我正在尝试编写一个生成器函数,该函数从数据库中获取行并一次返回一个。但是,我不确定下面标记**的清理代码是否按照我的想法执行。如果没有,那么将清理代码放在最后一个yield语句之后执行的生成器本身的最佳方法是什么?我看着捕获StopIteration,但这似乎是从调用者完成的,而不是在生成器中。
def MYSQLSelectGenerator(stmt):
...
try:
myDB = MySQLdb.connect(host=..., port=..., user=..., passwd=..., db=...)
dbc=myDB.cursor()
dbc.execute(stmt)
d = "asdf"
while d is not None:
d = dbc.fetchone() #can also use fetchmany() to be more efficient
yield d
dbc.close() #** DOES THIS WORK AS I INTEND, MEANING AS SOON AS d = "None"
except MySQLdb.Error, msg:
print("MYSQL ERROR!")
print msg
答案 0 :(得分:3)
您的版本会在dbc.close()
后立即运行d is None
,但如果引发异常则不会。你需要一个finally
clause。即使引发异常,此版本也保证运行dbc.close()
:
try:
myDB = MySQLdb.connect(host=..., port=..., user=..., passwd=..., db=...)
dbc = myDB.cursor()
dbc.execute(stmt)
d = "asdf"
while d is not None:
d = dbc.fetchone() #can also use fetchmany() to be more efficient
yield d
except MySQLdb.Error, msg:
print("MYSQL ERROR!")
print msg
finally:
dbc.close()
答案 1 :(得分:3)
您可以做的一件事是使用finally
子句。另一种选择(在这里可能有些过分但是有用的东西要知道)是创建一个与with
语句一起使用的类:
class DatabaseConnection:
def __init__(self, statement):
self.statemet = statement
def __enter__(self):
self.myDB = MySQLdb.connect(host=..., port=...,user=...,passwd=...,db=...)
self.dbc = myDB.cursor()
self.dbc.execute(self.statement)
self.d = "asdf"
def __exit__(self, exc_type, exc_value, traceback):
self.dbc.close()
def __iter__(self):
while self.d is not None:
self.d = self.dbc.fetchone()
yield self.d
with DatabaseConnection(stmnt) as dbconnection:
for i in dbconnection:
print(i)
答案 2 :(得分:1)
您可以使用上下文管理器和with
语句。 contextlib
提供closing
:
from contextlib import closing
myDB = MySQLdb.connect(host=..., port=..., user=..., passwd=..., db=...)
with closing(myDB.cursor()) as dbc:
dbc.execute(stmt)
d = "asdf"
while d is not None:
d = dbc.fetchone() #can also use fetchmany() to be more efficient
yield d
这将在close()
块的末尾自动调用dbc
上的with
,即使已经引发了异常。