Flask后端实践 连载六 基于Flask与SQLAlchemy的单表接口
基于Flask与SQLAlchemy的单表接口
tips:
- 本文主要介绍基于Flask与SQLAlchemy的单表接口
- 本文基于python3编写
- 本文适合有一定Flask项目的朋友阅读
- 代码仓库
项目场景
一日，项目经理A找到我，先是表扬最近项目重构的不错，然后，提出一个单表接口想法。经过和他的仔细探讨，单表接口主要是实现一个数据表的增删改查，不会对其他的数据表造成干扰。
技术细节
- 实现http请求分发处理
- 实现请求参数识别和处理
- 实现数据库操作
实现逻辑
Flask MethodView
- Flask提供MethodView对每个 HTTP 方法执行不同的函数，即只需要修改不同函数处理方法。详见官方文档
- 官方例子
-
定义接口方法
RL HTTP 方法 描述 /users/ GET 获得全部用户的列表 /users/ POST 创建一个新用户 /users/< id > GET 显示某个用户 /users/< id > PUT 更新某个用户 /users/< id > DELETE 删除某个用户 -
定义不同方法的处理逻辑
from flask.views import MethodView class UserAPI(MethodView): def get(self, user_id): if user_id is None: # return a list of users pass else: # expose a single user pass def post(self): # create a new user pass def delete(self, user_id): # delete a single user pass def put(self, user_id): # update a single user pass -
注册到路由
user_view = UserAPI.as_view('user_api') app.add_url_rule('/users/', defaults={'user_id': None}, view_func=user_view, methods=['GET',]) app.add_url_rule('/users/', view_func=user_view, methods=['POST',]) app.add_url_rule('/users/
编写字段解析类
-
参数定义
用法 operator_modelFiled ，即 下划线 前部分为操作方法，后部分为model的键。因此所有的键都应与数据库表字段对应。符号 含义 用法 gt 大于 gt_id=1 ge 大于等于 ge_id=1 lt 小于 lt_id=1 le 小于等于 le_id=1 ne 不等于 ne_id=1 eq 等于 eq_id=1 ic 包含 ic_id=1 ni 不包含 ni_id=1 in 查询多个相同字段的值 in_id=1|2 by 排序(0:正序,1:倒序) by_id=1 -
实现代码(base.py)
class BaseParse(object): """ 识别要查询的字段 """ __model__ = None __request__ = request by = frozenset(['by']) query = frozenset(['gt', 'ge', 'lt', 'le', 'ne', 'eq', 'ic', 'ni', 'in']) def __init__(self): self._operator_funcs = { 'gt': self.__gt_model, 'ge': self.__ge_model, 'lt': self.__lt_model, 'le': self.__le_model, 'ne': self.__ne_model, 'eq': self.__eq_model, 'ic': self.__ic_model, 'ni': self.__ni_model, 'by': self.__by_model, 'in': self.__in_model, } def _parse_page_size(self): """ 获取页码和获取每页数据量 :return: page 页码 page_size 每页数据量 """ default_page = current_app.config['DEFAULT_PAGE_INDEX'] default_size = current_app.config['DEFAULT_PAGE_SIZE'] page = self.__request__.args.get("page",default_page) page_size = self.__request__.args.get("size",default_size ) page = int(page) - 1 page_size = int(page_size) return page, page_size def _parse_query_field(self): """ 解析查询字段 :return: query_field 查询字段 by_field 排序字段 """ args = self.__request__.args query_field = list() by_field = list() for query_key, query_value in args.items(): key_split = query_key.split('_', 1) if len(key_split)!= 2: continue operator, key = key_split if not self._check_key(key=key): continue if operator in self.query: data = self._operator_funcs[operator](key=key, value=query_value) query_field.append(data) elif operator in self.by: data = self._operator_funcs[operator](key=key, value=query_value) by_field.append(data) return query_field, by_field def _parse_create_field(self): """ 检查字段是否为model的字段,并过滤无关字段. 1.list(dict) => list(dict) 2. dict => list(dict) :return: """ obj = self.__request__.get_json(force=True) if isinstance(obj, list): create_field = list() for item in obj: if isinstance(item, dict): base_dict = self._parse_field(obj=item) create_field.append(base_dict) return create_field elif isinstance(obj, dict): return [self._parse_field(obj=obj)] else: return list() def _parse_field(self, obj=None): """ 检查字段模型中是否有，并删除主键值 :param obj: :return: """ obj = obj if obj is not None else self.__request__.get_json(force=True) field = dict() # 获取model主键字段 primary_key = map(lambda x: x.name, inspect(self.__model__).primary_key) for key, value in obj.items(): if key in primary_key: continue if self._check_key(key): field[key] = value return field def _check_key(self, key): """ 检查model是否存在key :param key: :return: """ if hasattr(self.__model__, key): return True else: return False def __gt_model(self, key, value): """ 大于 :param key: :param value: :return: """ return getattr(self.__model__, key) > value def __ge_model(self, key, value): """ 大于等于 :param key: :param value: :return: """ return getattr(self.__model__, key) > value def __lt_model(self, key, value): """ 小于 :param key: :param value: :return: """ return getattr(self.__model__, key) < value def __le_model(self, key, value): """ 小于等于 :param key: :param value: :return: """ return getattr(self.__model__, key) <= value def __eq_model(self, key, value): """ 等于 :param key: :param value: :return: """ return getattr(self.__model__, key) == value def __ne_model(self, key, value): """ 不等于 :param key: :param value: :return: """ return getattr(self.__model__, key) != value def __ic_model(self, key, value): """ 包含 :param key: :param value: :return: """ return getattr(self.__model__, key).like('%{}%'.format(value)) def __ni_model(self, key, value): """ 不包含 :param key: :param value: :return: """ return getattr(self.__model__, key).notlike('%{}%'.format(value)) def __by_model(self, key, value): """ :param key: :param value: 0:正序,1:倒序 :return: """ try: value = int(value) except ValueError as e: logger.error(e) return getattr(self.__model__, key).asc() else: if value == 1: return getattr(self.__model__, key).asc() elif value == 0: return getattr(self.__model__, key).desc() else: return getattr(self.__model__, key).asc() def __in_model(self, key, value): """ 查询多个相同字段的值 :param key: :param value: :return: """ value = value.split('|') return getattr(self.__model__, key).in_(value)
编写数据库查询类
-
定义
根据项目实际需求，单表接口主要涉及到数据表的增删改查，主要有新建数据、批量新建数据、删除、更新、主键查询、多条件查询、分页、排序等数据库操作方法。 -
实现代码(base.py)
class BaseQuery(object): """ 查询方法 """ __model__ = None def _find(self, query): """ 根据查询参数获取内容 """ return self.__model__.query.filter(*query).all() def _find_by_page(self, page, size, query, by): """ 根据查询参数，分页，排序获取内容 """ base = self.__model__.query.filter(*query).order_by(*by) cnt = base.count() data = base.slice(page * size, (page + 1) * size).all() return cnt, data def _get(self, key): """ 根据主键ID获取数据 """ return self.__model__.query.get(key) def _create(self, args): """ 创建新的一条数据或批量新建 """ for base in args: model = self.__model__() for k, v in base.items(): setattr(model, k, v) db.session.add(model) try: db.session.commit() return True except Exception as e: logger.error(e) return False def _update(self, key, kwargs): """ 更新数据 """ model = self._get(key) if model: for k, v in kwargs.items(): setattr(model, k, v) try: db.session.add(model) db.session.commit() return True except Exception as e: logger.error(e) return False else: return False def _delete(self, key): """ 删除数据 """ model = self._get(key) if model: try: db.session.delete(model) db.session.commit() return True except Exception as e: logger.error(e) return False else: return False def parse_data(self, data): """ 解析查询的数据 """ if data: if isinstance(data, (list, tuple)): data = list(map(lambda x: {p.key: getattr(x, p.key) for p in self.__model__.__mapper__.iterate_properties }, data)) else: data = {p.key: getattr(data, p.key) for p in self.__model__.__mapper__.iterate_properties} return data
编写Service类
- 定义
前面我们已经知道了Flask实现了不同http请求的分发(MethodView)，因此只需要自定义不同http请求处理逻辑。通过前面以及解决了参数解析或数据库操作。那么将BaseParse, BaseQuery, MethodView组合起来，就可以完成单表接口。 - 实现代码(base.py)
def view_route(f): """ 路由设置,统一返回格式 :param f: :return: """ def decorator(*args, **kwargs): rv = f(*args, **kwargs) if isinstance(rv, (int, float)): res = ResMsg() res.update(data=rv) return jsonify(res.data) elif isinstance(rv, tuple): if len(rv) >= 3: return jsonify(rv[0]), rv[1], rv[2] else: return jsonify(rv[0]), rv[1] elif isinstance(rv, dict): return jsonify(rv) elif isinstance(rv, bytes): rv = rv.decode('utf-8') return jsonify(rv) else: return jsonify(rv) return decorator class Service(BaseParse, BaseQuery, MethodView): __model__ = None # 装饰器控制数据返回格式 decorators = [view_route] def get(self, key=None): """ 获取列表或单条数据 :param key: :return: """ res = ResMsg() if key is not None: data = self.parse_data(self._get(key=key)) if data: res.update(data=data) else: res.update(code=ResponseCode.NO_RESOURCE_FOUND) else: query, by = self._parse_query_field() page, size = self._parse_page_size() cnt, data = self._find_by_page(page=page, size=size, query=query, by=by) data = self.parse_data(data) if data: res.update(data=data) else: res.update(code=ResponseCode.NO_RESOURCE_FOUND) res.add_field(name='total', value=cnt) res.add_field(name='page', value=page + 1) res.add_field(name='size', value=size) res.update(data=data) return res.data def post(self): """ 创建数据 1.单条 2.多条 :return: """ res = ResMsg() data = self._parse_create_field() if data: if not self._create(args=data): res.update(code=ResponseCode.FAIL) else: res.update(code=ResponseCode.INVALID_PARAMETER) return res.data def put(self, key=None): """ 更新某个数据 :return: """ res = ResMsg() if key is None: res.update(code=ResponseCode.INVALID_PARAMETER) else: data = self._parse_field() if not self._update(key=key, kwargs=data): res.update(code=ResponseCode.FAIL) return res.data def delete(self, key=None): """ 删除某个数据 :return: """ res = ResMsg() if key is None: res.update(code=ResponseCode.INVALID_PARAMETER) elif not self._delete(key=key): res.update(code=ResponseCode.FAIL) return res.data
单表接口使用
- 数据库表(model.py)
class Article(db.Model): """ 文章表 """ __tablename__ = 'article' id = db.Column(db.Integer, autoincrement=True, primary_key=True) title = db.Column(db.String(20), nullable=False) # 文章标题 body = db.Column(db.String(255), nullable=False) # 文章内容 - 接口服务(service.py)
from base import Service from models import * class ArticleAPI(Service): """ 文章单表接口 """ __model__ = Article service_name = 'article' - 注册服务(app.py)
from flask import Flask from service import ArticleAPI app = Flask(__name__) article_view = ArticleAPI.as_view('article_api') app.add_url_rule('/article/', defaults={'key': None}, view_func=article_view , methods=['GET',]) app.add_url_rule('/article/', view_func=article_view , methods=['POST',]) app.add_url_rule('/article/ - 测试
- 单条新建 POST http://127.0.0.1:5000/article/
发送数据：{"title":"测试1","body":"测试1"}
返回：{"code":0,"data":null,"lang":"zh_CN","msg":"成功"} - 批量新建 POST http://127.0.0.1:5000/article/
发送数据：[{"title":"测试2","body":"测试2"},{"title":"测试3","body":"测试3"}]
返回：{"code":0,"data":null,"lang":"zh_CN","msg":"成功"} - 查询 GET http://127.0.0.1:5000/article/?eq_title=测试1
返回：{"code":0,"data":[{"body":"测试1","id":1,"title":"测试1"}],"lang":"zh_CN","msg":"成功", "page":1,"size":10,"total":1} - 查询 GET http://127.0.0.1:5000/article/1
返回：{"code":0, "data":{"body":"测试1", "id":1,"title":"测试1" },"lang":"zh_CN","msg":"成功"} - 更新 PUT http://127.0.0.1:5000/article/1
发送数据：{"title":"测试111","body":"测试111"}
返回：{"code":0,"data":null,"lang":"zh_CN","msg":"成功"} - 删除 DELETE http://127.0.0.1:5000/article/1
返回：{"code":0,"data":null,"lang":"zh_CN","msg":"成功"}
- 单条新建 POST http://127.0.0.1:5000/article/
总结
- 文中涉及到了字符串操作解析方面、类的定义、SQLAlchemy使用，Flask MethodView使用
- 通过类继承的方法，可以快速扩展单表接口，大大的提高了开发效率。如果有其他实现需求，也可以通过重写的方法完成定制。
- 下一篇文章将介绍reids在Flask里面的使用