Files
ragflow/internal/agent/component/list_operations_test.go

521 lines
17 KiB
Go
Raw Normal View History

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
//
// Copyright 2026 The InfiniFlow Authors. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
//
package component
import (
"context"
"reflect"
"sort"
feat(agent): align Go agent behavior with Python (except retrieval component) (#16225) ## Summary Aligns the **Go agent runtime/canvas/components/tools** behavior with the **Python `agent/` implementation** so the same stored canvas DSL produces the same execution result on either side. Every component, tool, and runtime primitive in `internal/agent/` is now driven by the same semantics as its Python counterpart — variable resolution, template substitution, control flow, error reporting, retry/cancel, and stream event shapes. The **retrieval component is the one explicit exception** in this PR. It is being reworked in a separate change and is excluded from this alignment pass; the wrapper slot (`universe_a_wrappers.go → newRetrievalComponent`) is preserved. ## Scope of alignment ### Components (all aligned with `agent/component/`) `Begin` · `Message` · `LLM` (incl. ChatTemplateKwargs, MessageHistoryWindowSize, VisualFiles, Cite, OutputStructure, JSONOutput, TopP, MaxRetries, DelayAfterError, credentials) · `Agent` (react + tool artifact capture + `Reset()` interface-assert) · `Switch` (12/12 operators, Python-equivalent semantics) · `Categorize` · `Invoke` · `Iteration` · `Loop` (macro-expansion through `workflowx.AddLoopNode`) · `UserFillUp` (Python-equivalent interrupt/resume via eino `compose.Interrupt`/`ResumeWithData`) · `FillUp` · `DataOperations` · `ListOperations` · `StringTransform` · `VariableAggregator` · `VariableAssigner` · `Browser` (full stagehand runtime parity) · `DocsGenerator` · `ExcelProcessor`. ### Tools (all aligned with `agent/tools/`) `Retrieval` (wrapper slot only — logic out of scope) · `MCPToolAdapter` (streamable-HTTP) · `CodeExec` (sandbox bridge with `code_exec_contract.go` matching Python contract) · `AkShare` · `ArXiv` · `Crawler` · `DeepL` · `DuckDuckGo` · `Email` · `ExeSQL` · `GitHub` · `Google` · `GoogleScholar` · `Jin10` · `PubMed` · `QWeather` · `SearXNG` · `Tavily` · `Tushare` · `Wencai` · `Wikipedia` · `YahooFinance` — uniform `eino tool.InvokableTool` interface, SSRF protection, shared HTTP client. ### Canvas execution engine (`internal/agent/canvas/`) Aligned with Python's `agent/canvas.py`: - **Scheduler** (`scheduler.go`): state pre/post handlers, node lambdas, per-component timeout resolver (4-level: per-class env → per-class table → uniform env → 600s fallback), `legacyNoOpNames`. - **Loop subgraph** (`loop_subgraph.go`): Python-equivalent `AddLoopNode` macro expansion + condition translation. - **Multibranch** (`multibranch.go`): `Switch` / `Categorize` routing via `compose.NewGraphMultiBranch` — same branch selection semantics as Python. - **Parallel subgraph** (`parallel_subgraph.go`): matches Python's parallel fan-out contract. - **Interrupt/Resume** (`interrupt_resume.go`): `UserFillUpNodeBody` / `IsInterruptError` / `ExtractInterruptContexts` — replaces the deprecated Python sentinel chain with eino's native interrupt API, preserving the same external behavior. - **Checkpoint** (`checkpoint_store.go`): `RedisCheckPointStore` Get/Set/Delete, with business metadata (status / canvas_id / parent_run_id) on a parallel Redis Hash. - **RunTracker** (`run_tracker.go`): Start / MarkSucceeded / MarkFailed / MarkCancelled / AttachCheckpoint — same lifecycle as the Python run record. - **Cancel** (`cancel.go`): Redis pub/sub watch. - **Stream** (`stream.go`): SSE channel with `messages` / `waiting` / `errors` / `done` events, same shape as Python's `agent.canvas.RunEvent` payload. ### DSL bridge (`internal/agent/dsl/`) - `normalize.go`: v1↔v2 collapsed into a single wire format — Python and Go consume the same stored JSON. - `reset.go`: per-run state reset matches Python's `Canvas.reset()` semantics. - Testdata mirrors Python's `agent_msg.json` / `all.json` / etc. ### Runtime (`internal/agent/runtime/`) - `CanvasState` / `NewCanvasState` / `GetVar` / `SetVar` / `ReadVars`: same `{{cpn_id@param}}` resolution model. - `ResolveTemplate` (regex fast path + gonja fallback) — Python Jinja-style semantics. - `selector.go`, `metrics.go`, `component.go`: shared runtime contracts. ## Out of scope (intentionally) - **`Retrieval` component logic** — wrapped only; full parity lands in a follow-up PR. - **Frontend** — only minor dsl-bridge / canvas UX fixes ride along. - **CLI / admin / model registry** — orthogonal to agent behavior. ## How alignment is verified `internal/service/agent_run_e2e_test.go` exercises the **full production chain** against real Python-shaped DSL fixtures: ``` loadCanvasForUser → versionDAO.GetLatest → decodeCanvasFromDSL → canvas.Compile → cc.Workflow.Invoke → answer extraction ``` using in-memory SQLite + miniredis (no Docker). Covers: - `TestRunAgent_RealCanvas_BeginMessage` — happy path, `{{sys.query}}` resolution - `TestRunAgent_RealCanvas_WaitForUserResume` — two-run resume cycle (Python-equivalent) - `TestRunAgent_RealCanvas_CompileFails` — unknown component name → sanitized error (Python-equivalent) - `TestRunAgent_RealCanvas_InvokeFails` — unresolvable template ref (Python-equivalent) - `TestRunAgent_RunTracker_AttachCheckpoint_CallSequence` — Start→AttachCheckpoint→MarkSucceeded lifecycle `internal/handler/agent_test.go` — SSE streaming parity (`Content-Type: text/event-stream`, `data: {…}\n\n`, trailing `data: [DONE]\n\n`, OpenAI-compatible non-stream `choices`). `internal/agent/canvas/fixture_compile_test.go` + per-component tests pin the Python-equivalent outputs. ``` go test -count=1 -v -run 'TestRunAgent_RealCanvas|TestRunAgent_RunTracker' ./internal/service/ ``` ## Design reference `docs/develop/agent-go-port-design.md` (1329 lines, last cross-checked 2026-06-17) — module layout, per-component / per-tool inventory, corner-case catalogue, and the actionable backlog (Section 14, including the retrieval alignment follow-up). --------- Co-authored-by: Claude <noreply@anthropic.com>
2026-06-22 11:58:29 +08:00
"strings"
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
"testing"
"ragflow/internal/agent/canvas"
)
// TestListOperations_Head: [1,2,3,4,5] op=head n=3 → [1,2,3], first=1, last=3.
func TestListOperations_Head(t *testing.T) {
c, err := NewListOperationsComponent(map[string]any{
"query": "cpn_0@xs",
"operations": "head",
"n": 3,
})
if err != nil {
t.Fatalf("NewListOperationsComponent: %v", err)
}
state := canvas.NewCanvasState("run-1", "task-1")
state.Outputs["cpn_0"] = map[string]any{"xs": []any{1, 2, 3, 4, 5}}
ctx := canvas.WithState(context.Background(), state)
out, err := c.Invoke(ctx, nil)
if err != nil {
t.Fatalf("Invoke: %v", err)
}
got, _ := out["result"].([]any)
want := []any{1, 2, 3}
if !reflect.DeepEqual(got, want) {
t.Errorf("result: got %v, want %v", got, want)
}
if got, want := out["first"], 1; got != want {
t.Errorf("first: got %v, want %v", got, want)
}
if got, want := out["last"], 3; got != want {
t.Errorf("last: got %v, want %v", got, want)
}
}
feat(agent): align Go agent behavior with Python (except retrieval component) (#16225) ## Summary Aligns the **Go agent runtime/canvas/components/tools** behavior with the **Python `agent/` implementation** so the same stored canvas DSL produces the same execution result on either side. Every component, tool, and runtime primitive in `internal/agent/` is now driven by the same semantics as its Python counterpart — variable resolution, template substitution, control flow, error reporting, retry/cancel, and stream event shapes. The **retrieval component is the one explicit exception** in this PR. It is being reworked in a separate change and is excluded from this alignment pass; the wrapper slot (`universe_a_wrappers.go → newRetrievalComponent`) is preserved. ## Scope of alignment ### Components (all aligned with `agent/component/`) `Begin` · `Message` · `LLM` (incl. ChatTemplateKwargs, MessageHistoryWindowSize, VisualFiles, Cite, OutputStructure, JSONOutput, TopP, MaxRetries, DelayAfterError, credentials) · `Agent` (react + tool artifact capture + `Reset()` interface-assert) · `Switch` (12/12 operators, Python-equivalent semantics) · `Categorize` · `Invoke` · `Iteration` · `Loop` (macro-expansion through `workflowx.AddLoopNode`) · `UserFillUp` (Python-equivalent interrupt/resume via eino `compose.Interrupt`/`ResumeWithData`) · `FillUp` · `DataOperations` · `ListOperations` · `StringTransform` · `VariableAggregator` · `VariableAssigner` · `Browser` (full stagehand runtime parity) · `DocsGenerator` · `ExcelProcessor`. ### Tools (all aligned with `agent/tools/`) `Retrieval` (wrapper slot only — logic out of scope) · `MCPToolAdapter` (streamable-HTTP) · `CodeExec` (sandbox bridge with `code_exec_contract.go` matching Python contract) · `AkShare` · `ArXiv` · `Crawler` · `DeepL` · `DuckDuckGo` · `Email` · `ExeSQL` · `GitHub` · `Google` · `GoogleScholar` · `Jin10` · `PubMed` · `QWeather` · `SearXNG` · `Tavily` · `Tushare` · `Wencai` · `Wikipedia` · `YahooFinance` — uniform `eino tool.InvokableTool` interface, SSRF protection, shared HTTP client. ### Canvas execution engine (`internal/agent/canvas/`) Aligned with Python's `agent/canvas.py`: - **Scheduler** (`scheduler.go`): state pre/post handlers, node lambdas, per-component timeout resolver (4-level: per-class env → per-class table → uniform env → 600s fallback), `legacyNoOpNames`. - **Loop subgraph** (`loop_subgraph.go`): Python-equivalent `AddLoopNode` macro expansion + condition translation. - **Multibranch** (`multibranch.go`): `Switch` / `Categorize` routing via `compose.NewGraphMultiBranch` — same branch selection semantics as Python. - **Parallel subgraph** (`parallel_subgraph.go`): matches Python's parallel fan-out contract. - **Interrupt/Resume** (`interrupt_resume.go`): `UserFillUpNodeBody` / `IsInterruptError` / `ExtractInterruptContexts` — replaces the deprecated Python sentinel chain with eino's native interrupt API, preserving the same external behavior. - **Checkpoint** (`checkpoint_store.go`): `RedisCheckPointStore` Get/Set/Delete, with business metadata (status / canvas_id / parent_run_id) on a parallel Redis Hash. - **RunTracker** (`run_tracker.go`): Start / MarkSucceeded / MarkFailed / MarkCancelled / AttachCheckpoint — same lifecycle as the Python run record. - **Cancel** (`cancel.go`): Redis pub/sub watch. - **Stream** (`stream.go`): SSE channel with `messages` / `waiting` / `errors` / `done` events, same shape as Python's `agent.canvas.RunEvent` payload. ### DSL bridge (`internal/agent/dsl/`) - `normalize.go`: v1↔v2 collapsed into a single wire format — Python and Go consume the same stored JSON. - `reset.go`: per-run state reset matches Python's `Canvas.reset()` semantics. - Testdata mirrors Python's `agent_msg.json` / `all.json` / etc. ### Runtime (`internal/agent/runtime/`) - `CanvasState` / `NewCanvasState` / `GetVar` / `SetVar` / `ReadVars`: same `{{cpn_id@param}}` resolution model. - `ResolveTemplate` (regex fast path + gonja fallback) — Python Jinja-style semantics. - `selector.go`, `metrics.go`, `component.go`: shared runtime contracts. ## Out of scope (intentionally) - **`Retrieval` component logic** — wrapped only; full parity lands in a follow-up PR. - **Frontend** — only minor dsl-bridge / canvas UX fixes ride along. - **CLI / admin / model registry** — orthogonal to agent behavior. ## How alignment is verified `internal/service/agent_run_e2e_test.go` exercises the **full production chain** against real Python-shaped DSL fixtures: ``` loadCanvasForUser → versionDAO.GetLatest → decodeCanvasFromDSL → canvas.Compile → cc.Workflow.Invoke → answer extraction ``` using in-memory SQLite + miniredis (no Docker). Covers: - `TestRunAgent_RealCanvas_BeginMessage` — happy path, `{{sys.query}}` resolution - `TestRunAgent_RealCanvas_WaitForUserResume` — two-run resume cycle (Python-equivalent) - `TestRunAgent_RealCanvas_CompileFails` — unknown component name → sanitized error (Python-equivalent) - `TestRunAgent_RealCanvas_InvokeFails` — unresolvable template ref (Python-equivalent) - `TestRunAgent_RunTracker_AttachCheckpoint_CallSequence` — Start→AttachCheckpoint→MarkSucceeded lifecycle `internal/handler/agent_test.go` — SSE streaming parity (`Content-Type: text/event-stream`, `data: {…}\n\n`, trailing `data: [DONE]\n\n`, OpenAI-compatible non-stream `choices`). `internal/agent/canvas/fixture_compile_test.go` + per-component tests pin the Python-equivalent outputs. ``` go test -count=1 -v -run 'TestRunAgent_RealCanvas|TestRunAgent_RunTracker' ./internal/service/ ``` ## Design reference `docs/develop/agent-go-port-design.md` (1329 lines, last cross-checked 2026-06-17) — module layout, per-component / per-tool inventory, corner-case catalogue, and the actionable backlog (Section 14, including the retrieval alignment follow-up). --------- Co-authored-by: Claude <noreply@anthropic.com>
2026-06-22 11:58:29 +08:00
// TestListOperations_TopNLegacyAlias pins the legacy DSL alias used by
// all.json: operations=topN should behave like head with n items so old
// imported workflows keep compiling and running unchanged.
func TestListOperations_TopNLegacyAlias(t *testing.T) {
c, err := NewListOperationsComponent(map[string]any{
"query": "cpn_0@xs",
"operations": "topN",
"n": 2,
})
if err != nil {
t.Fatalf("NewListOperationsComponent: %v", err)
}
state := canvas.NewCanvasState("run-topn", "task-topn")
state.Outputs["cpn_0"] = map[string]any{"xs": []any{1, 2, 3, 4}}
ctx := canvas.WithState(context.Background(), state)
out, err := c.Invoke(ctx, nil)
if err != nil {
t.Fatalf("Invoke: %v", err)
}
got, _ := out["result"].([]any)
want := []any{1, 2}
if !reflect.DeepEqual(got, want) {
t.Errorf("result: got %v, want %v", got, want)
}
}
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
// TestListOperations_Filter: items ["foo", "bar", "foobar"], op=filter,
// operator="contains", value="bar" → ["bar", "foobar"].
func TestListOperations_Filter(t *testing.T) {
c, _ := NewListOperationsComponent(map[string]any{
"query": "cpn_0@xs",
"operations": "filter",
"filter": map[string]any{"operator": "contains", "value": "bar"},
})
state := canvas.NewCanvasState("run-2", "task-2")
state.Outputs["cpn_0"] = map[string]any{"xs": []any{"foo", "bar", "foobar"}}
ctx := canvas.WithState(context.Background(), state)
out, err := c.Invoke(ctx, nil)
if err != nil {
t.Fatalf("Invoke: %v", err)
}
got, _ := out["result"].([]any)
want := []any{"bar", "foobar"}
if !reflect.DeepEqual(got, want) {
t.Errorf("filter: got %v, want %v", got, want)
}
}
// TestListOperations_DropDuplicates: [{"k":1},{"k":1},{"k":2}] → [{"k":1},{"k":2}].
func TestListOperations_DropDuplicates(t *testing.T) {
c, _ := NewListOperationsComponent(map[string]any{
"query": "cpn_0@xs",
"operations": "drop_duplicates",
})
state := canvas.NewCanvasState("run-3", "task-3")
state.Outputs["cpn_0"] = map[string]any{"xs": []any{
map[string]any{"k": 1},
map[string]any{"k": 1},
map[string]any{"k": 2},
}}
ctx := canvas.WithState(context.Background(), state)
out, err := c.Invoke(ctx, nil)
if err != nil {
t.Fatalf("Invoke: %v", err)
}
got, _ := out["result"].([]any)
if len(got) != 2 {
t.Fatalf("expected 2 elements, got %d: %v", len(got), got)
}
if !reflect.DeepEqual(got[0], map[string]any{"k": 1}) {
t.Errorf("got[0]: %v, want {k:1}", got[0])
}
if !reflect.DeepEqual(got[1], map[string]any{"k": 2}) {
t.Errorf("got[1]: %v, want {k:2}", got[1])
}
}
// TestListOperations_Tail: [1,2,3,4,5] op=tail n=2 → [4,5].
func TestListOperations_Tail(t *testing.T) {
c, _ := NewListOperationsComponent(map[string]any{
"query": "cpn_0@xs",
"operations": "tail",
"n": 2,
})
state := canvas.NewCanvasState("run-4", "task-4")
state.Outputs["cpn_0"] = map[string]any{"xs": []any{1, 2, 3, 4, 5}}
ctx := canvas.WithState(context.Background(), state)
out, err := c.Invoke(ctx, nil)
if err != nil {
t.Fatalf("Invoke: %v", err)
}
got, _ := out["result"].([]any)
want := []any{4, 5}
if !reflect.DeepEqual(got, want) {
t.Errorf("tail: got %v, want %v", got, want)
}
}
// TestListOperations_NthPositive: [a,b,c,d] n=3 → [c].
func TestListOperations_NthPositive(t *testing.T) {
c, _ := NewListOperationsComponent(map[string]any{
"query": "cpn_0@xs",
"operations": "nth",
"n": 3,
})
state := canvas.NewCanvasState("run-5", "task-5")
state.Outputs["cpn_0"] = map[string]any{"xs": []any{"a", "b", "c", "d"}}
ctx := canvas.WithState(context.Background(), state)
out, err := c.Invoke(ctx, nil)
if err != nil {
t.Fatalf("Invoke: %v", err)
}
got, _ := out["result"].([]any)
want := []any{"c"}
if !reflect.DeepEqual(got, want) {
t.Errorf("nth: got %v, want %v", got, want)
}
}
// TestListOperations_SortDesc: numeric sort with sort_method=desc.
func TestListOperations_SortDesc(t *testing.T) {
c, _ := NewListOperationsComponent(map[string]any{
"query": "cpn_0@xs",
"operations": "sort",
"sort_method": "desc",
})
state := canvas.NewCanvasState("run-6", "task-6")
state.Outputs["cpn_0"] = map[string]any{"xs": []any{3, 1, 4, 1, 5, 9, 2, 6}}
ctx := canvas.WithState(context.Background(), state)
out, err := c.Invoke(ctx, nil)
if err != nil {
t.Fatalf("Invoke: %v", err)
}
got, _ := out["result"].([]any)
sortedAsc := append([]any{}, got...)
sort.Slice(sortedAsc, func(i, j int) bool {
af, _ := sortedAsc[i].(int)
bf, _ := sortedAsc[j].(int)
return af < bf
})
// Reverse for desc
for i, j := 0, len(sortedAsc)-1; i < j; i, j = i+1, j-1 {
sortedAsc[i], sortedAsc[j] = sortedAsc[j], sortedAsc[i]
}
if !reflect.DeepEqual(got, sortedAsc) {
t.Errorf("sort desc: got %v, want %v", got, sortedAsc)
}
}
feat(agent): align Go agent behavior with Python (except retrieval component) (#16225) ## Summary Aligns the **Go agent runtime/canvas/components/tools** behavior with the **Python `agent/` implementation** so the same stored canvas DSL produces the same execution result on either side. Every component, tool, and runtime primitive in `internal/agent/` is now driven by the same semantics as its Python counterpart — variable resolution, template substitution, control flow, error reporting, retry/cancel, and stream event shapes. The **retrieval component is the one explicit exception** in this PR. It is being reworked in a separate change and is excluded from this alignment pass; the wrapper slot (`universe_a_wrappers.go → newRetrievalComponent`) is preserved. ## Scope of alignment ### Components (all aligned with `agent/component/`) `Begin` · `Message` · `LLM` (incl. ChatTemplateKwargs, MessageHistoryWindowSize, VisualFiles, Cite, OutputStructure, JSONOutput, TopP, MaxRetries, DelayAfterError, credentials) · `Agent` (react + tool artifact capture + `Reset()` interface-assert) · `Switch` (12/12 operators, Python-equivalent semantics) · `Categorize` · `Invoke` · `Iteration` · `Loop` (macro-expansion through `workflowx.AddLoopNode`) · `UserFillUp` (Python-equivalent interrupt/resume via eino `compose.Interrupt`/`ResumeWithData`) · `FillUp` · `DataOperations` · `ListOperations` · `StringTransform` · `VariableAggregator` · `VariableAssigner` · `Browser` (full stagehand runtime parity) · `DocsGenerator` · `ExcelProcessor`. ### Tools (all aligned with `agent/tools/`) `Retrieval` (wrapper slot only — logic out of scope) · `MCPToolAdapter` (streamable-HTTP) · `CodeExec` (sandbox bridge with `code_exec_contract.go` matching Python contract) · `AkShare` · `ArXiv` · `Crawler` · `DeepL` · `DuckDuckGo` · `Email` · `ExeSQL` · `GitHub` · `Google` · `GoogleScholar` · `Jin10` · `PubMed` · `QWeather` · `SearXNG` · `Tavily` · `Tushare` · `Wencai` · `Wikipedia` · `YahooFinance` — uniform `eino tool.InvokableTool` interface, SSRF protection, shared HTTP client. ### Canvas execution engine (`internal/agent/canvas/`) Aligned with Python's `agent/canvas.py`: - **Scheduler** (`scheduler.go`): state pre/post handlers, node lambdas, per-component timeout resolver (4-level: per-class env → per-class table → uniform env → 600s fallback), `legacyNoOpNames`. - **Loop subgraph** (`loop_subgraph.go`): Python-equivalent `AddLoopNode` macro expansion + condition translation. - **Multibranch** (`multibranch.go`): `Switch` / `Categorize` routing via `compose.NewGraphMultiBranch` — same branch selection semantics as Python. - **Parallel subgraph** (`parallel_subgraph.go`): matches Python's parallel fan-out contract. - **Interrupt/Resume** (`interrupt_resume.go`): `UserFillUpNodeBody` / `IsInterruptError` / `ExtractInterruptContexts` — replaces the deprecated Python sentinel chain with eino's native interrupt API, preserving the same external behavior. - **Checkpoint** (`checkpoint_store.go`): `RedisCheckPointStore` Get/Set/Delete, with business metadata (status / canvas_id / parent_run_id) on a parallel Redis Hash. - **RunTracker** (`run_tracker.go`): Start / MarkSucceeded / MarkFailed / MarkCancelled / AttachCheckpoint — same lifecycle as the Python run record. - **Cancel** (`cancel.go`): Redis pub/sub watch. - **Stream** (`stream.go`): SSE channel with `messages` / `waiting` / `errors` / `done` events, same shape as Python's `agent.canvas.RunEvent` payload. ### DSL bridge (`internal/agent/dsl/`) - `normalize.go`: v1↔v2 collapsed into a single wire format — Python and Go consume the same stored JSON. - `reset.go`: per-run state reset matches Python's `Canvas.reset()` semantics. - Testdata mirrors Python's `agent_msg.json` / `all.json` / etc. ### Runtime (`internal/agent/runtime/`) - `CanvasState` / `NewCanvasState` / `GetVar` / `SetVar` / `ReadVars`: same `{{cpn_id@param}}` resolution model. - `ResolveTemplate` (regex fast path + gonja fallback) — Python Jinja-style semantics. - `selector.go`, `metrics.go`, `component.go`: shared runtime contracts. ## Out of scope (intentionally) - **`Retrieval` component logic** — wrapped only; full parity lands in a follow-up PR. - **Frontend** — only minor dsl-bridge / canvas UX fixes ride along. - **CLI / admin / model registry** — orthogonal to agent behavior. ## How alignment is verified `internal/service/agent_run_e2e_test.go` exercises the **full production chain** against real Python-shaped DSL fixtures: ``` loadCanvasForUser → versionDAO.GetLatest → decodeCanvasFromDSL → canvas.Compile → cc.Workflow.Invoke → answer extraction ``` using in-memory SQLite + miniredis (no Docker). Covers: - `TestRunAgent_RealCanvas_BeginMessage` — happy path, `{{sys.query}}` resolution - `TestRunAgent_RealCanvas_WaitForUserResume` — two-run resume cycle (Python-equivalent) - `TestRunAgent_RealCanvas_CompileFails` — unknown component name → sanitized error (Python-equivalent) - `TestRunAgent_RealCanvas_InvokeFails` — unresolvable template ref (Python-equivalent) - `TestRunAgent_RunTracker_AttachCheckpoint_CallSequence` — Start→AttachCheckpoint→MarkSucceeded lifecycle `internal/handler/agent_test.go` — SSE streaming parity (`Content-Type: text/event-stream`, `data: {…}\n\n`, trailing `data: [DONE]\n\n`, OpenAI-compatible non-stream `choices`). `internal/agent/canvas/fixture_compile_test.go` + per-component tests pin the Python-equivalent outputs. ``` go test -count=1 -v -run 'TestRunAgent_RealCanvas|TestRunAgent_RunTracker' ./internal/service/ ``` ## Design reference `docs/develop/agent-go-port-design.md` (1329 lines, last cross-checked 2026-06-17) — module layout, per-component / per-tool inventory, corner-case catalogue, and the actionable backlog (Section 14, including the retrieval alignment follow-up). --------- Co-authored-by: Claude <noreply@anthropic.com>
2026-06-22 11:58:29 +08:00
// TestListOperations_SortByFieldList: with sort_by="score" the sort
// key is the value of the "score" map key, not the full hashable
// tuple. desc + score picks the row with the highest score first
// regardless of id ordering. Empty sort_by falls back to the legacy
// hashableKey (alphabetically first key) so existing DSLs keep
// working.
func TestListOperations_SortByFieldList(t *testing.T) {
state := canvas.NewCanvasState("run-sort-by", "task-sort-by")
state.Outputs["cpn_0"] = map[string]any{
"rows": []any{
map[string]any{"id": 1, "score": 0.91, "title": "Alpha"},
map[string]any{"id": 2, "score": 0.88, "title": "Beta"},
map[string]any{"id": 3, "score": 0.76, "title": "Gamma"},
},
}
ctx := canvas.WithState(context.Background(), state)
// sort_by="score", desc → Alpha(0.91), Beta(0.88), Gamma(0.76)
cSortDesc, err := NewListOperationsComponent(map[string]any{
"query": "cpn_0@rows",
"operations": "sort",
"sort_method": "desc",
"sort_by": "score",
})
if err != nil {
t.Fatalf("NewListOperationsComponent: %v", err)
}
out, err := cSortDesc.Invoke(ctx, nil)
if err != nil {
t.Fatalf("Invoke sort_by=score desc: %v", err)
}
got, _ := out["result"].([]any)
wantOrder := []any{
map[string]any{"id": 1, "score": 0.91, "title": "Alpha"},
map[string]any{"id": 2, "score": 0.88, "title": "Beta"},
map[string]any{"id": 3, "score": 0.76, "title": "Gamma"},
}
if !reflect.DeepEqual(got, wantOrder) {
t.Errorf("sort_by=score desc: got %v, want %v", got, wantOrder)
}
// sort_by="" — falls back to hashableKey (alphabetically first
// field = id). desc → Gamma(id:3), Beta(id:2), Alpha(id:1).
cLegacy, err := NewListOperationsComponent(map[string]any{
"query": "cpn_0@rows",
"operations": "sort",
"sort_method": "desc",
"sort_by": "",
})
if err != nil {
t.Fatalf("NewListOperationsComponent (legacy): %v", err)
}
out, err = cLegacy.Invoke(ctx, nil)
if err != nil {
t.Fatalf("Invoke sort_by=\"\" desc: %v", err)
}
got, _ = out["result"].([]any)
wantLegacy := []any{
map[string]any{"id": 3, "score": 0.76, "title": "Gamma"},
map[string]any{"id": 2, "score": 0.88, "title": "Beta"},
map[string]any{"id": 1, "score": 0.91, "title": "Alpha"},
}
if !reflect.DeepEqual(got, wantLegacy) {
t.Errorf("sort_by=\"\" desc: got %v, want %v", got, wantLegacy)
}
// sort_by="score,title" — primary score, tiebreak title.
cTiebreak, err := NewListOperationsComponent(map[string]any{
"query": "cpn_0@rows",
"operations": "sort",
"sort_method": "asc",
"sort_by": "score,title",
})
if err != nil {
t.Fatalf("NewListOperationsComponent (tiebreak): %v", err)
}
out, err = cTiebreak.Invoke(ctx, nil)
if err != nil {
t.Fatalf("Invoke sort_by=score,title asc: %v", err)
}
got, _ = out["result"].([]any)
wantTiebreak := []any{
map[string]any{"id": 3, "score": 0.76, "title": "Gamma"},
map[string]any{"id": 2, "score": 0.88, "title": "Beta"},
map[string]any{"id": 1, "score": 0.91, "title": "Alpha"},
}
if !reflect.DeepEqual(got, wantTiebreak) {
t.Errorf("sort_by=score,title asc: got %v, want %v", got, wantTiebreak)
}
}
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
// TestListOperations_NotAList: returns a clear error.
func TestListOperations_NotAList(t *testing.T) {
c, _ := NewListOperationsComponent(map[string]any{
"query": "cpn_0@x",
"operations": "head",
"n": 1,
})
state := canvas.NewCanvasState("run-7", "task-7")
state.Outputs["cpn_0"] = map[string]any{"x": "not-a-list"}
ctx := canvas.WithState(context.Background(), state)
_, err := c.Invoke(ctx, nil)
if err == nil {
t.Fatal("expected error for non-list input, got nil")
}
}
// TestListOperations_ParamCheck: empty query rejected.
func TestListOperations_ParamCheck(t *testing.T) {
_, err := NewListOperationsComponent(map[string]any{
"query": "",
"operations": "head",
})
if err == nil {
t.Fatal("expected error for empty query, got nil")
}
}
// TestListOperations_Registered: factory lookup.
func TestListOperations_Registered(t *testing.T) {
c, err := New("ListOperations", map[string]any{
"query": "sys.x",
"operations": "head",
"n": 1,
})
if err != nil {
t.Fatalf("registry lookup: %v", err)
}
if c.Name() != "ListOperations" {
t.Errorf("Name()=%q, want ListOperations", c.Name())
}
}
feat(agent): align Go agent behavior with Python (except retrieval component) (#16225) ## Summary Aligns the **Go agent runtime/canvas/components/tools** behavior with the **Python `agent/` implementation** so the same stored canvas DSL produces the same execution result on either side. Every component, tool, and runtime primitive in `internal/agent/` is now driven by the same semantics as its Python counterpart — variable resolution, template substitution, control flow, error reporting, retry/cancel, and stream event shapes. The **retrieval component is the one explicit exception** in this PR. It is being reworked in a separate change and is excluded from this alignment pass; the wrapper slot (`universe_a_wrappers.go → newRetrievalComponent`) is preserved. ## Scope of alignment ### Components (all aligned with `agent/component/`) `Begin` · `Message` · `LLM` (incl. ChatTemplateKwargs, MessageHistoryWindowSize, VisualFiles, Cite, OutputStructure, JSONOutput, TopP, MaxRetries, DelayAfterError, credentials) · `Agent` (react + tool artifact capture + `Reset()` interface-assert) · `Switch` (12/12 operators, Python-equivalent semantics) · `Categorize` · `Invoke` · `Iteration` · `Loop` (macro-expansion through `workflowx.AddLoopNode`) · `UserFillUp` (Python-equivalent interrupt/resume via eino `compose.Interrupt`/`ResumeWithData`) · `FillUp` · `DataOperations` · `ListOperations` · `StringTransform` · `VariableAggregator` · `VariableAssigner` · `Browser` (full stagehand runtime parity) · `DocsGenerator` · `ExcelProcessor`. ### Tools (all aligned with `agent/tools/`) `Retrieval` (wrapper slot only — logic out of scope) · `MCPToolAdapter` (streamable-HTTP) · `CodeExec` (sandbox bridge with `code_exec_contract.go` matching Python contract) · `AkShare` · `ArXiv` · `Crawler` · `DeepL` · `DuckDuckGo` · `Email` · `ExeSQL` · `GitHub` · `Google` · `GoogleScholar` · `Jin10` · `PubMed` · `QWeather` · `SearXNG` · `Tavily` · `Tushare` · `Wencai` · `Wikipedia` · `YahooFinance` — uniform `eino tool.InvokableTool` interface, SSRF protection, shared HTTP client. ### Canvas execution engine (`internal/agent/canvas/`) Aligned with Python's `agent/canvas.py`: - **Scheduler** (`scheduler.go`): state pre/post handlers, node lambdas, per-component timeout resolver (4-level: per-class env → per-class table → uniform env → 600s fallback), `legacyNoOpNames`. - **Loop subgraph** (`loop_subgraph.go`): Python-equivalent `AddLoopNode` macro expansion + condition translation. - **Multibranch** (`multibranch.go`): `Switch` / `Categorize` routing via `compose.NewGraphMultiBranch` — same branch selection semantics as Python. - **Parallel subgraph** (`parallel_subgraph.go`): matches Python's parallel fan-out contract. - **Interrupt/Resume** (`interrupt_resume.go`): `UserFillUpNodeBody` / `IsInterruptError` / `ExtractInterruptContexts` — replaces the deprecated Python sentinel chain with eino's native interrupt API, preserving the same external behavior. - **Checkpoint** (`checkpoint_store.go`): `RedisCheckPointStore` Get/Set/Delete, with business metadata (status / canvas_id / parent_run_id) on a parallel Redis Hash. - **RunTracker** (`run_tracker.go`): Start / MarkSucceeded / MarkFailed / MarkCancelled / AttachCheckpoint — same lifecycle as the Python run record. - **Cancel** (`cancel.go`): Redis pub/sub watch. - **Stream** (`stream.go`): SSE channel with `messages` / `waiting` / `errors` / `done` events, same shape as Python's `agent.canvas.RunEvent` payload. ### DSL bridge (`internal/agent/dsl/`) - `normalize.go`: v1↔v2 collapsed into a single wire format — Python and Go consume the same stored JSON. - `reset.go`: per-run state reset matches Python's `Canvas.reset()` semantics. - Testdata mirrors Python's `agent_msg.json` / `all.json` / etc. ### Runtime (`internal/agent/runtime/`) - `CanvasState` / `NewCanvasState` / `GetVar` / `SetVar` / `ReadVars`: same `{{cpn_id@param}}` resolution model. - `ResolveTemplate` (regex fast path + gonja fallback) — Python Jinja-style semantics. - `selector.go`, `metrics.go`, `component.go`: shared runtime contracts. ## Out of scope (intentionally) - **`Retrieval` component logic** — wrapped only; full parity lands in a follow-up PR. - **Frontend** — only minor dsl-bridge / canvas UX fixes ride along. - **CLI / admin / model registry** — orthogonal to agent behavior. ## How alignment is verified `internal/service/agent_run_e2e_test.go` exercises the **full production chain** against real Python-shaped DSL fixtures: ``` loadCanvasForUser → versionDAO.GetLatest → decodeCanvasFromDSL → canvas.Compile → cc.Workflow.Invoke → answer extraction ``` using in-memory SQLite + miniredis (no Docker). Covers: - `TestRunAgent_RealCanvas_BeginMessage` — happy path, `{{sys.query}}` resolution - `TestRunAgent_RealCanvas_WaitForUserResume` — two-run resume cycle (Python-equivalent) - `TestRunAgent_RealCanvas_CompileFails` — unknown component name → sanitized error (Python-equivalent) - `TestRunAgent_RealCanvas_InvokeFails` — unresolvable template ref (Python-equivalent) - `TestRunAgent_RunTracker_AttachCheckpoint_CallSequence` — Start→AttachCheckpoint→MarkSucceeded lifecycle `internal/handler/agent_test.go` — SSE streaming parity (`Content-Type: text/event-stream`, `data: {…}\n\n`, trailing `data: [DONE]\n\n`, OpenAI-compatible non-stream `choices`). `internal/agent/canvas/fixture_compile_test.go` + per-component tests pin the Python-equivalent outputs. ``` go test -count=1 -v -run 'TestRunAgent_RealCanvas|TestRunAgent_RunTracker' ./internal/service/ ``` ## Design reference `docs/develop/agent-go-port-design.md` (1329 lines, last cross-checked 2026-06-17) — module layout, per-component / per-tool inventory, corner-case catalogue, and the actionable backlog (Section 14, including the retrieval alignment follow-up). --------- Co-authored-by: Claude <noreply@anthropic.com>
2026-06-22 11:58:29 +08:00
// TestListOperations_StrictMode_ReturnsError pins Change #2 (panic →
// recoverable error). Strict mode with n=0 must surface a *listOpPanic
// error from Invoke() (not a goroutine-level panic). The Python
// reference raises ValueError, which the canvas framework catches.
func TestListOperations_StrictMode_ReturnsError(t *testing.T) {
c, err := NewListOperationsComponent(map[string]any{
"query": "cpn_0@xs",
"operations": "nth",
"n": 0,
"strict": true,
})
if err != nil {
t.Fatalf("NewListOperationsComponent: %v", err)
}
state := canvas.NewCanvasState("run-strict", "task-strict")
state.Outputs["cpn_0"] = map[string]any{"xs": []any{1, 2, 3}}
ctx := canvas.WithState(context.Background(), state)
_, err = c.Invoke(ctx, nil)
if err == nil {
t.Fatal("expected error from strict-mode nth with n=0, got nil")
}
if !strings.Contains(err.Error(), "strict mode") {
t.Errorf("error %q should mention 'strict mode'", err)
}
}
// TestListOperations_StrictStringCoercion pins Change #3: passing
// "strict" as a string ("true"/"1"/"yes"/"on") must be coerced to a
// true bool, matching Python's _is_strict accept-list.
func TestListOperations_StrictStringCoercion(t *testing.T) {
for _, v := range []string{"true", "TRUE", "True", "1", "yes", "on"} {
c, err := NewListOperationsComponent(map[string]any{
"query": "cpn_0@xs",
"operations": "nth",
"n": 0,
"strict": v,
})
if err != nil {
t.Fatalf("[strict=%q] NewListOperationsComponent: %v", v, err)
}
state := canvas.NewCanvasState("run-str-"+v, "task-str-"+v)
state.Outputs["cpn_0"] = map[string]any{"xs": []any{1, 2, 3}}
ctx := canvas.WithState(context.Background(), state)
_, err = c.Invoke(ctx, nil)
if err == nil {
t.Errorf("[strict=%q] expected error from strict-mode nth with n=0, got nil", v)
}
}
}
// TestListOperations_StrictFalseStringsIgnored pins the negative case
// for Change #3: strings other than the accept-list must coerce to
// false (no strict-mode error).
func TestListOperations_StrictFalseStringsIgnored(t *testing.T) {
for _, v := range []string{"false", "FALSE", "0", "no", "off", "random"} {
c, err := NewListOperationsComponent(map[string]any{
"query": "cpn_0@xs",
"operations": "nth",
"n": 99, // out-of-range, but non-strict → empty
"strict": v,
})
if err != nil {
t.Fatalf("[strict=%q] NewListOperationsComponent: %v", v, err)
}
state := canvas.NewCanvasState("run-strf-"+v, "task-strf-"+v)
state.Outputs["cpn_0"] = map[string]any{"xs": []any{1, 2, 3}}
ctx := canvas.WithState(context.Background(), state)
out, err := c.Invoke(ctx, nil)
if err != nil {
t.Errorf("[strict=%q] expected non-error (coerced to false), got %v", v, err)
}
got, _ := out["result"].([]any)
if len(got) != 0 {
t.Errorf("[strict=%q] expected empty result, got %v", v, got)
}
}
}
// TestListOperations_CoerceNBool pins Change #4: toInt must follow
// Python's int() semantics for booleans (true→1, false→0).
func TestListOperations_CoerceNBool(t *testing.T) {
if got := toInt(true); got != 1 {
t.Errorf("toInt(true) = %d, want 1", got)
}
if got := toInt(false); got != 0 {
t.Errorf("toInt(false) = %d, want 0", got)
}
}
// TestListOperations_FilterEqBool pins Change #5: normValue must render
// Go's bool as Python's str(bool) ("True"/"False") so filter `=` matches
// the Python DSL contract.
func TestListOperations_FilterEqBool(t *testing.T) {
c, err := NewListOperationsComponent(map[string]any{
"query": "cpn_0@xs",
"operations": "filter",
"filter": map[string]any{"operator": "=", "value": "True"},
})
if err != nil {
t.Fatalf("NewListOperationsComponent: %v", err)
}
state := canvas.NewCanvasState("run-feq", "task-feq")
state.Outputs["cpn_0"] = map[string]any{"xs": []any{true, false, true, "True"}}
ctx := canvas.WithState(context.Background(), state)
out, err := c.Invoke(ctx, nil)
if err != nil {
t.Fatalf("Invoke: %v", err)
}
got, _ := out["result"].([]any)
// Both true values plus the "True" string should match.
if len(got) != 3 {
t.Errorf("expected 3 matches (true, true, 'True'), got %d: %v", len(got), got)
}
}
// TestListOperations_UnknownOp_ReturnsError pins Change #1: the
// defensive default: branch in Invoke() must surface an explicit
// error for any operation name that bypasses the allowlist. We
// construct the struct directly (same package) to skip Check().
func TestListOperations_UnknownOp_ReturnsError(t *testing.T) {
c := &ListOperationsComponent{
name: "ListOperations",
param: listOperationsParam{
Query: "cpn_0@xs",
Operations: "bogus",
N: 1,
},
}
state := canvas.NewCanvasState("run-bogus", "task-bogus")
state.Outputs["cpn_0"] = map[string]any{"xs": []any{1, 2, 3}}
ctx := canvas.WithState(context.Background(), state)
_, err := c.Invoke(ctx, nil)
if err == nil {
t.Fatal("expected error for unknown operation, got nil")
}
if !strings.Contains(err.Error(), "unsupported operation") {
t.Errorf("error %q should mention 'unsupported operation'", err)
}
if !strings.Contains(err.Error(), "bogus") {
t.Errorf("error %q should mention the bad op name 'bogus'", err)
}
}
// TestListOperations_InputsDocMatchesAllowlist pins Change #6: the
// Inputs() docstring must not claim support for operations that are
// not in the Check() allowlist (slice/shuffle/take/reverse/deduplicate).
// A bug here misleads the editor dropdown and the agent developer.
func TestListOperations_InputsDocMatchesAllowlist(t *testing.T) {
c, err := NewListOperationsComponent(map[string]any{
"query": "cpn_0@xs",
})
if err != nil {
t.Fatalf("NewListOperationsComponent: %v", err)
}
doc := c.Inputs()
for _, banned := range []string{"slice", "shuffle", "reverse"} {
if strings.Contains(doc["operations"], banned) {
t.Errorf("Inputs()[operations] doc must not mention %q (not in allowlist): %q", banned, doc["operations"])
}
}
// The doc must mention the six operations that are actually supported.
for _, op := range []string{"nth", "head", "tail", "filter", "sort", "drop_duplicates"} {
if !strings.Contains(doc["operations"], op) {
t.Errorf("Inputs()[operations] doc must mention %q: %q", op, doc["operations"])
}
}
}