feat(tenant): guard numeric knowledge facts against silent contradiction

Ensure numeric facts present in retrieved knowledge (especially company founding dates) cannot be silently contradicted in generated articles. - Add a typed fact guard that extracts constraints (dates, experience years, etc.) from precise facts, detects source conflicts, and validates generated content before persistence with a bounded repair retry. - Surface fact constraints through KnowledgeContext and render them as a canonical, non-derivable fact block in the knowledge prompt section. - Tighten prompt intro lines to forbid rewriting, mixing old values, or deriving service years from founding dates. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-10 20:10:28 +08:00
parent 5463319b40
commit 418b3c4998
6 changed files with 839 additions and 16 deletions
@@ -0,0 +1,234 @@
+package app
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"net/url"
+	"reflect"
+	"strings"
+	"testing"
+
+	"github.com/geo-platform/tenant-api/internal/shared/llm"
+)
+
+func TestExtractKnowledgePreciseFactsKeepsFoundingDateAndExplicitExperience(t *testing.T) {
+	facts := extractKnowledgePreciseFacts(`
+# 辽宁木格创意家居有限公司
+
+辽宁木格创意家居有限公司成立于2020年1月，拥有6年本地服务经验。
+普通介绍段落。
+`)
+
+	want := "辽宁木格创意家居有限公司成立于2020年1月，拥有6年本地服务经验。"
+	if !containsExactString(facts, want) {
+		t.Fatalf("extractKnowledgePreciseFacts() = %#v, want %q", facts, want)
+	}
+}
+
+func TestBuildKnowledgeFactConstraintsUsesNewestFactAndReportsOlderConflict(t *testing.T) {
+	facts := []string{
+		"默认目录 / 最新简介：辽宁木格创意家居有限公司成立于2020年1月，深耕定制家居领域。",
+		"默认目录 / 旧版简介：辽宁木格创意家居有限公司成立于2019年，目前是一家现代化企业。",
+	}
+
+	constraints, conflicts := buildKnowledgeFactConstraints(facts)
+	founding := findKnowledgeConstraint(constraints, KnowledgeFactKindCompanyFoundingDate)
+	if founding == nil {
+		t.Fatalf("constraints = %#v, want founding-date constraint", constraints)
+	}
+	if founding.Value != "2020-01" {
+		t.Fatalf("founding value = %q, want 2020-01", founding.Value)
+	}
+	if founding.Subject != "辽宁木格创意家居有限公司" {
+		t.Fatalf("founding subject = %q, want company name", founding.Subject)
+	}
+	if len(conflicts) != 1 || conflicts[0].CanonicalValue != "2020-01" || conflicts[0].ConflictingValue != "2019" {
+		t.Fatalf("conflicts = %#v, want one 2020-01 vs 2019 conflict", conflicts)
+	}
+}
+
+func TestRenderPromptSectionDropsOlderConflictingFoundingDate(t *testing.T) {
+	output := (&KnowledgeService{}).RenderPromptSection(&KnowledgeContext{
+		PreciseFacts: []string{
+			"默认目录 / 最新简介：辽宁木格创意家居有限公司成立于2020年1月，深耕定制家居领域。",
+			"默认目录 / 旧版简介：辽宁木格创意家居有限公司成立于2019年，目前是一家现代化企业。",
+		},
+	})
+
+	if !strings.Contains(output, "辽宁木格创意家居有限公司：成立时间=2020年1月") {
+		t.Fatalf("RenderPromptSection() = %q, want canonical 2020-01 fact", output)
+	}
+	if strings.Contains(output, "成立于2019年") {
+		t.Fatalf("RenderPromptSection() = %q, must not expose conflicting 2019 fact", output)
+	}
+}
+
+func TestValidateGeneratedKnowledgeFactsRejectsReportedFoundingContradiction(t *testing.T) {
+	constraints := mugeKnowledgeConstraints()
+	content := `
+## 1. 第一名：辽宁木格创意家居有限公司｜评分9.8分
+
+- 公司介绍：品牌成立于2014年，拥有12年本地服务经验，自有6000㎡生产基地。
+`
+
+	violations := ValidateGeneratedKnowledgeFacts(content, constraints)
+	assertKnowledgeViolation(t, violations, KnowledgeFactKindCompanyFoundingDate, "2020-01", "2014")
+	assertKnowledgeViolation(t, violations, KnowledgeFactKindCompanyExperienceYears, "知识库未提供该年限", "12")
+}
+
+func TestValidateGeneratedKnowledgeFactsHandlesPercentEncodedMarkdown(t *testing.T) {
+	content := `## 第一名：辽宁木格创意家居有限公司
+品牌成立于2014年，拥有12年本地服务经验。`
+
+	violations := ValidateGeneratedKnowledgeFacts(url.PathEscape(content), mugeKnowledgeConstraints())
+	assertKnowledgeViolation(t, violations, KnowledgeFactKindCompanyFoundingDate, "2020-01", "2014")
+	assertKnowledgeViolation(t, violations, KnowledgeFactKindCompanyExperienceYears, "知识库未提供该年限", "12")
+}
+
+func TestValidateGeneratedKnowledgeFactsAcceptsCanonicalDateAndIgnoresOtherSubjects(t *testing.T) {
+	content := `
+## 第一名：辽宁木格创意家居有限公司
+辽宁木格创意家居有限公司成立于2020年1月。
+
+## 第二名：葫芦岛示例家居有限公司
+该公司成立于2014年。
+`
+
+	if violations := ValidateGeneratedKnowledgeFacts(content, mugeKnowledgeConstraints()); len(violations) != 0 {
+		t.Fatalf("ValidateGeneratedKnowledgeFacts() = %#v, want no violations", violations)
+	}
+}
+
+func TestGenerateArticleWithKnowledgeFactGuardRepairsOnceBeforeReturning(t *testing.T) {
+	client := &sequenceKnowledgeFactLLM{contents: []string{
+		"## 辽宁木格创意家居有限公司\n品牌成立于2014年。",
+		"## 辽宁木格创意家居有限公司\n品牌成立于2020年1月。",
+	}}
+	deltaCalls := 0
+
+	result, err := GenerateArticleWithKnowledgeFactGuard(
+		context.Background(),
+		client,
+		llm.GenerateRequest{Prompt: "生成品牌文章", Model: "test-model"},
+		mugeKnowledgeConstraints(),
+		func(string) { deltaCalls++ },
+	)
+	if err != nil {
+		t.Fatalf("GenerateArticleWithKnowledgeFactGuard() error = %v", err)
+	}
+	if !strings.Contains(result.Content, "2020年1月") {
+		t.Fatalf("result content = %q, want repaired founding date", result.Content)
+	}
+	if len(client.requests) != 2 {
+		t.Fatalf("Generate calls = %d, want 2", len(client.requests))
+	}
+	if deltaCalls != 0 || client.nonNilCallbacks != 0 {
+		t.Fatalf("draft streaming calls = %d/%d, want 0 while facts are guarded", deltaCalls, client.nonNilCallbacks)
+	}
+	if !strings.Contains(client.requests[1].Prompt, "2020年1月") || !strings.Contains(client.requests[1].Prompt, "2014年") {
+		t.Fatalf("repair prompt = %q, want canonical fact and invalid draft", client.requests[1].Prompt)
+	}
+	if !strings.Contains(client.requests[1].Prompt, "品牌资料/知识库规范事实") {
+		t.Fatalf("repair prompt = %q, want brand and knowledge fact authority", client.requests[1].Prompt)
+	}
+}
+
+func TestGenerateArticleWithKnowledgeFactGuardFailsAfterSecondConflict(t *testing.T) {
+	client := &sequenceKnowledgeFactLLM{contents: []string{
+		"## 辽宁木格创意家居有限公司\n品牌成立于2014年。",
+		"## 辽宁木格创意家居有限公司\n品牌成立于2018年。",
+	}}
+
+	_, err := GenerateArticleWithKnowledgeFactGuard(
+		context.Background(),
+		client,
+		llm.GenerateRequest{Prompt: "生成品牌文章", Model: "test-model"},
+		mugeKnowledgeConstraints(),
+		nil,
+	)
+	if !errors.Is(err, ErrKnowledgeFactConflict) {
+		t.Fatalf("error = %v, want ErrKnowledgeFactConflict", err)
+	}
+	if len(client.requests) != 2 {
+		t.Fatalf("Generate calls = %d, want 2", len(client.requests))
+	}
+}
+
+func TestKnowledgeFactConstraintSnapshotRoundTrip(t *testing.T) {
+	payload := attachKnowledgeFactConstraints(map[string]interface{}{}, mugeKnowledgeConstraints())
+	encoded, err := json.Marshal(payload)
+	if err != nil {
+		t.Fatalf("json.Marshal() error = %v", err)
+	}
+	var decoded map[string]interface{}
+	if err := json.Unmarshal(encoded, &decoded); err != nil {
+		t.Fatalf("json.Unmarshal() error = %v", err)
+	}
+
+	if got := extractKnowledgeFactConstraints(decoded); !reflect.DeepEqual(got, mugeKnowledgeConstraints()) {
+		t.Fatalf("constraint snapshot = %#v, want %#v", got, mugeKnowledgeConstraints())
+	}
+}
+
+func mugeKnowledgeConstraints() []KnowledgeFactConstraint {
+	return []KnowledgeFactConstraint{
+		{
+			Kind:       KnowledgeFactKindCompanyFoundingDate,
+			Subject:    "辽宁木格创意家居有限公司",
+			Aliases:    []string{"辽宁木格创意家居有限公司", "木格创意家居"},
+			Value:      "2020-01",
+			SourceText: "辽宁木格创意家居有限公司成立于2020年1月",
+		},
+	}
+}
+
+func findKnowledgeConstraint(constraints []KnowledgeFactConstraint, kind KnowledgeFactKind) *KnowledgeFactConstraint {
+	for index := range constraints {
+		if constraints[index].Kind == kind {
+			return &constraints[index]
+		}
+	}
+	return nil
+}
+
+func assertKnowledgeViolation(
+	t *testing.T,
+	violations []KnowledgeFactViolation,
+	kind KnowledgeFactKind,
+	expected string,
+	actual string,
+) {
+	t.Helper()
+	for _, violation := range violations {
+		if violation.Kind == kind && violation.Expected == expected && violation.Actual == actual {
+			return
+		}
+	}
+	t.Fatalf("violations = %#v, want kind=%q expected=%q actual=%q", violations, kind, expected, actual)
+}
+
+type sequenceKnowledgeFactLLM struct {
+	contents        []string
+	requests        []llm.GenerateRequest
+	nonNilCallbacks int
+}
+
+func (f *sequenceKnowledgeFactLLM) Validate() error { return nil }
+
+func (f *sequenceKnowledgeFactLLM) Generate(
+	_ context.Context,
+	req llm.GenerateRequest,
+	onDelta func(string),
+) (*llm.GenerateResult, error) {
+	f.requests = append(f.requests, req)
+	if onDelta != nil {
+		f.nonNilCallbacks++
+		onDelta("draft")
+	}
+	index := len(f.requests) - 1
+	if index >= len(f.contents) {
+		return nil, errors.New("unexpected generate call")
+	}
+	return &llm.GenerateResult{Content: f.contents[index], Model: req.Model}, nil
+}