Compares two files to find differences, then fixes the second file to be identical with the first. Useful when copying a large file between two storage devices.

make_files_identical.cpp

#include <algorithm>
#include <cstring>
#include <fstream>
#include <iostream>

template<typename _Stream>
size_t GetFileSize(_Stream &stream) {
    auto nCurPos = stream.tellg();
    stream.seekg(0, std::ios::end);
    auto nFileSize = stream.tellg();
    stream.seekg(nCurPos, std::ios::beg);
    return nFileSize;
}

int main(int nArgCnt, char *ppArgs[]) {
    std::ifstream src(ppArgs[1], std::ios::binary);
    auto nFileSize = GetFileSize(src);

    std::fstream dst(ppArgs[2], std::ios::binary | std::ios::in | std::ios::out);
    if (nFileSize != GetFileSize(dst)) {
        std::cout << "Size not equal" << std::endl;
        return -1;
    }

    size_t nChunkSize = 1024 * 1024 * 1024;
    size_t nChunks = (nFileSize + nChunkSize - 1) / nChunkSize;
    std::string strSrcBuf, strDstBuf;
    for (size_t i = 0; i < nChunks; ++i) {
        size_t nReadBeg = i * nChunkSize;
        size_t nReadEnd = std::min(nFileSize, (i + 1) * nChunkSize);
        size_t nReadSize = nReadEnd - nReadBeg;
        strSrcBuf.resize(nReadSize);
        strDstBuf.resize(nReadSize);

        std::cout << "Reading chunk " << i << std::endl;
        dst.seekg(nReadBeg, std::ios::beg);
        if (!src.read(const_cast<char*>(strSrcBuf.data()), strSrcBuf.size()) ||
            !dst.read(const_cast<char*>(strDstBuf.data()), strDstBuf.size())) {
            std::cout << "Read Failed!" << std::endl;
            return -1;
        }

        std::cout << "Comparing chunk " << i << std::endl;
        if (std::memcmp(strSrcBuf.data(), strDstBuf.data(), nReadSize) != 0) {
            for (size_t j = 0; j < nChunkSize; ++j) {
                if (strSrcBuf.data()[j] != strDstBuf.data()[j]) {
                    auto nPos = nReadBeg + j;
                    std::cout << "Make same at [" << nPos << "] " << (int)strSrcBuf[j]
                              << " vs " << (int)strDstBuf[j] << std::endl;
                    dst.seekp(nPos);
                    dst.write(&strSrcBuf[j], 1);
                }
            }
        }
    }
}