[Patch] Fix calculating starttime of repeated timers with start date

[udo_richter at gmx.de (Udo Richter)]

C.Y.M wrote:
> The problem I am experiencing with this starttime patch has to do with a new
> function in the text2skin plugin that was introduced by Enigma.
> 
> Enigma will list the upcoming timers in the main menu and thats what sends vdr
> into a lockup.

I *think* I found the reason in this code piece from status.c, 
cText2SkinStatus::UpdateEvents():

if (!dummy.IsSingleEvent()) // add 4 additional rep. timer
{
     do
     {
         dummy.Skip();
         dummy.Matches(); // Refresh start- and end-time
     } while (!dummy.DayMatches(dummy.Day()));
}

With my patch, StartTime() no longer points to days that are no 
recording days, and as a consequence, Skip() will correctly skip one 
week forward each time, to the day after the next recording. The above 
loop seems to expect Skip() to jump in 1-day steps and expects the timer 
to match the day mask finally. Since Skip() jumps to the day after the 
next timer, the day mask never matches.

The loop can now probably be avoided completely, since Skip() and 
StartTime() will deliver all future timers directly.

Cheers,

Udo