【SQL】spark sql 不等值 join

products一個商品價格變化的表，orders商品訂單，記錄每次購買商品和日期
基于Spark SQL中的不等值join實現(xiàn)orders和products的匹配，統(tǒng)計每個訂單中商品對應當時的價格

緩慢變化的商品價格表
旺仔牛奶，發(fā)生過一次價格變更

scala> val products = sc.parallelize(Array(
     |   ("旺仔牛奶", "2017-01-01", "2018-01-01", 4),
     |   ("旺仔牛奶", "2018-01-02", "2020-01-01", 5),
     |   ("王老吉", "2017-01-02", "2019-01-01", 5),
     |   ("衛(wèi)龍辣條", "2010-01-01", "2020-01-01", 2)
     | )).toDF("name", "startDate", "endDate", "price")
products: org.apache.spark.sql.DataFrame = [name: string, startDate: string ... 2 more fields]

scala> products.show();
+----+----------+----------+-----+
|name| startDate|   endDate|price|
+----+----------+----------+-----+
|旺仔牛奶|2017-01-01|2018-01-01|    4|
|旺仔牛奶|2018-01-02|2020-01-01|    5|
|   王老吉|2017-01-02|2019-01-01|    5|
|衛(wèi)龍辣條|2010-01-01|2020-01-01|    2|
+----+----------+----------+-----+

訂單表（商品名稱，訂單日期）
旺仔牛奶在不同價格時段分別發(fā)生了一次訂單

scala> val orders = sc.parallelize(Array(
     |   ("2017-06-01", "旺仔牛奶"),
     |   ("2017-07-01", "王老吉"),
     |   ("2018-03-01", "旺仔牛奶")
     | )).toDF("date", "product")
orders: org.apache.spark.sql.DataFrame = [date: string, product: string]

scala> orders.show
+----------+-------+
|      date|product|
+----------+-------+
|2017-06-01|旺仔牛奶|
|2017-07-01|    王老吉|
|2018-03-01|旺仔牛奶|
+----------+-------+

通過不等值連接，計算每個訂單當時的商品價格
查看出旺仔牛奶，兩個訂單在不同時間段上對應的價格

scala> orders.join(products, $"product" === $"name" && $"date" >= $"startDate" && $"date" <= $"endDate").show()
+-----------+------------+----------+------------+-------------+-----+
|     date      |  product   |    name  |   startDate |   endDate   | price|
+-----------+------------+----------+------------+-------------+-----+
|2017-07-01|    王老吉  | 王老吉   |2017-01-02|2019-01-01 |    5  |
|2017-06-01|  旺仔牛奶 |旺仔牛奶|2017-01-01|2018-01-01 |    4  |
|2018-03-01|  旺仔牛奶 |旺仔牛奶|2018-01-02|2020-01-01 |    5  |
+-----------+------------+----------+------------+-------------+-----+