Python 为什么在更改完成之前中断线程

我正在尝试创建python模块，用于通过IP地址获取MAC地址

def getMACs(addressesList):
    def _processArp(pkt):
        spa = _inet_ntoa(pkt.spa) 
        if pkt.op == dpkt.arp.ARP_OP_REPLY and spa in _cache.macTable:
            lock.acquire()
            try:
                _cache.macTable[spa] = _packedToMacStr(pkt.sha)
                _cache.notFilledMacs -= 1
            finally:
                lock.release()
            if _cache.notFilledMacs == 0:
                thrd.stop()

    addresses = _parseAddresses(addressesList)

    _cache.registerCacheEntry("macTable", {})
    _cache.registerCacheEntry("notFilledMacs", 0)
    _events.arpPacket += _processArp

    lock = threading.Lock()

    thrd = _CaptureThread(promisc=False, timeout_ms=30, filter="arp")
    thrd.start()
    for addr in addresses:
        if _sendArpQuery(addr):
            _cache.macTable[str(addr)] = None
            _cache.notFilledMacs += 1
    thrd.join(125)
    thrd.stop()
    return _cache.macTable

if __name__ == "__main__":
    macTable = getMACs([IPAddress("192.168.1.1"), IPAddress("192.168.1.3")])
    _pprint.pprint(macTable)

当我运行这个模块时，我得到

{'192.168.1.1': '00:11:95:9E:25:B1', '192.168.1.3': None}

当我一步一步地调试processArp时，我得到

{'192.168.1.1': '00:11:95:9E:25:B1', '192.168.1.3': '00:21:63:78:98:8E'}

课程名称如下：

class CaptureThread(threading.Thread):
    def __init__ (self, name=None, snaplen=65535, promisc=True, timeout_ms=0, immediate=False, filter=None):
        threading.Thread.__init__(self)
        self.__running = True
        self.__name = name 
        self.__snaplen = snaplen 
        self.__promisc = promisc 
        self.__timeout_ms = timeout_ms 
        self.__immediate = immediate 
        self.__filter = filter 

    def stop(self):
        self.__running = False

    def run(self):
        self.__pc = pcap.pcap(self.__name, self.__snaplen, self.__promisc, self.__timeout_ms, self.__immediate)
        if self.__filter:
            self.__pc.setfilter(self.__filter)

        while self.__running:
            self.__pc.dispatch(1, self.__processPacket)

    def __processPacket(self, timestamp, pkt):
        peth = dpkt.ethernet.Ethernet(pkt)
        if isinstance(peth.data, dpkt.arp.ARP):
            _events.arpPacket(peth.data)

---

愚蠢的错误。与使用线程时一样-因为线程同步

---

我中断线程的条件之一是“\u cache.notFilledMacs==0”。在主线程中，当CAPTURAREAD值减小时，cache.notFilledMacs没有时间获取2的值。

\u此uuuu代码uuu是uuuu非常uuuuuu硬uuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuu。可以我希望它们是模块级的，因为这可以解释用下划线开始它们的重要性。原因是什么？双下划线并不能取代私人密码。父类使用双下划线，以防止名称与其子类冲突，尤其是当父类由不同于子类的开发人员编写时。因此，一般来说，在python中，不允许不分青红皂白地使用双下划线或单下划线。更重要的是，您应该提供人们可以运行的代码，而不是不可命名的片段，期望人们只是眯着眼睛看它。请隔离您正在询问的问题，并提供一个自包含的代码块，我们可以将其复制到文件并进行测试。