TLDs that are also file extensions

There has recently been some concerns expressed around new TLDs colliding with file extensions. While I’m the first to raise concerns around collisions, I think that this particular ship has sailed. As an example, the DEC VAX and CP/M used .COM as a file extension. This was then carried over into MS-DOS (and Windows) - a good example of this being command.com. While this confusion was briefly abused, but mitigations were quickly put in place. This is a great example of the contextual important of naming - command.com on your computer is a different context to https://www.command.com - the 3M Command Adhesive website.

However, before completely dismissing the threat, I figured I’d quickly look and see how many collisions there are between well known file extensions and delegated TLDs. I used the Webopedia List Of Data Formats & File Extensions and wrote a small Python program to compare these with the IANA list of delegated TLDs

Here are the results:

Total: 139

Quick-n-dirty Python to generate this:

#!/usr/bin/env python3

"""
PoC to find file extensions that are also valid TLDs
"""
import sys

def read_filelist(filename):
  file_types = {}
  with open(filename) as f:
    lines = f.readlines()
    for line in lines:
      if line.startswith('.'):
        (ext, desc) = line.split(' ', 1)
        ext = ext.strip('.')
        file_types[ext] = desc.strip()
  return file_types

def read_tlds(filename):
  tlds = []
  with open(filename) as f:
    lines = f.readlines()
    for line in lines:
      tlds.append(line.lower().strip())
  return tlds

def main():
  """pylint FTW"""
  tlds = read_tlds('tlds-alpha-by-domain.txt')
  extensions = read_filelist('file_extension_tlds.txt')
  print ("| Extension | Description |")
  print ("| --------- | ----------- |")
  for ext in extensions:
    if ext in tlds:
      print ("| %s | %s |" % (ext, extensions[ext]))


if __name__ == "__main__":
    main()
    sys.exit(0)