Question

I am developing a plugin for NSIS (Unicode) and I am trying to use InternetCrackUrl() to get the hostname of a URL (ie: http://www.google.com/test.html -> http://www.google.com) but instead of lpszHostName just returning “www.google.com”, it returns “www.google.com/test.html”.

Here is my code:

void __declspec(dllexport) Example(HWND hwndParent, int string_size, TCHAR *variables, stack_t **stacktop, extra_parameters *extra) {
    g_hwndParent=hwndParent;

    EXDLL_INIT();

    LPWSTR szURI = new WCHAR[string_size];
    URL_COMPONENTS urlComp;

    // Sets szURI to "http://www.xyz.com/test.html"
    popstring(szURI);

    wstring strUri = szURI;

    ZeroMemory(&urlComp, sizeof(urlComp));
    urlComp.dwStructSize = sizeof(urlComp);

    // Set required component lengths to non-zero so that they are cracked.
    urlComp.dwHostNameLength = static_cast<DWORD>(-1);
    urlComp.dwSchemeLength = static_cast<DWORD>(-1);
    urlComp.dwUrlPathLength = static_cast<DWORD>(-1);
    urlComp.dwExtraInfoLength = static_cast<DWORD>(-1);

    if (!InternetCrackUrlW(strUri.c_str(), strUri.length(), 0, &urlComp)) {
        return _T("InternetCrackUrl failed");
    }

    // urlComp.lpszHostName = www.xyz.com/test.html
}

Any ideas?

Answer 1

If you don’t provide your own buffer InternetCrackUrl will return pointers to characters in the original string you pass as input. It doesn’t copy the string.

So, lpszHostName will point to the first character, and dwHostNameLength will give you the number of chars that make the host name.