Standardize outputs of DeviceRedirectionMITM and FileCrawlerMITM

[xshill]

Instead of being in pyrdp_output/files, they should be in pyrdp_output/crawler to make it easier to differentiate which files were intercepted and which files were crawled.

[xshill]

On second thought, it would be much nicer to have a unified system for both components. Essentially:

• Both DeviceRedirectionMITM and FileCrawler should recreate the victim's filesystem in pyrdp_output/filesystems/
• Actual files should be written in pyrdp_output/files/ and renamed to the sha1 hash of their contents
• Files in filesystems/ should only link to the corresponding file in files/. On Windows, this could be a .lnk file or just a text file containing the path or the hash of the file (e.g: virus.exe would actually be called virus.exe.txt, it's content would be [pyrdp_output_path]/files/sha1(virus.exe)).
• All files should be logged in pyrdp_output/mapping.json (this implies changing the path attributes to arrays)

[xshill]

One more note about mapping.json, we don't really need the remote_path attribute anymore since the filesystem is recreated. We can just have an array of local_paths were each file was saved.

[xshill]

Actually I removed the mapping.json file since all the information from it can be gathered elsewhere and it makes the code much simpler.