Ctypes-String pointers

Question

This python code calls a dll(Created in LabVIEW) which returns the size of the string passed as argument. Every time I try passing string with different size, it returns the size as "1". I tried calling the dll in different programming languages like C,LabVIEW, it's working fine.

download link for dll file: https://drive.google.com/open?id=1pnYI6-SfUY3Cn6EeD4mWUO6dlEBH3DmN

    import ctypes
    from ctypes import windll, cdll,\
        c_wchar, c_size_t, c_ulonglong, c_wchar_p, c_void_p,\
        sizeof,\
        WinError
    try:
        dllhandle = ctypes.CDLL("Strlen.dll")
        print("Library Loaded")
    except Exception as e:
        print("Can't open DLL:",e)
    dllhandle.Strlen.argtypes = [ctypes.c_wchar_p]
    dllhandle.Strlen.restype = ctypes.c_int32
    data = "Hello"
    data_ptr = ctypes.c_wchar_p(data)
    print("Data Pointer:",data_ptr)
    length = dllhandle.Strlen(data_ptr)
    print("Data:",data)
    print("String Length:",length)

Can anyone help me solve this?.Thanks in advance.

What does `SetExcursionFreeExecutionSetting` do? Also do you have the functions headers? I have an idea why it behaves like this but I want to test before making it public. Although I'm a bit reluctant executing unknown code on my *PC*. How do you call the func from other languages? — CristiFati, Feb 28 '18 at 13:04

CristiFati · Accepted Answer · 2019-03-09T02:25:32.383

Notes:

The .dll is unusable (its DllMain which in this case is LVDLLStatus, returns 0 - triggering [MS.Docs]: LoadLibraryW function to fail with ERROR_DLL_INIT_FAILED) without LabVIEW Run-Time Engine
I downloaded [NI]: LabVIEW Run-Time Engine 2017 - (64-bit) - Windows - Windows 10/8.1 64-bit/7 (SP1) 64-bit/Server 2012 R2 64-bit/Server 2008 R2 64-bit, and after that I was able to load it

My guess was correct: the function expects a char * (8bit), while wchar\t * is 16bit.

Test code (code.py):

#!/usr/bin/env python3

import sys
from ctypes import CDLL,\
    c_int, c_char_p, c_wchar_p


WORDS = [
    "",
    "a",
    "q z",
    "asd\n",
    "12345",
    "qwertyuiop",
]


if __name__ == "__main__":
    print("Python {:s} on {:s}\n".format(sys.version, sys.platform))
    strlen_dll = CDLL("Strlen")
    strlen_func = strlen_dll.Strlen
    strlen_func.restype = c_int
    for word in WORDS:
        print("{:s}:\n    len: {:d}".format(repr(word), len(word)))
        strlen_func.argtypes = [c_char_p]
        print("    Strlen.Strlen(c_char_p): {:d}".format(strlen_func(word.encode())))
        strlen_func.argtypes = [c_wchar_p]
        print("    Strlen.Strlen(c_wchar_p): {:d}".format(strlen_func(word)))

Output:

(py35x64_test) E:\Work\Dev\StackOverflow\q049030000>"e:\Work\Dev\VEnvs\py35x64_test\Scripts\python.exe" code.py
Python 3.5.4 (v3.5.4:3f56838, Aug  8 2017, 02:17:05) [MSC v.1900 64 bit (AMD64)] on win32

'':
    len: 0
    Strlen.Strlen(c_char_p): 0
    Strlen.Strlen(c_wchar_p): 0
'a':
    len: 1
    Strlen.Strlen(c_char_p): 1
    Strlen.Strlen(c_wchar_p): 1
'q z':
    len: 3
    Strlen.Strlen(c_char_p): 3
    Strlen.Strlen(c_wchar_p): 1
'asd\n':
    len: 4
    Strlen.Strlen(c_char_p): 4
    Strlen.Strlen(c_wchar_p): 1
'12345':
    len: 5
    Strlen.Strlen(c_char_p): 5
    Strlen.Strlen(c_wchar_p): 1
'qwertyuiop':
    len: 10
    Strlen.Strlen(c_char_p): 10
    Strlen.Strlen(c_wchar_p): 1

Explanation:

wchar_t (or unsigned short) is used by Win when working with Wide Characters. According to [MS.Docs]: Working with Strings

Windows represents Unicode characters using UTF-16 encoding, in which each character is encoded as a 16-bit value.
A Google search on UTF-16 will yield lots of results, here are a couple of them:
Let's take character A (hex 0x41) for example. Since it'a a plain ASCII, it will be represented as a wchar_t like:
- 0x00 - this byte is not required, and no code page, you can look at it like at a padding byte
- 0x41 - the value itself
As seen, the 1^st byte of the wchar_t is 0, which is the same as the NULL character terminator ([Wikipedia]: Null-terminated string). So, without knowing, you're passing a string that contains such chars, and when the function encounters the 1^st one, it stops
But why does it return 1 (and not 0 as the 1^st wide char in the string starts with a 0x00)? It's because [Wikipedia]: Endianness (and the fact that Intel CPUs and the OSes running on them are little-endian). You should also check [SO]: Python struct.pack() behavior (@CristiFati's answer) (endianness related section)

Ctypes-String pointers

1 Answers1