Files
ragflow/internal/agent/canvas/canvas_test.go
Zhichang Yu 3fa15c0e2f feat(agent): Go port — canvas engine, 22 components, DSL v2, 13 endpoints (#15952)
Ports the agent canvas subsystem from Python to Go.

## What's included

### Canvas Engine (Phase 0/1)
- State engine, scheduler, variable resolver, Redis checkpoint store,
cancel protocol
- **209 tests** across canvas / component / io packages

### 22 Components (P0–P4)
| Tier | Components |
|---|---|
| P0 T1+T2+T3 | LLM, Agent, ExitLoop, Switch, Categorize, Begin,
Message, Invoke |
| P1 T3 | VariableAggregator, VariableAssigner, StringTransform,
ListOperations, DataOperations |
| P2 T3 | Iteration, IterationItem, Loop, LoopItem |
| P3 T3 | UserFillUp, Fillup |
| P4 T5 | Browser, ExcelProcessor, DocsGenerator |

### DSL v2 Schema (Phase 2.5)
- Typed v2 in-memory model with v1-to-v2 auto-detect converter
- v1 legacy field stripping per plan §2.11.7

### HTTP Endpoints & Bug Fixes (Plans PR1–PR3)
- **DELETE SQL bug fix**: gorm v2 `Where("id = ?", id).Delete(...)`
pattern
- **CreateAgent validation**: title/DSL required, duplicate check, 103
envelope
- **13 new endpoints**: templates, prompts, tags, sessions CRUD,
chat/completions (SSE + non-stream stubs), rerun, test_db_connection,
logs, webhook/logs
- **756 Go unit tests** (745 → 756, +18)
- **17 → 0 Python integration test failures** (test_agents.py +
test_session_management/)

### Tools
21 eino tools: HTTPHelper, search tools, financial/data tools, mandatory
stubs

### Infrastructure
OTel observability, NATS message queue, DeepDoc gRPC client, SSRF
guards, IDOR mitigation
2026-06-12 22:58:28 +08:00

93 lines
3.3 KiB
Go

// Package canvas — Begin → Message e2e smoke test (Worker A, Phase 1).
//
// The simplest end-to-end compile+run path. Verifies:
//
// 1. BuildWorkflow returns a non-nil Workflow for a 2-node DSL.
// 2. Compile returns a CompiledCanvas.
// 3. The compiled Runnable.Invoke runs to completion (no eino wiring error).
// 4. The Message node's "{{sys.query}}" reference resolves against state
// that was seeded into Sys — even though our placeholder lambda doesn't
// actually emit a string, we exercise the variable resolution path by
// writing into Outputs via SetVar before Invoke.
//
// Real Begin/Message component bodies land in Phase 2 P0. Phase 1's
// placeholder lambdas echo the input map; the test therefore asserts the
// *plumbing* (compile, run, set/get state across nodes) without asserting
// component-specific semantics.
package canvas
import (
"context"
"testing"
)
// TestBeginToMessage_Smoke builds a Begin → Message DSL, seeds sys.query
// into state, and confirms the compiled workflow runs without error and
// the per-cpn Outputs bucket gets populated (proving the statePre/statePost
// handler chain works end-to-end).
func TestBeginToMessage_Smoke(t *testing.T) {
dsl := &Canvas{
Version: 1,
Components: map[string]CanvasComponent{
"begin_0": {
Obj: CanvasComponentObj{ComponentName: "Begin", Params: map[string]any{}},
Downstream: []string{"message_0"},
Upstream: []string{},
},
"message_0": {
Obj: CanvasComponentObj{ComponentName: "Message", Params: map[string]any{
"text": "hello {{sys.query}}",
}},
Downstream: []string{},
Upstream: []string{"begin_0"},
},
},
Path: []string{"begin_0", "message_0"},
}
cc, err := Compile(context.Background(), dsl)
if err != nil {
t.Fatalf("Compile: %v", err)
}
if cc.Workflow == nil {
t.Fatal("compiled Workflow is nil")
}
// Pre-seed state to mirror what the Begin node would normally inject.
// In Phase 1 we did this directly because no Begin body existed yet.
// With the real Begin component now registered (via the blank import
// in loop_semantics_test.go), Begin reads inputs["query"] and writes
// it into state.Sys["query"] itself — so we pass the query through
// the input map instead of seeding it directly, and Begin propagates
// it into the context-attached state.
runState := NewCanvasState("run-smoke", "task-smoke")
runState.SetVar("begin_0", "request", map[string]any{"q": "world"})
// Stash runState on the context so a hypothetical runner (Phase 5) can
// extract it via GetStateFromContext.
ctx := withState(context.Background(), runState)
// Invoke with the seed input. The "query" key flows into Begin's
// Invoke and is written to state.Sys["query"], where Message's
// ResolveTemplate of "{{sys.query}}" will read it.
in := map[string]any{"query": "world"}
out, err := cc.Workflow.Invoke(ctx, in)
if err != nil {
t.Fatalf("Invoke: %v", err)
}
if out == nil {
t.Fatal("Invoke returned nil output")
}
// Variable resolution: ResolveTemplate against the seeded state must
// produce "hello world" — this is what the real Message component will
// emit in Phase 2 P0.
got, err := ResolveTemplate("hello {{sys.query}}", runState)
if err != nil {
t.Fatalf("ResolveTemplate: %v", err)
}
if got != "hello world" {
t.Fatalf("template resolve: got %q want %q", got, "hello world")
}
}