sqlalchemy查询json

小石潭记丶2025-04-08 8:02

第一种：字段op是json格式：

复制代码

{
        "uid": "cxb123456789",
        "role": 2,
        "op_start_time": 1743513707504,
        "op_end_time": 1743513707504,
        "op_start_id": "op_001",
        "op_end_id": "op_002",
        "op": {
          "ability_node": [{"id": "2c6adc22-e05a-4fe9-835f-ba5d7f8b303", "name": "连接数限制"}],
          "skill_node": [{"id": "cvlp6o324te32pj24teg", "name": "yyyy"}],
          "knowledge_node": [{"id": "cvlp6o324te32pj24teg", "name": "端到端加密"}],
          "op_id": "9131c441-4304-4101-b064-362352d26789",
          "op_name": "Initial Data Import",
          "pre_op": "",
          "next_op": "op_002",
          "dep_relationship": "",
          "op_result": 1,
          "op_score": 95.5,
          "op_count": 1
        }
      }

要求：查询op的json字段里面的op_id为传入的op_id

复制代码

from sqlalchemy.future import select

db_query = select(KnowledgeAiReport)
op_id = report.op.op_id
db_query = db_query.filter(KnowledgeAiReport.uid == report.uid).filter(KnowledgeAiReport.op.op('->>')('op_id') == op_id)
item = await db_session.execute(db_query)
report_db = item.scalars().first()

第二种：person_type也是json字段，

要求：存储的值是 ["1","2"]，要求传入 "1"能查出对应的数据

实现代码：

复制代码

from sqlalchemy.dialects.postgresql import JSONB  # 使用JSONB，因为它支持全文搜索
from sqlalchemy import cast
from sqlalchemy.future import select


db_query = select(TbResident)
db_query = db_query.filter(cast(TbResident.illness_type, JSONB).op('@>')([f'{illness_type}']))
db_query = db_query.offset(page_no * page_size).limit(page_size).order_by(TbResident.created_at.desc())
    item = await db_session.execute(db_query)
    residents = item.scalars().all()