Orchestration Python

Workflows are written in Artificial’s Orchestration Python (OP), a strict subset of the Python programming language.

✅ Orchestration Python supports the following:

• Most built-in functions

• math

• enum: Enum

• typing: List, Dict, Set

• dataclasses-json and json: these are imported, but wfgen will not try to load or introspect these.

• dataclasses: @dataclass and field() (including the default_factory parameter)

• Any module that is parseable by wfgen

• Unpacking return values - e.g. foo, bar = baz()

❌ Orchestration Python does not support the following:

• del

• async for and async with

• raise

• assert

• global

• nonlocal

• lambda

• set literals

• list, dict, and set comprehensions and generator expressions

• yield and yield from

• tuples (except unpacking return values as noted above)

• the * and ** splat operators

• the @ operator

• range

• instantiating plain Python classes (dataclasses are supported)

Finally, most of the Python object model is unsupported. You can access fields of dataclasses and elements of lists, but you cannot call methods on objects.