我正试图用pig和twitter的 elephant-bird 库解析一个相当简单的json文件，但它变成了一个非常痛苦的调试过程。
json具有以下结构：

oid_id: (oid:chararray), 
bookmarks: {(
  oid_id:(oid:chararray),
  id:chararray,
  creator: chararray,
  position:chararray,
  creationdate:($ate:chararray)
  )},
lastaction:(date:chararray),
settings:(preferredlanguage:chararray),
userid:chararray

行的示例：
{“oid\u id”：{“oid”：“573239f905474a686e2333f0”}，“bookmarks”：[{“id”：“legoninx106w0079264”，“creator”：“player”，“position”：96，“creationdate”：{“date”：“2016-12-26t09:37:36.916z”}，“oid id”：{“oid”：“5860e4e0ca6baf9032edc0d0”}，“id”：“onepercentmw0128677”，“creator”：“player”，“position”：0.08，“creationdate”：{“date”：“2018-12-18t15:42:33.956z”}，“oid\u id”：{“oid”：“5c191569faf8474953758930”}]，“lastaction”：{“date”：“2018-12-18t15:42:28.107z”}，“settings”：{“preferredlanguage”：“vf”，“preferredvideoquality”：“hd”}，“userid”：“ocs\u 32a6ad6dd242d5e3842f9211fd236723\u 1461773211”}
以下是我的代码（受本教程启发：https://acadgild.com/blog/determining-popular-hashtags-in-twitter-using-pig)

register /path/to/json-simple-1.1.1.jar 
register /path/to/elephant-bird-core-4.17.jar
register /path/to/elephant-bird-pig-4.17.jar
register /path/to/elephant-bird-hadoop-compat-4.17.jar
define JsonLoaderEB com.twitter.elephantbird.pig.load.JsonLoader;
A = LOAD 'file.json' USING JsonLoaderEB('-nestedLoad=true') as myMap;
describe A;

输入表：{mymap:bytearray}

B = foreach A generate flatten(myMap#'bookmarks') as (bookmark:map[]);
describe B;

b:{书签：Map[]}
当我们转储上述关系时，我们可以看到所有数据都已成功加载。
（[{“oid\u id”：{“oid”：“5860e4e0ca6baf9032edc0d0”}，“creator”：“player”，“creationdate”：{“date”：“2016-12-26t09:37:36.916z”}，“id”：“legoninx106w0079264”，“position”：96}，“oid id”：{“5c191569faf8475538930”}，“creator”：“player”，“creationdate”：{“date”：“2018-12-18t15:42:33.956z”}，“id”：“onepercentmw0128677”，“position”：0.08}]）
现在我们从书签中提取creationdate、creator、id和position。

C = foreach B generate bookmark#'creationdate' as date_fact, bookmark#'creator' as creator, bookmark#'id' as id, bookmark#'position' as position;

c:{日期{fact:bytearray，创建者：bytearray，id:bytearray，位置：bytearray}
转储表会产生以下错误：

清管器堆放痕迹

错误1066:无法打开别名c的迭代器。后端错误：vertex失败，vertexname=scope-41，vertexid=vertex\u 1542613138136\u 6721 88\u 2\u 00，diagnostics=[任务失败，taskid=task\u 1542613138136\u 672188\u 2\u 00\u000000，diagnostics=[任务尝试0失败，info=[错误：运行任务时出错（失败）：尝试\u 1542613138136 \u 672188 \u 2 \u 00 \u000000 \u 0:org.apache.pig.backend.executionengine.executeexception:错误0:执行时异常（名称：c:store）(hdfs://sandbox/tmp/temp-1543074195/tmp277240455：org.apache.pig.impl.io.interstorage）-sc ope-40运算符键：scope-40）：org.apache.pig.backend.executionengine.executexception:错误0:执行[pomaplookup（名称：pomaplookup[bytearray]-scope-28运算符键：scope-28]时出现异常子项：null位于[null[4，31]]]：java.lang.classcastexception:java.lan g.string不能强制转换为org.apache.pig.backend.hadoop.executionengine.physicallayer.physicaloperator.processinput（physicaloperator）上的java.util.map。java:315)在org.apache.pig.backend.hadoop.executionengine.tez.plan.operator.postoretez.getnexttuple（postoretez。java:123)在org.apache.pig.backend.hadoop.executionengine.tez.runtime.pigprocessor.runpipeline（pigprocessor。java:376)在org.apache.pig.backend.hadoop.executionengine.tez.runtime.pigprocessor.run（pigprocessor。java:241)位于org.apache.tez.runtime.logicalioprocessorruntimetask.run（logicalioprocessorruntimetask）。java:370)在org.apache.tez.runtime.task.taskrunner2callable$1.run（taskrunner2callable。java:73)在org.apache.tez.runtime.task.taskrunner2callable$1.run（taskrunner2callable。java:61)位于javax.security.auth.subject.doas（subject）的java.security.accesscontroller.doprivileged（本机方法）。java:422)在org.apache.hadoop.security.usergroupinformation.doas（用户组信息。java:1698)在org.apache.tez.runtime.task.taskrunner2callable.callinternal（taskrunner2callable。java:61)在org.apache.tez.runtime.task.taskrunner2callable.callinternal（taskrunner2callable。java:37)在org.apache.tez.common.callablewithndc.call（callablewithndc。java:36)在java.util.concurrent.futuretask.run（futuretask。java:266)位于java.util.concurrent.threadpoolexecutor.runworker（threadpoolexecutor。java:1149)在java.util.concurrent.threadpoolexecutor$worker.run（threadpoolexecutor。java:624)在java.lang.thread.run（线程。java:748)原因：org.apache.pig.backend.executionengine.executexception:错误0:异常执行[pomaplookup（name:pomaplookup[byt earray]-scope-28 operator key:scope-28）时，children:在[null[4]处为null，31]]]：java.lang.classcastexception:java.lang.string不能在org.apache.pig.backend.hadoop.executionengine.physicallayer.physicaloperator.getnext（physicaloperator）上转换为java.util.map。java:364)在org.apache.pig.backend.hadoop.executionengine.physicallayer.relationaloperators.poforeach.processplan（poforeach。java:406)在org.apache.pig.backend.hadoop.executionengine.physicallayer.relationaloperators.poforeach.getnexttuple（poforeach）。java:323)在org.apache.pig.backend.hadoop.executionengine.physicallayer.physicaloperator.processinput（physicaloperator。java:305)1,9顶部