# coding: utf-8

# In[60]:

import pymysql
pymysql.install_as_MySQLdb()
import MySQLdb
import sys
import math
def similarityRelation(c,n):
    c = list(c)
    c = np.array(c,dtype=np.int8)+1
    union = sum(c)
    mul =1;
    for i in c:
        mul *= i
    result = math.log((n*union)/(mul))/math.log(n)
    return result


# In[61]:
host = "localhost"
user = "0187cs151057"
password = "noor@896247"
db = "0187cs151057"

conn = MySQLdb.connect(host,user,password,db)

cursor = conn.cursor()

#searchString = "HACKVEDA ONE2ONE"
searchString = " ".join(sys.argv[1:])

#sql = "select * from similarityrelation where searchString like '%s'" %(searchString)
sql = "select * from similarityrelation where searchString like '%s'" %(searchString)     ###For command line argument

cursor.execute(sql)

result = cursor.fetchall()

sql = "desc similarityrelation"
cursor.execute(sql)
name = cursor.fetchall()

head = []
for i in name:
    head.append(i[0])
import pandas as pd
import numpy as np

df = pd.DataFrame(np.array(result),columns=head);


# In[62]:

#sql = "select id,link from query where searchString like '%s'" %(searchString)
sql = "select id,link from query where searchString like '%s'" %(searchString)

cursor.execute(sql)

result = cursor.fetchall()
sql = "desc query"
cursor.execute(sql)
name = cursor.fetchall()

head = []
for i in name:
    head.append(i[0])

df4 = pd.DataFrame(np.array(result),columns=["id","link"]);


# In[63]:

searchEngineValue = df["searchEngineValue"].unique()

df3 = pd.DataFrame(columns=["searchEngineValue","searchId","similarity"])

for engineValue in searchEngineValue:
    df1 = df[:][df["searchEngineValue"]==engineValue]

    searchId = df1["searchId"].unique()
    n = len(searchId)
    st = []
    ss = []
    su = []
    for i in searchId:
        c = df1["titleCount"][df1["searchId"] == i]
        st.append(similarityRelation(c,n))
        c = df1["summaryCount"][df1["searchId"] == i]
        ss.append(similarityRelation(c,n))
        c = df1["linkCount"][df1["searchId"] == i]
        su.append(similarityRelation(c,n))
    st = np.array(st)
    ss = np.array(ss)
    su = np.array(su)
    s = st+su+ss
    d = pd.Series(s)
    a = len(d.unique())
    b = len(searchId)
    s = pd.Series(s, dtype=np.float64)
    dict1 = {'searchId':searchId, "similarity":s,"searchEngineValue":engineValue}
    df2 =pd.DataFrame(dict1, dtype=np.float32)
    df3 = df3.append(df2,ignore_index=True)


# In[ ]:

df3 = df3.join(df4)


# In[85]:



# In[86]:

df5 = df3.drop_duplicates("link")
df5 = df5.sort_values("similarity", ascending = False)
#print(df5["link"])
for link in df5["link"]:
    print("<a href=\"" + link + "\">" + link + "</a><br >")