README
¶
CPN Batch Processing Example
Demonstrates the Colored Petri Net (CPN) features of the workflow engine using a
batch of orders that flow pending → processing → done. Each order is a
token carrying its own data (id, amount).
Run it:
go run ./examples/cpn_batch_processing/
What it shows
| Feature | API |
|---|---|
| Seed data-carrying tokens | wf.CreateTokens("pending", []workflow.TokenData{...}) |
| Start a place from a clean slate | wf.ClearPlace("pending") |
| Aggregate over a numeric field | wf.AggregateTokens(nil, "amount") |
| Rewrite matching tokens' data | wf.TransformTokens(place, pred, fn) |
| Advance one token out of a batch | wf.ApplyTransitionForToken(ctx, "start", id) |
| Advance every token at once | wf.ApplyTransition("start") |
| Read tokens at a place | wf.GetTokens("done") |
A plain (boolean) workflow is just the special case where places hold a single
uncolored token — the same ApplyTransition calls work unchanged. See
docs/guides/CPN_GUIDE.md for the full model.
Declarative equivalent
The same initial batch can be declared in YAML with initial_marking (its map
form seeds colored tokens per place):
workflow:
name: batch
initial_marking:
pending:
- {id: "A", amount: 100}
- {id: "B", amount: 250}
- {id: "C", amount: 40}
places: [{name: pending}, {name: processing}, {name: done}]
transitions:
- {name: start, from: [pending], to: [processing]}
- {name: finish, from: [processing], to: [done]}
Documentation
¶
Overview ¶
Command cpn_batch_processing demonstrates Colored Petri Net (CPN) features of the workflow engine: data-carrying tokens, whole-batch and per-token firing, and token queries / aggregation / transformation.
The workflow models a batch of orders flowing pending -> processing -> done, where each order is a token carrying its own data (id, amount).
Source Files