Hi, I've been calling ftdi_usb_open_string() simply for the purpose of validating device strings for use later, and then discarding my context.
Unfortunately, sometimes the call to libusb_init() fails and ftdi_new then returns 0. When this happens, it is hard to find a way to reliably validate one of these strings. ftdi_usb_open_string() is documented as possibly changing what strings it accepts in the future, so doing the validation by hand is unideal. It seems ftdi_usb_open_string() returns -12 if the ftdi pointer is null, _before_ it validates the string, and contrariwise dereferences the passed pointer if it is not null, preventing a user from fudging validation with a fake pointer. It would be great if this check returning -12 were simply moved after the basic string validation, so the -11 return value could be used even without an ftdi context. This could probably be done robustly by handing off the port opening to the other functions that don't take strings, after the string has been parsed. It would alternatively be great if the string validation code were moved into a new public function like ftdi_valid_usb_string(const char *), so that what the user is accomplishing can be detected without setting up all the infrastructure. Thank you for reading. -- libftdi - see http://www.intra2net.com/en/developer/libftdi for details. To unsubscribe send a mail to libftdi+unsubscr...@developer.intra2net.com
