Flask-RESTful起步
Flask-RESTful是用于快速构建REST API的Flask扩展。
1 安装
pip install flask-restful
2 Hello World
from flask import Flask
from flask_restful import Resource, Apiapp = Flask(__name__)
api = Api(app)class HelloWorldResource(Resource):def get(self):return {'hello': 'world'}def post(self):return {'msg': 'post hello world'}api.add_resource(HelloWorldResource, '/')# 此处启动对于1.0之后的Flask可有可无
if __name__ == '__main__':app.run(debug=True)
关于视图
1 为路由起名
通过endpoint参数为路由起名
api.add_resource(HelloWorldResource, '/', endpoint='HelloWorld')
2 蓝图中使用
from flask import Flask, Blueprint
from flask_restful import Api, Resourceapp = Flask(__name__)user_bp = Blueprint('user', __name__)user_api = Api(user_bp)class UserProfileResource(Resource):def get(self):return {'msg': 'get user profile'}user_api.add_resource(UserProfileResource, '/users/profile')app.register_blueprint(user_bp)
3 装饰器
使用method_decorators
添加装饰器
- 为类视图中的所有方法添加装饰器
def decorator1(func):def wrapper(*args, **kwargs):print('decorator1')return func(*args, **kwargs)return wrapperdef decorator2(func):def wrapper(*args, **kwargs):print('decorator2')return func(*args, **kwargs)return wrapperclass DemoResource(Resource):method_decorators = [decorator1, decorator2]def get(self):return {'msg': 'get view'}def post(self):return {'msg': 'post view'}
- 为类视图中不同的方法添加不同的装饰器
class DemoResource(Resource):method_decorators = {'get': [decorator1, decorator2],'post': [decorator1]}# 使用了decorator1 decorator2两个装饰器def get(self):return {'msg': 'get view'}# 使用了decorator1 装饰器def post(self):return {'msg': 'post view'}# 未使用装饰器def put(self):return {'msg': 'put view'}
关于请求处理
Flask-RESTful 提供了RequestParser
类,用来帮助我们检验和转换请求数据。
from flask_restful import reqparseparser = reqparse.RequestParser()
parser.add_argument('rate', type=int, help='Rate cannot be converted', location='args')
parser.add_argument('name')
args = parser.parse_args()
使用步骤:
-
创建
RequestParser
对象 -
向
RequestParser
对象中添加需要检验或转换的参数声明 -
使用
parse_args()
方法启动检验处理 -
检验之后从检验结果中获取参数时可按照字典操作或对象属性操作
args.rate 或 args['rate']
参数说明
1 required
描述请求是否一定要携带对应参数,默认值为False
-
True 强制要求携带
若未携带,则校验失败,向客户端返回错误信息,状态码400
-
False 不强制要求携带
若不强制携带,在客户端请求未携带参数时,取出值为None
class DemoResource(Resource):def get(self):rp = RequestParser()rp.add_argument('a', required=False)args = rp.parse_args()return {'msg': 'data={}'.format(args.a)}
2 help
参数检验错误时返回的错误描述信息
rp.add_argument('a', required=True, help='missing a param')
3 action
描述对于请求参数中出现多个同名参数时的处理方式
action='store'
保留出现的第一个, 默认action='append'
以列表追加保存所有同名参数的值
rp.add_argument('a', required=True, help='missing a param', action='append')
4 type
描述参数应该匹配的类型,可以使用python的标准数据类型string、int,也可使用Flask-RESTful提供的检验方法,还可以自己定义
-
标准类型
rp.add_argument('a', type=int, required=True, help='missing a param', action='append')
-
Flask-RESTful提供
检验类型方法在
flask_restful.inputs
模块中-
url
-
regex(指定正则表达式)
from flask_restful import inputs rp.add_argument('a', type=inputs.regex(r'^\d{2}&'))
-
natural
自然数0、1、2、3... -
positive
正整数 1、2、3... -
int_range(low ,high)
整数范围rp.add_argument('a', type=inputs.int_range(1, 10))
-
boolean
-
-
自定义
def mobile(mobile_str):"""检验手机号格式:param mobile_str: str 被检验字符串:return: mobile_str"""if re.match(r'^1[3-9]\d{9}$', mobile_str):return mobile_strelse:raise ValueError('{} is not a valid mobile'.format(mobile_str))rp.add_argument('a', type=mobile)
5 location
描述参数应该在请求数据中出现的位置
# Look only in the POST body
parser.add_argument('name', type=int, location='form')# Look only in the querystring
parser.add_argument('PageSize', type=int, location='args')# From the request headers
parser.add_argument('User-Agent', location='headers')# From http cookies
parser.add_argument('session_id', location='cookies')# From json
parser.add_argument('user_id', location='json')# From file uploads
parser.add_argument('picture', location='files')
也可指明多个位置
parser.add_argument('text', location=['headers', 'json'])
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。
from flask_restful import Resource, fields, marshal_withresource_fields = {'name': fields.String,'address': fields.String,'user_id': fields.Integer
}class Todo(Resource):@marshal_with(resource_fields, envelope='resource')def get(self, **kwargs):return db_get_todo()
也可以不使用装饰器的方式
class Todo(Resource):def get(self, **kwargs):data = db_get_todo()return marshal(data, resource_fields)
示例
# 用来模拟要返回的数据对象的类
class User(object):def __init__(self, user_id, name, age):self.user_id = user_idself.name = nameself.age = ageresoure_fields = {'user_id': fields.Integer,'name': fields.String}class Demo1Resource(Resource):@marshal_with(resoure_fields, envelope='data1')def get(self):user = User(1, 'itcast', 12)return userclass Demo2Resource(Resource):def get(self):user = User(1, 'itcast', 12)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
{"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
class DemoResource(Resource):def get(self):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?
{"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
api = Api(app)@api.representation('application/json')
def handle_json(data, code, headers):# TODO 此处添加自定义处理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
from flask import make_response, current_app
from flask_restful.utils import PY3
from json import dumpsdef output_json(data, code, headers=None):"""Makes a Flask response with a JSON encoded body"""settings = current_app.config.get('RESTFUL_JSON', {})# If we're in debug mode, and the indent is not set, we set it to a# reasonable value here. Note that this won't override any existing value# that was set. We also set the "sort_keys" value.if current_app.debug:settings.setdefault('indent', 4)settings.setdefault('sort_keys', not PY3)# always end the json dumps with a new line# see https://github.com/mitsuhiko/flask/pull/1262dumped = dumps(data, **settings) + "\n"resp = make_response(dumped, code)resp.headers.extend(headers or {})return resp
为满足需求,做如下改动即可
@api.representation('application/json')
def output_json(data, code, headers=None):"""Makes a Flask response with a JSON encoded body"""# 此处为自己添加***************if 'message' not in data:data = {'message': 'OK','data': data}# **************************settings = current_app.config.get('RESTFUL_JSON', {})# If we're in debug mode, and the indent is not set, we set it to a# reasonable value here. Note that this won't override any existing value# that was set. We also set the "sort_keys" value.if current_app.debug:settings.setdefault('indent', 4)settings.setdefault('sort_keys', not PY3)# always end the json dumps with a new line# see https://github.com/mitsuhiko/flask/pull/1262dumped = dumps(data, **settings) + "\n"resp = make_response(dumped, code)resp.headers.extend(headers or {})return resp