neo4j+python（2）

导入csv文件

文件位置
–PyNeo
-----data
--------math.csv
-----run.py

代码

# -*- coding:utf-8 -*-
# 读取csv文件的三元组

import os
import csv
import codecs
from py2neo import Node,Relationship,Graph,NodeMatcher

data_path = "./data"
file_name = 'math.csv'
file = os.path.join(data_path, file_name)

def load_csv(file):
    data_list = []
    with codecs.open(file, encoding='ANSI') as f:
        reader = csv.DictReader(f)
        for line in reader:
            data = list(line.values())
            data_list.append(data)
        return data_list

tri_list = load_csv(file) # [ [head，relation，tail],[head，relation，tail],[head，relation，tail]...]

head_entities = []
relations = []
tail_entities = []

for s in tri_list:
    head_entities.append(s[0])
    relations.append(s[1])
    tail_entities.append(s[2])

graph = Graph('http://localhost:7474', username='neo4j', password='0208')
graph.delete_all()
matcher = NodeMatcher(graph)

entities = set(head_entities + tail_entities)
for n in entities:
    graph.create(Node('Concept', name = n))  # 创建所有实体节点

for i in range(len(relations)):
    head = head_entities[i]
    head_node = matcher.match('Concept').where("_.name="+"'"+head+"'").first() #去掉重复节点

    tail = tail_entities[i]
    tail_node = matcher.match('Concept').where("_.name="+"'"+tail+"'").first()

    relation = relations[i]
    graph.create(Relationship(head_node, relation, tail_node))  #创建关系

最大收获：

head_node = matcher.match('Concept').where("_.name="+"'"+head+"'").first()

可以去掉重复节点，参考：https://blog.csdn.net/zsh1184528359/article/details/106003346/

其他注意事项写在前一篇了