Converting python string object to c char* using ctypes

Question

I am trying to send 2 strings from Python (3.2) to C using ctypes. This is a small part of my project on my Raspberry Pi. To test if the C function received the strings correctly, I place one of them in a text file.

Python code

string1 = "my string 1"
string2 = "my string 2"

# create byte objects from the strings
b_string1 = string1.encode('utf-8')
b_string2 = string2.encode('utf-8')

# send strings to c function
my_c_function(ctypes.create_string_buffer(b_string1),
              ctypes.create_string_buffer(b_string2))

C code

void my_c_function(const char* str1, const char* str2)
{
    // Test if string is correct
    FILE *fp = fopen("//home//pi//Desktop//out.txt", "w");
    if (fp != NULL)
    {
        fputs(str1, fp);
        fclose(fp);
    }

    // Do something with strings..
}

The problem

Only the first letter of the string appears in the text file.

I've tried many ways to convert the Python string object with ctypes.

ctypes.c_char_p
ctypes.c_wchar_p
ctypes.create_string_buffer

With these conversions I keep getting the error "wrong type" or "bytes or integer address expected instead of str instance".

I hope someone can tell me where it goes wrong. Thanks in advance.

Set my_c_function.argtypes = [ctypes.c_char_p, ctypes.c_char_p]. Then, because the parameters are const, simply call it as my_c_function(b_string1, b_string2). — Eryk Sun
– Eryk Sun, Commented Nov 25, 2014 at 14:38
FYI, a literal backslash character needs to be escaped as "\\", but it's not required for a forward slash. It's just "/home/pi/Desktop/out.txt". — Eryk Sun
– Eryk Sun, Commented Nov 25, 2014 at 14:42
@eryksun Thanks for your reply. It works now, I totally forgot I had argtypes still set on c_wchar_p. About the slashes, I always get them mixed up. — MorphieBlossom
– MorphieBlossom, Commented Nov 25, 2014 at 15:03
Only use buf = ctypes.create_string_buffer(bstr) when the function modifies the string. It's equivalent to buf = (ctypes.c_char * (len(bstr) + 1))(); buf.value = bstr. — Eryk Sun
– Eryk Sun, Commented Nov 25, 2014 at 15:12

TheLizzard · Accepted Answer · 2024-08-12 20:38:22Z

50

Thanks to @ErykSun the solution:

Python code

string1 = "my string 1"
string2 = "my string 2"

# create byte objects from the strings
b_string1 = string1.encode('utf-8')
b_string2 = string2.encode('utf-8')

# send strings to c function
my_c_function.argtypes = [ctypes.c_char_p, ctypes.c_char_p]
my_c_function(b_string1, b_string2)

edited Aug 12, 2024 at 20:38

TheLizzard

7,8002 gold badges16 silver badges43 bronze badges

answered Nov 25, 2014 at 15:07

MorphieBlossom

1,1171 gold badge12 silver badges22 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

hola Over a year ago

my_c_function.argtypes = [ctypes.c_char_p, ctypes_char_p] You mean my_c_function.argtypes = [ctypes.c_char_p, ctypes.char_p] (notice . instead of _)?

MorphieBlossom Over a year ago

Haha, after 5 years someone noticed or felt bothered to point it out. Thanks for it, I edited it in the answer.

hola Over a year ago

You know, Obsessive-Compulsive Disorder!

a11apurva Over a year ago

my_c_function.argtypes = [ctypes.c_char_p, ctypes.char_p] Do you mean my_c_function.argtypes = [ctypes.c_char_p, ctypes.c_char_p] ?

Gaslight Deceive Subvert Over a year ago

This is incorrect. encode does not create NULL-terminated strings which your C function expects.

Aksel · Accepted Answer · 2016-06-01 17:57:46Z

20

I think you just need to use c_char_p() instead of create_string_buffer().

string1 = "my string 1"
string2 = "my string 2"

# create byte objects from the strings
b_string1 = string1.encode('utf-8')
b_string2 = string2.encode('utf-8')

# send strings to c function
my_c_function(ctypes.c_char_p(b_string1),
              ctypes.c_char_p(b_string2))

If you need mutable strings then use create_string_buffer() and cast those to c_char_p using ctypes.cast().

answered Jun 1, 2016 at 17:57

Aksel

5334 silver badges12 bronze badges

Comments

user4113344 · Accepted Answer · 2014-11-25 15:54:46Z

1

Have you considered using SWIG? I haven't tried it myself but here's what it would look like, without changing your C source:

/*mymodule.i*/

%module mymodule
extern void my_c_function(const char* str1, const char* str2);

This would make your Python source as simple as (skipping compilation):

import mymodule

string1 = "my string 1"
string2 = "my string 2"
my_c_function(string1, string2)

Note I'm not certain .encode('utf-8') is necessary if your source file is already UTF-8.

answered Nov 25, 2014 at 15:54

user4113344

1 Comment

Eryk Sun Over a year ago

For Python 3 remember to use swig's -py3 option. The wrapper encodes a Python 3 string as UTF-8, using PyUnicode_AsUTF8String and then PyBytes_AsStringAndSize. Passing bytes raises a TypeError.

Collectives™ on Stack Overflow

Converting python string object to c char* using ctypes

3 Answers 3

5 Comments

Comments

1 Comment

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

5 Comments

Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Linked

Related