I requested a copy of all of my Facebook data, in HTML format. It gave me five zip files ranging from 2.8GB to 4.7GB to download, totaling 19GB. Those archives contain sub-folders named messages, posts etc.
I made the assumption that they should be merged since there was only one index.html as an entry point and there were folders with the same name (like messages for example, which was present in all of the archives).When I tried to extract all of them, I noticed that the extractor (Winzip) told me that there were duplicate files and asked how I wanted to handle them (keep the old copy or replace). From the few files I checked, the size was exactly the same so it probably was exactly the same file.
I couldn't find any pattern on the files I checked (after the prompt), and at this point manually reviewing to search for missing data is pointless (I can't exactly remember what I posted or what messages I sent years ago).
When the extraction was completed, the unzipped files were around 90% the size of the archives, so I am assuming that the amount of duplicates was quite large.
Why would this happen? Some of my hypotheses are that this has to do with facebook trying to split the data into five zip file for easier download or because their database is convoluted, or that my data is simply corrupted somehow.
Is there a way to be sure that this is normal behavior and my data is downloaded correctly?
index.htmlfile in the top folder? Comparing the archive contents/folder structure (DIF) may help you as well. It's unclear the method you used to extract & merge the files but FB could easily build their archives to trigger it automatically if it was the intended behavior.