nzb backup directory syncronisation with queue

[jack42]

When I upgraded from B2 to B4 I had about 60 items in my queue, but had to rebuild it all. Unfortunately I had about 130 nzbs in my backup directory. It took me ages to carefully go through them all and figure out which nzbs should be in the new queue, which ones I had cancelled and which ones had already downloaded. Now that we hopefully won't have to rebuild queues when upgrading this is less of a problem, but there may still be occasions when it's best to rebuild the queue and clear out the cache.

My suggestion that would make it much easier to do a queue rebuild in these cases is that under the nzb directory you have some subdirectories: cancelled, completed, failed. When a job completes and the downloaded file gets moved to the done directory you also move the corresponding nzb to the completed or failed subdirectory as appropriate. When the user clicks the delete button in the interface the corresponding nzb gets moved to the cancelled directory. With these adjustments it means that only the nzbs that correspond to currently queued jobs will still be in the main nzb directory. So when the user wants to rebuild his queue he can just drag these files to the blackhole and he'll get the same queue back he had before (but not necessarily in the same order, of course). You could even add a command line option to be used with the -c option to automatically add the nzbs currently in the backup directory to the new empty queue. (This is a much less important bonus item though.)

Having sab keep the nzb directory in order is also usefull if you want to migrate some or all of your download queue onto another computer/operating system/download tool.