Introduces a range of and_then and after function that will handle callbacks

[Conni2461]

Builds on top of your previous work for and_then. We might also wanna change chain i my opinion it has the even better interface to do that.

None of these additions are necessary but after i saw @tami5 code it might be helpful. tami is currently in the on_exit callback hell which makes maintaining his work a nightmare, no offense tami :). This might help writing jobs that go from job to job easier.
Example at the end. For me that is easier to read and i actually know when which job does what and what the conditions are for that case to happen. Does this make sense? We do not need to integrate it :)
I only want that sync returns the result code. Thats the only thing that interests me tbh 🤣 Saves me one line in telescope.

Adds:

• sync will now return code as second return value
• and_then(job): will always run job
• and_then_wrap(job): will always run job (schedule_wrapped)
• and_then_on_success(job): will only run job if code == 0
• and_then_on_success_wrap(job): will only run job if code == 0 (schedule_wrapped)
• and_then_on_failure(job): will only run job if code ~= 0
• and_then_on_failure_wrap(job): will only run job if code ~= 0 (schedule_wrapped)
• after(fn): will always run function
• after_success(fn): will only run function if code == 0
• after_failure(fn): will only run function if code ~= 0

Example:

local gen_j1 = function()
  -- only generate the job outline no callbacks
end
-- ... gen_j2 -> gen_j4 exist

local run = function()
  local j1 = gen_j1()
  local j2 = gen_j2()
  local j3 = gen_j3()
  
  j1:after_failure(function(j, code) log.trace(j:result, code) end)
  j1:and_then_on_success(j2)
  j2:after_failure(function() ... end)
  j2:and_then_on_success(j3)
  -- etc 
  j3:start()
end

[kkharji]

Oh yah this soooooooo useful. no more callbacks

[tjdevries]

Cool 👍