Cyrene_v2/internal/services/gateway/session_document_e2e_test.go

package gateway

import (
	"context"
	"crypto/sha256"
	"encoding/hex"
	"encoding/json"
	"net/http"
	"net/http/httptest"
	"os"
	"path/filepath"
	"sync"
	"testing"
	"time"

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

// 本文件是 S3.5 在 **HTTP 层**的端到端回归（design/24 §6.1、§9.7）：一次 `chat`
// 带上 3 个附件 → 全部落库 → 回合结束后抽取结论回写 → 历史接口按 `documents`
// 数组把引用与结论一起交回客户端。它补齐单测覆盖不到的那一段接线：REST 入口 →
// 编排器 → 真实 history 存储 → 历史读端点。

// docStatusRunner 是端到端的假回合驱动器：它把这一回合用户消息里的**每一个**
// 文档部件都当成「已抽取」的结论回给网关（正是真实抽取层 reshape 后的形状），
// 并记录它实际看到的文档部件数。
type docStatusRunner struct {
	mu    sync.Mutex
	docs  []model.Content
	calls int
}

func (r *docStatusRunner) Run(_ context.Context, _ string, msgs []model.Message) (*loop.Result, error) {
	r.mu.Lock()
	defer r.mu.Unlock()
	r.calls++
	refs := make([]history.DocumentRef, 0, len(r.docs))
	for _, m := range msgs {
		for _, c := range m.Content {
			if c.Type != model.ContentTypeDocument {
				continue
			}
			r.docs = append(r.docs, c)
			refs = append(refs, history.DocumentRef{
				Name: c.DocumentName, MIME: c.DocumentMIME, Size: c.DocumentSize,
				Chars: 12, Status: history.DocumentStatusExtracted,
			})
		}
	}
	return &loop.Result{Text: "读完了", Documents: refs, Steps: 1, StopReason: "stop"}, nil
}

// documentCount 返回该回合被驱动时看到的文档部件数。
func (r *docStatusRunner) documentCount() int {
	r.mu.Lock()
	defer r.mu.Unlock()
	return len(r.docs)
}

// TestChatDocumentsEndToEndCarriesEveryAttachmentAndConclusion 是 S3.5 两个缺口的
// 端到端回归：3 个附件全部进入模型视图、全部落库，且落库的每一行都带回
// status/chars/truncated（此前只有第一个附件落库，且结论永远到不了客户端）。
func TestChatDocumentsEndToEndCarriesEveryAttachmentAndConclusion(t *testing.T) {
	t.Parallel()
	dataDir := t.TempDir()
	jwt := auth.New(config.AuthConfig{JWTSecret: testSecret})
	cfg := &config.Config{Auth: config.AuthConfig{AdminUser: "admin", AdminPass: "secret", JWTSecret: testSecret}}
	runner := &docStatusRunner{}
	st := newGatewayHistoryStore(t)
	srv, err := New(cfg, Deps{Auth: jwt, DataDir: dataDir, Loop: runner, History: st})
	if err != nil {
		t.Fatalf("构造网关失败: %v", err)
	}
	ts := httptest.NewServer(srv.Handler())
	t.Cleanup(ts.Close)
	t.Cleanup(srv.Close)
	tok, err := jwt.Sign("acct", auth.RoleOwner, time.Hour)
	if err != nil {
		t.Fatalf("签发 token 失败: %v", err)
	}

	names := []string{
		storeGatewayDocument(t, dataDir, "第一份", ".txt"),
		storeGatewayDocument(t, dataDir, "第二份", ".md"),
		storeGatewayDocument(t, dataDir, "第三份", ".csv"),
	}
	refs := make([]chatDocumentRef, 0, len(names))
	for _, name := range names {
		refs = append(refs, chatDocumentRef{Name: name})
	}
	resp, body := doReq(t, http.MethodPost, ts.URL+"/api/v1/chat", map[string]any{
		"sessionId": history.MainSessionID, "text": "这三份都看看", "documents": refs,
	}, tok)
	if resp.StatusCode != http.StatusOK {
		t.Fatalf("chat status = %d, want 200 (body=%s)", resp.StatusCode, body)
	}
	if got := runner.documentCount(); got != 3 {
		t.Fatalf("进入模型可见视图的文档部件 = %d 个, want 3（缺口①：附件不能丢）", got)
	}

	resp, body = doReq(t, http.MethodGet, ts.URL+"/api/v1/sessions/"+history.MainSessionID+"/messages", nil, tok)
	if resp.StatusCode != http.StatusOK {
		t.Fatalf("messages status = %d, want 200 (body=%s)", resp.StatusCode, body)
	}
	var envelope struct {
		Data []struct {
			ID        string                `json:"id"`
			Role      string                `json:"role"`
			Content   string                `json:"content"`
			Documents []history.DocumentRef `json:"documents"`
		} `json:"data"`
	}
	if err := json.Unmarshal(body, &envelope); err != nil {
		t.Fatalf("unmarshal: %v (body=%s)", err, body)
	}
	var userDocs []history.DocumentRef
	for _, m := range envelope.Data {
		if m.Role == history.RoleUser && m.Content == "这三份都看看" {
			userDocs = m.Documents
		}
	}
	if len(userDocs) != 3 {
		t.Fatalf("历史接口返回的 documents = %+v, want 3 条（缺口①）", userDocs)
	}
	for i, name := range names {
		if userDocs[i].Name != name {
			t.Fatalf("documents[%d].Name = %q, want %q（顺序必须与发送一致）", i, userDocs[i].Name, name)
		}
		if userDocs[i].Status != history.DocumentStatusExtracted || userDocs[i].Chars != 12 {
			t.Fatalf("documents[%d] = %+v, want extracted/chars=12（缺口②）", i, userDocs[i])
		}
	}
}

// storeGatewayDocument 按 S1 的内容寻址规则（sha256 前 32 位 + 扩展名）在
// `<DataDir>/uploads/documents` 下落一份文档并返回其存储名（与 chat_documents_test
// 的 helper 同规则；这里在 HTTP 端到端用例里独立使用，避免依赖那边的 env）。
func storeGatewayDocument(t *testing.T, dataDir, content, ext string) string {
	t.Helper()
	sum := sha256.Sum256([]byte(content))
	name := hex.EncodeToString(sum[:])[:documentHashHexLen] + ext
	if !validDocumentName(name) {
		t.Fatalf("构造的存储名 %q 不是合法形态", name)
	}
	dir := filepath.Join(dataDir, "uploads", documentSubDir)
	if err := os.MkdirAll(dir, 0o700); err != nil {
		t.Fatalf("创建文档目录失败: %v", err)
	}
	if err := os.WriteFile(filepath.Join(dir, name), []byte(content), 0o600); err != nil {
		t.Fatalf("写入文档失败: %v", err)
	}
	return name
}