mirror of
https://gitcode.com/JianFeeeee/HomeAgent.git
synced 2026-11-01 06:47:29 +00:00
b850feaef2
记忆层： - MemoryBlock 承载 FrameTS/PageSeq/PageKind/BlockRole/CallID - coldcontext 归档与回填按完整页传递，工具调用参数和配对键不丢失 - 召回先估算总 token，超预算时按全局页相似度标记 inactive - inactive 仅是 context 渲染期状态，不落盘、不进入 coldcontext，Prune 后重置 - Peek/确认/Remove 两阶段召回，失败保留冷上下文副本 - stablemem 记录 last_recalled_at，删块连带清理边 命名： - internal/memory/document → internal/memory/coldcontext - Doc/DocEvent/DocHit → ColdDoc/ColdEvent/ColdHit - ContextToColdContext/RecentColdContexts/FindColdContexts - 保留既有 JSON 标签、doc_ 文件前缀与 doc_count 状态键，兼容存量数据 安全治理： - 工具策略门禁、安全配置、事件总线发布者身份与入站脱敏 - 声明式 SecurityAPI 与 QQ owner/guest profile - 日志保留策略与相关 e2e/回归判据 验证：go build ./...、go vet ./...、go test ./...（62 包）； 生产副本 e2e：7013 块迁移后 ID/向量/指纹 0 变化、0 缺失。
293 lines
8.6 KiB
Go
293 lines
8.6 KiB
Go
package core
|
||
|
||
import (
|
||
"testing"
|
||
|
||
"github.com/JianFeeeee/HologramAgent/internal/memory"
|
||
"github.com/JianFeeeee/HologramAgent/internal/memory/coldcontext"
|
||
)
|
||
|
||
func needJieba() bool {
|
||
return memory.GetJieba() != nil
|
||
}
|
||
|
||
func TestDocToTriplesEmpty(t *testing.T) {
|
||
doc := &coldcontext.ColdDoc{
|
||
Summary: "empty doc",
|
||
Content: "",
|
||
Source: "test",
|
||
}
|
||
triples := coldDocToTriples(doc, nil)
|
||
if len(triples) < 2 {
|
||
t.Fatalf("expected at least 2 triples (主题+来源), got %d", len(triples))
|
||
}
|
||
|
||
if triples[0].Subject != "文档" || triples[0].Relation != "主题" || triples[0].Object != "empty doc" {
|
||
t.Errorf("first triple mismatch: %+v", triples[0])
|
||
}
|
||
|
||
last := triples[len(triples)-1]
|
||
if last.Subject != "文档" || last.Relation != "来源" || last.Object != "test" {
|
||
t.Errorf("last triple mismatch: %+v", last)
|
||
}
|
||
}
|
||
|
||
func TestDocToTriplesConversation(t *testing.T) {
|
||
doc := &coldcontext.ColdDoc{
|
||
Summary: "测试对话 (qq) 涉及: 天气",
|
||
Content: "[15:04] qq: 今天天气怎么样\n[15:05] agent: 今天天气很好",
|
||
Source: "qq",
|
||
}
|
||
triples := coldDocToTriples(doc, nil)
|
||
|
||
if len(triples) < 2 {
|
||
t.Errorf("expected at least 2 triples (主题+来源), got %d", len(triples))
|
||
}
|
||
|
||
for i, tr := range triples {
|
||
if tr.Subject == "" || tr.Relation == "" || tr.Object == "" {
|
||
t.Errorf("triple[%d] has empty field: %+v", i, tr)
|
||
}
|
||
if tr.Confidence <= 0 {
|
||
t.Errorf("triple[%d] has non-positive confidence: %+v", i, tr)
|
||
}
|
||
}
|
||
}
|
||
|
||
func TestDocToTriplesMultiLine(t *testing.T) {
|
||
doc := &coldcontext.ColdDoc{
|
||
Summary: "多轮对话",
|
||
Content: "[10:00] user: 你好\n[10:01] agent: 你好，有什么可以帮助你的\n[10:02] user: 今天天气如何\n[10:03] agent: 今天天气很好",
|
||
Source: "qq",
|
||
}
|
||
triples := coldDocToTriples(doc, nil)
|
||
if len(triples) < 2 {
|
||
t.Fatalf("expected at least 2 triples, got %d", len(triples))
|
||
}
|
||
|
||
if triples[0].Subject != "文档" || triples[0].Relation != "主题" {
|
||
t.Errorf("first triple should be 主题, got %+v", triples[0])
|
||
}
|
||
last := triples[len(triples)-1]
|
||
if last.Subject != "文档" || last.Relation != "来源" {
|
||
t.Errorf("last triple should be 来源, got %+v", last)
|
||
}
|
||
}
|
||
|
||
func TestDocToTriplesEmptyContent(t *testing.T) {
|
||
doc := &coldcontext.ColdDoc{
|
||
Summary: "空内容",
|
||
Content: "",
|
||
Source: "test",
|
||
}
|
||
triples := coldDocToTriples(doc, nil)
|
||
if len(triples) != 2 {
|
||
t.Fatalf("expected exactly 2 triples (主题+来源) for empty content, got %d", len(triples))
|
||
}
|
||
}
|
||
|
||
// Phase 2: 归档上下文文档不得产出模板垃圾（context_archived 来源/主题模板三元组）
|
||
func TestDocToTriplesArchivedContext(t *testing.T) {
|
||
doc := &coldcontext.ColdDoc{
|
||
Summary: "来自 2 个来源的 5 条对话 (qq, webui) 涉及: 天气, 测试",
|
||
Content: "[15:04] qq: 今天天气怎么样\n[15:05] agent: 今天天气很好",
|
||
Source: "context_archived",
|
||
Meta: map[string]string{"is_archived_context": "true"},
|
||
}
|
||
triples := coldDocToTriples(doc, nil)
|
||
|
||
for _, tr := range triples {
|
||
if tr.Subject == "文档" && tr.Relation == "来源" && tr.Object == "context_archived" {
|
||
t.Errorf("archived context must not write 来源 triple: %+v", tr)
|
||
}
|
||
if tr.Subject == "文档" && tr.Relation == "主题" {
|
||
t.Errorf("archived context must not write 主题 template triple: %+v", tr)
|
||
}
|
||
}
|
||
}
|
||
|
||
// Phase 2: 模板化摘要（summarizeEntries 生成）不得作为主题写入
|
||
func TestDocToTriplesTemplateSummary(t *testing.T) {
|
||
doc := &coldcontext.ColdDoc{
|
||
Summary: "来自 3 个来源的 10 条对话 (a, b, c) 涉及: 关键词1, 关键词2, 关键词3",
|
||
Content: "[10:00] a: 你好",
|
||
Source: "manual",
|
||
}
|
||
triples := coldDocToTriples(doc, nil)
|
||
|
||
for _, tr := range triples {
|
||
if tr.Subject == "文档" && tr.Relation == "主题" {
|
||
t.Errorf("template summary must not be written as 主题 triple: %+v", tr)
|
||
}
|
||
}
|
||
// 但非归档来源仍保留 来源 三元组
|
||
foundSource := false
|
||
for _, tr := range triples {
|
||
if tr.Subject == "文档" && tr.Relation == "来源" && tr.Object == "manual" {
|
||
foundSource = true
|
||
}
|
||
}
|
||
if !foundSource {
|
||
t.Errorf("non-archived source should still produce 来源 triple")
|
||
}
|
||
}
|
||
|
||
// Phase 2: 过长摘要不得写入主题
|
||
func TestDocToTriplesLongSummary(t *testing.T) {
|
||
long := ""
|
||
for i := 0; i < 100; i++ {
|
||
long += "很长的摘要内容片段重复拼接"
|
||
}
|
||
doc := &coldcontext.ColdDoc{
|
||
Summary: long,
|
||
Content: "[10:00] a: 你好",
|
||
Source: "test",
|
||
}
|
||
triples := coldDocToTriples(doc, nil)
|
||
for _, tr := range triples {
|
||
if tr.Subject == "文档" && tr.Relation == "主题" {
|
||
t.Errorf("overlong summary must not be written as 主题 triple")
|
||
}
|
||
}
|
||
}
|
||
|
||
func TestIsTemplateSummary(t *testing.T) {
|
||
if !isTemplateSummary("来自 2 个来源的 5 条对话 (qq, webui) 涉及: 天气") {
|
||
t.Errorf("template summary not recognized")
|
||
}
|
||
if isTemplateSummary("今天天气很好") {
|
||
t.Errorf("plain summary wrongly recognized as template")
|
||
}
|
||
if !isTemplateSummary("") {
|
||
t.Errorf("empty summary should be treated as template")
|
||
}
|
||
}
|
||
|
||
func TestTruncateStr(t *testing.T) {
|
||
tests := []struct {
|
||
input string
|
||
max int
|
||
want string
|
||
}{
|
||
{"hello", 10, "hello"},
|
||
{"hello world", 5, "hello..."},
|
||
{"你好世界", 2, "你好..."},
|
||
{"", 5, ""},
|
||
{"abc", 3, "abc"},
|
||
}
|
||
for _, tt := range tests {
|
||
got := truncateStr(tt.input, tt.max)
|
||
if got != tt.want {
|
||
t.Errorf("truncateStr(%q, %d) = %q, want %q", tt.input, tt.max, got, tt.want)
|
||
}
|
||
}
|
||
}
|
||
|
||
func TestGetString(t *testing.T) {
|
||
m := map[string]interface{}{
|
||
"name": "张三",
|
||
"age": 30,
|
||
}
|
||
|
||
if got := getString(m, "name"); got != "张三" {
|
||
t.Errorf("expected '张三', got %q", got)
|
||
}
|
||
if got := getString(m, "age"); got != "" {
|
||
t.Errorf("expected empty for int, got %q", got)
|
||
}
|
||
if got := getString(m, "nonexistent"); got != "" {
|
||
t.Errorf("expected empty for missing key, got %q", got)
|
||
}
|
||
if got := getString(nil, "key"); got != "" {
|
||
t.Errorf("expected empty for nil map, got %q", got)
|
||
}
|
||
}
|
||
|
||
func TestGetFloat(t *testing.T) {
|
||
m := map[string]interface{}{
|
||
"count": 42.5,
|
||
"score": 100,
|
||
"name": "test",
|
||
}
|
||
|
||
if got := getFloat(m, "count"); got != 42.5 {
|
||
t.Errorf("expected 42.5, got %f", got)
|
||
}
|
||
if got := getFloat(m, "score"); got != 100.0 {
|
||
t.Errorf("expected 100.0, got %f", got)
|
||
}
|
||
if got := getFloat(m, "name"); got != 0 {
|
||
t.Errorf("expected 0 for string, got %f", got)
|
||
}
|
||
if got := getFloat(m, "nonexistent"); got != 0 {
|
||
t.Errorf("expected 0 for missing key, got %f", got)
|
||
}
|
||
if got := getFloat(nil, "key"); got != 0 {
|
||
t.Errorf("expected 0 for nil map, got %f", got)
|
||
}
|
||
}
|
||
|
||
func TestGetFloatInt(t *testing.T) {
|
||
m := map[string]interface{}{
|
||
"top_k": float64(5),
|
||
}
|
||
if got := getFloat(m, "top_k"); got != 5.0 {
|
||
t.Errorf("expected 5.0, got %f", got)
|
||
}
|
||
}
|
||
|
||
// TestDocToTriplesDropsNoiseEntities 钉住 doc→graph 的噪音闸门。
|
||
//
|
||
// 背景：doc→graph 在 1cb3e87 从「CutExact 滑窗词链」换成 NLP 依存提取器后，
|
||
// 唯一还拦常用词的那层（CutExact：去停用词 + validEntityName）失去调用点，
|
||
// 闸门只剩 validEntityName——它只管名字像不像名字，不管名字是不是常用词。
|
||
// 实测生产库里因此攒下「文档 --主题--> 来自 N 个来源的 M 条对话 …」这类
|
||
// 模板回声，以及 context_archived 这个内部标记。
|
||
func TestDocToTriplesDropsNoiseEntities(t *testing.T) {
|
||
doc := &coldcontext.ColdDoc{
|
||
Summary: "来自 1 个来源的 2 条对话 (agent) 涉及: qq, 通道",
|
||
Content: "",
|
||
Source: "context_archived",
|
||
}
|
||
triples := coldDocToTriples(doc, nil)
|
||
|
||
for _, tr := range triples {
|
||
if memory.IsNoiseEntity(tr.Subject) || memory.IsNoiseEntity(tr.Object) {
|
||
t.Errorf("coldDocToTriples 漏出噪音实体: %+v", tr)
|
||
}
|
||
}
|
||
|
||
// 模板摘要不当「主题」、context_archived 不当「来源」：两条模板三元组都该被拦下。
|
||
for _, tr := range triples {
|
||
if tr.Relation == "主题" {
|
||
t.Errorf("模板摘要被写成主题: %+v", tr)
|
||
}
|
||
if tr.Relation == "来源" && tr.Object == "context_archived" {
|
||
t.Errorf("归档内部标记被写成来源: %+v", tr)
|
||
}
|
||
}
|
||
}
|
||
|
||
// TestDocToTriplesKeepsTemplateAnchors 保证闸门没把正常的模板三元组一起误杀。
|
||
func TestDocToTriplesKeepsTemplateAnchors(t *testing.T) {
|
||
doc := &coldcontext.ColdDoc{
|
||
Summary: "多轮对话",
|
||
Content: "",
|
||
Source: "qq",
|
||
}
|
||
triples := coldDocToTriples(doc, nil)
|
||
|
||
var hasTopic, hasSource bool
|
||
for _, tr := range triples {
|
||
if tr.Subject == "文档" && tr.Relation == "主题" && tr.Object == "多轮对话" {
|
||
hasTopic = true
|
||
}
|
||
if tr.Subject == "文档" && tr.Relation == "来源" && tr.Object == "qq" {
|
||
hasSource = true
|
||
}
|
||
}
|
||
if !hasTopic || !hasSource {
|
||
t.Errorf("正常模板三元组被误杀: topic=%v source=%v, triples=%+v", hasTopic, hasSource, triples)
|
||
}
|
||
}
|