yuetsh
eed0e630e6
阶段 0 盘点时判为 CUT 的 17 条，当时只在代码里标了 DEPRECATED（2026-05-26），
没删。这次删掉：15 个视图类整体移除，2 条只删 URL。
## 为什么有两条只能删 URL
ProblemSetProblemAPI 和 ProblemSetProgressAPI **各有一条在用的路径**：
ProblemSetProblemAPI /problemset/<id>/problems 在用
/problemset/<id>/problems/<pid> 废弃
ProblemSetProgressAPI /problemset/progress 在用
/problemset/<id>/progress 废弃
删类会直接搞挂"看题单题目"和"加入题单"。清单里记的是**路径**不是视图类，
这个区别不看一眼就动手会出事。
## 核实过程
不敢只信清单 —— 今天已经因为"读过文档还是漏了盲点"栽过一次（前端 /api2 前缀
漏改 5 处）。所以：
1. 用 ast（不是正则）建「视图类 → 它的所有 URL」映射，判定哪些类的全部 URL
都废弃。正则版把 B 组算成 0 个，而我明知至少有 2 个 —— 那种"看起来干净"的
结果最危险。
2. DEPRECATED 检测最初往前多看一行，把上一条的行内注释算到了下一条头上，
于是 /hitokoto、/profile、/reset_password 被误判成废弃。这三个明显在用，
才发现范围错了。
3. 按端点字符串精确搜整个 ojnext（覆盖 http.get、原生 fetch、独立 axios、
模板插值），16/16 确认无调用。**并且用 5 个明确在用的端点反测了检测器本身** ——
未经验证的检测器报"全部无调用"是没有意义的。
## 验证
起本地 Django + PostgreSQL 实跑 33 条断言，全过：
- 保留的端点还能用（13 条，含那两个"类保留"的路径，加入题单确认真的成功）
- 删掉的端点确实 404（17 条）
- reverse() 也解析不到了（3 条）
`ruff check .` 全绿（顺手清掉 44 个因此变成未使用的 import）。
## 一个差点造成的意外
脚本无条件重写所有 urls.py，把 class_pk/ 和 message/ 下三个 **CRLF** 文件
静默转成了 LF —— 内容没变但整文件显示为改动。已原样恢复，最终 diff 是
纯删除，零意外改动。
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
179 lines
7.3 KiB
Python
179 lines
7.3 KiB
Python
from datetime import datetime, timedelta
|
||
from ipaddress import ip_network
|
||
|
||
from django.utils.timezone import now
|
||
|
||
from account.decorators import ensure_created_by, teacher_admin_required
|
||
from problem.models import Problem
|
||
from utils.api import APIView, validate_serializer
|
||
|
||
from ..models import ACMContestRank, Contest
|
||
from ..serializers import (
|
||
ACMContesHelperSerializer,
|
||
ContestAdminSerializer,
|
||
ContestCloneSerializer,
|
||
CreateConetestSeriaizer,
|
||
EditConetestSeriaizer,
|
||
)
|
||
|
||
|
||
class ContestAPI(APIView):
|
||
@validate_serializer(CreateConetestSeriaizer)
|
||
@teacher_admin_required
|
||
def post(self, request):
|
||
data = request.data
|
||
# DRF 的 DateTimeField 已经校验并规范化过，这里拿到的必定是 ISO 8601 字符串
|
||
data["start_time"] = datetime.fromisoformat(data["start_time"])
|
||
data["end_time"] = datetime.fromisoformat(data["end_time"])
|
||
data["created_by"] = request.user
|
||
if data["end_time"] <= data["start_time"]:
|
||
return self.error("Start time must occur earlier than end time")
|
||
if data.get("password") and data["password"] == "":
|
||
data["password"] = None
|
||
for ip_range in data["allowed_ip_ranges"]:
|
||
try:
|
||
ip_network(ip_range, strict=False)
|
||
except ValueError:
|
||
return self.error(f"{ip_range} is not a valid cidr network")
|
||
contest = Contest.objects.create(**data)
|
||
return self.success(ContestAdminSerializer(contest).data)
|
||
|
||
@validate_serializer(EditConetestSeriaizer)
|
||
@teacher_admin_required
|
||
def put(self, request):
|
||
data = request.data
|
||
try:
|
||
contest = Contest.objects.get(id=data.pop("id"))
|
||
except Contest.DoesNotExist:
|
||
return self.error("Contest does not exist")
|
||
ensure_created_by(contest, request.user)
|
||
# DRF 的 DateTimeField 已经校验并规范化过，这里拿到的必定是 ISO 8601 字符串
|
||
data["start_time"] = datetime.fromisoformat(data["start_time"])
|
||
data["end_time"] = datetime.fromisoformat(data["end_time"])
|
||
if data["end_time"] <= data["start_time"]:
|
||
return self.error("Start time must occur earlier than end time")
|
||
if not data["password"]:
|
||
data["password"] = None
|
||
for ip_range in data["allowed_ip_ranges"]:
|
||
try:
|
||
ip_network(ip_range, strict=False)
|
||
except ValueError:
|
||
return self.error(f"{ip_range} is not a valid cidr network")
|
||
for k, v in data.items():
|
||
setattr(contest, k, v)
|
||
contest.save()
|
||
return self.success(ContestAdminSerializer(contest).data)
|
||
|
||
@teacher_admin_required
|
||
def get(self, request):
|
||
contest_id = request.GET.get("id")
|
||
if contest_id:
|
||
try:
|
||
contest = Contest.objects.get(id=contest_id)
|
||
ensure_created_by(contest, request.user)
|
||
return self.success(ContestAdminSerializer(contest).data)
|
||
except Contest.DoesNotExist:
|
||
return self.error("Contest does not exist")
|
||
|
||
contests = Contest.objects.all().order_by("-create_time")
|
||
if not request.user.is_super_admin():
|
||
contests = contests.filter(created_by=request.user)
|
||
|
||
keyword = request.GET.get("keyword")
|
||
if keyword:
|
||
contests = contests.filter(title__contains=keyword)
|
||
return self.success(self.paginate_data(request, contests, ContestAdminSerializer))
|
||
|
||
|
||
class ACMContestHelper(APIView):
|
||
@teacher_admin_required
|
||
def get(self, request):
|
||
contest_id = request.GET.get("contest_id")
|
||
if not contest_id:
|
||
return self.error("Parameter error, contest_id is required")
|
||
try:
|
||
contest = Contest.objects.get(id=contest_id, visible=True)
|
||
except Contest.DoesNotExist:
|
||
return self.error("Contest does not exist")
|
||
ensure_created_by(contest, request.user)
|
||
|
||
problems = Problem.objects.filter(contest=contest).values("id", "_id")
|
||
problem_id_map = {str(p["id"]): p["_id"] for p in problems}
|
||
|
||
ranks = ACMContestRank.objects.filter(contest=contest, accepted_number__gt=0).values("id", "user__username", "user__userprofile__real_name", "submission_info")
|
||
results = []
|
||
for rank in ranks:
|
||
for problem_id, info in rank["submission_info"].items():
|
||
if info["is_ac"]:
|
||
results.append(
|
||
{
|
||
"id": rank["id"],
|
||
"username": rank["user__username"],
|
||
"real_name": rank["user__userprofile__real_name"],
|
||
"problem_id": problem_id,
|
||
"problem_display_id": problem_id_map.get(problem_id, problem_id),
|
||
"ac_info": info,
|
||
"checked": info.get("checked", False),
|
||
}
|
||
)
|
||
results.sort(key=lambda x: -x["ac_info"]["ac_time"])
|
||
return self.success(results)
|
||
|
||
@teacher_admin_required
|
||
@validate_serializer(ACMContesHelperSerializer)
|
||
def put(self, request):
|
||
data = request.data
|
||
# 原来只按 pk 取，contest_id 明明在序列化器里、客户端一直在传，却完全没用上。
|
||
# 于是任何老师都能改别人比赛里的检查标记 —— teacher_admin_required 只保证
|
||
# "是老师"。同类的 get 是有 ensure_created_by 的，这里补齐。
|
||
try:
|
||
rank = ACMContestRank.objects.get(pk=data["rank_id"], contest_id=data["contest_id"])
|
||
except ACMContestRank.DoesNotExist:
|
||
return self.error("Rank id does not exist")
|
||
ensure_created_by(rank.contest, request.user)
|
||
problem_rank_status = rank.submission_info.get(data["problem_id"])
|
||
if not problem_rank_status:
|
||
return self.error("Problem id does not exist")
|
||
problem_rank_status["checked"] = data["checked"]
|
||
rank.save(update_fields=("submission_info",))
|
||
return self.success()
|
||
|
||
|
||
class ContestCloneAPI(APIView):
|
||
@validate_serializer(ContestCloneSerializer)
|
||
@teacher_admin_required
|
||
def post(self, request):
|
||
try:
|
||
original = Contest.objects.get(id=request.data["contest_id"])
|
||
except Contest.DoesNotExist:
|
||
return self.error("Contest does not exist")
|
||
|
||
duration = original.end_time - original.start_time
|
||
new_start = now() + timedelta(minutes=10)
|
||
new_end = new_start + duration
|
||
|
||
new_contest = Contest.objects.create(
|
||
title=original.title,
|
||
description=original.description,
|
||
tag=original.tag,
|
||
password=original.password,
|
||
visible=False,
|
||
allowed_ip_ranges=original.allowed_ip_ranges,
|
||
start_time=new_start,
|
||
end_time=new_end,
|
||
created_by=request.user,
|
||
)
|
||
|
||
for problem in Problem.objects.filter(contest=original):
|
||
tags = problem.tags.all()
|
||
problem.pk = None
|
||
problem.contest = new_contest
|
||
problem.submission_number = 0
|
||
problem.accepted_number = 0
|
||
problem.statistic_info = {}
|
||
problem.created_by = request.user
|
||
problem.save()
|
||
problem.tags.set(tags)
|
||
|
||
return self.success(ContestAdminSerializer(new_contest).data)
|