Python线程实际上并未启动

Question

I wrote a python GUI program to deal with some scientific calculations. One of the functions was extremely time consuming so that I wish to put it into a separate thread so that running it will not block the entire program.

The problem is that this thread does not actually starts (or queue.get never obtains a result).

However, if I passed the option block = True into the self.PLUQ = self.queue.get(block = False) statement, the function will execute but will also block the entire program.

    Class PLUQ_GUI_Dialog():
        #somefunctions to initiatialize the class 

        def guess_button_click_cb(self):
            self.selected_peaks = self.session.selected_peaks()
            if (not self.is_ready_to_guess()):
                return
            self.input_frequencies = self.get_input_frequencies(self.selected_peaks)

            self.queue = Queue.Queue()
            thread = threading.Thread(target = self.guess, args = (self.queue, self.input_frequencies))
            thread.setDaemon(True)
            thread.start()
            self.top.after(100, self.process_queue)

        def guess(self, queue, inputs):
            PLUQ = pluq.PLUQ_interface() 
            #pluq is a separate python module
            PLUQ.query(inputs)
            #query does some calculations and put results in the PLUQ object
            queue.put(PLUQ)

        def process_queue(self):
            try:
                self.PLUQ = self.queue.get(block = False)
                self.update_pluq_result() # This will display the result
            except Queue.Empty:
                self.top.after(100, self.process_queue)

        #some helper function

Answer 1

.setDaemon() is part of the Python 2.6 and earlier API. You should now be setting your daemon threads like so:

self.queue = Queue.Queue()
thread = threading.Thread(target = self.guess, args = (self.queue, self.input_frequencies))
thread.daemon = True # like this
thread.start()
self.top.after(100, self.process_queue)

You can read more about it here .

Answer 2

The problem could be GIL related. I got around the problem by using multiprocessing.Process instead of threading.Thread.