关于响应处理

1 序列化数据

Flask-RESTful 提供了marshal工具，用来帮助我们将数据序列化为特定格式的字典数据，以便作为视图的返回值。

  1. from flask_restful import Resource, fields, marshal_with
  2. resource_fields = {
  3.     'name': fields.String,
  4.     'address': fields.String,
  5.     'user_id': fields.Integer
  6. }
  7. class Todo(Resource):
  8.     @marshal_with(resource_fields, envelope='resource')
  9.     def get(self, **kwargs):
  10.         return db_get_todo()

也可以不使用装饰器的方式

  1. class Todo(Resource):
  2.     def get(self, **kwargs):
  3.         data = db_get_todo()
  4.         return marshal(data, resource_fields)

示例

  1. # 用来模拟要返回的数据对象的类
  2. class User(object):
  3.     def __init__(self, user_id, name, age):
  4.         self.user_id = user_id
  5.         self.name = name
  6.         self.age = age
  7. resoure_fields = {
  8.         'user_id': fields.Integer,
  9.         'name': fields.String
  10.     }
  11. class Demo1Resource(Resource):
  12.     @marshal_with(resoure_fields, envelope='data1')
  13.     def get(self):
  14.         user = User(1, 'itcast', 12)
  15.         return user
  16. class Demo2Resource(Resource):
  17.     def get(self):
  18.         user = User(1, 'itcast', 12)
  19.         return marshal(user, resoure_fields, envelope='data2')

2 定制返回的JSON格式

需求

想要接口返回的JSON数据具有如下统一的格式

  1. {"message": "描述信息", "data": {要返回的具体数据}}

在接口处理正常的情况下， message返回ok即可，但是若想每个接口正确返回时省略message字段

  1. class DemoResource(Resource):
  2.     def get(self):
  3.         return {'user_id':1, 'name': 'itcast'}

对于诸如此类的接口，能否在某处统一格式化成上述需求格式？

  1. {"message": "OK", "data": {'user_id':1, 'name': 'itcast'}}

解决

Flask-RESTful的Api对象提供了一个representation的装饰器，允许定制返回数据的呈现格式

  1. api = Api(app)
  2. @api.representation('application/json')
  3. def handle_json(data, code, headers):
  4.     # TODO 此处添加自定义处理
  5.     return resp

Flask-RESTful原始对于json的格式处理方式如下：
代码出处：flask_restful.representations.json

  1. from flask import make_response, current_app
  2. from flask_restful.utils import PY3
  3. from json import dumps
  4. def output_json(data, code, headers=None):
  5.     """Makes a Flask response with a JSON encoded body"""
  6.     settings = current_app.config.get('RESTFUL_JSON', {})
  7.     # If we're in debug mode, and the indent is not set, we set it to a
  8.     # reasonable value here.  Note that this won't override any existing value
  9.     # that was set.  We also set the "sort_keys" value.
  10.     if current_app.debug:
  11.         settings.setdefault('indent', 4)
  12.         settings.setdefault('sort_keys', not PY3)
  13.     # always end the json dumps with a new line
  14.     # see https://github.com/mitsuhiko/flask/pull/1262
  15.     dumped = dumps(data, **settings) + "\n"
  16.     resp = make_response(dumped, code)
  17.     resp.headers.extend(headers or {})
  18.     return resp

为满足需求，做如下改动即可

  1. @api.representation('application/json')
  2. def output_json(data, code, headers=None):
  3.     """Makes a Flask response with a JSON encoded body"""
  4.     # 此处为自己添加***************
  5.     if 'message' not in data:
  6.         data = {
  7.             'message': 'OK',
  8.             'data': data
  9.         }
  10.     # **************************
  11.     settings = current_app.config.get('RESTFUL_JSON', {})
  12.     # If we're in debug mode, and the indent is not set, we set it to a
  13.     # reasonable value here.  Note that this won't override any existing value
  14.     # that was set.  We also set the "sort_keys" value.
  15.     if current_app.debug:
  16.         settings.setdefault('indent', 4)
  17.         settings.setdefault('sort_keys', not PY3)
  18.     # always end the json dumps with a new line
  19.     # see https://github.com/mitsuhiko/flask/pull/1262
  20.     dumped = dumps(data, **settings) + "\n"
  21.     resp = make_response(dumped, code)
  22.     resp.headers.extend(headers or {})
  23.     return resp