截断JD
This commit is contained in:
+66
-8
@@ -1,11 +1,14 @@
|
||||
"""
|
||||
投递记录路由:上报投递结果、查询历史、统计
|
||||
"""
|
||||
import re
|
||||
from datetime import datetime, timedelta
|
||||
from typing import Optional
|
||||
from pathlib import Path
|
||||
from typing import Optional, Tuple
|
||||
from fastapi import APIRouter, Depends, HTTPException, Query
|
||||
from sqlalchemy.orm import Session
|
||||
from sqlalchemy import func
|
||||
from loguru import logger
|
||||
|
||||
from ..auth import get_current_user
|
||||
from ..database import get_db
|
||||
@@ -15,10 +18,33 @@ from ..schemas import ApplicationIn, ApplicationOut, Message, StatsOut
|
||||
|
||||
router = APIRouter(prefix="/api/jobs", tags=["投递记录"])
|
||||
|
||||
# jd_text 归档上限(字符):控制写入包体积,避免触发 MySQL max_allowed_packet(1153);
|
||||
# 正常 JD 全文远低于此值,超限基本意味着抓取异常(如抓到整页文本)
|
||||
# jd_text 入库上限(字符):超过则落盘为文件(库内只存路径),
|
||||
# 控制写入包体积,避免触发 MySQL max_allowed_packet(1153)
|
||||
JD_TEXT_LIMIT = 20000
|
||||
|
||||
# 归档目录:项目根目录下的 jd_archive/(本文件位于 server/routers/,上溯两级即项目根)
|
||||
PROJECT_ROOT = Path(__file__).resolve().parents[2]
|
||||
JD_DIR_NAME = "jd_archive"
|
||||
|
||||
|
||||
def _archive_jd(user_id: int, job_id: str, jd_text: Optional[str]) -> Tuple[Optional[str], Optional[str]]:
|
||||
"""JD 存储分流:≤ 上限直接入库;超限写入 jd_archive/ 文件。
|
||||
返回 (jd_text, jd_file):两者必有一空。文件名按 用户_职位 去重,重复上报覆盖"""
|
||||
if not jd_text or len(jd_text) <= JD_TEXT_LIMIT:
|
||||
return jd_text, None
|
||||
safe_job = re.sub(r"[^\w\-]", "_", job_id or "")[:48] or "unknown"
|
||||
rel = f"{JD_DIR_NAME}/{user_id}_{safe_job}.txt"
|
||||
try:
|
||||
target = PROJECT_ROOT / rel
|
||||
target.parent.mkdir(parents=True, exist_ok=True)
|
||||
target.write_text(jd_text, encoding="utf-8")
|
||||
logger.info(f"超长 JD 已归档到文件({len(jd_text)} 字符):{rel}")
|
||||
return None, rel
|
||||
except OSError as e:
|
||||
# 落盘失败也不能把巨包塞进数据库:截断到上限入库,保底不丢全部信息且不炸库
|
||||
logger.error(f"JD 归档落盘失败,截断入库:{e}")
|
||||
return jd_text[:JD_TEXT_LIMIT], None
|
||||
|
||||
|
||||
def _clip(value: Optional[str], limit: int) -> Optional[str]:
|
||||
"""按列长度截断文本,防止超长写入导致 1153 / Data too long"""
|
||||
@@ -26,7 +52,8 @@ def _clip(value: Optional[str], limit: int) -> Optional[str]:
|
||||
|
||||
|
||||
def _sanitize(req: ApplicationIn) -> ApplicationIn:
|
||||
"""上报字段统一裁剪(与 models.py 列长度保持一致)"""
|
||||
"""上报字段统一裁剪(与 models.py 列长度保持一致);
|
||||
jd_text 不在此截断——由 _archive_jd 按上限分流到数据库/文件"""
|
||||
req.job_title = _clip(req.job_title, 128)
|
||||
req.company = _clip(req.company, 128)
|
||||
req.salary = _clip(req.salary, 32)
|
||||
@@ -35,7 +62,6 @@ def _sanitize(req: ApplicationIn) -> ApplicationIn:
|
||||
req.ai_reason = _clip(req.ai_reason, 255)
|
||||
req.skip_reason = _clip(req.skip_reason, 255)
|
||||
req.greeting = _clip(req.greeting, 500)
|
||||
req.jd_text = _clip(req.jd_text, JD_TEXT_LIMIT)
|
||||
return req
|
||||
|
||||
|
||||
@@ -51,6 +77,8 @@ def report_application(
|
||||
同一用户 + 同一职位 ID 去重,已存在则更新。
|
||||
"""
|
||||
req = _sanitize(req)
|
||||
# JD 分流:短文本入库,超长落盘为文件(库存路径)
|
||||
jd_text, jd_file = _archive_jd(current_user.id, req.job_id, req.jd_text)
|
||||
existing = db.query(JobApplication).filter(
|
||||
JobApplication.user_id == current_user.id,
|
||||
JobApplication.job_id == req.job_id,
|
||||
@@ -65,7 +93,9 @@ def report_application(
|
||||
# 否则记录停留在首次上报时间,看起来像“漏报”
|
||||
existing.applied_at = datetime.utcnow()
|
||||
if req.jd_text:
|
||||
existing.jd_text = req.jd_text
|
||||
# 新 JD 覆盖旧值;文本与归档文件二者只留其一,避免读到陈旧数据
|
||||
existing.jd_text = jd_text
|
||||
existing.jd_file = jd_file
|
||||
if req.greeting:
|
||||
existing.greeting = req.greeting
|
||||
if session_id:
|
||||
@@ -84,7 +114,8 @@ def report_application(
|
||||
ai_score=req.ai_score,
|
||||
ai_reason=req.ai_reason,
|
||||
skip_reason=req.skip_reason,
|
||||
jd_text=req.jd_text,
|
||||
jd_text=jd_text,
|
||||
jd_file=jd_file,
|
||||
greeting=req.greeting,
|
||||
)
|
||||
db.add(record)
|
||||
@@ -208,6 +239,7 @@ def page_applications(
|
||||
rows = query.order_by(JobApplication.applied_at.desc()) \
|
||||
.offset((page - 1) * page_size).limit(page_size).all()
|
||||
items = [{
|
||||
"id": r.id,
|
||||
"job_title": r.job_title,
|
||||
"company": r.company,
|
||||
"salary": r.salary,
|
||||
@@ -217,8 +249,10 @@ def page_applications(
|
||||
"ai_reason": r.ai_reason,
|
||||
"skip_reason": r.skip_reason,
|
||||
"greeting": r.greeting,
|
||||
# JD 全文截断,供客户端悬停摘要展示,控制分页响应体积
|
||||
# JD 全文截断,供客户端悬停摘要展示,控制分页响应体积;
|
||||
# 完整内容(含归档文件)通过 /{app_id}/jd 接口按需获取,has_jd 标记有无 JD 可看
|
||||
"jd_text": (r.jd_text or "")[:500],
|
||||
"has_jd": bool(r.jd_text or r.jd_file),
|
||||
"applied_at": (r.applied_at + timedelta(hours=8)).strftime("%Y-%m-%d %H:%M:%S"),
|
||||
} for r in rows]
|
||||
return {
|
||||
@@ -256,3 +290,27 @@ def get_stats(
|
||||
total_skipped=count(apply_type="skipped"),
|
||||
today_applied=today_applied,
|
||||
)
|
||||
|
||||
|
||||
@router.get("/{app_id}/jd", response_model=dict)
|
||||
def get_application_jd(
|
||||
app_id: int,
|
||||
current_user: User = Depends(get_current_user),
|
||||
db: Session = Depends(get_db),
|
||||
):
|
||||
"""查看单条投递记录的完整 JD:先取 jd_text 字段,为空则读 jd_file 归档文件"""
|
||||
record = db.query(JobApplication).filter(
|
||||
JobApplication.id == app_id,
|
||||
JobApplication.user_id == current_user.id,
|
||||
).first()
|
||||
if not record:
|
||||
raise HTTPException(status_code=404, detail="投递记录不存在")
|
||||
if record.jd_text:
|
||||
return {"id": app_id, "source": "db", "jd_text": record.jd_text}
|
||||
if record.jd_file:
|
||||
target = (PROJECT_ROOT / record.jd_file).resolve()
|
||||
# 防路径穿越:归档文件必须位于项目根目录内,防止库内路径被篡改后读到任意文件
|
||||
if target.is_file() and str(target).startswith(str(PROJECT_ROOT)):
|
||||
return {"id": app_id, "source": "file", "jd_text": target.read_text(encoding="utf-8")}
|
||||
raise HTTPException(status_code=410, detail="JD 归档文件不存在或已丢失")
|
||||
raise HTTPException(status_code=404, detail="该记录未包含 JD")
|
||||
|
||||
Reference in New Issue
Block a user