Re: How to convert unicode string to unsigned char *

[Simon Posnjak]

On Mon, May 5, 2008 at 4:16 PM, Jean-Paul Calderone <ex*****@divmod.comwrote:

>
On Mon, 5 May 2008 16:05:08 +0200, Simon Posnjak <sp******@gmail.comwrote:

On Mon, May 5, 2008 at 3:48 PM, Jean-Paul Calderone <ex*****@divmod.com>

wrote:

On Mon, 5 May 2008 15:41:08 +0200, Simon Posnjak <sp******@gmail.com>

wrote:

>
Hi!

I have a C module for which I created a wrapper with swig. The

function

def is:

C:

int some_thing(unsigned char * the_str);

eg:

Python:

some_module.some_thing (the_str)

Now I would like to feed it with a UTF-8 formatted string:

test = u'Make \u0633\u0644\u0627\u0645, not war.'

>
`test´ is not a UTF-8 encoded string. It's a unicode string.
>
To get a UTF-8 encoded string from a unicode string, use the `encode´
method:
>
some_module.some_thing(test.encode('utf-8'))
>

Yes you are correct. It is unicode string. But still if I use encode I
get the same error:

TypeError with message: in method 'some_thing', argument 1 of type
'unsigned char *'

So I am looking for a way to "cast" unicode string to unsigned char *.

You need to provide some more information about `some_module.some_thing´.
How is it implemented? What Python type does it expect? If it doesn't
take a unicode string and it doesn't take a byte string, I don't know
what kind of string it does take.

some_module.some_thing(the_string) function is a swig generated
function from a C lib. The C lib function expects unsigned char *.

The generated function is:

SWIGINTERN PyObject *_wrap_some_thing(PyObject *SWIGUNUSEDPARM(self),
PyObject *args) {
PyObject *resultobj = 0;
unsigned char *arg1 = (unsigned char *) 0 ;
unsigned char result;
void *argp1 = 0 ;
int res1 = 0 ;
PyObject * obj0 = 0 ;

if (!PyArg_ParseTuple(args,(char
*)"O:cpot_printer_simple_printf",&obj0)) SWIG_fail;
res1 = SWIG_ConvertPtr(obj0, &argp1,SWIGTYPE_p_unsigned_char, 0 | 0 );
if (!SWIG_IsOK(res1)) {
SWIG_exception_fail(SWIG_ArgError(res1), "in method '"
some_thing"" "', argument " "1"" of type '" "unsigned char *""'");
}
arg1 = (unsigned char *)(argp1);
result = (unsigned char)some_thing(arg1);
resultobj = SWIG_From_unsigned_SS_char((unsigned char)(result));
return resultobj;
fail:
return NULL;
}

[=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=]

some_module.some_thing(the_string) function is a swig generated

function from a C lib. The C lib function expects unsigned char *.

The generated function is:

If you don't want to change the generated function, I recommend to
put a wrapper around it, as Jean-Paul suggested:

def some_thing(s):
return some_thing_real(s.encode("utf-8"))

Regards,
Martin