diff --git a/bin_search.md b/bin_search.md
index 610be05..662840e 100644
--- a/bin_search.md
+++ b/bin_search.md
@@ -33,7 +33,7 @@ list.index()无法应对大规模数据的查询，需要用其它方法解决
 	def binarySearch(lst, value,low,high):          #low,high是lst的查找范围
 	    if high < low:
 	        return -1
-	    mid = (low + high)/2
+	    mid = int((low + high)/2)
 	    if lst[mid] > value:
 	        return binarySearch(lst, value, low, mid-1)
 	    elif lst[mid] < value:
@@ -46,7 +46,7 @@ list.index()无法应对大规模数据的查询，需要用其它方法解决
 	def bsearch(l, value):
 	    lo, hi = 0, len(l)-1
 	    while lo <= hi:
-	        mid = (lo + hi) / 2
+	        mid = int((lo + hi) / 2)
 	        if l[mid] < value:
 	            lo = mid + 1
 	        elif value < l[mid]:
@@ -57,8 +57,8 @@ list.index()无法应对大规模数据的查询，需要用其它方法解决
 	 
 	if __name__ == '__main__':
 	    l = range(50)
-	    print binarySearch(l,10,0,49)
-	    print bsearch(l,10)
+	    print(binarySearch(l,10,0,49))
+	    print(bsearch(l,10))
 
 对于python，不能忽视其强大的标准库。经查阅，发现标准库中就有一个模块，名为：bisect。其文档中有这样一句话：
 
diff --git a/test.py b/test.py
new file mode 100644
index 0000000..9a23168
--- /dev/null
+++ b/test.py
@@ -0,0 +1,43 @@
+def get_kNN_result(name):
+    # Analyze start
+    new_list = []
+    new_vector = []
+    cos_list = []
+    result = []
+    # Chinese participle
+    new_cut = jieba.cut_for_search(name)
+    # Generate new words list
+    for cut in new_cut:
+        if not (cut.strip() in stop_list):
+            new_list.append(cut)
+    # Generate vector by mapping
+    for c_name in corpus_names:
+        if not (c_name[0].strip() in new_list):
+            new_vector.append(0)
+        else:
+            new_vector.append(1)
+    # Get the list of Cosine Similarity
+    for fid, vector in failure_names:
+        cos = get_cos_similar(new_vector, list(eval(vector)))
+        cos_list.append((cos, fid))
+    # Sort the list of last step and get top5
+    k_nn5 = sorted(cos_list)[-5:]
+    k_nn5.reverse()  # big to small
+    # Get id and similarity of top5
+    case_collect5 = []
+    for cos, fid in k_nn5:
+        case_collect5.append(fid)
+    print(case_collect5)
+    # Query the top5 cases
+    cursor.execute('select id,name,industry,mode from failurecase where id in ' + str(tuple(case_collect5)))
+    value_cases = cursor.fetchall()
+    print(value_cases)
+    # Sort the list(value_cases)
+    case_dict = {}
+    for item2 in value_cases:
+        case_dict[item2[0]] = {'name': item2[1], 'industry': item2[2], 'mode': item2[3]}
+    for item1 in case_collect5:
+        dict_item = case_dict[item1]
+        dict_item['id'] = item1
+        result.append(dict_item)
+    return result