1/dao/wl_statistics_dao.py

#统计分析 数据访问层（动态查询 + 聚合）
from os.path import join
from typing import Optional

from sqlalchemy import and_, case, func
from sqlalchemy.orm import Session

from model.wl_class_model import Class_
from model.wl_emp_model import Emp
from model.wl_score_model import Score
from model.wl_student_model import Student
from scheme.wl_statistics_scheme import AgeCompareOp, SortOrder

# stu_gender 字段的实际取值，确认数据库里存的到底是什么后改这两行即可
MALE_VALUE = "男"
FEMALE_VALUE = "女"




# ---- 2.6.1.1 动态年龄范围查询 ----
# 比较条件 -> SQLAlchemy 表达式工厂，动态拼接 where 条件
_AGE_OP_FACTORY = {
    AgeCompareOp.gt: lambda v1, v2: Student.stu_age > v1,
    AgeCompareOp.ge: lambda v1, v2: Student.stu_age >= v1,
    AgeCompareOp.lt: lambda v1, v2: Student.stu_age < v1,
    AgeCompareOp.le: lambda v1, v2: Student.stu_age <= v1,
    AgeCompareOp.eq: lambda v1, v2: Student.stu_age == v1,
    AgeCompareOp.between: lambda v1, v2: Student.stu_age.between(v1, v2),
}

class StatisticDao:

    @staticmethod
    def search_students_by_age(db: Session, op: AgeCompareOp,
                               value: int, value2: Optional[int] = None):
        if op == AgeCompareOp.between:
            if value2 is None:
                raise ValueError("区间查询（between）需要同时提供上界 value2")
            if value > value2:
                raise ValueError("区间查询的下界不能大于上界")

        condition = _AGE_OP_FACTORY[op](value, value2)
        return db.query(Student) \
                 .filter(Student.is_deleted == 0, condition) \
                 .order_by(Student.stu_age) \
                 .all()
    @staticmethod
    def statistic_classes_count(db: Session):
        """每个班级的总人数 + 按性别细分的分布（含 0 人的空班级）"""
        result=[]
        for c in db.query(Class_).filter(Class_.is_deleted==0).all():
            students=[s for s in c.students if s.is_deleted == 0]
            result.append({
                "class_id":c.class_id,
                "total":len(students),
                "male_count":sum(1 for s in students if s.stu_gender==MALE_VALUE),
                "female_count":sum(1 for s in students if s.stu_gender==FEMALE_VALUE)
            })
        return result
        # 另一个版本的写法，效率更高
        # male = func.sum(case((Student.stu_gender == MALE_VALUE, 1), else_=0))
        # female = func.sum(case((Student.stu_gender == FEMALE_VALUE, 1), else_=0))
        #
        # return db.query(
        #     WlClass.class_id.label("class_id"),
        #     func.count(Student.stu_id).label("total"),
        #     male.label("male_count"),
        #     female.label("female_count"),
        # ).select_from(WlClass) \
        #  .outerjoin(
        #      Student,
        #      # 逻辑删除的过滤条件必须写在 ON 里，写进 where 会把空班级整行滤掉
        #      and_(Student.class_id == WlClass.class_id, Student.is_deleted == 0),
        #  ) \
        #  .filter(WlClass.is_deleted == 0) \
        #  .group_by(WlClass.class_id) \
        #  .order_by(WlClass.class_id) \
        #  .all()

    @staticmethod
    def query_score_by_line(db:Session,score_line:float):
        result=[]
        for s in db.query(Student).filter(Student.is_deleted==0).all():
            if all(sc.score>score_line for sc in s.scores):
                result.append({
                    "stu_no":s.stu_no,
                    "stu_name":s.stu_name,
                    "scores": [{"exam_order":i.exam_order,"score":i.score} for i in s.scores]
                })
        return result

    @staticmethod
    def query_by_fail_count(db:Session,fail_count:int):
        result=[]
        students=db.query(Student).filter(Student.is_deleted==0).all()
        for stu in students:
            if len([sc for sc in stu.scores if sc.score<60])>fail_count:
                result.append({
                    "stu_name":stu.stu_name,
                    "class_id":stu.class_id,
                    "scores":[{"exam_order":i.exam_order,"score":i.score} for i in stu.scores]
                })
        return result

    @staticmethod
    def get_class_avg_score(db: Session, sort_order: SortOrder | None = None):
        """每次考试每个班级的平均分，支持按平均分动态升/降序"""
        avg_score = func.avg(Score.score).label("avg_score")

        q = db.query(
            Student.class_id.label("class_id"),
            Score.exam_order.label("exam_order"),
            avg_score,
        ).select_from(Score) \
         .join(Score.student) \
         .filter(Student.is_deleted == 0, Student.class_id.isnot(None)) \
         .group_by(Student.class_id, Score.exam_order)

        # 动态排序：传了方向就按平均分排，没传则按班级 + 考试序次排
        if sort_order == SortOrder.desc:
            q = q.order_by(avg_score.desc())
        elif sort_order == SortOrder.asc:
            q = q.order_by(avg_score.asc())
        else:
            q = q.order_by(Student.class_id, Score.exam_order)

        return q.all()

    @staticmethod
    def get_info_by_rank(db:Session,rank:int):
        q=(db.query(
            Student.stu_name,
            Student.class_id,
            Emp.offer_time,
            Emp.company_name
        ).select_from(Student) \
         .join(Student.emp)) \
         .filter(Student.is_deleted==0, Emp.is_deleted==0) \
         .order_by(Emp.salary.desc())\
        .limit(rank)

        return q.all()
    @staticmethod
    def get_emp_time(db:Session):
        """每个学生的就业时长（天）= offer下发时间 - 就业开放时间
        """
        emp_total_time = func.coalesce(
            # 未就业学生左外连接后 Emp 列为 NULL，DATEDIFF 也返回 NULL，统一兜成 0
            func.datediff(Emp.offer_time, Emp.emp_open_time),
            0,
        ).label('emp_total_time')

        # 逻辑删除条件写在 ON 里，写进 where 会变成 INNER JOIN，未就业学生会被滤掉
        q=db.query(
            Student.stu_name,
            emp_total_time,
        ).select_from(Student) \
        .outerjoin(Emp, and_(Student.stu_no == Emp.stu_no, Emp.is_deleted == 0)) \
        .filter(Student.is_deleted==0)

        return q.all()

    @staticmethod
    def get_class_avg_emp_time(db:Session, sort_order: SortOrder | None = None):
        """每个班级的平均就业时长（天）= offer下发时间 - 就业开放时间

        口径：
        - 以班级表为主表左外连接，保证没有任何就业学生的班级也出现在结果里
        - 只统计已进入就业阶段（emp_open_time 非空）且已拿到 offer（offer_time 非空）的学生
        - 该班可统计的人数为 0 时 AVG 返回 NULL，出参里 avg_emp_time 给 None，即"无就业学生"
        """
        emp_time = func.datediff(Emp.offer_time, Emp.emp_open_time)
        # AVG 和 COUNT(表达式) 都会自动忽略 NULL，所以没 offer 时间的学生不参与分子也不占分母
        avg_emp_time = func.avg(emp_time).label("avg_emp_time")
        # 用 COUNT(表达式) 而不是 COUNT(*)，它正好是平均值的分母
        emp_stu_count = func.count(emp_time).label("emp_stu_count")

        q = db.query(
            Class_.class_id.label("class_id"),
            avg_emp_time,
            emp_stu_count,
        ).select_from(Class_) \
         .outerjoin(
             Student,
             # 逻辑删除的过滤条件必须写在 ON 里，写进 where 会把空班级整行滤掉
             and_(Student.class_id == Class_.class_id, Student.is_deleted == 0),
         ) \
         .outerjoin(
             Emp,
             and_(
                 Emp.stu_no == Student.stu_no,
                 Emp.is_deleted == 0,
                 Emp.emp_open_time.isnot(None),
                 Emp.offer_time.isnot(None),
             ),
         ) \
         .group_by(Class_.class_id)

        # 动态排序逻辑与 get_class_avg_score 保持一致
        if sort_order == SortOrder.desc:
            # MySQL 里 DESC 时 NULL 排最后，"无就业学生"的班级自然落到末尾
            q = q.order_by(avg_emp_time.desc())
        elif sort_order == SortOrder.asc:
            q = q.order_by(avg_emp_time.asc())
        else:
            q = q.order_by(Class_.class_id)

        return [
            {
                "class_id": r.class_id,
                # AVG 出来是 Decimal，转 float 并按天保留 1 位小数；None 表示该班无就业学生
                "avg_emp_time": round(float(r.avg_emp_time), 1)
                                if r.avg_emp_time is not None else None,
                "emp_stu_count": r.emp_stu_count,
            }
            for r in q.all()
        ]