[fix][evaluation] add OpenAPI extra_output field and fix two horizontal privilege escalations

[jamesonics]

What

本 PR 合并了 evaluation 模块的三项改动（一次发布）：

1. feat: ListExperimentResultOApi 响应新增 extra_output 字段

• domain_openapi/evaluator.thrift 新增 EvaluatorExtraOutputContent struct 及 extra_output 字段，重新生成
  kitex_gen
• eval_openapi_app.go 注入 fileProvider，对 TOS URI 签名为可访问 URL（fillExtraOutputURLs）
• 补充 convertor 与单测

2. fix: BatchGetExptAggrResult 水平越权校验

• expt_result_aggr_impl.go：取出实验后按 expt.SpaceID == spaceID 过滤出
  validExptIDs，下游聚合/评估器引用/标签查询全部改用过滤后的 ID，防止跨 workspace 读取他人实验聚合结果

3. fix: UpdateExperiment 水平越权校验

• experiment_app.go：在写库前补 got.SpaceID != req.GetWorkspaceID() 校验（对齐 DeleteExperiment
  既有范式），防止用本 workspace 的 id 覆盖/搬移他人实验的 space_id

Why

#2、#3 是同类水平越权（horizontal privilege escalation）修复：攻击者持有 A workspace
的合法凭证，却能读取/修改属于 B workspace 的实验资源。修复对齐已有 DeleteExperiment 校验范式，复用
errno.CommonBadRequestCode，不新增错误码。

关联 Meego: https://meego.larkoffice.com/fornax/story/detail/7344006097 （水平校验修复）

单元测试（gomock + testify，本地 GOTOOLCHAIN=go1.24.6 go test 全绿）：

• TestExperimentApplication_UpdateExperiment 新增 "workspace mismatch with experiment space" 用例：mock 返回
  SpaceID 不匹配的实验，断言报错且不调用 Update（先校验后写库）；红→绿变异验证通过
• TestExptAggrResultServiceImpl_BatchGetExptAggrResultByExperimentIDs 新增 cross-space filter 用例：混入跨
  space 实验，断言下游查询只对合法 exptID 发起；移除过滤逻辑后用例如期 FAIL
• evaluation application + domain/service 全模块回归通过

PPE 泳道端到端验证（ppe_test 泳道，跨 workspace 越权场景）：

• 越权读（A 查 B 的实验聚合结果）→ 静默过滤返回空结果，无穿透报错（对比线上 baseline 601200702）
• 正向读（A 查自己实验）→ 正常返回完整聚合结果，未误伤
• 越权写（A 改 B 的实验）→ 返回 601200201 "expt ... not found in space ..."，DB 未写入
• LogID 日志确认请求落在 ppe 泳道（_env=ppe）

Scope note

本 PR 仅含 evaluation/openapi 改动（16 文件，其中 3 个为 kitex_gen 生成代码）。仅修水平越权校验，未改 DAO 层
WHERE（纵深防御另议）；notification_conf / pool.go 经核查当前代码已一致，不在本次范围。

[codecov]

Codecov Report

❌ Patch coverage is 95.65217% with 3 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
...modules/evaluation/application/eval_openapi_app.go	93.75%	2 Missing and 1 partial ⚠️

@@            Coverage Diff             @@
##             main     #564      +/-   ##
==========================================
+ Coverage   77.60%   78.12%   +0.51%     
==========================================
  Files         670      670              
  Lines       75995    76214     +219     
==========================================
+ Hits        58979    59543     +564     
+ Misses      13565    13238     -327     
+ Partials     3451     3433      -18     

Flag	Coverage Δ
unittests	78.12% <95.65%> (+0.51%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
...luation/application/convertor/evaluator/openapi.go	96.53% <100.00%> (+4.63%)	⬆️
...uation/application/convertor/experiment/openapi.go	89.82% <100.00%> (+6.25%)	⬆️
...d/modules/evaluation/application/experiment_app.go	85.25% <100.00%> (+0.90%)	⬆️
...evaluation/domain/service/expt_result_aggr_impl.go	79.75% <100.00%> (+0.08%)	⬆️
...modules/evaluation/application/eval_openapi_app.go	92.33% <93.75%> (+0.05%)	⬆️

... and 26 files with indirect coverage changes

---

Continue to review full report in Codecov by Harness.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 016641a...0f3db7a. Read the comment docs.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.