Not sure if I can help either way, but I think for clarity’s sake, you should disambiguate your antecedents. There are too many references to “the project”. For instance, this is my estimation of what you’re saying:
… I’m probably getting some of these references wrong. I’m almost certain I have.
Also, what do you mean by messing up the binder? Are your folders and documents just out of order, or is it something more serious?
In theory you should be okay doing that, although it may be easier to replace the .scrivx alone, from the backup. That’s the part that needs to be fixed, so instead of copying a bunch of RTF and TXT files over, do something like this:
Zip the current project and set it aside for safe-keeping.
If necessary, extract a copy of the project from the last good automated backup.
Open both project folders in Explorer.
Drag the .scrivx file from the backup into the copy we’re attempting to repair.
Open the repaired project and update the search index via Tools.
Now I’d just expand the entire Binder outline and quickly DownArrow through the whole list looking for blank spots.
Thanks, AmberV. It occurred to me that it might work to just replace the .scrivx file. But short of a confirmation of that option, the other approach, though not as elegant, seemed possibly safer in terms of project integrity. I knew the binder order was controlled by the scrivx, but not the extent the scrivx kept track of indexes, etc. Knowing that (luckily) no files had been created or merged since the last backup, and that Scrivener is able to rebuild/reconcile index discrepancies, it seemed that swapping in a set of identically-named files wouldn’t screw things up too badly.
As far as I can tell, it worked. I believe I might have lost some document name changes (which makes sense, as I understand how S works), but the files themselves matched up.
It was an interesting exercise, though not one I’d volunteer to repeat.
Glad to hear it worked out. And yes the file name is stored in the .scrivx, as well as pretty much everything else about a file, other than the raw data itself. So stuff like labels, keywords and references will also revert.