结合使用saiku、mondrian workbentch建立多维查询报表

1、简介

前几篇博客已经介绍了saiku、mondrian、MDX和模式文件他们之间的关系，那么如何将它们串联起来，供产品人员使用哪？下面咱们一步一步的实现

2、建立数据表

建表语句参考：http://www.cnblogs.com/liqiu/p/5202708.html

模式图如下：

这是一个典型的事实表＋维度表的星型数据结构！

3、建立模式文件

这里就不说具体细节啦，直接给出xml文件内容，感兴趣的同学可以直接载入进去即可

<Schema name="SaleSchema">
  <Cube name="SalesCube" visible="true" cache="true" enabled="true">
    <Table name="sale" schema="public">
    </Table>
    <Dimension type="StandardDimension" visible="true" foreignKey="cusid" highCardinality="false" name="dimCustomer">
      <Hierarchy visible="true" hasAll="true" allMemberName="allCustomer" primaryKey="cusid">
        <Table name="customer" schema="public">
        </Table>
        <Level name="gender" visible="true" column="gender" type="String" uniqueMembers="true" levelType="Regular" hideMemberIf="Never">
        </Level>
      </Hierarchy>
    </Dimension>
    <Dimension type="StandardDimension" visible="true" foreignKey="proid" highCardinality="false" name="dimProductType">
      <Hierarchy visible="true" hasAll="true" allMemberName="allProduct" allMemberCaption="allProductTitle" primaryKey="proid" primaryKeyTable="product">
        <Join leftKey="proid" rightKey="protypeid">
          <Table name="product" schema="public">
          </Table>
          <Table name="producttype" schema="public">
          </Table>
        </Join>
        <Level name="productId" visible="true" table="product" column="proid" nameColumn="proname" type="String" uniqueMembers="true" levelType="Regular" hideMemberIf="Never">
        </Level>
        <Level name="productTypeId" visible="true" table="producttype" column="protypeid" nameColumn="protypename" type="String" uniqueMembers="true" levelType="Regular" hideMemberIf="Never">
        </Level>
      </Hierarchy>
    </Dimension>
    <Measure name="saleNumber" column="num" aggregator="sum" caption="NumTitle" visible="true">
    </Measure>
    <Measure name="saleAmount" aggregator="sum" caption="allSaleTitle" visible="true">
      <MeasureExpression>
        <SQL dialect="generic">
          <![CDATA[(unitprice*num)]]>
        </SQL>
      </MeasureExpression>
    </Measure>
    <CalculatedMember name="avgPrice" caption="avgPriceTitle" dimension="Measures" visible="true">
      <Formula>
        <![CDATA[[Measures].saleAmount/[Measures].saleNumber]]>
      </Formula>
      <CalculatedMemberProperty name="formatString" value="$#,##0.00">
      </CalculatedMemberProperty>
    </CalculatedMember>
  </Cube>
</Schema>

4、查询数据

写如下MDX语句验证数据准确性

select
       {[Measures].saleNumber,[Measures].saleAmount,[Measures].avgPrice}
on columns,
       {([dimProductType].[allProduct],[dimCustomer].[allCustomer])}
on rows
from [salesCube]