diff --git a/backend/app/api/compare.py b/backend/app/api/compare.py index 44bf258e..16fbf75f 100644 --- a/backend/app/api/compare.py +++ b/backend/app/api/compare.py @@ -1,11 +1,12 @@ """运行对比:同一个工作流的两次运行按「变化三分法」对照——数据变了、口径变了、还是模板 / SQL 变了(engine/compare.py)。 -GET /api/runs/{run_id}/compare/base:默认拿哪一次来比,{base_run_id, reason, base}。规则在 pick_base,定时任务结束后 -的自动对比(scheduler/auto_compare.py)用的是同一个函数——触发历史里的「与上一次对比」和运行详情里的同名入口必须拿到 -同一个基准,计数才对得上;触发历史里存着的那一份,事后基准不再是这里挑出来的那一次(中间失败的那次被接着运行成功 -了)就不再给出(auto_compare.outdated)。这次运行是定时任务发起的,取同一定时任务、同一工作流、比它先发起的运行里 +GET /api/runs/{run_id}/compare/base:默认拿哪一次来比,{base_run_id, reason, base, previous}。规则在 pick_base,定时 +任务结束后的自动对比(scheduler/auto_compare.py)用的是同一个函数——触发历史里的「与上一次对比」和运行详情里的同名入口 +必须拿到同一个基准,计数才对得上;触发历史里存着的那一份,事后基准不再是这里挑出来的那一次(中间失败的那次被接着运行 +成功了)就不再给出(auto_compare.outdated)。这次运行是定时任务发起的,取同一定时任务、同一工作流、比它先发起的运行里 成功结束并封存的最近一次(reason: same_schedule);不是、或者同一任务之前没有成功的,取同一工作流、同一运行类别里 -同样条件的最近一次(reason: same_workflow)。找不到 404。 +同样条件的最近一次(reason: same_workflow)。previous 说基准是不是紧挨着的上一次(同自动对比记下的 previous):不是时 +两个入口都写「与上一次成功的运行对比」。找不到 404。 GET /api/runs/{run_id}/compare?base={base_id}:{target, base, summary, template, metrics, queries, notes},每个字段 的形状见 engine/compare.py。两次运行都必须已结束、封存且封存核对一致,封存判定照搬证据接口(api/evidence._sealed): @@ -62,11 +63,13 @@ BASE = "作为基准的运行" @router.get("/{run_id}/compare/base") async def compare_base(run_id: str) -> dict[str, Any]: - """默认拿来对比的那一次运行:{base_run_id, reason: same_schedule | same_workflow, base}。""" + """默认拿来对比的那一次运行:{base_run_id, reason: same_schedule | same_workflow, base, previous}。previous 是 + pick_base 算的「是不是紧挨着的上一次」:运行详情「更多」里的入口按它定叫法(不是时写「与上一次成功的运行对比」), + 和触发历史里同名的链接(自动对比记下的 previous)说法一致。""" found = await default_base(run_id) if found is None: raise CodedHTTPException(404, NO_BASE_TEXT, NO_BASE) - return {"base_run_id": found.run.id, "reason": found.reason, "base": run_ref(found.run)} + return {"base_run_id": found.run.id, "reason": found.reason, "base": run_ref(found.run), "previous": found.previous} @router.get("/{run_id}/compare") diff --git a/backend/tests/test_compare_api.py b/backend/tests/test_compare_api.py index 2257f070..541f87b7 100644 --- a/backend/tests/test_compare_api.py +++ b/backend/tests/test_compare_api.py @@ -7,7 +7,8 @@ - 对比只读:两次运行的 last_seq、事件、工件引用、工件文件一样不多 - 遮罩列(查询当时记下的、数据源现在设的)的值不出现在结果里 - 默认基准:同一定时任务优先,否则同一工作流、同一运行类别;只挑比这次先发起、成功结束并封存的运行,按发起先后取 - 最近一次(失败、取消的跳过,和定时任务的自动对比挑的是同一次,见 test_scheduler_compare.py) + 最近一次(失败、取消的跳过,和定时任务的自动对比挑的是同一次,见 test_scheduler_compare.py);previous 说它是不是 + 紧挨着的上一次 - 真跑:同一条 SQL 数据变了是 data_only;只是代进去的日期不同是 params_only,图一样 """ from __future__ import annotations @@ -335,6 +336,49 @@ async def test_the_default_base_follows_launch_order(client): assert (await get(client, f"/api/runs/{waiting}/compare/base"))["base_run_id"] == later +async def test_the_default_base_says_whether_it_is_the_previous_run(client): + """previous:基准是不是紧挨着的上一次,语义同 pick_base——同一范围(同一定时任务,或同一工作流、同一运行类别)里比 + 这次先发起、比基准后发起的运行一次都没有。失败、取消、还没结束、没封存的都算隔着;别的运行类别、别的定时任务(这次 + 是定时任务发起的)、比这次晚发起的不在范围里,不算。运行详情「更多」里的入口按它写「与上一次对比」还是「与上一次 + 成功的运行对比」,和触发历史里同名的链接一个说法""" + day = timedelta(days=1) + + async def pick(run_id: str) -> tuple[str, str, bool]: + body = await get(client, f"/api/runs/{run_id}/compare/base") + return body["base_run_id"], body["reason"], body["previous"] + + # 同一定时任务:紧挨着的上一次就是成功的 + workflow = await new_workflow() + first = await bare(workflow, T0, schedule_id="sched-weekly") + await bare(workflow, T0 + day) # 手动发起的:不在同一定时任务的范围里 + await bare(workflow, T0 + day, schedule_id="sched-daily", status="failed") # 别的定时任务失败的:同上 + second = await bare(workflow, T0 + 2 * day, schedule_id="sched-weekly") + await bare(workflow, T0 + 3 * day, schedule_id="sched-weekly", status="failed") # 比 second 晚发起 + assert await pick(second) == (first, "same_schedule", True) + + # 中间隔着一次没成功的(失败、取消、还没结束、结束了没封存):基准是再往前成功的那一次,不是紧挨着的上一次 + for between in ({"status": "failed"}, {"status": "cancelled"}, {"status": "running", "sealed": False}, + {"status": "interrupted", "sealed": False}, {"sealed": False}): + workflow = await new_workflow() + first = await bare(workflow, T0, schedule_id="sched-weekly") + await bare(workflow, T0 + day, schedule_id="sched-weekly", **between) + current = await bare(workflow, T0 + 2 * day, schedule_id="sched-weekly") + assert await pick(current) == (first, "same_schedule", False), between + + # 同一工作流、同一运行类别(手动发起的):别的运行类别不算隔着,同一类别里定时任务发起的失败的那次算 + workflow = await new_workflow() + first = await bare(workflow, T0) + await bare(workflow, T0 + day, run_class="exploratory", status="failed") + second = await bare(workflow, T0 + 2 * day) + assert await pick(second) == (first, "same_workflow", True) + await bare(workflow, T0 + 3 * day, schedule_id="sched-weekly", status="failed") + third = await bare(workflow, T0 + 4 * day) + assert await pick(third) == (second, "same_workflow", False) + # 定时任务的第一次退回同一工作流:隔没隔着按退回的那个范围算 + lone = await bare(workflow, T0 + 5 * day, schedule_id="sched-monthly") + assert await pick(lone) == (third, "same_workflow", True) + + # -------------------------------------------------------------------------- # 真跑:虚构的 SQLite 库,两次运行之间改了数、换了日期参数 # -------------------------------------------------------------------------- diff --git a/backend/tests/test_scheduler_compare.py b/backend/tests/test_scheduler_compare.py index d1bcb71e..1f850159 100644 --- a/backend/tests/test_scheduler_compare.py +++ b/backend/tests/test_scheduler_compare.py @@ -256,6 +256,9 @@ async def test_finished_run_is_compared_with_the_previous_run_of_its_schedule(re items = {f["id"]: f for f in (await client.get(f"/api/schedules/{sid}/fires")).json()["items"]} assert items[fire_id]["compare"] == {"base_run_id": base_id, "summary": SUMMARY, "previous": True} assert items[base_fire]["compare"] is None + # 运行详情的默认基准:同一次、同样是紧挨着的上一次(中间手动的、别的定时任务的两次不在同一定时任务的范围里) + base = (await client.get(f"/api/runs/{target_id}/compare/base")).json() + assert (base["base_run_id"], base["previous"]) == (base_id, True) [listed] = [s for s in (await client.get("/api/schedules")).json()["items"] if s["id"] == sid] assert listed["last_fire"]["id"] == fire_id and listed["last_fire"]["compare"]["base_run_id"] == base_id @@ -321,9 +324,10 @@ async def test_after_a_failed_run_the_base_is_the_last_successful_one(recorder, assert (body["compare"]["base_run_id"], body["compare"]["previous"]) == (first, False) items = {f["id"]: f for f in (await client.get(f"/api/schedules/{sid}/fires")).json()["items"]} assert items[fire_id]["compare"] == {"base_run_id": first, "summary": SUMMARY, "previous": False} - # 运行详情「与上一次对比」的默认基准挑的是同一次(同一个函数):两个同名入口比的是同一对运行、计数一致 + # 运行详情「与上一次对比」的默认基准挑的是同一次(同一个函数):两个同名入口比的是同一对运行、计数一致; + # previous 也一样,运行详情的菜单项和触发历史的链接都写「与上一次成功的运行对比」 base = (await client.get(f"/api/runs/{target_id}/compare/base")).json() - assert (base["base_run_id"], base["reason"]) == (first, "same_schedule") + assert (base["base_run_id"], base["reason"], base["previous"]) == (first, "same_schedule", False) async def fires_of(client, sid: str) -> dict[str, dict]: @@ -396,6 +400,9 @@ async def test_a_stored_compare_stays_while_its_base_is_still_the_pick(recorder, assert (await client.delete(f"/api/runs/{second}", params={"force": "true"})).status_code == 204 assert (await fires_of(client, sid))[third_fire]["compare"] == shown assert await default_base_of(client, third) == (first, "same_schedule") + # 运行详情的默认基准按现在的运行记录算:中间那次删掉了,第 1 周就是紧挨着的上一次(菜单写「与上一次对比」, + # 触发历史照当时记下的写「与上一次成功的运行对比」,两句都属实,比的是同一对运行) + assert (await client.get(f"/api/runs/{third}/compare/base")).json()["previous"] is True async def test_without_an_earlier_success_there_is_no_base(recorder, client):