我自己封装了一个函数GetMd5(LPCWSTR FileDirectory)。其中调用了windows api：CryptAcquireContext，CryptCreateHash,CryptHashData等。要计算不同文件的md5值，只需在调用此函数时传入不同文件的全路径即可，不需要额外的dll。示例代码在vs2008中测试通过：
#include "stdafx.h"
#include <iostream>
#include "windows.h"
using namespace std;void GetMd5(LPCWSTR FileDirectory)
{HANDLE hFile=CreateFile(FileDirectory,GENERIC_READ,FILE_SHARE_READ,NULL,OPEN_EXISTING,NULL,NULL);if (hFile==INVALID_HANDLE_VALUE) //如果CreateFile调用失败{cout<<"CreateFile go wrong :"<<GetLastError()<<endl; //提示CreateFile调用失败，并输出错误号。visual studio中可在“工具”>“错误查找”中利用错误号得到错误信息。CloseHandle(hFile);}HCRYPTPROV hProv=NULL;if(CryptAcquireContext(&hProv,NULL,NULL,PROV_RSA_FULL,CRYPT_VERIFYCONTEXT)==FALSE) //获得CSP中一个密钥容器的句柄{cout<<"CryptAcquireContext go wrong:"<<GetLastError()<<endl;}HCRYPTPROV hHash=NULL;if(CryptCreateHash(hProv,CALG_MD5,0,0,&hHash)==FALSE) //初始化对数据流的hash，创建并返回一个与CSP的hash对象相关的句柄。这个句柄接下来将被CryptHashData调用。{cout<<"CryptCreateHash go wrong:"<<GetLastError()<<endl;}DWORD dwFileSize=GetFileSize(hFile,0); //获取文件的大小if (dwFileSize==0xFFFFFFFF) //如果获取文件大小失败{cout<<"GetFileSize go wrong:"<<GetLastError()<<endl;}byte* lpReadFileBuffer=new byte[dwFileSize];DWORD lpReadNumberOfBytes;if (ReadFile(hFile,lpReadFileBuffer,dwFileSize,&lpReadNumberOfBytes,NULL)==0) //读取文件{cout<<"ReadFile go wrong:"<<GetLastError()<<endl;}if(CryptHashData(hHash,lpReadFileBuffer,lpReadNumberOfBytes,0)==FALSE) //hash文件{cout<<"CryptHashData go wrong:"<<GetLastError()<<endl;}delete[] lpReadFileBuffer;CloseHandle(hFile); //关闭文件句柄BYTE *pbHash;DWORD dwHashLen=sizeof(DWORD);//以下注释掉的代码不用使用，因为已经知道md5值就占32个字节，没有必要通过CryptGetHashParam函数来得到字节数。/*BYTE *pbHashSize;if (!(pbHashSize=(byte*)malloc(dwHashLen))) //为pbHashSize分配内存{cout<<"memory allocation failed:"<<GetLastError()<<endl;}//将第二个参数的值设为HP_HASHSIZE。dwHashLen中存放着hash值的字节数。这个调用必须在将第三个参数设置为HP_HASHVAL的调用前，这样才能分配正确数量的内存。if (CryptGetHashParam(hHash,HP_HASHSIZE,pbHashSize,&dwHashLen,0)) {free(pbHashSize);}else{cout<<"get size go wrong"<<GetLastError()<<endl;}*/if (CryptGetHashParam(hHash,HP_HASHVAL,NULL,&dwHashLen,0)) //我也不知道为什么要先这样调用CryptGetHashParam，这块是参照的msdn {}else{cout<<"get length wrong"<<endl;}if(pbHash=(byte*)malloc(dwHashLen)){}else{cout<<"allocation failed"<<endl;}if(CryptGetHashParam(hHash,HP_HASHVAL,pbHash,&dwHashLen,0)) //获得md5值{for(DWORD i=0;i<dwHashLen;i++) //输出md5值{printf("%02x",pbHash[i]);}cout<<endl;}//善后工作if(CryptDestroyHash(hHash)==FALSE) //销毁hash对象{cout<<"CryptDestroyHash go wrong:"<<GetLastError()<<endl;}if(CryptReleaseContext(hProv,0)==FALSE){cout<<"CryptReleaseContext go wrong:"<<GetLastError()<<endl;}
}int _tmain(int argc, _TCHAR* argv[])
{GetMd5(L"C:\\Windows\\System32\\kernel32.dll");system("pause");return 0;
}
