Python 2.7、多处理和 Tkinter [英] Python 2.7, Multiprocessing and Tkinter
问题描述
我正在尝试创建一个类,该类使用多处理从新的 python 实例启动 tkinter 表单.
I am trying to create a class that launches a tkinter form from a new instance of python using multiprocessing.
我的问题是:将 Tkinter 类对象传递给多处理函数的推荐方法是什么?当我通常将对象传递给多处理池时,我必须将对象转换为字符串.这是否意味着我必须腌制课程?有没有更好的方法来做到这一点?
My question is this: What would be the recommended way of passing the Tkinter class object to the multiprocessing function? When I normally pass objects to the multiprocessing pool, I have to convert the objects to string. Does this mean I have to pickle the class? Is there a better way to do this?
我不能使用线程,因为我必须使用的一些 python 包不是线程安全的.. :(
I cannot using threading because some of the python packages I have to use are not thread safe.. :(
示例代码:
import Tkinter as Tk
########################################################################
class TkinterApp(object):
""""""
#----------------------------------------------------------------------
def __init__(self, parent):
"""Constructor"""
self.root = parent
self.root.title("Main frame")
self.frame = Tk.Frame(parent)
self.frame.pack()
btn = Tk.Button(self.frame, text="Close Frame", command=self.hide)
btn.pack()
#----------------------------------------------------------------------
def hide(self):
""""""
self.root.withdraw()
系统规格:
- Windows 7
- Python 2.7.5 32 位
非常欢迎任何建议或反馈!
Any suggestions or feedback is very welcome!
谢谢
推荐答案
在池外创建 Tkinter 实例以及类实例,然后从不同的进程调用函数时,我没有遇到任何问题.下面是一个简单"的例子
I haven't encountered any problems creating the Tkinter instance outside of the pool, and the class instance as well, and then calling the functions from separate processes. A "simple" example is below
from multiprocessing import Process
import time
try:
import Tkinter as tk ## Python 2.x
except:
import tkinter as tk ## Python 3.x
class ProgressBar():
def __init__(self, root):
self.root=root
self.root.geometry("75x50+900+100")
self.ctr=25
def mainloop(self):
self.root.mainloop()
def start_countdown(self):
""" a separate process in a separate GUI
"""
self.root.withdraw()
self.top_count=tk.Toplevel(self.root)
self.top_count.geometry("75x50+750+50")
self.label_ctr = tk.IntVar()
self.label_ctr.set(self.ctr)
label = tk.Label(self.top_count, textvariable=self.label_ctr)
label.pack()
if self.ctr > 0:
self.top_count.after(750, self.update)
def start_running(self):
""" create the progress bar widget
"""
self.top=tk.Toplevel(self.root, takefocus=True)
self.top.title("Progress Bar")
self.top.geometry("+700+200")
canvas = tk.Canvas(self.top, width=261, height=60, bg='lightgray')
canvas.pack()
rc2 = canvas.create_rectangle(15, 20, 243, 50, outline='blue', \
fill='lightblue')
rc1 = canvas.create_rectangle(24, 20, 34, 50, outline='white', \
fill='blue')
total=100
x = 5
while self.ctr: ## move the small rectangle +5 or -5 units
total += x
if total > 311:
x = -5
elif total < 100:
x = 5
time.sleep(0.2)
canvas.move(rc1, x, 0)
canvas.update()
def update(self):
self.ctr -= 1
self.label_ctr.set(self.ctr)
if self.ctr > 0:
self.top_count.after(750, self.update)
else:
## sleep to allow any remaining after() to execute
## can also use self.root.after_cancel(id)
self.top_count.after(500, self.root.destroy) ## destroy root when zero is reached
root = tk.Tk()
PB=ProgressBar(root)
pr1=Process(target=PB.start_countdown(), args=())
pr1.start()
pr2=Process(target=PB.start_running(), args=())
pr2.start()
## start mainloop in a separate process as a function of the class
## don't know if this is really necessary or not
## the theory is, it is detached from the other 2 processes and so
## can react to both independently
## also the mainloop() process can be killed=shut down properly
pr3=Process(target=PB.mainloop(), args=())
pr3.start()
## safety clean up
pr1.terminate()
pr2.terminate()
pr3.terminate()
这篇关于Python 2.7、多处理和 Tkinter的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!