使用MQL5

Question

我正在尝试从 Train() -ed DLL调用 #import 函数。

函数在那里，在DLL文件中，我已经确认它是通过DLL-viewer导出的。

我在MQL5中尝试了以下代码来调用DLL函数：

#import "NNModel.dll"
string Train(
   double &inpTrain[], // Input training data (1D array carrying 2D data, old first)
   double &outTarget[],// Output target data for training (2D data as 1D array, oldest 1st)
   double &outTrain[], // Output 1D array to hold net outputs from training
   int     ntr,        // # of training sets
   int     UEW,        // Use Ext. Weights for initialization (1=use extInitWt, 0=use rnd)
   double &extInitWt[],// Input 1D array to hold 3D array of external initial weights
   double &trainedWt[],// Output 1D array to hold 3D array of trained weights
   int     numLayers,  // # of layers including input, hidden and output
   int    &lSz[],      // # of neurons in layers. lSz[0] is # of net inputs
   int     AFT,        // Type of neuron activation function (0:sigm, 1:tanh, 2:x/(1+x))
   int     OAF,        // 1 enables activation function for output layer; 0 disables
   int     nep,        // Max # of training epochs
   double  maxMSE      // Max MSE; training stops once maxMSE is reached
   );

string Test(
   double &inpTest[],  // Input test data (2D data as 1D array, oldest first)
   double &outTest[],  // Output 1D array to hold net outputs from training (oldest first)
   int     ntt,        // # of test sets
   double &extInitWt[],// Input 1D array to hold 3D array of external initial weights
   int     numLayers,  // # of layers including input, hidden and output
   int    &lSz[],      // # of neurons in layers. lSz[0] is # of net inputs
   int     AFT,        // Type of neuron activation function (0:sigm, 1:tanh, 2:x/(1+x))
   int     OAF         // 1 enables activation function for output layer; 0 disables
   );
#import

在C文件中，我将函数声明为：

MT5_EXPFUNC char* __stdcall Train(
const double* inpTrain, // Input training data (2D data as 1D array, oldest first)
const double* outTarget,// Output target data for training (2D data as 1D array, oldest first)
      double* outTrain, // Output 1D array to hold net outputs from training
const int     ntr,      // # of training sets
const int     UEW,      // Use External Weights for initialization (1=use extInitWt, 0=use rnd)
const double* extInitWt,// Input 1D array to hold 3D array of external initial weights
      double* trainedWt,// Output 1D array to hold 3D array of trained weights
const int     numLayers,// # of net layers including input, hidden and output
const int*    lSz,      // # of neurons in layers. lSz[0] is # of net inputs (nin)
const int     AFT,      // Type of neuron activation function (0:sigm, 1:tanh, 2:x/(1+x))
const int     OAF,      // 1 enables activation function for output layer neurons; 0 disables
const int     nep,      // Max # of training epochs
const double  maxMSE    // Max MSE; training stops once maxMSE is reached
)

和

MT5_EXPFUNC char* __stdcall Test(
const double* inpTest,  // Input test data (2D data as 1D array, oldest first)
      double* outTest,  // Net outputs from testing
const int     ntt,      // # of test sets
const double* extInitWt,// Input 1D array to hold 3D array of external initial weights
const int     numLayers,// # of net layers including input, hidden and output
const int*    lSz,      // # of neurons in layers. lSz[0] is # of net inputs (nin)
const int     AFT,      // Type of neuron activation function (0:sigm, 1:tanh, 2:x/(1+x))
const int     OAF       // 1 enables activation function for output layer neurons; 0 disables
)

调用指标后，我收到以下错误：

Cannot find 'Train' in 'NNModel.dll'
unresolved import function call

我检查了查看器，DLL中有可用的Train函数。

见图：
  

这是DLL：NNModel.dll

Answer 1

新手可能会对MQL4 / 5中的 string 不是 string

感到惊讶

这就是the documentation is clear in this，因为几年前开始的 New -MQL4.56789的早期设计版本，补救措施应从调用接口验证开始。

对于刚刚暂时有效的实际脏 #pragma -decorated struct{} - 布局的逆向工程搜索是可能的，但在生产级软件工程中存在永久风险且非常低效...

最好的下一步 - 保持C端最佳实践的安全性：

设计一个简单的界面 - 就像 int aDllCallInterfaceVALIDATOR( <params, ...> ); 一样简单，这样任何项目都不适用于MQL4 / 5-side string -s。

一旦您的调用接口代码集成适用于您的DLL端意图的所有参数，并且还可以将合理的数据/结果返回给MQL4 / 5代码，请返回到预期的功能。

走向简单意味着变得聪明。

（在 string -s开始在技术上 struct{} -s时，还有很多头发被撕掉了 - 仍然是New-MQL4.56789重新设计的蠕动岩浆。如果没有必要，不需要重复这种疼痛，是吗？是 :o) ）