaining errors? Please which compiler version are you using.
Thus I downloaded the pygsl-0.9.0 tarball from sourceforge and run python setup.py build in there -- simply worked for me (This is on a PPC machine).
Since 0.3.2 is quite old and also does not support numpy, IIUC, maybe we can get a 0.9.0 port?
What was your problem with the 0.9.0 version? - Would the new simanmodule.c help? - Could it be you don't have py-numpy installed, but that y-numpy should be a dependency for py-gsl, instead of py-numeric?
Maybe sending the output to Pierre or the pygsl ml would immediately show them the problem?
Oops, resending with my susbscribed address ....... Hi Jochen, Since installing numpy requires installing a lengthy gcc install, I won't be able to try that as a dependency for py-gsl for awhile. You can try it yourself though. Just edit the py-gsl Portfile and replace the old version and checksum with these. version 0.9.0 checksum c69c3deed25525f153f2f03942e99040 Then uninstall py-gsl and reinstall it and you'll get py-gsl 0.9.0. If it turns out that py-numpy is required to install it, then just let us know and the port can be updated. You could also consider becaoming maintainer of the port if you want since it has no maintainer now. Mark
participants (2)
-
Jochen Küpper
-
markd＠macports.org