python代码转java_有没有大神帮忙把python代码转换成java

该楼层疑似违规已被系统折叠 隐藏此楼查看此楼

#!/usr/bin/pythonimport randomimport numpy#global all_setserver1_ip = '192.168.140.1'server2_ip = '192.168.140.2'server3_ip = '192.168.140.3'server4_ip = '192.168.140.4'server5_ip = '192.168.140.5'server6_ip = '192.168.140.6'server7_ip = '192.168.140.7'server8_ip = '192.168.140.8'struct_of_fault_domains={ 'room1':{ 'carbinet1':{ 'power1_1':[ server1_ip, server2_ip ], 'power1_2':[ server3_ip, server4_ip ] }, 'carbinet2':{ 'power2_1':[ server5_ip, server6_ip ], 'power2_2':[ server7_ip, server8_ip ] } }}'''获取最高级非平凡故障域数量如果当前故障域有多个则故障域非平凡，该故障域为最高级非平凡故障域，直接返回该故障域的值。如果当前故障域只有一个，比如room1，则向下递归，直到找到故障域大于1为止。'''def get_domain_num(fault_domains): if len(fault_domains) > 1: top_domain = fault_domains num_of_domain = len(fault_domains) return top_domain,num_of_domain elif len(fault_domains) == 1: return get_domain_num(fault_domains[list(fault_domains)[0]])'''整数拆分函数将一个整数拆分为N个整数，比如5可以拆分为[1,1,1,1,1]/[1,1,1,2]/[1,2,2]/[1,1,3]/[2,3]'''def SpliteUnit(lens, step, arr, index, results, seg=3): if lens == 0: if seg == index: all_set.append(arr[:index]) results.append(arr[:index]) for i in range(step, lens + 1, 1): arr[index] = i SpliteUnit(lens - i, i, arr, index + 1, results, seg)'''选取方差最小整数拆分方案如果有5个consul server要分布在3个最高级非平凡故障域上则需要将5这个数字拆成三份，那么可能会有[1,1,3]/[1,2,2]两种方案，但是为了将故障的重叠率降到最低，则需要部署consulserver分布在尽量分散的故障域上，如此[1,2,2]会优于[1,1,3]，此时需要利用列表方差值来选取最优方案'''def get_best_set(consul_server_num, seg): num = consul_server_num result = [] tmp_arr = [0] * num global all_set all_set = [] SpliteUnit(num, 1, tmp_arr, 0, result, seg) if len(all_set) > 1: variance = numpy.var(all_set[0]) best = all_set[0] for i in all_set[1:]: if numpy.var(i) < variance: variance = numpy.var(i) best = i else: best = all_set[0] return best#获取consulserver部署节点def get_consul_node(top_domain, domain_num, consul_server_num=2): consul_servers = [] ''' 如果最高级非平凡故障域数量大于consul server数量，则随机选取与consul server数量相等的最高级非平凡故障域。 向下查找直到找到最终的服务器IP，随机抽取其中一个作为consul server ''' if domain_num >= consul_server_num: consul_domains = {} cur_domain = {} if isinstance(top_domain, dict): for i in random.sample(top_domain, consul_server_num): consul_domains[i]=top_domain[i] for key in consul_domains.keys(): cur_domain = consul_domains[key] #@print cur_domain while isinstance(cur_domain, dict): cur_domain = cur_domain[str(random.sample(cur_domain, 1)[0])] consul_server = random.sample(cur_domain, 1)[0] #print consul_server consul_servers.append(consul_server) else: consul_servers = random.sample(top_domain, consul_server_num) return consul_servers ''' 如果最高级非平凡故障域数量小于consul server数量则将consul_server_num进行拆分成与故障域数量相同的数字(调用get_best_set方法获取最优)。 轮询拆分后的数组，如果下级的故障域数量大于等于上级故障域分配的consul数量则上面的代码会执行并返回consul_server节点，否则继续向下递归。 ''' else: consul_sets = get_best_set(consul_server_num, domain_num) i = 0 for key in top_domain.keys(): consul_in_domain = consul_sets[i] cur_domain,subdomain_num = get_domain_num(top_domain[key]) #print cur_domain, subdomain_num, consul_in_domain consul_servers.append(get_consul_node(cur_domain, subdomain_num, consul_in_domain))#向下递归 i = i+1 return consul_servers# else#def get_server_num():# ...#print random.sample(struct_of_fault_domains['room1'], 1)consul_server_num = 2server_num = 51if server_num > 50: consul_server_num = 5else: consul_server_num = 3top_domain,domain_num = get_domain_num(struct_of_fault_domains)print get_consul_node(top_domain,domain_num, consul_server_num)