Json 配置单元：如何将多个数组的结果放入一个数组结果集中

我有一个配置单元表，用于存储以下示例定义的json对象：

{bpy:[{bpyNr:string，anr:string}]

我加载了3个文件，如下所示：

文件1

{"bpy":[
{"bpyNr":"","anr":"BESHPR","al1Te":"123 BELWAFFLE","al2Te":"ADDR LINE2","al3Te":"ADDR LINE3","ctcNa":"FROMBE LAGO","cnyCd":"ES","cnySdvCd":"","divNa":"BRUSSELS","emlTe":"","faxNr":"","munNa":"","na":"BELGUIM SHIPPER","phnNr":"18889991234","pslCd":"1000","bpyRlCd":"01"}]}

文件2：

{"bpy": [
{"bpyNr":"","anr":"00005483WW","al1Te":"14 RICH WAY","al2Te":"REC ADDR2","al3Te":"REC ADDR3","ctcNa":"IAMRECEIVER","cnyCd":"ES","cnySdvCd":"","divNa":"PORT ELIZABETH","emlTe":"CONS-EMAIL@YAHOO.NET","faxNr":"7747747774","munNa":"","na":"AFRICAN SHIPPERS","phnNr":"2018289999","pslCd":"6001","bpyRlCd":"02"},   {"bpyNr":"","anr":"UPS11111","al1Te":"","al2Te":"","al3Te":"3","ctcNa":"","cnyCd":"US","cnySdvCd":"","divNa":"","emlTe":"","faxNr":"","munNa":"","na":"THIRD PARTY SHIPPXX","phnNr":"2018289999","pslCd":"","bpyRlCd":"05"}
]}

它们被加载到一个蜂箱外部表businessparty\u tbl 当我查询表时，我得到这个结果

hive> select bpy.anr from businessparty_tbl;
OK

NULL
NULL
NULL
NULL
["BESHPR"]
["00005483WW","UPS11111"]

有没有办法得到这样的结果：

[NULL, NULL, NULL, NULL, "BESHPR", "00005483WW","UPS11111"]

更新：增加了业务方定义

ADD JAR /opt/cloudera/parcels/CDH/jars/hive-hcatalog-core-1.1.0-cdh5.4.7.jar;

DROP TABLE IF EXISTS brokerage.businessparty_tbl;

CREATE EXTERNAL TABLE IF NOT EXISTS brokerage.businessparty_tbl (

  bpy  array<struct<bpyNr    :STRING COMMENT      "businessPartyNumber",
    anr         :STRING COMMENT      "accountNumber",
    al1Te       :STRING COMMENT      "addressLine1Text",
    al2Te       :STRING COMMENT      "addressLine2Text",
    al3Te       :STRING COMMENT      "addressLine3Text",
    crrCd       :STRING COMMENT      "carrierCode",
    ctcEmlAdTe  :STRING COMMENT      "contactEmailAddressText",
    ctcFaxLinNr :STRING COMMENT      "contactFaxLineNumber",
    ctcNa       :STRING COMMENT      "contactName",
    ctcPhnNr    :STRING COMMENT      "contactPhoneNumber",
    ctcDepNa    :STRING COMMENT      "contactDepartmentName",
    ctcDivNa    :STRING COMMENT      "contactDivisionName",
    cnyCd       :STRING COMMENT      "countryCode",
    cnySdvCd    :STRING COMMENT      "countrySubdivisionCode",
    dunsNr      :STRING COMMENT      "dunsNumber",
    depNa       :STRING COMMENT      "departmentName",
    divNa       :STRING COMMENT      "divisionName",
    emlTe       :STRING COMMENT      "emailText",
    faxNr       :STRING COMMENT      "faxNumber",
    munNa       :STRING COMMENT      "municipalityName",
    na          :STRING COMMENT      "name",
    phnNr       :STRING COMMENT      "telephoneNumber",
    pslCd       :STRING COMMENT      "postalCode",
    sutTe       :STRING COMMENT      "suiteText",
    crrNa       :STRING COMMENT      "carrierName",
    shrCrCrdNr  :STRING COMMENT      "shipperCreditCardNumber",
    shrCrCrdExpDt       :STRING COMMENT      "shipperCreditCardExpirationDate" 
  >> 
) 
        ROW FORMAT SERDE 'org.apache.hive.hcatalog.data.JsonSerDe'; 
        load data local inpath '/data/5/sys1deh/brokerage/testdata/BAS/SHIPMENTS/' OVERWRITE into table brokerage.businessparty_tbl;

添加JAR/opt/cloudera/parcels/CDH/jars/hive-hcatalog-core-1.1.0-cdh5.4.7.JAR；
如果存在brokerage.businessparty\u tbl，则删除表格；
如果不存在，则创建外部表brokerage.businessparty\u tbl(
bpy阵列
) 
行格式SERDE'org.apache.hive.hcatalog.data.JsonSerDe'；
将路径“/data/5/sys1deh/broker/testdata/BAS/shippings/”中的本地数据加载到表brokerage.businessparty\u tbl中；

share businessparty\u tbl table DDL statement也请？@squid，我在原始帖子中添加了定义。谢谢！