使用原生 python 造轮子搭建博客

这篇用来 记录一个 从零开始的 博客搭建，希望坚持下去，因为python 开发效率令人发指，所以会原生从零写 ORM ,Web 框架

前提是打好 异步 io 的基础， 使用异步，有一点要谨记，一旦开始异步，层层异步，从 http 到 数据库层都要用异步框架写异步函数，所谓开弓没有回头箭

# -*- coding: utf-8 -*-
import asyncio
import time
from functools import wraps
__author__ = 'Frank Li'

def time_count(func):
    @wraps(func)
    def inner_func(*args,**kw):
        start = time.time()
        result = func(*args,**kw)
        end = time.time()
        print('{} cost {:.1f} s...'.format(func.__name__,end-start))
        return result
    return inner_func

async def do_some_work(n):
    await asyncio.sleep(n)

@asyncio.coroutine
def task_io_01():
    print('{} start to run ...'.format(task_io_01.__name__))
    n = 3
    yield from do_some_work(n)
    print('{} continue to work {} seconds later...'.format(task_io_01.__name__,n))
    return task_io_01.__name__

@asyncio.coroutine
def task_io_02():
    print('{} start to run ...'.format(task_io_02.__name__))
    n = 5
    yield from do_some_work(n)
    print('{} continue to do the work in {} seconds'.format(task_io_02.__name__,n))
    return task_io_02.__name__

@time_count
def main():
    tasks = [task_io_02(),task_io_01()]
    loop = asyncio.get_event_loop()
    done, pending = loop.run_until_complete(asyncio.wait(tasks))
    for d in done:
        print('协程无序返回值：{}'.format(d.result()))
    loop.close()

if __name__ == '__main__':
    main()

环境准备 flask 都不要，惊掉下巴

python 3.7

pip install aiohttp

pip install jinja2

pip install aiomysql

pycharm( python 开发 IDE)  + git bash(练习命令) + github(远程仓库)

---

在 pycharm 里 新建 project

并 cd 进入 project 目录

1. 执行 git init 初始化本地仓库

2. 登录 github 创建自己的远程仓库，因为没给钱是public 的仓库，放心，这点儿代码没人偷，而且python开发者首先信奉开源

3. 执行命令 git remote add origin git@github.com:FrankLi99/awesome-python3-webapp.git 关联本地与远程仓库

4. 创建如下图所示的 工程目录
   
   

5. 添加 .gitignore 参考

6. 执行 git add . , git commit -m "init commit" 添加到暂存区，并提交到本地仓库的 master 分支，因为就我一个人开发，其实 master + dev 两个分支就已足够

7. 执行 git push -u origin master 推送本地仓库 master 分支 到远程 master 分支， -u 参数 还可以将 两个 master 分支 关联起来，下次拉取推送 就直接 git pull/push origin master 了

上面这么麻烦，完全可以使用 git clone repo-addr 来解决，只是 为了练习下另一种方法

day2 接下来 ，添加第一个 webserver --》 app.py

# -*- coding: utf-8 -*-
__author__ = 'Frank Li'

import logging; logging.basicConfig(level=logging.INFO)

import asyncio, os, json, time
from datetime import datetime

from aiohttp import web

def index(request):
    return web.Response(body=b'<h1>Awesome</h1>',content_type='text/html')

@asyncio.coroutine
def init(loop):
    app = web.Application(loop=loop)
    app.router.add_route('GET', '/', index)
    srv = yield from loop.create_server(app.make_handler(), '127.0.0.1', 9000)
    logging.info('server started at http://127.0.0.1:9000...')
    return srv

loop = asyncio.get_event_loop()
loop.run_until_complete(init(loop))
loop.run_forever()

自此，我们 Web 骨架搭好了，可以进行深入开发了。。。

1. 添加并提交到本地仓库 git add www/app.py
2. 推送到远程仓库 git push origin master

day03 编写 orm 框架 ==> 使用 元类 + mysql 异步库 aiomysql 编写 orm 框架，看起来复杂，不过总要为之。

# 第一次写这么复杂的代码，我建议先把主要的 结构搭建出来，然后在熟读源码的基础上，再来给每个pass部分做替换(这部分真的很综合，我认为是进阶)

import asyncio
import logging;logging.basicConfig(level=logging.info)
import aiomysql

def log(sql,args):
    logging.info('SQL: {sql} , other args: {args}'.format(sql=sql,args=args))

@asyncio.coroutine
def create_pool(loop,**db_info):
    logging.info('start to create aiomysql database connection pool...')
    global __pool
    __pool = yield from aiomysql.create_pool(host=db_info.get('host','localhost'),
                                             port=db_info.get('port',3306),
                                             db=db_info.get('db'),
                                             user=db_info.get('user'),
                                             password=db_info.get('password'),
                                             charset=db_info.get('charset','utf-8'),
                                             autocommit=db_info.get('autocommit',True),
                                             minsize=db_info.get('minsize',1),
                                             maxsize=db_info.get('maxsize',10),
                                             loop=loop)

@asyncio.coroutine
def select(sql,args,size=None):
    log(sql,args)
    global __pool
    with (yield from __pool) as conn:
        csr = yield from conn.cursor(aiomysql.DictCursor)
        yield from csr.execute(sql.replace('?','%s'),args or ())
        rs = csr.fetchmany(size) if size else csr.fetchall()
        return rs

@asyncio.coroutine
def execute(sql,args,autocommit=True):
    log(sql,args)
    global __pool
    with (yield from __pool) as conn:
        if not autocommit:
            yield from conn.begin()
        try:
            csr = yield from conn.cursor()
            yield from csr.execute(sql.replace('?','%s'),args or ())
            affectedRow = csr.rowcount
            if not autocommit:
                yield from conn.commit()
        except BaseException as e:
            raise
            if not autocommit:
                yield from conn.rollback()
        finally:
            yield from csr.close()
        return affectedRow

# 根据长度构造 占位符
def create_args_string(num):
    return ','.join('?'*num)

# 定义 Field 字段类
class Field(object):
    pass
class IntegerField(Field):
    pass
class StringField(Field):
    pass
class FloatField(Field):
    pass
class BooleanField(Field):
    pass
class TextField(Field):
    pass

# 定义实体类的元类，先有类来后有天，元类更在类之前
class ModelMetaclass(type):
    pass

class Model(dict,metaclass=ModelMetaclass):
    pass

完善版 orm

# -*- coding: utf-8 -*-
__author__ = 'Frank Li'

import asyncio
import logging;logging.basicConfig(level=logging.DEBUG)
import aiomysql

def log(sql,args):
    logging.info('SQL:{sql} , other ARGS:{args}'.format(sql=sql,args=args))

# 创建数据库连接池 -- 全局
@asyncio.coroutine
def create_pool(loop,**db_info):
    logging.info('start to create global database connection pool...')
    global __pool
    __pool = yield from aiomysql.create_pool(host=db_info.get('host','localhost'),
                                            port=db_info.get('port',3306),
                                            db=db_info.get('db'),
                                            user=db_info.get('user'),
                                            password=db_info.get('password'),
                                            charset=db_info.get('charset','utf8'),
                                            autocommit=db_info.get('autocommit',True),
                                            minsize=db_info.get('minsize',1),
                                            maxsize=db_info.get('maxsize',10),
                                            loop=loop)

@asyncio.coroutine
def select(sql,args,size=None):
    log(sql,args)
    global __pool
    with (yield from __pool) as conn:
        csr = yield from conn.cursor(aiomysql.DictCursor)
        yield from csr.execute(sql.replace('?','%s'),args or ())
        rs = csr.fetchmany(size) if size else csr.fetchall()
        yield from csr.close()
        # logging.info('result rowcount: {}'.format(len(rs)))
        return rs

@asyncio.coroutine
def execute(sql,args,autocommit=True):
    log(sql,args)
    global __pool
    with (yield from __pool) as conn:
        if not autocommit:
            yield from conn.begin()
        try:
            csr = yield from conn.cursor()
            yield from csr.execute(sql.replace('?','%s'),args or ())
            affctedRow = csr.rowcount
            logging.info('execute sql affcted row: {}'.format(affctedRow))
            if not autocommit:
                yield from conn.commit()

        except BaseException:
            if not autocommit:
                yield from conn.rollback()
            raise
        finally:
            try:
                yield from csr.close()
            except BaseException as e:
                logging.info(e)
        return affctedRow

# 开始定义 字段类 Field
class Field(object):
    def __init__(self,name,column_type,is_pk,default):
        self.name = name
        self.column_type = column_type
        self.is_pk = is_pk
        self.default = default
    def __repr__(self):
        return '<{},{}:{}>'.format(self.__class__.__name__,self.column_type,self.name)
    __str__ = __repr__

class IntegerField(Field):
    def __init__(self,name=None,column_type='bigint',is_pk=False,default=0):
        super(IntegerField,self).__init__(name,column_type,is_pk,default)

class StringField(Field):
    def __init__(self,name=None,column_type='varchar(256)',is_pk=False,default=None):
        super(StringField,self).__init__(name,column_type,is_pk,default)

class FloatField(Field):
    def __init__(self,name=None,column_type='real',is_pk=False,default=0.0):
        super(FloatField,self).__init__(name,column_type,is_pk,default)

class BooleanField(Field):
    def __init__(self,name=None,column_type='boolean',default=False):
        super(BooleanField,self).__init__(name,column_type,False,default)

class TextField(Field):
    def __init__(self,name=None,column_type='text',default=None):
        super(TextField,self).__init__(name,column_type,False,default)

# 根据参数个数构造 ？ 占位符
def create_args_string(num):
    return ','.join('?'*num)

# 开始写 元类 ModelMetaclass  主要作用就是 对 所有 Model 构造 select ,insert,delete,update 语句
# 先有类来后有天，元类更在类之前
class ModelMetaclass(type):
    def __new__(cls,name,bases,attrs):
        # 排除掉 Model 这个 父类本身
        if name=='Model':
            return type.__new__(cls,name,bases,attrs)

        tb_name = attrs.get('__table__',str.lower(name))

        # 如果是其子类
        mappings = {}
        fields = []
        primaryKey = None

        for k,v in attrs.items():

            if isinstance(v,Field):
                logging.info('found field mapping {} <==> {}'.format(k, v))
                mappings[k] = v
                if v.is_pk:
                    if primaryKey:
                        raise RuntimeError('Duplicated primary key for field: {}'.format(k))
                    primaryKey = k
                else:
                    fields.append(k)

        if not primaryKey:
            raise RuntimeError('Primary key not found...')

        # 去除 类中 与 实例中 变量同名的 类变量
        for k in mappings.keys():
            attrs.pop(k)

        escape_fields = list(map(lambda f: '`{}`'.format(f),fields))

        # 将获取到的值 放入 类属性中
        attrs['__mappings__'] = mappings
        attrs['__table__'] = tb_name
        attrs['__fields__'] = fields
        attrs['__primary_key__'] = primaryKey

        # 构造 select , update ,delete ,insert 语句
        attrs['__select__'] = 'select `{primaryKey}`,{escape_fields} from `{tb_name}` '.format(primaryKey=primaryKey,escape_fields=','.join(escape_fields),tb_name=tb_name)
        attrs['__insert__'] = 'insert into `{tb_name}`({escape_fields},`{primaryKey}`) values({args})'.format(tb_name=tb_name,escape_fields=','.join(escape_fields),primaryKey=primaryKey,args=create_args_string(len(fields)+1))
        attrs['__delete__'] = 'delete from `{tb_name}` where `{primaryKey}`=?'.format(tb_name=tb_name,primaryKey=primaryKey)
        attrs['__update__'] = 'update `tb_name` set {set_cols} where `{primaryKey}`=?'.format(tb_name=tb_name,set_cols=','.join(list(map(lambda f:'`{}`=?'.format(mappings.get(f).name or f),fields))),primaryKey=primaryKey)

        # 返回 改造好的 类模板
        return type.__new__(cls,name,bases,attrs)

# 造实体类的 模板
class Model(dict,metaclass=ModelMetaclass):

    def __init__(self,*args,**kw):
        super(Model,self).__init__(*args,**kw)

    # 方便 dict 对象 如同 属性一般调用 dict.name , dict.id 等等
    def __getattr__(self,key):
        try:
            return self[key]
        except KeyError:
            raise AttributeError(r"'Model' object has no attribute {}".format(key))
    def __setattr__(self,key,value):
        self[key] = value

    def getValue(self,key):
        return getattr(self,key,None)

    def getValueOrDefault(self,key):
        value = self.getValue(key)
        if value is None:
            field = self.__mappings__[key]
            if field is not None:
                value = field.default()  if callable(field.default) else field.default
                logging.info('using default for {} : {}'.format(key,str(value)))
                setattr(self,key,value)
        return value

    @classmethod
    @asyncio.coroutine
    def findAll(cls,where=None,args=None,**kw):
        'find object by where clause'
        select_sql = [cls.__select__]
        if where:
            select_sql.append('where')
            select_sql.append(where)

        if args is None:
            args=[]

        orderBy = kw.get('orderbBy',None)
        if orderBy:
            select_sql.append('order by ')
            select_sql.append(orderBy)

        limit = kw.get('limit',None)
        if limit:
            select_sql.append('limit')
            if isinstance(limit,int):
                select_sql.append('?')
                args.append(limit)
            elif isinstance(limit,tuple):

                select_sql.append(create_args_string(2)) #  ?,?
                args.extend(limit)
            else:
                raise ValueError('Invalid limit value: {}'.format(str(limit)))

        rs = yield from select(' '.join(select_sql),args)
        return [cls(**r) for r in rs]

    @classmethod
    @asyncio.coroutine
    def findNumber(cls,selectField,where=None,args=None):
        ''' find number by select and where'''
        select_sql = ['select {selectField} _num_ from `tb_name`'.format(selectField=selectField,tb_name=cls.__table__)]
        if where:
            select_sql.append('where')
            select_sql.append(where)
        rs = yield from select(' '.join(select_sql),args,1)
        if len(rs) ==0:
            return None
        return rs[0]['_num_']

    @classmethod
    @asyncio.coroutine
    def find(cls,pk):
        '''find object by primary key'''
        rs = yield from select('{select_sql} where `{pk_field}`=?'.format(select_sql=cls.__select__,pk_field=cls.__primary_key__),[pk],size=1)
        return None if len(rs)==0 else cls(**rs[0])

    # save
    @asyncio.coroutine
    def save(self):
        args = list(map(self.getValueOrDefault,self.__fields__))
        args.append(self.getValueOrDefault(self.__primary_key__))
        rows = yield from execute(self.__insert__,args,True)
        if rows != 1:
            logging.warn('failed to insert record: affcted rows: {rowCount}'.format(rowCount=rows))
        else:
            logging.warn('affcted rows: {rowCount}'.format(rowCount=rows))
    # update
    @asyncio.coroutine
    def update(self):
        args = list(map(self.getValueOrDefault,self.__fields__))
        args.append(self.getValueOrDefault(self.__primary_key__))
        rows = yield from execute(self.__update__,args)
        if rows !=1:
            logging.warn('failed to update by primary key: affected {rowCount}'.format(rowCount=rows))

    # delete
    @asyncio.coroutine
    def remove(self):
        args = list(map(self.getValueOrDefault,self.__fields__))
        args.append(self.getValueOrDefault(self.__primary_key__))
        rows = yield from execute(self.__delete__,args)
        if rows != 1:
            logging.info('failed to delete by primary key affected rows: {rowCount}'.format(rowCount=rows))

model 模块

# -*- coding: utf-8 -*-
__author__ = 'Frank Li'
from www.orm2 import Model,StringField,IntegerField,BooleanField,FloatField,create_pool
import logging;logging.basicConfig(level=logging.DEBUG)
import time
from www.models import User
import asyncio
import uuid

def next_id():
    return '%015d%s000' % (int(time.time() * 1000), uuid.uuid4().hex)

class User(Model):
    __table__ = 'users'

    id = StringField(is_pk=True, default=next_id, column_type='varchar(50)')
    email = StringField(column_type='varchar(50)')
    passwd = StringField(column_type='varchar(50)')
    admin = BooleanField()
    name = StringField(column_type='varchar(50)')
    image = StringField(column_type='varchar(500)')
    created_at = FloatField(default=time.time)

@asyncio.coroutine
def test(loop):
    yield from  create_pool(loop=loop,user='www-data', password='www-data', db='awesome')
    u = User(name='Frank', email='Frank@example.com', passwd='1234567890', image='about:blank')
    yield from u.save()
    logging.info(u.__select__)
    logging.info(u.__insert__)
    logging.info(u.__delete__)
    logging.info(u.__update__)

    users = yield from User.findAll()
    for user in users:
        logging.info('user info:\n  '+str(user))
if __name__ == '__main__':
    loop = asyncio.get_event_loop()
    loop.run_until_complete(test(loop))
    loop.run_forever()

编写 异步 web 框架， 这里真的好难。。。比 ORM 还难很多，初次看来是这样，可能需要 细细钻研源码。。。

遍寻网络终于发现注释版的代码

这篇博客不错，终还是自己见识太浅薄

# 首先来个最简单的吧， 构造 两个 装饰器用来对应 前台页面传来的 两种请求 ，get / post
from functools import wraps,partial

def handler_decorator(path,*,method='GET'):
    def decorator(func):
        @wraps(func) # 更正函数签名
        def wrapper(*args,**kw):
            result = func(*args,**kw)
            return result
        wrapper.__method__ = method
        wrapper.__urlpath__ = path
        return wrapper
    return decorator

get=partial(handler_decorator,method='GET')
post=partial(handler_decorator,method='POST')

# 使用时候 就方便了
@get('/home/index')
def home_index(request):
    pass

### 定义 RequestHandler 用来解析 request 中的 信息

import inspect,asyncio
from web_app.APIError import APIError
from aiohttp import web
from urllib import parse

#运用inspect模块，创建几个函数用以获取URL处理函数与request参数之间的关系
def get_required_kw_args(fn): #收集没有默认值的命名关键字参数
    args = []
    params = inspect.signature(fn).parameters #inspect模块是用来分析模块，函数
    for name, param in params.items():
        if str(param.kind) == 'KEYWORD_ONLY' and param.default == inspect.Parameter.empty:
            args.append(name)
    return tuple(args)

def get_named_kw_args(fn):  #获取命名关键字参数
    args = []
    params = inspect.signature(fn).parameters
    for name,param in params.items():
        if str(param.kind) == 'KEYWORD_ONLY':
            args.append(name)
    return tuple(args)

def has_named_kw_arg(fn): #判断有没有命名关键字参数
    params = inspect.signature(fn).parameters
    for name,param in params.items():
        if str(param.kind) == 'KEYWORD_ONLY':
            return True

def has_var_kw_arg(fn): #判断有没有关键字参数
    params = inspect.signature(fn).parameters
    for name,param in params.items():
        if str(param.kind) == 'VAR_KEYWORD':
            return True

def has_request_arg(fn): #判断是否含有名叫'request'参数，且该参数是否为最后一个参数
    params = inspect.signature(fn).parameters
    sig = inspect.signature(fn)
    found = False
    for name,param in params.items():
        if name == 'request':
            found = True
            continue #跳出当前循环，进入下一个循环
        if found and (str(param.kind) != 'VAR_POSITIONAL' and str(param.kind) != 'KEYWORD_ONLY' and str(param.kind != 'VAR_KEYWORD')):
            raise ValueError('request parameter must be the last named parameter in function: %s%s'%(fn.__name__,str(sig)))
    return found

#定义RequestHandler,正式向request参数获取URL处理函数所需的参数，发现 请求信息有问题的 raise APIError ，没有问题 则放行 ，注意所有这些 fn 都是指的实际 handler.py 中的函数

class RequestHandler(object):

    def __init__(self,app,fn):#接受app参数
        self._app = app
        self._fn = fn
        self._required_kw_args = get_required_kw_args(fn)
        self._named_kw_args = get_named_kw_args(fn)
        self._has_named_kw_arg = has_named_kw_arg(fn)
        self._has_var_kw_arg = has_var_kw_arg(fn)
        self._has_request_arg = has_request_arg(fn)

    async def __call__(self,request): #__call__这里要构造协程
        kw = None
        if self._has_named_kw_arg or self._has_var_kw_arg:
            if request.method == 'POST': #判断客户端发来的方法是否为POST
                if not request.content_type: #查询有没提交数据的格式（EncType）
                    return web.HTTPBadRequest(text='Missing Content_Type.')#这里被廖大坑了，要有text
                ct = request.content_type.lower() #小写
                if ct.startswith('application/json'): #startswith
                    params = await request.json() #Read request body decoded as json.
                    if not isinstance(params,dict):
                        return web.HTTPBadRequest(text='JSON body must be object.')
                    kw = params
                elif ct.startswith('application/x-www-form-urlencoded') or ct.startswith('multipart/form-data'):
                    params = await request.post() # reads POST parameters from request body.If method is not POST, PUT, PATCH, TRACE or DELETE or content_type is not empty or application/x-www-form-urlencoded or multipart/form-data returns empty multidict.
                    kw = dict(**params)
                else:
                    return web.HTTPBadRequest(text='Unsupported Content_Tpye: %s'%(request.content_type))
            if request.method == 'GET':
                qs = request.query_string #The query string in the URL
                if qs:
                    kw = dict()
                    for k,v in parse.parse_qs(qs,True).items(): #Parse a query string given as a string argument.Data are returned as a dictionary. The dictionary keys are the unique query variable names and the values are lists of values for each name.
                        kw[k] = v[0]
        if kw is None:
            kw = dict(**request.match_info)
        else:
            if not self._has_var_kw_arg and self._named_kw_args: #当函数参数没有关键字参数时，移去request除命名关键字参数所有的参数信息
                copy = dict()
                for name in self._named_kw_args:
                    if name in kw:
                        copy[name] = kw[name]
                kw = copy
            for k,v in request.match_info.items(): #检查命名关键参数
                if k in kw:
                    logging.warning('Duplicate arg name in named arg and kw args: %s' % k)
                kw[k] = v
        if self._has_request_arg:
            kw['request'] = request
        if self._required_kw_args: #假如命名关键字参数(没有附加默认值)，request没有提供相应的数值，报错
            for name in self._required_kw_args:
                if name not in kw:
                    return web.HTTPBadRequest(text='Missing argument: %s'%(name))
        logging.info('call with args: %s' % str(kw))

        try:
            r = await self._fn(**kw)
            return r
        except APIError as e: #APIError另外创建
            return dict(error=e.error, data=e.data, message=e.message)

# 注册 url  对应处理函数 相当于 spring中的  @RequestMapping  注解

import inspect,asyncio

#编写一个add_route函数，用来注册一个URL处理函数
def add_route(app,fn):
    method = getattr(fn,'__method__',None)
    path = getattr(fn,'__route__',None)
    if method is None or path is None:
        return ValueError('@get or @post not defined in %s.'%str(fn))
    if not asyncio.iscoroutinefunction(fn) and not inspect.isgeneratorfunction(fn): #判断是否为协程且生成器,不是使用isinstance
        fn = asyncio.coroutine(fn)
    logging.info('add route %s %s => %s(%s)'%(method,path,fn.__name__,','.join(inspect.signature(fn).parameters.keys())))
    app.router.add_route(method,path,RequestHandler(app,fn))#别忘了RequestHandler的参数有两个

# 上面注册如果 一个一个 弄 岂不是会烦死，下面来一个 模块导入，批量注册
#直接导入文件，批量注册一个URL处理函数
def add_routes(app,module_name):
    n = module_name.rfind('.')
    if n == -1:
        mod = __import__(module_name,globals(),locals())
    else:
        name = module_name[n+1:]
        mod = getattr(__import__(module_name[:n],globals(),locals(),[name],0),name)#第一个参数为文件路径参数，不能掺夹函数名，类名
    for attr in dir(mod):
        if attr.startswith('_'):
            continue
        fn = getattr(mod,attr)
        if callable(fn):
            method = getattr(fn,'__method__',None)
            path = getattr(fn,'__route__',None)
            if path and method: #这里要查询path以及method是否存在而不是等待add_route函数查询，因为那里错误就要报错了

### 说到底 还是要 顺着 aiohttp 提供的 接口来，给他 构造出他要的参数
import os,logging

#添加静态文件夹的路径
def add_static(add):
    path = os.path.join(os.path.dirname(os.path.abspath(__file__)),'static')#输出当前文件夹中'static'的路径
    app.router.add_static('/static/',path)#prefix (str) – URL path prefix for handled static files
    logging.info('add static %s => %s'%('/static/',path))

from jinja2 import Environment, FileSystemLoader
from datetime import datetime
import json, time
import logging

#初始化jinja2，以便其他函数使用jinja2模板
def init_jinja2(app, **kw):
    logging.info('init jinja2...')
    options = dict(
        autoescape = kw.get('autoescape', True),
        block_start_string = kw.get('block_start_string', '{%'),
        block_end_string = kw.get('block_end_string', '%}'),
        variable_start_string = kw.get('variable_start_string', '{{'),
        variable_end_string = kw.get('variable_end_string', '}}'),
        auto_reload = kw.get('auto_reload', True)
    )
    path = kw.get('path', None)
    if path is None:
        path = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'templates')
    logging.info('set jinja2 template path: %s' % path)
    env = Environment(loader=FileSystemLoader(path), **options)
    filters = kw.get('filters', None)
    if filters is not None:
        for name, f in filters.items():
            env.filters[name] = f
    app['__templating__'] = env

def datetime_filter(t):
    delta = int(time.time() - t)
    if delta < 60:
        return u'1分钟前'
    if delta < 3600:
        return u'%s分钟前' % (delta // 60)
    if delta < 86400:
        return u'%s小时前' % (delta // 3600)
    if delta < 604800:
        return u'%s天前' % (delta // 86400)
    dt = datetime.fromtimestamp(t)
    return u'%s年%s月%s日' % (dt.year, dt.month, dt.day)