（三）Neo4j自带northwind案例--Cypher语言应用

0、概述

通过该案例，应用Cypher查询语言，感受Neo4j套路。官方的用此案例的用意：

The Northwind Graph demonstrates how to migrate（迁移） from a relational database to Neo4j（把一个负责的多表关系数据库关系等价转化为图数据中，并且查询表现出巨大的优势）. The transformation is iterative and deliberate, emphasizing the conceptual shift from relational tables to the nodes and relationships of a graph.

This guide will show you how to:

Load: create data from external CSV files（加载数据）
Index: index nodes based on label
Relate: transform foreign key references into data relationships
Promote: transform join records into relationships

1、找到位置

:play start

弹出如下界面：

点击中间的“Jump into code”，进入到如下界面：

这里有Movie Graph和Northwind Graph两个案例，这里演示第二个案例。

2、 Load: create data from external products CSV files

LOAD CSV WITH HEADERS FROM "http://data.neo4j.com/northwind/products.csv" AS row
CREATE (n:Product)
SET n = row,
n.unitPrice = toFloat(row.unitPrice),
n.unitsInStock = toInteger(row.unitsInStock), n.unitsOnOrder = toInteger(row.unitsOnOrder),
n.reorderLevel = toInteger(row.reorderLevel), n.discontinued = (row.discontinued <> "0")-------------------------------------------------------------------------
1、LOAD CSV WITH HEADERS FROM "http://data.neo4j.com/northwind/products.csv" AS row return row，这个row到底是什么呢？
{"reorderLevel": "10","unitsInStock": "39","unitPrice": "18.00","supplierID": "1","productID": "1","discontinued": "0","quantityPerUnit": "10 boxes x 20 bags","categoryID": "1","unitsOnOrder": "0","productName": "Chai"
}
{"reorderLevel": "25","unitsInStock": "17","unitPrice": "19.00","supplierID": "1","productID": "2","discontinued": "0","quantityPerUnit": "24 - 12 oz bottles","categoryID": "1","unitsOnOrder": "40","productName": "Chang"
}
.....(后面还有)
2、SET n = row 是建立了77 labels，对应到关系数据库就是用上面的属性结构和数据建立了77条纪律；
3、n.unitPrice = toFloat(row.unitPrice), 有些属性要用Neo4j的字段类型进行强制说明，以免出错；

3、 Load: create data from external categories CSV files

LOAD CSV WITH HEADERS FROM "http://data.neo4j.com/northwind/categories.csv" AS row
CREATE (n:Category)
SET n = row

4、 Load: create data from external suppliers CSV files

LOAD CSV WITH HEADERS FROM "http://data.neo4j.com/northwind/suppliers.csv" AS row
CREATE (n:Supplier)
SET n = row

5、Create indexes

# 以下三个分开执行
CREATE INDEX ON :Product(productID)
CREATE INDEX ON :Category(categoryID)
CREATE INDEX ON :Supplier(supplierID)

6、Create data relationships

# Product和Category建立PART_OF关系，一个Category可以有多个Product，
# 这个是很自然的隶属关系（PART_OF），注意where的这种查询方式
MATCH (p:Product),(c:Category)
WHERE p.categoryID = c.categoryID
CREATE (p)-[:PART_OF]->(c)# 查询所建立的PART_OF关系
MATCH p=()-[r:PART_OF]->() RETURN p LIMIT 80000

形成的图关系如下所示，里面显示的数量可以简单分析一下，所有的Product(77)都找到了隶属的类，一共隶属8个Category(8)，所以下图中共有8组图，共有77＋8=85（nodes）。

# 建立Products和Supplier之间的关系
MATCH (p:Product),(s:Supplier)
WHERE p.supplierID = s.supplierID
CREATE (s)-[:SUPPLIES]->(p)# 查询关系如下图
MATCH p=()-[r:SUPPLIES]->() RETURN p LIMIT 2555

7、Query using patterns

# 查询关联上的nodes
MATCH (s:Supplier)-->(:Product)-->(c:Category)
RETURN s.companyName as Company, collect(distinct c.categoryName) as Categories# collect(distinct c.categoryName)是做了去重处理，一个s:Supplier可以
# 有很多Product，每一个Product只对应一个c:Category，所有一个s:Supplier
# 的两件不同的Product可能对应相同的c:Category

查询结果如下：

这样的查询时不直观的，既然建立了三种节点之间的两种关系，那么在图谱中是什么样子呢？

# 查询PART_OF和SUPPLIES两种关系的节点
MATCH p=()-[r1:PART_OF]-()-[r2:SUPPLIES]-() RETURN p LIMIT 2500

# 一个s:Supplier可以对应多个Product
# 一个Product属于一个categoryName
# 查询可以提供Produce类别Product的s:Supplier，返回结果去重
MATCH (c:Category {categoryName:"Produce"})<--(:Product)<--(s:Supplier)
RETURN DISTINCT s.companyName as ProduceSuppliers

8、 Load: create data from external Customer、Orders CSV files

LOAD CSV WITH HEADERS FROM "http://data.neo4j.com/northwind/customers.csv" AS row
CREATE (n:Customer)
SET n = rowLOAD CSV WITH HEADERS FROM "http://data.neo4j.com/northwind/orders.csv" AS row
CREATE (n:Order)
SET n = rowCREATE INDEX ON :Customer(customerID)CREATE INDEX ON :Order(orderID)

9、Create data relationships

MATCH (c:Customer),(o:Order)
WHERE c.customerID = o.customerID
CREATE (c)-[:PURCHASED]->(o)MATCH p=()-[r:PURCHASED]->() RETURN p LIMIT 25

10、 Load: create data from external order-details CSV files

# 到这一步实体对象就多了，之前有:
# Sublier-[:SUBLIES]->Product-[:PART_OF]->Category
# Customer-[:PURCHASED]->Order
# 上面两种图之间是不联通的，下面还要建立Order-[details:ORDERS]->p:Product)
# 这样所有的实体都联通了，注意多次执行不会覆盖，会建立重名ORDERS关系，本质上是两个系统id
LOAD CSV WITH HEADERS FROM "http://data.neo4j.com/northwind/order-details.csv" AS row
MATCH (p:Product), (o:Order)
WHERE p.productID = row.productID AND o.orderID = row.orderID
CREATE (o)-[details:ORDERS]->(p)
SET details = row,
details.quantity = toInteger(row.quantity)

11、Query using patterns

# 图数据库的匹配是一种结构匹配，而不是属性值匹配
# 譬如(cust:Customer)-[:PURCHASED]->(:Order)-[o:ORDERS]->(p:Product)-
# [:PART_OF]->(c:Category {categoryName:"Produce"})就是一种结构
# 下面的意思是说：找到买Produce类的Customer的名字，并且计算该用户订单上
# 所有Product价格的总和，这在电商中是非常有意义的
# 和关系数据库相比，这就体现了图数据库多个表链接查询的重大优势
MATCH (cust:Customer)-[:PURCHASED]->(:Order)-[o:ORDERS]->(p:Product)-[:PART_OF]->(c:Category {categoryName:"Produce"})
RETURN DISTINCT cust.contactName as CustomerName, SUM(o.quantity) AS TotalProductsPurchased# 官网给出的是如下的方式，和上面等价的
MATCH (cust:Customer)-[:PURCHASED]->(:Order)-[o:ORDERS]->(p:Product),(p)-[:PART_OF]->(c:Category {categoryName:"Produce"})
RETURN DISTINCT cust.contactName as CustomerName, SUM(o.quantity) AS TotalProductsPurchased

12、我们的图谱长什么样子呢？

# 图关系要拆分成具有线性关系的r1,r2，链接起立展示
MATCH r1=(cust:Customer)-[:PURCHASED]->(:Order)-[o:ORDERS]->(p:Product),r2=(sup:Supplier)-[:SUPPLIES]-(p)-[:PART_OF]->(c:Category {categoryName:"Produce"})
RETURN r1, r2 limit 1