Support #5774
closed
Xerces bug: java.lang.ArrayIndexOutOfBoundsException when reading a document with doc() function
0%
Description
We use Saxon-EE 11.4 to perform a large number of transformations every day via a batch process. Normally, this happens smoothly until we recently bumped into a Xerces bug (java.lang.ArrayIndexOutOfBoundsException) when reading a file with the doc() function. A simple restart of the process solved the problem. The full error message can be found in attachment.
Files
Updated by Michael Kay almost 2 years ago
Strange, but I don't think there's much we can do about it, as it's not our product. Especially if it's transient.
Updated by Johan Gheys almost 2 years ago
Indeed. Apparently, this is a long-known Xerces bug that is still unresolved (see e.g. [[https://stackoverflow.com/questions/1919687/arrayindexoutofboundsexception-in-xerces-parsing]] ).
Updated by Michael Kay almost 2 years ago
Is this Apache Xerces or the version of Xerces in the JDK?
We used to recommend Apache as being the more reliable, but I think this is no longer the case; I get the impression bugs are being fixed in the JDK version but not in Apache.
Updated by Johan Gheys almost 2 years ago
Sorry for the confusion, but you are right because our bug has more to do with the version of Xerces in the JDK I suppose.
Updated by Johan Gheys almost 2 years ago
We have scheduled to migrate to IBM's latest JDK version on our servers. This ticket may therefore be closed.
Please register to edit this issue
.