Problem:

The documentation flow (`gen_vimdoc.py`) has several issues:
- it's not very versatile
- depends on doxygen
- doesn't work well with Lua code as it requires an awkward filter script to convert it into pseudo-C.
- The intermediate XML files and filters makes it too much like a rube goldberg machine.

Solution:

Re-implement the flow using Lua, LPEG and treesitter.

- `gen_vimdoc.py` is now replaced with `gen_vimdoc.lua` and replicates a portion of the logic.
- `lua2dox.lua` is gone!
- No more XML files.
- Doxygen is now longer used and instead we now use:
  - LPEG for comment parsing (see `scripts/luacats_grammar.lua` and `scripts/cdoc_grammar.lua`).
  - LPEG for C parsing (see `scripts/cdoc_parser.lua`)
  - Lua patterns for Lua parsing (see `scripts/luacats_parser.lua`).
  - Treesitter for Markdown parsing (see `scripts/text_utils.lua`).
- The generated `runtime/doc/*.mpack` files have been removed.
   - `scripts/gen_eval_files.lua` now instead uses `scripts/cdoc_parser.lua` directly.
- Text wrapping is implemented in `scripts/text_utils.lua` and appears to produce more consistent results (the main contributer to the diff of this change).

Luarocks is no longer needed after
25e51d393a420765d5efd44c1b4be823a5cf280a.

Co-authored-by: Brede Yabo Sherling Kristensen <bredeyabo@hotmail.com>
Co-authored-by: zeertzjq <zeertzjq@outlook.com>
Co-authored-by: István Donkó <istvan.donko@gmail.com>
Co-authored-by: Julian Berman <Julian@GrayVines.com>
Co-authored-by: bryant <bryant@users.noreply.github.com>
Co-authored-by: Michael Lingelbach <m.j.lbach@gmail.com>
Co-authored-by: nlueb <9465658+nlueb@users.noreply.github.com>
Co-authored-by: Leonhard Saam <leonhard.saam@yahoo.com>
Co-authored-by: Jesse Wertheim <jaawerth@gmail.com>
Co-authored-by: dm1try <me@dmitry.it>
Co-authored-by: Jakub Łuczyński <doubleloop@o2.pl>
Co-authored-by: Louis Lebrault <louis.lebrault@gmail.com>
Co-authored-by: Brede Yabo Sherling Kristensen <bredeyabo@hotmail.com>
Co-authored-by: zeertzjq <zeertzjq@outlook.com>
Co-authored-by: István Donkó <istvan.donko@gmail.com>
Co-authored-by: Julian Berman <Julian@GrayVines.com>
Co-authored-by: bryant <bryant@users.noreply.github.com>
Co-authored-by: Michael Lingelbach <m.j.lbach@gmail.com>
Co-authored-by: nlueb <9465658+nlueb@users.noreply.github.com>
Co-authored-by: Leonhard Saam <leonhard.saam@yahoo.com>
Co-authored-by: Jesse Wertheim <jaawerth@gmail.com>
Co-authored-by: dm1try <me@dmitry.it>
Co-authored-by: Jakub Łuczyński <doubleloop@o2.pl>
Co-authored-by: Louis Lebrault <louis.lebrault@gmail.com>

Also update doc.

Using it takes 30+ additional seconds for me with a ccache-enabled build
(43s vs. 12s).

While it certainly makes sense to use DEBUG during development,
bisecting etc, it should be made clearer what causes this.

This was discouraged (as an option) in 5b5d353151 [1], not enabled/used by
default, and not working according to the comment in local.mk.example.

Taken out of https://github.com/neovim/neovim/pull/10395.

1: https://github.com/neovim/neovim/pull/2465

There was never any investigation done to determine whether using
jemalloc was actually a net benefit for nvim.  It has been a portability
limitation and adds another factor to consider when triaging issues.

We have two ways to disable third-party/

  1. make USE_BUNDLED_DEPS=OFF
  2. cmake USE_BUNDLED=OFF

The example used the make option in a cmake context.