是否可以将数据从void_pointer写入python中的文件？
是否可以将数据从void_pointer写入python中的文件？
提问于 2013-07-01 23:58:59
我有一个C程序，它从我的相机中获取图像并将二进制数据写入文件。
到目前为止，我正在尝试编写一个python脚本来模仿C代码，但没有成功。
在我尝试将其写入文件之前，Python示例似乎没有错误地执行。换句话说，在Python中，fwrite无法在这个实例中写入要文件的数据。fwrite返回一个值0，指示已写入0文件。
所以，我要问的是:由于我有一个外部的c-函数，它读取二进制数据，大概是到内存中的某个缓冲区，所以我要问的是：的实例是否可能根本无法读取对这个内存位置的访问？如果是这样的话，python是否有可能获得访问内存中适当区域所需的权限？
作为参考，我包括了一个简化的C示例(按预期工作)，后面跟着下面的Python示例，它打开并创建了适当的文件，但没有写入它。
//Data Saving Sample
//The sample will open the first camera attached
//and acquire 3 frames. The 3 frames will be saved
//in a raw format and can be opened with applications
//such as image-j
#define NUM_FRAMES 3
#define NO_TIMEOUT -1
#include "stdio.h"
#include "picam.h"
#include <iostream>
using namespace std;
int main()
{
// The following structure contains the image data
AvailableData data;
int readoutstride = 2097152;
FILE *pFile;
Acquire( NUM_FRAMES, NO_TIMEOUT, &data, &errors );
pFile = fopen( "sampleX3.raw", "wb" );
if( pFile )
{
fwrite( data.initial_readout, 1, (NUM_FRAMES*readoutstride), pFile );
fclose( pFile );
}
}下面是我的Python版本：
""" Load the picam.dll """
picamDll = 'DLLs/Picam.dll'
picam = ctypes.cdll.LoadLibrary(picamDll)
libc = ctypes.cdll.msvcrt
fopen = libc.fopen
data = AvailableData()
readoutstride = ctypes.c_int(2097152)
"""
This is the C-structure form from the provided header file, followed by the Python syntax for creating the same structure type
typedef struct AvailableData
{
void* initial_readout;
int64 readout_count;
} AvailableData;
class AvailableData(ctypes.Structure):
_fields_ = [("initial_readout", ctypes.c_void_p), ("readout_count",ctypes.c_int64)]
"""
"""
Simplified C-Prototype for Acquire
Acquire(
int64 readout_count,
int readout_time_out,
AvailableData* available);
"""
picam.Acquire.argtypes = int64, int, ctypes.POINTER(AvailableData)
picam.Acquire.restype = int
print picam.Acquire( 3, -1, ctypes.byref(data), ctypes.byref(errors))
""" Write contents of Data to binary file """
fopen.argtypes = ctypes.c_char_p, ctypes.c_char_p
fopen.restype = ctypes.c_void_p
fwrite = libc.fwrite
fwrite.argtypes = ctypes.c_void_p, ctypes.c_size_t, ctypes.c_size_t, ctypes.c_void_p
fwrite.restype = ctypes.c_int
fclose = libc.fclose
fclose.argtypes = ctypes.c_void_p,
fclose.restype = ctypes.c_int
fp = fopen('PythonBinOutput.raw', 'wb')
print 'fwrite returns: ',fwrite(data.initial_readout, 3*readoutstride.value, 1, fp)
fclose(fp)回答 1
Stack Overflow用户
发布于 2013-07-02 02:22:30
创建一个用于常规Python文件的数组。例如：
获取一个空指针：
class AvailableData(Structure):
_fields_ = [
("initial_readout", c_void_p),
]
cglobals = CDLL(None) # POSIX dlopen
cglobals.malloc.restype = c_void_p
size = 100 # 3 * readoutstride.value
data = AvailableData(cglobals.malloc(size))现在将其用作数组：
buf = (c_char * size).from_address(data.initial_readout)
with open('PythonBinOutput.raw', 'wb') as f:
f.write(buf)页面原文内容由Stack Overflow提供。腾讯云小微IT领域专用引擎提供翻译支持
原文链接：
https://stackoverflow.com/questions/17415460
复制相关文章
相似问题
腾讯云开发者
