Fixing corrupted RTF translation file

I recently received an RTF file from a translation vendor which had been corrupted from editing in MS Word. This file showed the classic signs: Indents, "1." in front of every line upon publishing to HTML.. When I opened the file in Notepad I could see loads of extra 'stuff' in the file which is not there on the RTF file which Lectora exports.

When I mentioned the file corruption to the vendor, they asked for the corrupted RTF and my Lectora title and within an hour sent back an RTF which was free of corruption and associated formatting errors. When I asked how they managed to clean up the file, they said it was a proprietary process and were unwilling to share the details. They did mention that they imported and then exported the corrupt file from Lectora and did the editing in notepad.

Does anyone have any insight into how they might have 'uncorrupted' this file?