|
»
»
Octave as a module for Python
|
View:
New views
8 Messages
—
Rating Filter:
Alert me
|
 |
Octave as a module for Python
by David Grundberg
::
Rate this Message:
Announcing Pytave
An Octave module for Python =========================== Pytave enables Python scripts to use existing m-files (Octave/Matlab scripts) for numerical calculations. Pytave embeds the Octave language interpreter as a module to Python. This new revision allows dictionaries and lists to be passed to Octave. Example use =========== Calling Octave code in the interactive Python interpreter: >>> import pytave >>> pytave.feval(1, "cos", 0) (1.0,) Goals ===== Pytave strives to uphold these points * Good out of the box experience * Good-natured implicit type conversions, no strange PyApple -> octave_orange -> PyBanana chains Features ======== A short list of what Pytave is capable of * Implicit type conversions between Python and Octave. Supports all Numeric integer, real double (and possibly real float) matrices * Architecture independent - no assumption on endian type or integer sizes * Supports cell <-> list and struct <-> dict conversions. (NEW) Project homepage ================ https://launchpad.net/pytave Using/hacking ============= You need the Bazaar version control software (bzr). Branch from trunk with: $ bzr branch lp:pytave You will now have a directory called `pytave' with source code for the module. Read the INSTALL file for building instructions. _______________________________________________ Help-octave mailing list Help-octave@... https://www-old.cae.wisc.edu/mailman/listinfo/help-octave |
|
|
Re: Octave as a module for Python
by Jaroslav Hajek-2
::
Rate this Message:
On Tue, May 5, 2009 at 9:32 PM, David Grundberg <c04dgg@...> wrote:
> Announcing Pytave > > An Octave module for Python > =========================== > > Pytave enables Python scripts to use existing m-files (Octave/Matlab > scripts) for numerical calculations. Pytave embeds the Octave > language interpreter as a module to Python. > > This new revision allows dictionaries and lists to be passed to > Octave. > > Example use > =========== > > Calling Octave code in the interactive Python interpreter: > > >>> import pytave > >>> pytave.feval(1, "cos", 0) > (1.0,) > > Goals > ===== > > Pytave strives to uphold these points > > * Good out of the box experience > > * Good-natured implicit type conversions, no strange PyApple -> > octave_orange -> PyBanana chains > > Features > ======== > > A short list of what Pytave is capable of > > * Implicit type conversions between Python and Octave. Supports all > Numeric integer, real double (and possibly real float) matrices > > * Architecture independent - no assumption on endian type or integer > sizes > > * Supports cell <-> list and struct <-> dict conversions. (NEW) > > Project homepage > ================ > > https://launchpad.net/pytave > > Using/hacking > ============= > > You need the Bazaar version control software (bzr). Branch from trunk > with: > > $ bzr branch lp:pytave > > You will now have a directory called `pytave' with source code for > the module. Read the INSTALL file for building instructions. > > _______________________________________________ > Help-octave mailing list > Help-octave@... > https://www-old.cae.wisc.edu/mailman/listinfo/help-octave > Excellent! I just built pytave with the development Octave on OpenSUSE/x86_64, Python 2.5.2, and all seems OK. Great job. Hopefully no more pipes for calling Octave from Python... I must find time to inspect the code. cheers -- RNDr. Jaroslav Hajek computing expert & GNU Octave developer Aeronautical Research and Test Institute (VZLU) Prague, Czech Republic url: www.highegg.matfyz.cz _______________________________________________ Help-octave mailing list Help-octave@... https://www-old.cae.wisc.edu/mailman/listinfo/help-octave |
|
|
Re: Octave as a module for Python
by Michael Creel
::
Rate this Message:
Can .oct files also be called? Thanks, Michael |
|
|
Re: Octave as a module for Python
by Jaroslav Hajek-2
::
Rate this Message:
On Wed, May 6, 2009 at 10:02 AM, Michael Creel <michael.creel@...> wrote:
> > > > David Grundberg wrote: >> >> Announcing Pytave >> >> An Octave module for Python >> =========================== >> >> Pytave enables Python scripts to use existing m-files (Octave/Matlab >> scripts) for numerical calculations. Pytave embeds the Octave >> language interpreter as a module to Python. >> ... >> > > Can .oct files also be called? > Thanks, Michael > Yes, anything callable with "feval". -- RNDr. Jaroslav Hajek computing expert & GNU Octave developer Aeronautical Research and Test Institute (VZLU) Prague, Czech Republic url: www.highegg.matfyz.cz _______________________________________________ Help-octave mailing list Help-octave@... https://www-old.cae.wisc.edu/mailman/listinfo/help-octave |
|
|
Re: Octave as a module for Python
by Michael Creel
::
Rate this Message:
On Wed, May 6, 2009 at 10:25 AM, Jaroslav Hajek <highegg@...> wrote:
> On Wed, May 6, 2009 at 10:02 AM, Michael Creel <michael.creel@...> wrote: >> >> >> >> David Grundberg wrote: >>> >>> Announcing Pytave >>> >>> An Octave module for Python >>> =========================== >>> >>> Pytave enables Python scripts to use existing m-files (Octave/Matlab >>> scripts) for numerical calculations. Pytave embeds the Octave >>> language interpreter as a module to Python. >>> ... >>> >> >> Can .oct files also be called? >> Thanks, Michael >> > > Yes, anything callable with "feval". > > > -- > RNDr. Jaroslav Hajek > computing expert & GNU Octave developer > Aeronautical Research and Test Institute (VZLU) > Prague, Czech Republic > url: www.highegg.matfyz.cz > Thanks, that's what I suspected. M. _______________________________________________ Help-octave mailing list Help-octave@... https://www-old.cae.wisc.edu/mailman/listinfo/help-octave |
|
|
Re: Octave as a module for Python
by Jaroslav Hajek-2
::
Rate this Message:
On Tue, May 5, 2009 at 9:32 PM, David Grundberg <c04dgg@...> wrote:
> Announcing Pytave > > An Octave module for Python > =========================== > > Pytave enables Python scripts to use existing m-files (Octave/Matlab > scripts) for numerical calculations. Pytave embeds the Octave > language interpreter as a module to Python. > > This new revision allows dictionaries and lists to be passed to > Octave. > > Example use > =========== > > Calling Octave code in the interactive Python interpreter: > > >>> import pytave > >>> pytave.feval(1, "cos", 0) > (1.0,) > > Goals > ===== > > Pytave strives to uphold these points > > * Good out of the box experience > > * Good-natured implicit type conversions, no strange PyApple -> > octave_orange -> PyBanana chains > > Features > ======== > > A short list of what Pytave is capable of > > * Implicit type conversions between Python and Octave. Supports all > Numeric integer, real double (and possibly real float) matrices > > * Architecture independent - no assumption on endian type or integer > sizes > > * Supports cell <-> list and struct <-> dict conversions. (NEW) > > Project homepage > ================ > > https://launchpad.net/pytave > > Using/hacking > ============= > > You need the Bazaar version control software (bzr). Branch from trunk > with: > > $ bzr branch lp:pytave > > You will now have a directory called `pytave' with source code for > the module. Read the INSTALL file for building instructions. > > _______________________________________________ > Help-octave mailing list > Help-octave@... > https://www-old.cae.wisc.edu/mailman/listinfo/help-octave > I have wrapped also the "eval" function into pytave - see the attached patch (created using bzr diff). How should I go about it? Should contributions be reported as bugs using the launchpad tracker? Sorry I have no experience with neither bazaar nor launchpad. summary: "eval" is provided as a wrapper to eval_string (parse.h). In principle, this could be achieved more simply using feval("eval", but the advantages are: 1. faster (avoids double call and double conversion of code string) 2. explicit control of printing rather than implicitly with nargout (as in eval) 3. a separate exception class for parse error regards -- RNDr. Jaroslav Hajek computing expert & GNU Octave developer Aeronautical Research and Test Institute (VZLU) Prague, Czech Republic url: www.highegg.matfyz.cz [pytave_eval.diff] === modified file 'exceptions.cc' --- exceptions.cc 2008-10-18 10:28:00 +0000 +++ exceptions.cc 2009-05-06 09:21:59 +0000 @@ -25,6 +25,7 @@ PyObject *octave_error_exception::excclass = NULL; PyObject *value_convert_exception::excclass = NULL; PyObject *object_convert_exception::excclass = NULL; + PyObject *octave_parse_exception::excclass = NULL; } === modified file 'exceptions.h' --- exceptions.h 2009-05-03 18:24:48 +0000 +++ exceptions.h 2009-05-06 09:28:18 +0000 @@ -55,6 +55,26 @@ }; + class octave_parse_exception { + public: + static bool init() { + excclass = PyErr_NewException( + const_cast<char*>("pytave.ParseError"), + PyExc_RuntimeError, NULL); + return excclass != NULL; + }; + static void translate_exception(octave_parse_exception const &py_ex) { + PyErr_SetString(excclass, py_ex.error.c_str()); + } + static PyObject *excclass; + + octave_parse_exception(std::string err) { error = err; }; + + private: + std::string error; + + }; + class value_convert_exception { public: static bool init() { === modified file 'package/pytave.py' --- package/pytave.py 2009-05-03 18:50:25 +0000 +++ package/pytave.py 2009-05-06 09:46:58 +0000 @@ -96,6 +96,62 @@ return _pytave.feval(nargout, funcname, arguments) +def eval(nargout, code, silent=True): + + """Executes a given Octave code. + + The expression is expected to return nargout values. Returned + values are stored in a tuple. If the nargout argument is less than + or equal to 0, an empty tuple is returned. + + All normal scope and function search rules apply. If silent is + true (default), the result is not auto-printed, as if a semicolon + was appended. Otherwise, auto-printing is enabled. + + See also the Octave documentation for the builtin Octave function + eval. + + Type conversions + **************** + + The following type conversions are supported: + + Octave to Python + ================ + + Scalar values to objects: + bool bool + real scalar float (64-bit) + any string* str + struct dict + cell* list + + * Cell arrays must be one-dimensional (row vector) and + character matrices must only have one row. Any + other form will raise a ValueConvertError. + + Matrix values to Numeric arrays: + int64 LONG + int32, uint32 INT, UINT + int16, uint16 SHORT, USHORT + int8, unint8 SBYTE, UBYTE + + All other values causes a pytave.ValueConvertError to be + raised. This exception inherits TypeError. + + Errors + ****** + + Octave runtime errors are encapsulated into + pytave.OctaveError exceptions, base class RuntimeError. + + If the code can't be parsed, a + pytave.ParseError exception occurs. + + """ + + return _pytave.eval(nargout, code, silent) + def addpath(*arguments): """See Octave documentation""" return _pytave.feval(1, "addpath", arguments)[0] === modified file 'pytave.cc' --- pytave.cc 2009-05-03 18:24:48 +0000 +++ pytave.cc 2009-05-06 09:46:02 +0000 @@ -48,7 +48,8 @@ if (!octave_error_exception::init() || !value_convert_exception::init() - || !object_convert_exception::init()) { + || !object_convert_exception::init() + || !octave_parse_exception::init()) { PyErr_SetString(PyExc_ImportError, "_pytave: init failed"); return; } @@ -77,8 +78,40 @@ object_convert_exception::excclass))); } + string make_error_message (const Octave_map& map) { + ostringstream exceptionmsg; + string message = map.stringfield("message", ""); + string identifier = map.stringfield("identifier", ""); + Cell stackCell = map.contents("stack"); + + // Trim trailing new lines + message = message.substr(0, message.find_last_not_of("\r\n") + 1); + + if (!stackCell.is_empty() && stackCell(0).is_map()) { + // The struct element is called "stack" but only contain + // info about the top frame. + Octave_map stack = stackCell(0).map_value(); + string file = stack.stringfield("file", ""); + string name = stack.stringfield("name", ""); + int line = stack.intfield("line", 1); + int column = stack.intfield("column", 2); + + exceptionmsg << file << ":" << line << ":" << column << ": "; + if (!name.empty()) + exceptionmsg << "in '" << name << "': "; + } + + if (!identifier.empty()) { + exceptionmsg << "(identifier: " << identifier << ") "; + } + exceptionmsg << message; + + return exceptionmsg.str (); + } + + boost::python::tuple func_eval(const int nargout, - const std::string &funcname, + const string &funcname, const boost::python::tuple &arguments) { octave_value_list octave_args, retval; @@ -102,35 +135,52 @@ octave_value_list lasterror = eval_string("lasterror", true, parse_status, 1); if (!lasterror.empty() && lasterror(0).is_map()) { - ostringstream exceptionmsg; - Octave_map map = lasterror(0).map_value(); - string message = map.stringfield("message", ""); - string identifier = map.stringfield("identifier", ""); - Cell stackCell = map.contents("stack"); - - // Trim trailing new lines - message = message.substr(0, message.find_last_not_of("\r\n") + 1); - - if (!stackCell.is_empty() && stackCell(0).is_map()) { - // The struct element is called "stack" but only contain - // info about the top frame. - Octave_map stack = stackCell(0).map_value(); - string file = stack.stringfield("file", ""); - string name = stack.stringfield("name", ""); - int line = stack.intfield("line", 1); - int column = stack.intfield("column", 2); - - exceptionmsg << file << ":" << line << ":" << column << ": "; - if (!name.empty()) - exceptionmsg << "in '" << name << "': "; - } - - if (!identifier.empty()) { - exceptionmsg << "(identifier: " << identifier << ") "; - } - exceptionmsg << message; - - throw octave_error_exception(exceptionmsg.str()); + string exceptionmsg = make_error_message (lasterror(0).map_value ()); + throw octave_error_exception(exceptionmsg); + } else + throw octave_error_exception("No Octave error available"); + } + + if (nargout > 0) { + boost::python::tuple pytuple; + octlist_to_pytuple(pytuple, retval); + return pytuple; + } else { + // Return () if nargout <= 0. + return make_tuple(); + } + } + + boost::python::tuple str_eval(int nargout, + const string &code, + bool silent) { + + octave_value_list retval; + int parse_status; + + reset_error_handler(); + buffer_error_messages++; + + Py_BEGIN_ALLOW_THREADS + retval = eval_string(code, silent, parse_status, nargout); + Py_END_ALLOW_THREADS + + if (parse_status != 0 || error_state != 0) { +// error_state values: +// -2 error without traceback +// -1 traceback +// 1 general error + int parse_status1 = 0; + reset_error_handler(); + octave_value_list lasterror = eval_string("lasterror", + true, parse_status1, 1); + if (!lasterror.empty() && lasterror(0).is_map()) { + string exceptionmsg = make_error_message (lasterror(0).map_value ()); + + if (parse_status != 0) + throw octave_parse_exception(exceptionmsg); + else + throw octave_error_exception(exceptionmsg); } else throw octave_error_exception("No Octave error available"); } @@ -151,6 +201,7 @@ def("init", pytave::init); def("feval", pytave::func_eval); + def("eval", pytave::str_eval); def("get_exceptions", pytave::get_exceptions); register_exception_translator<pytave::pytave_exception>( @@ -159,6 +210,9 @@ register_exception_translator<pytave::octave_error_exception>( pytave::octave_error_exception::translate_exception); + register_exception_translator<pytave::octave_parse_exception>( + pytave::octave_parse_exception::translate_exception); + register_exception_translator<pytave::object_convert_exception>( pytave::object_convert_exception::translate_exception); _______________________________________________ Help-octave mailing list Help-octave@... https://www-old.cae.wisc.edu/mailman/listinfo/help-octave |
|
|
Re: Octave as a module for Python
by David Grundberg
::
Rate this Message:
Jaroslav Hajek skrev:
> On Tue, May 5, 2009 at 9:32 PM, David Grundberg <c04dgg@...> wrote: > >> Announcing Pytave > > Hi David, > > I have wrapped also the "eval" function into pytave - see the attached > patch (created using bzr diff). How should I go about it? Should > contributions be reported as bugs using the launchpad tracker? Sorry I > have no experience with neither bazaar nor launchpad. > > summary: > "eval" is provided as a wrapper to eval_string (parse.h). In > principle, this could be achieved more simply using feval("eval", but > the advantages are: > 1. faster (avoids double call and double conversion of code string) > 2. explicit control of printing rather than implicitly with nargout (as in eval) > 3. a separate exception class for parse error > > regards > > Glad to see you improving my software! As your patch is of a manageable size, you might as well open a bug report and attach your patch. I have some comments 1. As the code stands, it is not possible to catch pytave.ParseError. ([module pytave has] "no attribute 'ParseError'") In order to be correctly declared, the octave_parse_exception will have to be returned by the _pytave::get_exceptions function and correctly assigned to a symbol named ParseError in pytave.py. 2. It's unnecessary to duplicate the value conversion documentation, I think it'll be easier to maintain if it just referred to feval. I'm attaching a modified patch as a suggestion. I modified the semantics of the nargout parameter since 0 is valid input. Furthermore, I see that the builtin eval function accepts a second code string, CATCH. What are the purpose of this argument, and why did you not include it? sincerely, David === modified file 'exceptions.cc' --- exceptions.cc 2008-10-18 10:28:00 +0000 +++ exceptions.cc 2009-05-06 11:42:53 +0000 @@ -25,6 +25,7 @@ PyObject *octave_error_exception::excclass = NULL; PyObject *value_convert_exception::excclass = NULL; PyObject *object_convert_exception::excclass = NULL; + PyObject *octave_parse_exception::excclass = NULL; } === modified file 'exceptions.h' --- exceptions.h 2009-05-03 18:24:48 +0000 +++ exceptions.h 2009-05-06 12:13:29 +0000 @@ -55,6 +55,25 @@ }; + class octave_parse_exception { + public: + static bool init() { + excclass = PyErr_NewException( + const_cast<char*>("pytave.ParseError"), + PyExc_RuntimeError, NULL); + return excclass != NULL; + }; + static void translate_exception(octave_parse_exception const &py_ex) { + PyErr_SetString(excclass, py_ex.error.c_str()); + } + static PyObject *excclass; + + octave_parse_exception(std::string err) { error = err; }; + + private: + std::string error; + }; + class value_convert_exception { public: static bool init() { === modified file 'package/pytave.py' --- package/pytave.py 2009-05-03 18:50:25 +0000 +++ package/pytave.py 2009-05-06 12:24:27 +0000 @@ -22,16 +22,16 @@ import _pytave _pytave.init() -(OctaveError, ValueConvertError, ObjectConvertError) \ +(OctaveError, ValueConvertError, ObjectConvertError, ParseError) \ = _pytave.get_exceptions(); def feval(nargout, funcname, *arguments): """Executes an Octave function called funcname. - The function is set to return nargout values. Returned values are - stored in a tuple. If the nargout argument is less than or equal - to 0, an empty tuple is returned. + The function is set to return nargout values. Returned values + are stored in a tuple. If the nargout argument is less than 0, + an empty tuple is returned. M-files are searched for in the Octave path. @@ -96,6 +96,41 @@ return _pytave.feval(nargout, funcname, arguments) +def eval(nargout, code, silent=True): + + """Executes a given Octave code. + + The expression is expected to return nargout values. Returned + values are stored in a tuple. If the nargout argument is less + than 0, an empty tuple is returned. + + All normal scope and function search rules apply. If silent is + true (default), the result is not auto-printed, as if a + semicolon was appended. Otherwise, auto-printing is enabled. + + See also the Octave documentation for the builtin Octave + function eval. + + For information about returned value conversion, see + pytave.feval. + + Errors + ****** + + If the code cannot be parsed, a pytave.ParseError exception + occurs. + + Octave runtime errors are encapsulated into pytave.OctaveError + exceptions, base class RuntimeError. + + If the resulting values cannot be converted, a + pytave.ValueConvertError is raised. This exception inherits + TypeError. + + """ + + return _pytave.eval(nargout, code, silent) + def addpath(*arguments): """See Octave documentation""" return _pytave.feval(1, "addpath", arguments)[0] === modified file 'pytave.cc' --- pytave.cc 2009-05-03 18:24:48 +0000 +++ pytave.cc 2009-05-06 12:23:34 +0000 @@ -48,7 +48,8 @@ if (!octave_error_exception::init() || !value_convert_exception::init() - || !object_convert_exception::init()) { + || !object_convert_exception::init() + || !octave_parse_exception::init()) { PyErr_SetString(PyExc_ImportError, "_pytave: init failed"); return; } @@ -74,11 +75,45 @@ object(handle<PyObject>( value_convert_exception::excclass)), object(handle<PyObject>( - object_convert_exception::excclass))); - } - + object_convert_exception::excclass)), + object(handle<PyObject>( + octave_parse_exception::excclass))); + } + + string make_error_message (const Octave_map& map) { + ostringstream exceptionmsg; + string message = map.stringfield("message", ""); + string identifier = map.stringfield("identifier", ""); + Cell stackCell = map.contents("stack"); + + // Trim trailing new lines + message = message.substr(0, message.find_last_not_of("\r\n") + 1); + + if (!stackCell.is_empty() && stackCell(0).is_map()) { + // The struct element is called "stack" but only contain + // info about the top frame. + Octave_map stack = stackCell(0).map_value(); + string file = stack.stringfield("file", ""); + string name = stack.stringfield("name", ""); + int line = stack.intfield("line", 1); + int column = stack.intfield("column", 2); + + exceptionmsg << file << ":" << line << ":" << column << ": "; + if (!name.empty()) + exceptionmsg << "in '" << name << "': "; + } + + if (!identifier.empty()) { + exceptionmsg << "(identifier: " << identifier << ") "; + } + exceptionmsg << message; + + return exceptionmsg.str (); + } + + boost::python::tuple func_eval(const int nargout, - const std::string &funcname, + const string &funcname, const boost::python::tuple &arguments) { octave_value_list octave_args, retval; @@ -89,7 +124,7 @@ buffer_error_messages++; Py_BEGIN_ALLOW_THREADS - retval = feval(funcname, octave_args, nargout); + retval = feval(funcname, octave_args, (nargout >= 0) ? nargout : 0); Py_END_ALLOW_THREADS if (error_state != 0) { @@ -102,45 +137,63 @@ octave_value_list lasterror = eval_string("lasterror", true, parse_status, 1); if (!lasterror.empty() && lasterror(0).is_map()) { - ostringstream exceptionmsg; - Octave_map map = lasterror(0).map_value(); - string message = map.stringfield("message", ""); - string identifier = map.stringfield("identifier", ""); - Cell stackCell = map.contents("stack"); - - // Trim trailing new lines - message = message.substr(0, message.find_last_not_of("\r\n") + 1); - - if (!stackCell.is_empty() && stackCell(0).is_map()) { - // The struct element is called "stack" but only contain - // info about the top frame. - Octave_map stack = stackCell(0).map_value(); - string file = stack.stringfield("file", ""); - string name = stack.stringfield("name", ""); - int line = stack.intfield("line", 1); - int column = stack.intfield("column", 2); - - exceptionmsg << file << ":" << line << ":" << column << ": "; - if (!name.empty()) - exceptionmsg << "in '" << name << "': "; - } - - if (!identifier.empty()) { - exceptionmsg << "(identifier: " << identifier << ") "; - } - exceptionmsg << message; - - throw octave_error_exception(exceptionmsg.str()); - } else - throw octave_error_exception("No Octave error available"); - } - - if (nargout > 0) { - boost::python::tuple pytuple; - octlist_to_pytuple(pytuple, retval); - return pytuple; - } else { - // Return () if nargout <= 0. + string exceptionmsg = make_error_message(lasterror(0).map_value ()); + throw octave_error_exception(exceptionmsg); + } else + throw octave_error_exception("No Octave error available"); + } + + if (nargout >= 0) { + boost::python::tuple pytuple; + octlist_to_pytuple(pytuple, retval); + return pytuple; + } else { + // Return () if nargout < 0. + return make_tuple(); + } + } + + boost::python::tuple str_eval(int nargout, + const string &code, + bool silent) { + + octave_value_list retval; + int parse_status; + + reset_error_handler(); + buffer_error_messages++; + + Py_BEGIN_ALLOW_THREADS + retval = eval_string(code, silent, parse_status, + (nargout >= 0) ? nargout : 0); + Py_END_ALLOW_THREADS + + if (parse_status != 0 || error_state != 0) { +// error_state values: +// -2 error without traceback +// -1 traceback +// 1 general error + int parse_status1 = 0; + reset_error_handler(); + octave_value_list lasterror = eval_string("lasterror", + true, parse_status1, 1); + if (!lasterror.empty() && lasterror(0).is_map()) { + string exceptionmsg = make_error_message (lasterror(0).map_value ()); + + if (parse_status != 0) + throw octave_parse_exception(exceptionmsg); + else + throw octave_error_exception(exceptionmsg); + } else + throw octave_error_exception("No Octave error available"); + } + + if (nargout >= 0) { + boost::python::tuple pytuple; + octlist_to_pytuple(pytuple, retval); + return pytuple; + } else { + // Return () if nargout < 0. return make_tuple(); } } @@ -151,6 +204,7 @@ def("init", pytave::init); def("feval", pytave::func_eval); + def("eval", pytave::str_eval); def("get_exceptions", pytave::get_exceptions); register_exception_translator<pytave::pytave_exception>( @@ -159,6 +213,9 @@ register_exception_translator<pytave::octave_error_exception>( pytave::octave_error_exception::translate_exception); + register_exception_translator<pytave::octave_parse_exception>( + pytave::octave_parse_exception::translate_exception); + register_exception_translator<pytave::object_convert_exception>( pytave::object_convert_exception::translate_exception); === modified file 'test/test.py' --- test/test.py 2009-05-05 18:57:24 +0000 +++ test/test.py 2009-05-06 12:26:38 +0000 @@ -86,12 +86,28 @@ except Exception, e: print "FAIL", (value,), e +def testparseerror(*value): + try: + print pytave.eval(*value); + print "FAIL:", (value,) + except pytave.ParseError: + pass + except Exception, e: + print "FAIL", (value,), e + def testvalueok(*value): try: pytave.feval(1, *value); except Exception, e: print "FAIL", (value,), e +def testevalexpect(numargout, code, expectations): + try: + results = pytave.eval(numargout, code); + if results != expectations: + print "FAIL: eval: ", code, " because", results, " != ", expectations, "," + except Exception, e: + print "FAIL: eval:", code, ":", e def testcellinvariant(value): pass @@ -200,4 +216,7 @@ if result.shape != (3, 1): print "FAIL: expected 3x1 matrix" - +testparseerror(1, "endfunction") +testevalexpect(1, "2 + 2", (4,)) +testevalexpect(0, "{2}", ([2],)) +testevalexpect(2, "struct('foo', 2)", ({'foo': [2]},)) _______________________________________________ Help-octave mailing list Help-octave@... https://www-old.cae.wisc.edu/mailman/listinfo/help-octave |
|
|
Re: Octave as a module for Python
by Michael Creel
::
Rate this Message:
I have this built, and the tests work. I can't seem to get it to evaluate an .m file, though. Could I get an example of python code that evaluates the file hello.m which contains the single line
printf("Hello, world\n"); please? Thanks. If I get going a little bit I'll direct my questions to Lauchpad. Thanks again, Michael |
| Free embeddable forum powered by Nabble | Forum Help |
