S
swille
I'm sort of pondering a project in which I think a DSL would be a good
fit for user programmability to control transactional flow. I'm
thinking about how the DSL might look, and have come up with something
like this:
get :approval, :from =3D> [:manager, :system_owner]
if :approval =3D=3D
k
get :information, :request =3D> :clearance, :from =3D>
[
erson_with_role, "Security Role"]
modify
erson, :with =3D> [:clearance, :clearance_date =3D> :now]
notify :success =3D> [:requestee, :requestor]
else
notify :failure =3D> [:requestee, :requestor]
finish
end
# unrelated to the code above
push :change,
f =3D> :cn, :to =3D> :accounts, :where =3D> ":attribute
=3D=3D
ld_cn"
It would expand much deeper, but for now, that's what I have. Any
thoughts on how best to implement something like this, how I could
make it a little cleaner, and how I might deal with completing the
flow after a system failure would be much appreciated
fit for user programmability to control transactional flow. I'm
thinking about how the DSL might look, and have come up with something
like this:
get :approval, :from =3D> [:manager, :system_owner]
if :approval =3D=3D
get :information, :request =3D> :clearance, :from =3D>
[
modify
notify :success =3D> [:requestee, :requestor]
else
notify :failure =3D> [:requestee, :requestor]
finish
end
# unrelated to the code above
push :change,
=3D=3D
It would expand much deeper, but for now, that's what I have. Any
thoughts on how best to implement something like this, how I could
make it a little cleaner, and how I might deal with completing the
flow after a system failure would be much appreciated