def geneticoptimize(self):
mm = sensitive_path_info.build_m()
# 初始化种群
popsize = 30
pc = 0.85
pm = 0.1
datapop = []
dataM = []
searchFlag = 0
for i in range(popsize):
datapop.append(self.Initial_data(self.varType))
#评估初始种群
for data in datapop:
seq_to_script.runcase(self.tevent, self.pathT, data) # 执行当前序列
test_fit, coverage = cal_data_fit.array_spath(
) # 获取对各条路径的覆盖情况 test_fit，及完全覆盖路径的覆盖率
if test_fit[self.objectPathIndex] == 0.0:
print "满足结果的测试数据为", zip(self.pathT, data)
searchFlag = 1
return searchFlag, data
else:
if len(test_fit) == 0: # 构造覆盖矩阵dataM
dataM.append(mm)
else:
dataM.append(test_fit)
iteration = 0
while iteration <= 500:
temp = copy.deepcopy(datapop) # 保留父代
oldM = dataM
# 选择
selectpop = self.selection(datapop, dataM)
# 交叉
crosspop = self.crossover(selectpop, pc)
# 变异
mutanpop = self.mutantion(crosspop, pm)
# 评估新种群
dataM = []
for data in mutanpop:
seq_to_script.runcase(self.tevent, self.pathT, data) # 执行当前序列
test_fit, coverage = cal_data_fit.array_spath(
) # 获取对各条路径的覆盖情况 test_fit，及完全覆盖路径的覆盖率
if test_fit[self.objectPathIndex] == 0.0:
print "满足结果的测试数据为", data
searchFlag = 1
return searchFlag, zip(self.pathT, data)
else:
if len(test_fit) == 0: # 构造覆盖矩阵dataM
dataM.append(mm)
else:
dataM.append(test_fit)
# 更新
datapop = self.updata(temp, mutanpop, oldM, dataM, popsize)
iteration += 1
def array_spath():
pbdict, pcdict = hand_instru() # schoolmate,faqforge
# pbdict, pcdict = hand_instru_webchess() #webchess
path_fit = judge_new(pcdict, pbdict)
spath = sensitive_path_info.obtain_spath()
m = sensitive_path_info.build_m()
# print "m",m
for key in path_fit:
# print "path_fit_key",key
index = spath.index(key)
# print "index in spath"
# print index
m[index] = path_fit[key]
t= m.count(0.0)
covrage = t / len(m)
print m,t,"个体覆盖率",covrage
return m,covrage #返回该个体对各条路径的覆盖情况matrix，及完全覆盖路径的覆盖率
#encoding:utf-8
from __future__ import division
import recordFun
import generate_data_ga #数据生成器-----序列上的变量产生数据
import sensitive_path_info
mm = sensitive_path_info.build_m()
# uncp = [2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2]
def table_handle(M):
# 所有敏感路径入口点检测：pot = 1已完全覆盖所有的入口点；pot =0为未完全覆盖
# 方法：查看table中是不是至少每一列都有一个在[0,1]之间的个体
# 所有敏感路径完全覆盖检测：flag = 1为完全覆盖；flag=0为未完全覆盖。
# 方法：查看二维表中的0是否已全覆盖上所有的敏感路径,每一列至少有一个0
count = 0
count1 = 0
flag = 0
pot = 0
uncoverp = []
potlist = []
statistic = []
print "M ==",len(M)
for i in range(len(M)):
# print len(M[i])
print M[i]
for k in range(len(M[0])):
# print "第",k,"条路径"
for j in range(len(M)):
# print "第", j, "个测试用例的fit值",M[j][k]
# print "table[j][k]", table[j][k]
if M[j][k] == 0.0: #table里面的fit值都是浮点数
count = count + 1
if M[j][k] >=0.0 and M[j][k] < 2: # 这里实际上是<=2,cal_data_fit时，忘记归一化了，先这样