欢迎各位兄弟 发布技术文章

这里的技术是共享的

You are here

首页

Python修饰器的函数式编程

shiping1 的头像
星期二, 2014-12-16 21:52 — shiping1
给函数打日志

下面这个示例演示了一个logger的decorator，这个decorator输出了函数名，参数，返回值，和运行时间。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
from functools import wraps
def logger(fn):
    @wraps(fn)
    def wrapper(*args, **kwargs):
        ts = time.time()
        result = fn(*args, **kwargs)
        te = time.time()
        print "function      = {0}".format(fn.__name__)
        print "    arguments = {0} {1}".format(args, kwargs)
        print "    return    = {0}".format(result)
        print "    time      = %.6f sec" % (te-ts)
        return result
    return wrapper
 
@logger
def multipy(x, y):
    return x * y
 
@logger
def sum_num(n):
    s = 0
    for i in xrange(n+1):
        s += i
    return s
 
print multipy(2, 10)
print sum_num(100)
print sum_num(10000000)

上面那个打日志还是有点粗糙，让我们看一个更好一点的（带log level参数的）：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
import inspect
def get_line_number():
    return inspect.currentframe().f_back.f_back.f_lineno
 
def logger(loglevel):
    def log_decorator(fn):
        @wraps(fn)
        def wrapper(*args, **kwargs):
            ts = time.time()
            result = fn(*args, **kwargs)
            te = time.time()
            print "function   = " + fn.__name__,
            print "    arguments = {0} {1}".format(args, kwargs)
            print "    return    = {0}".format(result)
            print "    time      = %.6f sec" % (te-ts)
            if (loglevel == 'debug'):
                print "    called_from_line : " + str(get_line_number())
            return result
        return wrapper
    return log_decorator

但是，上面这个带log level参数的有两具不好的地方，
1） loglevel不是debug的时候，还是要计算函数调用的时间。
2） 不同level的要写在一起，不易读。

我们再接着改进：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
import inspect
 
def advance_logger(loglevel):
 
    def get_line_number():
        return inspect.currentframe().f_back.f_back.f_lineno
 
    def _basic_log(fn, result, *args, **kwargs):
        print "function   = " + fn.__name__,
        print "    arguments = {0} {1}".format(args, kwargs)
        print "    return    = {0}".format(result)
 
    def info_log_decorator(fn):
        @wraps(fn)
        def wrapper(*args, **kwargs):
            result = fn(*args, **kwargs)
            _basic_log(fn, result, args, kwargs)
        return wrapper
 
    def debug_log_decorator(fn):
        @wraps(fn)
        def wrapper(*args, **kwargs):
            ts = time.time()
            result = fn(*args, **kwargs)
            te = time.time()
            _basic_log(fn, result, args, kwargs)
            print "    time      = %.6f sec" % (te-ts)
            print "    called_from_line : " + str(get_line_number())
        return wrapper
 
    if loglevel is "debug":
        return debug_log_decorator
    else:
        return info_log_decorator

你可以看到两点，
1）我们分了两个log level，一个是info的，一个是debug的，然后我们在外尾根据不同的参数返回不同的decorator。
2）我们把info和debug中的相同的代码抽到了一个叫_basic_log的函数里，DRY原则。

一个MySQL的Decorator

下面这个decorator是我在工作中用到的代码，我简化了一下，把DB连接池的代码去掉了，这样能简单点，方便阅读。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
import umysql
from functools import wraps
 
class Configuraion:
    def __init__(self, env):
        if env == "Prod":
            self.host    = "coolshell.cn"
            self.port    = 3306
            self.db      = "coolshell"
            self.user    = "coolshell"
            self.passwd  = "fuckgfw"
        elif env == "Test":
            self.host   = 'localhost'
            self.port   = 3300
            self.user   = 'coolshell'
            self.db     = 'coolshell'
            self.passwd = 'fuckgfw'
 
def mysql(sql):
 
    _conf = Configuraion(env="Prod")
 
    def on_sql_error(err):
        print err
        sys.exit(-1)
 
    def handle_sql_result(rs):
        if rs.rows > 0:
            fieldnames = [f[0] for f in rs.fields]
            return [dict(zip(fieldnames, r)) for r in rs.rows]
        else:
            return []
 
    def decorator(fn):
        @wraps(fn)
        def wrapper(*args, **kwargs):
            mysqlconn = umysql.Connection()
            mysqlconn.settimeout(5)
            mysqlconn.connect(_conf.host, _conf.port, _conf.user, \
                              _conf.passwd, _conf.db, True, 'utf8')
            try:
                rs = mysqlconn.query(sql, {})
            except umysql.Error as e:
                on_sql_error(e)
 
            data = handle_sql_result(rs)
            kwargs["data"] = data
            result = fn(*args, **kwargs)
            mysqlconn.close()
            return result
        return wrapper
 
    return decorator
 
@mysql(sql = "select * from coolshell" )
def get_coolshell(data):
    ... ...
    ... ..
线程异步

下面量个非常简单的异步执行的decorator，注意，异步处理并不简单，下面只是一个示例。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
from threading import Thread
from functools import wraps
 
def async(func):
    @wraps(func)
    def async_func(*args, **kwargs):
        func_hl = Thread(target = func, args = args, kwargs = kwargs)
        func_hl.start()
        return func_hl
 
    return async_func
 
if __name__ == '__main__':
    from time import sleep
 
    @async
    def print_somedata():
        print 'starting print_somedata'
        sleep(2)
        print 'print_somedata: 2 sec passed'
        sleep(2)
        print 'print_somedata: 2 sec passed'
        sleep(2)
        print 'finished print_somedata'
 
    def main():
        print_somedata()
        print 'back in main'
        print_somedata()
        print 'back in main'
 
    main()

其它

关于更多的示例，你可以参看： Python Decorator Library

关于Python Decroator的各种提案，可以参看：Python Decorator Proposals

（全文完）

来自 http://coolshell.cn/articles/11265.html
普通分类: 
python
Powered by Drupal

友情链接

校园好文网

 www.shipingzhong.cn  个人技术网_前端_后台    备案号:苏ICP备18010659号-2