Cyrene_v2/internal/services/gateway/subagent_backfill_extract_test.go

package gateway

import (
	"context"
	"strings"
	"sync"
	"testing"

	"git.yeij.top/AskaEth/Cyrene_v2/internal/bus"
	"git.yeij.top/AskaEth/Cyrene_v2/internal/runtime/history"
	"git.yeij.top/AskaEth/Cyrene_v2/internal/runtime/memory"
	"git.yeij.top/AskaEth/Cyrene_v2/internal/runtime/model"
	"git.yeij.top/AskaEth/Cyrene_v2/internal/runtime/orchestration"
	"git.yeij.top/AskaEth/Cyrene_v2/internal/runtime/transcript"
)

// 本文件锁住 M32 的第二道红线：**子代理结果绝不进记忆抽取**。
//
// 它不是「提示词里让模型别提取」，而是一条元数据判据（role=system + 注记标记）：
// 抽取器的输入在到达内层之前就被剔干净。测试因此断言的是**内层实际收到了什么**
// —— 这正是缺陷的判定点（那条线刚查过：84% 记忆来自内部会话，不能再添污染源）。

// captureExtractor 记录每次抽取收到的对话，返回预设条目。
type captureExtractor struct {
	mu    sync.Mutex
	seen  [][]model.Message
	items []memory.MemoryItem
}

func (e *captureExtractor) Extract(_ context.Context, conversation []model.Message) ([]memory.MemoryItem, error) {
	e.mu.Lock()
	defer e.mu.Unlock()
	e.seen = append(e.seen, append([]model.Message(nil), conversation...))
	return e.items, nil
}

func (e *captureExtractor) last() []model.Message {
	e.mu.Lock()
	defer e.mu.Unlock()
	if len(e.seen) == 0 {
		return nil
	}
	return e.seen[len(e.seen)-1]
}

// captureAsyncExtractor 是带异步能力的捕获提取器（对应 memory.LLMExtractor）。
type captureAsyncExtractor struct {
	*captureExtractor
}

func (e *captureAsyncExtractor) ExtractAsync(ctx context.Context, conversation []model.Message) (<-chan memory.AsyncResult, error) {
	items, err := e.Extract(ctx, conversation)
	out := make(chan memory.AsyncResult, 1)
	out <- memory.AsyncResult{Items: items, Err: err}
	close(out)
	return out, nil
}

// noteMessage 返回一条与回填落库行**同源**的系统注记模型消息：它经过
// `transcript.Message` 的映射，因此走的就是下一回合真正播种的那条路。
func noteMessage(t *testing.T) model.Message {
	t.Helper()
	row := history.Message{
		ID:        "msg-note",
		SessionID: "web",
		Role:      history.RoleSystem,
		Content:   subagentNote(settledResult("web", orchestration.SpawnAgentResult{ID: "sub-a", Status: "done", Summary: "子代理的结论"})),
	}
	return transcript.Message(row)
}

// userMessage 构造一条普通主人输入。
func userMessage(text string) model.Message {
	return model.TextMessage(model.RoleUser, text)
}

// TestWithoutSubagentNotesStripsBackfilledNotes 锁住同步路径：注记在内层提取器看到
// 之前就被剔掉，其余对话逐字保留，提取结果原样透传。
func TestWithoutSubagentNotesStripsBackfilledNotes(t *testing.T) {
	t.Parallel()
	inner := &captureExtractor{items: []memory.MemoryItem{{Content: "记住的事"}}}
	wrapped := WithoutSubagentNotes(inner)
	conv := []model.Message{
		userMessage("你好"),
		noteMessage(t),
		model.TextMessage(model.RoleAssistant, "在的"),
	}
	items, err := wrapped.Extract(context.Background(), conv)
	if err != nil {
		t.Fatalf("Extract: %v", err)
	}
	if len(items) != 1 || items[0].Content != "记住的事" {
		t.Fatalf("items = %+v, want the inner extractor's result passed through", items)
	}
	seen := inner.last()
	if len(seen) != 2 {
		t.Fatalf("inner extractor saw %d messages, want 2 (the note must be removed): %+v", len(seen), seen)
	}
	for _, m := range seen {
		if strings.HasPrefix(modelText(m), subagentNoteMarker) {
			t.Fatalf("inner extractor saw the sub-agent note %q: it would be ingested as dialogue", modelText(m))
		}
	}
	if modelText(seen[1]) != "在的" {
		t.Fatalf("conversation order changed: %+v", seen)
	}
}

// TestWithoutSubagentNotesKeepsCompactionSummary 锁住**不过度剔除**：压缩摘要同样是
// system 角色（loop 的 summaryMarker），它是对话的合法压缩，必须继续参与抽取。
func TestWithoutSubagentNotesKeepsCompactionSummary(t *testing.T) {
	t.Parallel()
	inner := &captureExtractor{}
	wrapped := WithoutSubagentNotes(inner)
	summary := model.TextMessage(model.RoleSystem, "[前文摘要 · history summary]\n\n之前聊过口味")
	if _, err := wrapped.Extract(context.Background(), []model.Message{summary, userMessage("我喜欢淡一点")}); err != nil {
		t.Fatalf("Extract: %v", err)
	}
	if got := len(inner.last()); got != 2 {
		t.Fatalf("inner extractor saw %d messages, want both the summary and the user turn", got)
	}
}

// TestWithoutSubagentNotesKeepsAsync 锁住「装饰之后仍然异步」：慢提取绝不能从后台
// goroutine 掉回回合落定的同步路径上（那会直接拖慢回复）。
func TestWithoutSubagentNotesKeepsAsync(t *testing.T) {
	t.Parallel()
	inner := &captureAsyncExtractor{&captureExtractor{}}
	wrapped := WithoutSubagentNotes(inner)
	async, ok := wrapped.(memoryAsyncExtractor)
	if !ok {
		t.Fatal("the decorator dropped ExtractAsync: a slow LLM extraction would now block turn settlement")
	}
	ch, err := async.ExtractAsync(context.Background(), []model.Message{noteMessage(t), userMessage("你好")})
	if err != nil {
		t.Fatalf("ExtractAsync: %v", err)
	}
	<-ch
	seen := inner.last()
	if len(seen) != 1 || modelText(seen[0]) != "你好" {
		t.Fatalf("async extractor saw %+v, want only the user turn", seen)
	}
}

// TestWithoutSubagentNotesIsNoopWithoutNotes 锁住零成本路径：没有注记时原样返回入参
// （不分配、不改顺序），且 nil 提取器仍是 nil（未配置提取器的部署逐字不变）。
func TestWithoutSubagentNotesIsNoopWithoutNotes(t *testing.T) {
	t.Parallel()
	if got := WithoutSubagentNotes(nil); got != nil {
		t.Fatalf("WithoutSubagentNotes(nil) = %v, want nil", got)
	}
	conv := []model.Message{userMessage("你好")}
	got := dialogueOnly(conv)
	if len(got) != len(conv) {
		t.Fatalf("dialogueOnly changed a note-free conversation: %+v", got)
	}
	if &got[0] != &conv[0] {
		t.Fatal("a note-free conversation must be returned as-is (no copy on the hot path)")
	}
}

// TestBackfilledNoteStaysOutOfMemoryExtraction 是这条防线的**端到端**接缝断言：
// 真回填落库 → 转录播种 → 抽取。注记在整条链上看得到（历史与模型上下文都有它），
// 但抽取器**一次也看不到**它。
func TestBackfilledNoteStaysOutOfMemoryExtraction(t *testing.T) {
	t.Parallel()
	b := bus.New()
	t.Cleanup(func() { _ = b.Close() })
	store := &noteStore{}
	bf, err := NewSubagentBackfill(SubagentBackfillDeps{Bus: b, History: store})
	if err != nil {
		t.Fatalf("NewSubagentBackfill: %v", err)
	}
	if err := bf.Handle(context.Background(), bus.Event{
		Subject: orchestration.SubagentResultSubject,
		Payload: settledResult("web", orchestration.SpawnAgentResult{ID: "sub-a", Status: "done", Summary: "子代理的结论"}),
	}); err != nil {
		t.Fatalf("Handle: %v", err)
	}
	rows := store.rows()
	if len(rows) != 1 {
		t.Fatalf("rows = %+v, want one", rows)
	}
	// 下一回合的模型上下文 = 播种的转录行（这就是 loop 交给提取器的 msgs 的来源）。
	nextTurn := []model.Message{transcript.Message(rows[0]), userMessage("接着说")}
	inner := &captureExtractor{}
	if _, err := WithoutSubagentNotes(inner).Extract(context.Background(), nextTurn); err != nil {
		t.Fatalf("Extract: %v", err)
	}
	seen := inner.last()
	if len(seen) != 1 {
		t.Fatalf("extractor input = %+v, want only the user turn", seen)
	}
	if !strings.Contains(modelText(seen[0]), "接着说") {
		t.Fatalf("extractor lost the real dialogue: %+v", seen)
	}
}

// modelText 返回一条模型消息的正文（测试窄视图）。
func modelText(m model.Message) string {
	var b strings.Builder
	for _, c := range m.Content {
		if c.Type == model.ContentTypeText {
			b.WriteString(c.Text)
		}
	}
	return b.String()
}