article_list=open('topfourcites.txt').read()
article_list=article_list.replace('\n   ','\t')
article_list=article_list.split('\n')
cite_list=[ab[3:] for ab in article_list if ab[:2]=="CR"]
cite_list=[[c for  c in cite.split('\t')] for cite in cite_list]




cite_dict={}
edge_dict={}

for cites in cite_list:
    cite_list=[]
    for cite in cites:
        split=cite.split(', ')
        try:
            id=split[0].upper().replace(' ',' ').replace('.','')+' '+split[1]
            ids=id.split()
            for i in ids:
                if len(i)>2:
                    new=i[0].upper()+i[1:].lower() 
                    id=id.replace(i,new)
        except: 
            print 'Eror with ',split
        else:
            
            if id not in cite_list:
                if id in cite_dict:
                    cite_dict[id]=cite_dict[id]+1
                else:
                    cite_dict[id]=1
                    
                if len(cite_list)>0:
                    for cite in cite_list:
                        if (id,cite) in edge_dict:
                            edge_dict[(id,cite)]=edge_dict[(id,cite)]+1
                        elif (cite,id) in edge_dict:
                            edge_dict[(cite,id)]=edge_dict[(cite,id)]+1
                        else:
                            edge_dict[(id,cite)]=1
                cite_list.append(id)
                print id
'''

import csv
writer=csv.writer(open('cite_edge_list.csv','wb'))

writer.writerow(('Source','Target','Weight','Type'))
for edge in edge_dict:
    if edge_dict[edge]>2 and cite_dict[edge[0]]>9 and cite_dict[edge[1]]>9 :
        writer.writerow((edge[0],edge[1],edge_dict[edge],'Undirected'))
'''
import networkx as nx
from networkx.readwrite import d3_js
import re
G=nx.Graph()
counter=0
for edge in edge_dict:
    if edge_dict[edge]>3 and cite_dict[edge[0]]>=8 and cite_dict[edge[1]]>=8 :
        G.add_edge(edge[0],edge[1],weight=edge_dict[edge])
        counter=counter+1
        print counter
        
for node in G:    
    G.add_node(node,freq=cite_dict[node])

import community

partition=community.best_partition(G)

print len(partition)

for node in G:    
    G.add_node(node,freq=cite_dict[node], group=str(partition[node]))

d3_js.export_d3_js(G,files_dir="netweb",graphname="cites",node_labels=True,group="group")

fix=open('netweb/cites.json','rb').read()
for n in G:
    try:
        fix=re.sub(str(n)+'''"''',str(n)+'''" , "nodeSize":'''+str(cite_dict[n]),fix)
    except:
        print 'error with',n
f = open('netweb/cites.json', 'w+')
f.write(fix)
f.close()