Cyrene_v2/internal/services/gateway/history_document_test.go

package gateway

import (
	"encoding/json"
	"sort"
	"strings"
	"testing"

	"git.yeij.top/AskaEth/Cyrene_v2/internal/runtime/history"
	"git.yeij.top/AskaEth/Cyrene_v2/internal/runtime/model"
)

// 本文件是历史 DTO 的 document 对象与持久化取值的测试（design/24 §2.4、§6.1）：
// 引用与统计往返、**不回传抽取正文**、坏列可见，以及 session_persist 的
// messageTextImageDocument 取值。

// TestHistoryMessageDocumentCarriesReferenceOnly 验证 document 对象只带
// 引用与统计：字段集被逐个钉住，将来谁往 DocumentRef 里加一个正文/摘要字段，
// 这里就会红。
func TestHistoryMessageDocumentCarriesReferenceOnly(t *testing.T) {
	t.Parallel()
	raw, err := history.EncodeDocument(history.DocumentRef{
		Name:      "0123456789abcdef0123456789abcdef.txt",
		MIME:      "text/plain",
		Size:      12345,
		Chars:     8000,
		Truncated: true,
		Status:    history.DocumentStatusExtracted,
	})
	if err != nil {
		t.Fatalf("encode: %v", err)
	}
	wire := toHistoryMessage(history.Message{ID: "m1", SessionID: "web", Role: "user", Content: "总结这份报告", Document: raw})
	b, err := json.Marshal(wire)
	if err != nil {
		t.Fatalf("marshal: %v", err)
	}

	var decoded map[string]any
	if err := json.Unmarshal(b, &decoded); err != nil {
		t.Fatalf("unmarshal: %v", err)
	}
	doc, ok := decoded["document"].(map[string]any)
	if !ok {
		t.Fatalf("document object missing in %s", b)
	}
	keys := make([]string, 0, len(doc))
	for k := range doc {
		keys = append(keys, k)
	}
	sort.Strings(keys)
	if got := strings.Join(keys, ","); got != "chars,mime,name,size,status,truncated" {
		t.Fatalf("document keys = %q, want chars,mime,name,size,status,truncated", got)
	}
	if doc["name"] != "0123456789abcdef0123456789abcdef.txt" || doc["status"] != history.DocumentStatusExtracted {
		t.Fatalf("document = %v, want the persisted reference", doc)
	}
}

// TestHistoryMessageDocumentAbsentWhenNoColumn 验证没有附件的消息省略该字段：
// 空列与「带附件」必须可区分（省略 = 没有附件）。
func TestHistoryMessageDocumentAbsentWhenNoColumn(t *testing.T) {
	t.Parallel()
	wire := toHistoryMessage(history.Message{ID: "m1", Role: "user", Content: "没有附件"})
	b, err := json.Marshal(wire)
	if err != nil {
		t.Fatalf("marshal: %v", err)
	}
	if strings.Contains(string(b), `"document"`) {
		t.Fatalf("empty column should omit document, got %s", b)
	}
}

// TestHistoryMessageDocumentMalformedColumnIsVisible 验证列里是坏 JSON 时
// 不静默当成「没有附件」：回一个标着 status=unreadable 的最小对象，让客户端
// 看得见这条引用读不出来（design/24 §4.4 的「失败必须可见」）。
func TestHistoryMessageDocumentMalformedColumnIsVisible(t *testing.T) {
	t.Parallel()
	wire := toHistoryMessage(history.Message{ID: "m1", Role: "user", Content: "带附件", Document: "{not json"})
	if wire.Document == nil {
		t.Fatal("malformed document column was silently dropped")
	}
	if wire.Document.Status != history.DocumentStatusUnreadable {
		t.Fatalf("status = %q, want %q", wire.Document.Status, history.DocumentStatusUnreadable)
	}
}

// TestMessageTextImageDocument 验证持久化取值：文本、首个图片 URL 与首个文档
// 引用各归各列，且文档列是引用（JSON 小对象）而不是正文。
func TestMessageTextImageDocument(t *testing.T) {
	t.Parallel()
	doc := model.Document{Name: "aaaabbbbccccdddd.txt", MIME: "text/plain", URL: "/api/v1/documents/aaaabbbbccccdddd.txt", Size: 1234}
	m := model.DocumentMessage(model.RoleUser, "读一下", doc)
	m.Content = append(m.Content, model.Content{Type: model.ContentTypeImage, ImageURL: "data:image/png;base64,abc"})

	text, image, raw, err := messageTextImageDocument(m)
	if err != nil {
		t.Fatalf("messageTextImageDocument: %v", err)
	}
	if text != "读一下" {
		t.Fatalf("text = %q, want 读一下", text)
	}
	if image != "data:image/png;base64,abc" {
		t.Fatalf("image = %q", image)
	}
	ref, err := history.DecodeDocument(raw)
	if err != nil {
		t.Fatalf("decode document: %v", err)
	}
	if ref == nil || ref.Name != doc.Name || ref.MIME != doc.MIME || ref.Size != doc.Size {
		t.Fatalf("ref = %+v, want name/mime/size of %+v", ref, doc)
	}
}

// TestMessageTextImageDocumentWithoutDocument 验证没有文档部件时文档列为空
// （而不是 "{}"）：拼写空对象会让读取侧把它当成一条读不出来的附件。
func TestMessageTextImageDocumentWithoutDocument(t *testing.T) {
	t.Parallel()
	_, _, raw, err := messageTextImageDocument(model.TextMessage(model.RoleUser, "只有文本"))
	if err != nil {
		t.Fatalf("messageTextImageDocument: %v", err)
	}
	if raw != "" {
		t.Fatalf("document = %q, want empty", raw)
	}
}