Generated Seatbelt profile contains many duplicates

[dwt]

This is probably not problematic, but ti does make the --debug output much harder to read than necessary.

Some examples:

(deny file-write-unlink
  (subpath "/Users/dwt/.pypirc")
  (with message "CMD64_cGk=_END_cfe05cd74_SBX"))
(deny file-write-unlink
  (literal "/Users/dwt")
  (with message "CMD64_cGk=_END_cfe05cd74_SBX"))
(deny file-write-unlink
  (literal "/Users")
  (with message "CMD64_cGk=_END_cfe05cd74_SBX"))
(deny file-write-unlink
  (subpath "/Users/dwt/.netrc")
  (with message "CMD64_cGk=_END_cfe05cd74_SBX"))
(deny file-write-unlink
  (literal "/Users/dwt")
  (with message "CMD64_cGk=_END_cfe05cd74_SBX"))
(deny file-write-unlink
  (literal "/Users")

This looks like a generic "follow up the directory trail" logic generating the same entries regardless of whether they already exist. Perhaps adding the rules to a set before generating them could be a good workaround here?