* This is optimized for Groff, but works for Heirloom Troff and Neatroff as well.
  Currently, the Heirloom and Neatroff requests are just added ontop of the Groff
  ones. Theoretically, we could also try to separate the keyword lists into
  a base K&R set with Groff, Heirloom and Neatroff ontop.
* The lexer necessarily has many restrictions, as Troff is fundamentally unparseable
  (like classic TECO) and needs a lot of per-request knowledge.
* The "*.mm" extension has been removed from the lexers/cpp.tes.
  I don't know what language this was for, and I prefer `*.mm` files
  to be considered Troff.
* Temporarily changed the lexilla submodule URL.
  The corresponding Lexila lexer is in the process of being upstreamed.
  Once it is, I will probably revert the submodule to the official repository,
  as the "troff" branch is not stable (can be rebased).

* The lexer scripts are not installed if the LEXILLA option is disabled,
  so they need to be excluded from pkg-plist.

* it is supposed to be mainstreamed, but this did not yet happen
* even if it will eventually become part of the ports tree, I won't
  have access to this repo and could not keep it up to date.
  But I can keep this copy here up to date and it can serve as an
  upstream source for the ports tree maintainer.
* Also, it can be used to build FreeBSD binary packages even now
  without being part of the official ports tree.