Kasper Sørensen created METAMODEL-32:
----------------------------------------
Summary: Excel module: XlsxSpreadsheetReaderDelegate is not thread
safe
Key: METAMODEL-32
URL: https://issues.apache.org/jira/browse/METAMODEL-32
Project: Metamodel
Issue Type: Bug
Reporter: Kasper Sørensen
XlsxSpreadsheetReaderDelegate uses a HashMap to store references to table names
and internal IDs. It may occur that a new table is created, it's reference
added to this HashMap, and then queried shortly after by another thread. But
since HashMaps are not threadsafe or volatile, it may occur that the other
thread will get a dirty write which will result in a stack trace like this
because of a null reference passed on:
{code}
java.lang.NullPointerException: null
at java.util.TreeMap.getEntry(TreeMap.java:342) ~[na:1.7.0_45]
at java.util.TreeMap.get(TreeMap.java:273) ~[na:1.7.0_45]
at
org.apache.poi.openxml4j.opc.PackageRelationshipCollection.getRelationshipByID(PackageRelationshipCollection.java:282)
~[poi-ooxml-3.8.jar:3.8]
at
org.apache.poi.openxml4j.opc.PackagePart.getRelationship(PackagePart.java:379)
~[poi-ooxml-3.8.jar:3.8]
at org.apache.poi.xssf.eventusermodel.XSSFReader.getSheet(XSSFReader.java:139)
~[poi-ooxml-3.8.jar:3.8]
at
org.eobjects.metamodel.excel.XlsxRowPublisherAction.run(XlsxRowPublisherAction.java:56)
~[MetaModel-excel-3.4.7.jar:na]
at
org.eobjects.metamodel.excel.XlsxRowPublisherAction.run(XlsxRowPublisherAction.java:36)
~[MetaModel-excel-3.4.7.jar:na]
at
org.eobjects.metamodel.data.RowPublisherDataSet$1.run(RowPublisherDataSet.java:88)
~[MetaModel-core-3.4.7.jar:na]
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:471)
~[na:1.7.0_45]
at java.util.concurrent.FutureTask.run(FutureTask.java:262) ~[na:1.7.0_45]
at
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
[na:1.7.0_45]
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
[na:1.7.0_45]
at java.lang.Thread.run(Thread.java:744) [na:1.7.0_45]
{code}
I suggest simply replacing the HashMap with a ConcurrentHashMap, since that
will fix the issue. Furthermore we should check the used ID for null, so that
we fail fast and avoid this stack which initially got us thinking it was an
Apache POI issue.
--
This message was sent by Atlassian JIRA
(v6.1#6144)
