You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@pdfbox.apache.org by Damien Butaye <da...@gmail.com> on 2016/05/24 13:51:32 UTC
PDFBox*.tmp files not deleted by PDFParser
Dear all,
I'm trying to add a signature to a PDF using PDFBOX 2.0.1. During the
process, a tmp file (e.g: tmpPDFBoxXXX.pdf) is stored inside the /tmp
directory (RehHat server). This file is not deleted after completion.
After some checks, it seems that the object responsible of the file
creation is "RandomAccessBufferedFileInputStream(InputStream is)". This
object is used by the PDFParser object which doesn't close the stream after
completion.
The release note 2.0.0 [PDFBOX-2723] seems to handle this bug by adding the
following line (see https://issues.apache.org/jira/browse/PDFBOX-2723) in
the COSParser :
xrefStream.close(); // <--- *** NEW LINE ***
But, in debug mode, I saw this line is never reached so the stream is not
closed and the tmp file is not deleted. Has anybody a workaround to handle
this ?
Thanks for your help!
Re: PDFBox*.tmp files not deleted by PDFParser
Posted by Andreas Lehmkühler <an...@lehmi.de>.
> Damien Butaye <da...@gmail.com> hat am 25. Mai 2016 um 10:01
> geschrieben:
>
>
> Hello Tilman,
>
> Yes I did it. I verified in debug mode and this method (close() on
> SignatureOption) is well reached but the close() method of the object
> "RandomAccessBufferedFileInputStream" is never called, so the tmp file is
> never deleted.
The RandomAccessBufferedFileInputStream is used for the whole document and for
the SignatureOptions. You already confirmed that you close the latter one. What
about the document itself, did you close it as well?
> The patch [PDFBOX-2723] adds a line in the method parseXrefObjStream of the
> COSParser but this method is not called. It seems this method is called
> only in case of "xref stream" (line 286 COSParser) , why not in other case
> -> xref (line 218 COSParser)?
Neither a xrefstream nor a xref table uses its own
RandomAccessBufferedFileInputStream, so that this question doesn't seem to be
related to your problem.
BR
Andreas
> Damien.
>
> 2016-05-24 20:10 GMT+02:00 Tilman Hausherr <TH...@t-online.de>:
>
> > did you close the options object?
> >
> > Tilman
> >
> >
> > Am 24.05.2016 um 15:51 schrieb Damien Butaye:
> >
> >> Dear all,
> >>
> >> I'm trying to add a signature to a PDF using PDFBOX 2.0.1. During the
> >> process, a tmp file (e.g: tmpPDFBoxXXX.pdf) is stored inside the /tmp
> >> directory (RehHat server). This file is not deleted after completion.
> >> After some checks, it seems that the object responsible of the file
> >> creation is "RandomAccessBufferedFileInputStream(InputStream is)". This
> >> object is used by the PDFParser object which doesn't close the stream
> >> after
> >> completion.
> >>
> >> The release note 2.0.0 [PDFBOX-2723] seems to handle this bug by adding
> >> the
> >> following line (see https://issues.apache.org/jira/browse/PDFBOX-2723) in
> >> the COSParser :
> >>
> >> xrefStream.close(); // <--- *** NEW LINE ***
> >>
> >> But, in debug mode, I saw this line is never reached so the stream is not
> >> closed and the tmp file is not deleted. Has anybody a workaround to handle
> >> this ?
> >> Thanks for your help!
> >>
> >>
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: users-unsubscribe@pdfbox.apache.org
> > For additional commands, e-mail: users-help@pdfbox.apache.org
> >
> >
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@pdfbox.apache.org
For additional commands, e-mail: users-help@pdfbox.apache.org
Re: PDFBox*.tmp files not deleted by PDFParser
Posted by Tilman Hausherr <TH...@t-online.de>.
I did an implementation similar to what is done in PDDocument. I
remember that once (but don't remember why), the cached stream was
accessed after parsing so we shouldn't close it too soon. But thanks for
pointing me to the right direction!
https://issues.apache.org/jira/browse/PDFBOX-3363
Tilman
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@pdfbox.apache.org
For additional commands, e-mail: users-help@pdfbox.apache.org
Re: PDFBox*.tmp files not deleted by PDFParser
Posted by Tilman Hausherr <TH...@t-online.de>.
Hi,
Thanks, I'll research this. I think you have a good argument, I just
looked at PDDocument and there pdfSource is kept.
Tilman
Am 25.05.2016 um 12:11 schrieb Damien Butaye:
> Hello Tilman,
>
> For your information I found a workaround to avoid the problem of tmp file
> deletion. In the SignatureOptions object, I modified the method *public
> void setVisualSignature(InputStream is) *with the following content :
>
> RandomAccessBufferedFileInputStream tmp=new
> RandomAccessBufferedFileInputStream(is); --> NEW LINE
> PDFParser parser = new PDFParser(tmp); --> NEW LINE
> parser.parse();
> visualSignature = parser.getDocument();
> tmp.close(); --> NEW LINE
>
> Tmp files are well deleted and all unit tests passed without error on the
> pdfbox maven project.
> If you think it's not a good solution, don't hesitate to tell me.
>
> Thank you.
>
>
> 2016-05-25 10:01 GMT+02:00 Damien Butaye <da...@gmail.com>:
>
>> Hello Tilman,
>>
>> Yes I did it. I verified in debug mode and this method (close() on
>> SignatureOption) is well reached but the close() method of the object
>> "RandomAccessBufferedFileInputStream" is never called, so the tmp file is
>> never deleted.
>> The patch [PDFBOX-2723] adds a line in the method parseXrefObjStream of
>> the COSParser but this method is not called. It seems this method is called
>> only in case of "xref stream" (line 286 COSParser) , why not in other case
>> -> xref (line 218 COSParser)?
>>
>> Damien.
>>
>> 2016-05-24 20:10 GMT+02:00 Tilman Hausherr <TH...@t-online.de>:
>>
>>> did you close the options object?
>>>
>>> Tilman
>>>
>>>
>>> Am 24.05.2016 um 15:51 schrieb Damien Butaye:
>>>
>>>> Dear all,
>>>>
>>>> I'm trying to add a signature to a PDF using PDFBOX 2.0.1. During the
>>>> process, a tmp file (e.g: tmpPDFBoxXXX.pdf) is stored inside the /tmp
>>>> directory (RehHat server). This file is not deleted after completion.
>>>> After some checks, it seems that the object responsible of the file
>>>> creation is "RandomAccessBufferedFileInputStream(InputStream is)". This
>>>> object is used by the PDFParser object which doesn't close the stream
>>>> after
>>>> completion.
>>>>
>>>> The release note 2.0.0 [PDFBOX-2723] seems to handle this bug by adding
>>>> the
>>>> following line (see https://issues.apache.org/jira/browse/PDFBOX-2723)
>>>> in
>>>> the COSParser :
>>>>
>>>> xrefStream.close(); // <--- *** NEW LINE ***
>>>>
>>>> But, in debug mode, I saw this line is never reached so the stream is not
>>>> closed and the tmp file is not deleted. Has anybody a workaround to
>>>> handle
>>>> this ?
>>>> Thanks for your help!
>>>>
>>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: users-unsubscribe@pdfbox.apache.org
>>> For additional commands, e-mail: users-help@pdfbox.apache.org
>>>
>>>
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@pdfbox.apache.org
For additional commands, e-mail: users-help@pdfbox.apache.org
Re: PDFBox*.tmp files not deleted by PDFParser
Posted by Damien Butaye <da...@gmail.com>.
Hello Tilman,
For your information I found a workaround to avoid the problem of tmp file
deletion. In the SignatureOptions object, I modified the method *public
void setVisualSignature(InputStream is) *with the following content :
RandomAccessBufferedFileInputStream tmp=new
RandomAccessBufferedFileInputStream(is); --> NEW LINE
PDFParser parser = new PDFParser(tmp); --> NEW LINE
parser.parse();
visualSignature = parser.getDocument();
tmp.close(); --> NEW LINE
Tmp files are well deleted and all unit tests passed without error on the
pdfbox maven project.
If you think it's not a good solution, don't hesitate to tell me.
Thank you.
2016-05-25 10:01 GMT+02:00 Damien Butaye <da...@gmail.com>:
> Hello Tilman,
>
> Yes I did it. I verified in debug mode and this method (close() on
> SignatureOption) is well reached but the close() method of the object
> "RandomAccessBufferedFileInputStream" is never called, so the tmp file is
> never deleted.
> The patch [PDFBOX-2723] adds a line in the method parseXrefObjStream of
> the COSParser but this method is not called. It seems this method is called
> only in case of "xref stream" (line 286 COSParser) , why not in other case
> -> xref (line 218 COSParser)?
>
> Damien.
>
> 2016-05-24 20:10 GMT+02:00 Tilman Hausherr <TH...@t-online.de>:
>
>> did you close the options object?
>>
>> Tilman
>>
>>
>> Am 24.05.2016 um 15:51 schrieb Damien Butaye:
>>
>>> Dear all,
>>>
>>> I'm trying to add a signature to a PDF using PDFBOX 2.0.1. During the
>>> process, a tmp file (e.g: tmpPDFBoxXXX.pdf) is stored inside the /tmp
>>> directory (RehHat server). This file is not deleted after completion.
>>> After some checks, it seems that the object responsible of the file
>>> creation is "RandomAccessBufferedFileInputStream(InputStream is)". This
>>> object is used by the PDFParser object which doesn't close the stream
>>> after
>>> completion.
>>>
>>> The release note 2.0.0 [PDFBOX-2723] seems to handle this bug by adding
>>> the
>>> following line (see https://issues.apache.org/jira/browse/PDFBOX-2723)
>>> in
>>> the COSParser :
>>>
>>> xrefStream.close(); // <--- *** NEW LINE ***
>>>
>>> But, in debug mode, I saw this line is never reached so the stream is not
>>> closed and the tmp file is not deleted. Has anybody a workaround to
>>> handle
>>> this ?
>>> Thanks for your help!
>>>
>>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: users-unsubscribe@pdfbox.apache.org
>> For additional commands, e-mail: users-help@pdfbox.apache.org
>>
>>
>
Re: PDFBox*.tmp files not deleted by PDFParser
Posted by Damien Butaye <da...@gmail.com>.
Hello Tilman,
Yes I did it. I verified in debug mode and this method (close() on
SignatureOption) is well reached but the close() method of the object
"RandomAccessBufferedFileInputStream" is never called, so the tmp file is
never deleted.
The patch [PDFBOX-2723] adds a line in the method parseXrefObjStream of the
COSParser but this method is not called. It seems this method is called
only in case of "xref stream" (line 286 COSParser) , why not in other case
-> xref (line 218 COSParser)?
Damien.
2016-05-24 20:10 GMT+02:00 Tilman Hausherr <TH...@t-online.de>:
> did you close the options object?
>
> Tilman
>
>
> Am 24.05.2016 um 15:51 schrieb Damien Butaye:
>
>> Dear all,
>>
>> I'm trying to add a signature to a PDF using PDFBOX 2.0.1. During the
>> process, a tmp file (e.g: tmpPDFBoxXXX.pdf) is stored inside the /tmp
>> directory (RehHat server). This file is not deleted after completion.
>> After some checks, it seems that the object responsible of the file
>> creation is "RandomAccessBufferedFileInputStream(InputStream is)". This
>> object is used by the PDFParser object which doesn't close the stream
>> after
>> completion.
>>
>> The release note 2.0.0 [PDFBOX-2723] seems to handle this bug by adding
>> the
>> following line (see https://issues.apache.org/jira/browse/PDFBOX-2723) in
>> the COSParser :
>>
>> xrefStream.close(); // <--- *** NEW LINE ***
>>
>> But, in debug mode, I saw this line is never reached so the stream is not
>> closed and the tmp file is not deleted. Has anybody a workaround to handle
>> this ?
>> Thanks for your help!
>>
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe@pdfbox.apache.org
> For additional commands, e-mail: users-help@pdfbox.apache.org
>
>
Re: PDFBox*.tmp files not deleted by PDFParser
Posted by Tilman Hausherr <TH...@t-online.de>.
did you close the options object?
Tilman
Am 24.05.2016 um 15:51 schrieb Damien Butaye:
> Dear all,
>
> I'm trying to add a signature to a PDF using PDFBOX 2.0.1. During the
> process, a tmp file (e.g: tmpPDFBoxXXX.pdf) is stored inside the /tmp
> directory (RehHat server). This file is not deleted after completion.
> After some checks, it seems that the object responsible of the file
> creation is "RandomAccessBufferedFileInputStream(InputStream is)". This
> object is used by the PDFParser object which doesn't close the stream after
> completion.
>
> The release note 2.0.0 [PDFBOX-2723] seems to handle this bug by adding the
> following line (see https://issues.apache.org/jira/browse/PDFBOX-2723) in
> the COSParser :
>
> xrefStream.close(); // <--- *** NEW LINE ***
>
> But, in debug mode, I saw this line is never reached so the stream is not
> closed and the tmp file is not deleted. Has anybody a workaround to handle
> this ?
> Thanks for your help!
>
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@pdfbox.apache.org
For additional commands, e-mail: users-help@pdfbox.apache.org