Fix holes in the sales history for forecast preperation

I have to forecast about +/- 700 products for my company and I am planning on using some kind of smoothing method like Holt-Winter. The problem i have, is that the history has "holes" in it, meaning there are dips in the sales history due to out of stock situations. Before i can calculate a forecast i need to correct/fix the history where in some cases i only have a years' data.

Something i thought about was, the holes can be seen as outliers and these outliers can be fixed (in other words fix the points such that it isn't an outlier anymore). But i have difficulty finding examples or references on people doing something like this and because of the amount of products i don't have the capacity to fixe all the history by hand. I need some kind of mathematical calculation/engine that can fix the history automatically.

I am a bit lost for ideas, are there any known reference i can use? Attahced is an example of a product's history for a year on weekly sales with a big hole just after December that needs to be fixed for forecasting.