[ 
https://issues.apache.org/jira/browse/PDFBOX-3388?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15347764#comment-15347764
 ] 

Ankit Agarwal commented on PDFBOX-3388:
---------------------------------------

Yes there are instances of scratch file buffer still in the heap.
I saw the code for finalize method in that class, should i close the object
there just before the debug warning, or is it used somewhere else??(I don't
think that's possible as I want to close in the finalize method)
But if a reference to the object remains in memory then gc won't know to
call finalize and hence my code won't close the reference . Am I right?
Please post a workaround this.
I am sorry I am unable to post the code. But there are instances of
finalizer telated to scratchBufferClass in my heap dump




> PDFTextStripper - ScratchFileBuffer not closed!
> -----------------------------------------------
>
>                 Key: PDFBOX-3388
>                 URL: https://issues.apache.org/jira/browse/PDFBOX-3388
>             Project: PDFBox
>          Issue Type: Bug
>            Reporter: Roman Pichlik
>         Attachments: CloseablePDFParser.java, PDFStripperTest.java, test.pdf
>
>
> _PDFTextStripper_ or inherently used classes probably do not close all opened 
> streams under all circumstances. You can reproduce that by the following 
> snippet of code and the attached PDF file.
> {code}
> try (RandomAccessBuffer rab = new RandomAccessBuffer(is)) {
>             PDFParser parser = new PDFParser(rab);
>             parser.parse();
>             try (COSDocument cosDoc = parser.getDocument();PDDocument pdDoc = 
> new PDDocument(cosDoc);){
>                 PDFTextStripper pdfStripper = new PDFTextStripper();
>                 pdfStripper.getText(pdDoc);
>             }
>         } catch (IOException e) {
>             throw new RuntimeException(e);
>         }
> {code}



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to