[
https://issues.apache.org/jira/browse/PDFBOX-2225?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14066787#comment-14066787
]
Tilman Hausherr commented on PDFBOX-2225:
-----------------------------------------
This is indeed a weird piece of code. Would this make more sense?
{code}
catch (Exception e)
{
// if we are not ignoring exceptions, we'll re-throw this
if (!ignoreAcroFormErrors && e instanceof IOException)
{
throw new IOException(e);
}
}
{code}
This is what the author intended, although it isn't helpful either, as it would
swallow all language related exceptions, e.g. the mentioned
IndexOutOfBoundsException.
> ClassCastException in PDFMergerUtility.appendDocument
> -----------------------------------------------------
>
> Key: PDFBOX-2225
> URL: https://issues.apache.org/jira/browse/PDFBOX-2225
> Project: PDFBox
> Issue Type: Bug
> Components: Utilities
> Affects Versions: 1.8.6
> Environment: Windows 7 64-bit, JDK8, pdfbox 1.8.1 or 1.8.6
> Reporter: Brandon Lyon
> Priority: Minor
>
> Edit: Problem confirmed to exist in 1.8.1 and 1.8.6
> The following exception will occur if any type of Exception other than
> IOException is thrown within a particular try-catch in
> PDFMergerUtility.appendDocument:
> java.lang.ClassCastException: java.lang.IndexOutOfBoundsException cannot be
> cast to java.io.IOException
> For the version I am using (1.8.1), the problem occurs on line 294
> The problem is that the author mistakenly assumed that the only exception
> which could be caught was IOException. I recommend that the cast be removed,
> and that either the additional checked exceptions be thrown by appendDocument
> or wrapped with an unchecked exception
> Also, anyone seeing this may want to investigate why an
> IndexOutOfBoundsException is occurring here, though this is obviously a
> separate issue
--
This message was sent by Atlassian JIRA
(v6.2#6252)
