next release?

classic Classic list List threaded Threaded
15 messages Options
Reply | Threaded
Open this post in threaded view
|

next release?

Tim Allison
Hi All,
  We're getting pinged over on Tika for when the next release of POI will
be available.  Any plans?

https://issues.apache.org/jira/browse/TIKA-3017

Thank you!
Reply | Threaded
Open this post in threaded view
|

Re: next release?

kiwiwings
Hi,

I'm only waiting for feedback on my EMF stuff, which actually involved fixing some SL rendering issues ...
Otherwise I'm fixing sonar issues up till then ...

Andi


On 23.01.20 15:41, Tim Allison wrote:
> Hi All,
>   We're getting pinged over on Tika for when the next release of POI will
> be available.  Any plans?
>
> https://issues.apache.org/jira/browse/TIKA-3017
>
> Thank you!
>



signature.asc (499 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: next release?

Tim Allison
Sorry...waiting on me?  Should I run regression tests now?

On Thu, Jan 23, 2020 at 3:04 PM Andreas Beeker <[hidden email]> wrote:

> Hi,
>
> I'm only waiting for feedback on my EMF stuff, which actually involved
> fixing some SL rendering issues ...
> Otherwise I'm fixing sonar issues up till then ...
>
> Andi
>
>
> On 23.01.20 15:41, Tim Allison wrote:
> > Hi All,
> >   We're getting pinged over on Tika for when the next release of POI will
> > be available.  Any plans?
> >
> > https://issues.apache.org/jira/browse/TIKA-3017
> >
> > Thank you!
> >
>
>
>
Reply | Threaded
Open this post in threaded view
|

Re: next release?

kiwiwings
> Sorry...waiting on me?  Should I run regression tests now?
Nope ... I've implemented some changes for Ghazi.
Please wait with the regression tests.

On 23.01.20 22:47, Tim Allison wrote:

> Sorry...waiting on me?  Should I run regression tests now?
>
> On Thu, Jan 23, 2020 at 3:04 PM Andreas Beeker <[hidden email]> wrote:
>
>> Hi,
>>
>> I'm only waiting for feedback on my EMF stuff, which actually involved
>> fixing some SL rendering issues ...
>> Otherwise I'm fixing sonar issues up till then ...
>>
>> Andi
>>
>>
>> On 23.01.20 15:41, Tim Allison wrote:
>>> Hi All,
>>>   We're getting pinged over on Tika for when the next release of POI will
>>> be available.  Any plans?
>>>
>>> https://issues.apache.org/jira/browse/TIKA-3017
>>>
>>> Thank you!
>>>
>>
>>


signature.asc (499 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: next release?

Dominik Stadler
Hi,

I started an initial regression-run to verify some changes to the
regression-test-functionality itself and to get early initial feedback,
it's very easy for me to run those nowadays anyway.

Results compared to 4.1.1 are at
http://people.apache.org/~centic/poi_regression/reports/index410RC2to411RC1.html

A few interesting failures, each quite rare, between 19 and 1 time for more
than 2 mio documents, some maybe caused by more testing, but some might be
worth a quick look/fix.


Full results at
http://people.apache.org/~centic/poi_regression/reportsAll/index410RC2to411RC1.html

Thanks... Dominik.

On Fri, Jan 24, 2020 at 12:05 AM Andreas Beeker <[hidden email]>
wrote:

> > Sorry...waiting on me?  Should I run regression tests now?
> Nope ... I've implemented some changes for Ghazi.
> Please wait with the regression tests.
>
> On 23.01.20 22:47, Tim Allison wrote:
> > Sorry...waiting on me?  Should I run regression tests now?
> >
> > On Thu, Jan 23, 2020 at 3:04 PM Andreas Beeker <[hidden email]>
> wrote:
> >
> >> Hi,
> >>
> >> I'm only waiting for feedback on my EMF stuff, which actually involved
> >> fixing some SL rendering issues ...
> >> Otherwise I'm fixing sonar issues up till then ...
> >>
> >> Andi
> >>
> >>
> >> On 23.01.20 15:41, Tim Allison wrote:
> >>> Hi All,
> >>>   We're getting pinged over on Tika for when the next release of POI
> will
> >>> be available.  Any plans?
> >>>
> >>> https://issues.apache.org/jira/browse/TIKA-3017
> >>>
> >>> Thank you!
> >>>
> >>
> >>
>
>
>
Reply | Threaded
Open this post in threaded view
|

Re: next release?

PJ Fanning
Hi Dominik,

The links you sent as from 'Sat Oct 12 00:15:42 CEST 2019'. Do you have the links for the latest run?








On Monday 27 January 2020, 17:15:45 GMT, Dominik Stadler <[hidden email]> wrote:





Hi,

I started an initial regression-run to verify some changes to the
regression-test-functionality itself and to get early initial feedback,
it's very easy for me to run those nowadays anyway.

Results compared to 4.1.1 are at
http://people.apache.org/~centic/poi_regression/reports/index410RC2to411RC1.html

A few interesting failures, each quite rare, between 19 and 1 time for more
than 2 mio documents, some maybe caused by more testing, but some might be
worth a quick look/fix.


Full results at
http://people.apache.org/~centic/poi_regression/reportsAll/index410RC2to411RC1.html

Thanks... Dominik.

On Fri, Jan 24, 2020 at 12:05 AM Andreas Beeker <[hidden email]>
wrote:

> > Sorry...waiting on me?  Should I run regression tests now?
> Nope ... I've implemented some changes for Ghazi.
> Please wait with the regression tests.
>
> On 23.01.20 22:47, Tim Allison wrote:
> > Sorry...waiting on me?  Should I run regression tests now?
> >
> > On Thu, Jan 23, 2020 at 3:04 PM Andreas Beeker <[hidden email]>
> wrote:
> >
> >> Hi,
> >>
> >> I'm only waiting for feedback on my EMF stuff, which actually involved
> >> fixing some SL rendering issues ...
> >> Otherwise I'm fixing sonar issues up till then ...
> >>
> >> Andi
> >>
> >>
> >> On 23.01.20 15:41, Tim Allison wrote:
> >>> Hi All,
> >>>  We're getting pinged over on Tika for when the next release of POI
> will
> >>> be available.  Any plans?
> >>>
> >>> https://issues.apache.org/jira/browse/TIKA-3017
> >>>
> >>> Thank you!
> >>>
> >>
> >>
>
>
>

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

Reply | Threaded
Open this post in threaded view
|

Re: next release?

PJ Fanning
I found the latest results at http://people.apache.org/~centic/poi_regression/reports/


I fixed the NPE in

java.lang.NullPointerException
    at o.a.p.BaseIntegrationTest.testOneFile(BaseIntegrationTest.java:78)

But this will just allow us to find the exception that is being thrown and that has null error message.





On Monday 27 January 2020, 18:26:29 GMT, PJ Fanning <[hidden email]> wrote:





Hi Dominik,

The links you sent as from 'Sat Oct 12 00:15:42 CEST 2019'. Do you have the links for the latest run?








On Monday 27 January 2020, 17:15:45 GMT, Dominik Stadler <[hidden email]> wrote:





Hi,

I started an initial regression-run to verify some changes to the
regression-test-functionality itself and to get early initial feedback,
it's very easy for me to run those nowadays anyway.

Results compared to 4.1.1 are at
http://people.apache.org/~centic/poi_regression/reports/index410RC2to411RC1.html

A few interesting failures, each quite rare, between 19 and 1 time for more
than 2 mio documents, some maybe caused by more testing, but some might be
worth a quick look/fix.


Full results at
http://people.apache.org/~centic/poi_regression/reportsAll/index410RC2to411RC1.html

Thanks... Dominik.

On Fri, Jan 24, 2020 at 12:05 AM Andreas Beeker <[hidden email]>
wrote:

> > Sorry...waiting on me?  Should I run regression tests now?
> Nope ... I've implemented some changes for Ghazi.
> Please wait with the regression tests.
>
> On 23.01.20 22:47, Tim Allison wrote:
> > Sorry...waiting on me?  Should I run regression tests now?
> >
> > On Thu, Jan 23, 2020 at 3:04 PM Andreas Beeker <[hidden email]>
> wrote:
> >
> >> Hi,
> >>
> >> I'm only waiting for feedback on my EMF stuff, which actually involved
> >> fixing some SL rendering issues ...
> >> Otherwise I'm fixing sonar issues up till then ...
> >>
> >> Andi
> >>
> >>
> >> On 23.01.20 15:41, Tim Allison wrote:
> >>> Hi All,
> >>>  We're getting pinged over on Tika for when the next release of POI
> will
> >>> be available.  Any plans?
> >>>
> >>> https://issues.apache.org/jira/browse/TIKA-3017
> >>>
> >>> Thank you!

> >>>
> >>
> >>
>
>
>

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


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

Reply | Threaded
Open this post in threaded view
|

Re: next release?

Dominik Stadler
Ah, yes, sorry, good catch, the correct links are
*
http://people.apache.org/~centic/poi_regression/reports/index411RC2to412RC1.html
*
http://people.apache.org/~centic/poi_regression/reportsAll/index411RC2to412RC1.html

I'll do a run with all recent changes for a "RC2" soon to get the actual
exceptions.

Thanks... Dominik.


On Mon, Jan 27, 2020 at 7:32 PM PJ Fanning <[hidden email]>
wrote:

> I found the latest results at
> http://people.apache.org/~centic/poi_regression/reports/
>
>
> I fixed the NPE in
>
> java.lang.NullPointerException
>     at o.a.p.BaseIntegrationTest.testOneFile(BaseIntegrationTest.java:78)
>
> But this will just allow us to find the exception that is being thrown and
> that has null error message.
>
>
>
>
>
> On Monday 27 January 2020, 18:26:29 GMT, PJ Fanning
> <[hidden email]> wrote:
>
>
>
>
>
> Hi Dominik,
>
> The links you sent as from 'Sat Oct 12 00:15:42 CEST 2019'. Do you have
> the links for the latest run?
>
>
>
>
>
>
>
>
> On Monday 27 January 2020, 17:15:45 GMT, Dominik Stadler <
> [hidden email]> wrote:
>
>
>
>
>
> Hi,
>
> I started an initial regression-run to verify some changes to the
> regression-test-functionality itself and to get early initial feedback,
> it's very easy for me to run those nowadays anyway.
>
> Results compared to 4.1.1 are at
>
> http://people.apache.org/~centic/poi_regression/reports/index410RC2to411RC1.html
>
> A few interesting failures, each quite rare, between 19 and 1 time for more
> than 2 mio documents, some maybe caused by more testing, but some might be
> worth a quick look/fix.
>
>
> Full results at
>
> http://people.apache.org/~centic/poi_regression/reportsAll/index410RC2to411RC1.html
>
> Thanks... Dominik.
>
> On Fri, Jan 24, 2020 at 12:05 AM Andreas Beeker <[hidden email]>
> wrote:
>
> > > Sorry...waiting on me?  Should I run regression tests now?
> > Nope ... I've implemented some changes for Ghazi.
> > Please wait with the regression tests.
> >
> > On 23.01.20 22:47, Tim Allison wrote:
> > > Sorry...waiting on me?  Should I run regression tests now?
> > >
> > > On Thu, Jan 23, 2020 at 3:04 PM Andreas Beeker <[hidden email]>
> > wrote:
> > >
> > >> Hi,
> > >>
> > >> I'm only waiting for feedback on my EMF stuff, which actually involved
> > >> fixing some SL rendering issues ...
> > >> Otherwise I'm fixing sonar issues up till then ...
> > >>
> > >> Andi
> > >>
> > >>
> > >> On 23.01.20 15:41, Tim Allison wrote:
> > >>> Hi All,
> > >>>  We're getting pinged over on Tika for when the next release of POI
> > will
> > >>> be available.  Any plans?
> > >>>
> > >>> https://issues.apache.org/jira/browse/TIKA-3017
> > >>>
> > >>> Thank you!
>
> > >>>
> > >>
> > >>
> >
> >
> >
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>
>
Reply | Threaded
Open this post in threaded view
|

Re: next release?

kiwiwings
HI Dominik

thank you for providing the results!

I've asked now several times about the feedback, but to no avail.
So from my side, I can roll a release any time - is there anyone else who wants to be release manager this time?

Andi

On 28.01.20 00:02, Dominik Stadler wrote:
> Ah, yes, sorry, good catch, the correct links are
> *
> http://people.apache.org/~centic/poi_regression/reports/index411RC2to412RC1.html
> *
> http://people.apache.org/~centic/poi_regression/reportsAll/index411RC2to412RC1.html
>
> I'll do a run with all recent changes for a "RC2" soon to get the actual
> exceptions.
>



signature.asc (499 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: next release?

Dominik Stadler
Hi,

Updated run is at
http://people.apache.org/~centic/poi_regression/reports/index411RC2to412RC2.html

At least some of the failures look like being caused by more coverage in
the testing itself.

All failures are at
http://people.apache.org/~centic/poi_regression/reportsAll/index411RC2to412RC2.html

Dominik

On Tue, Jan 28, 2020, 19:25 Andreas Beeker <[hidden email]> wrote:

> HI Dominik
>
> thank you for providing the results!
>
> I've asked now several times about the feedback, but to no avail.
> So from my side, I can roll a release any time - is there anyone else who
> wants to be release manager this time?
>
> Andi
>
> On 28.01.20 00:02, Dominik Stadler wrote:
> > Ah, yes, sorry, good catch, the correct links are
> > *
> >
> http://people.apache.org/~centic/poi_regression/reports/index411RC2to412RC1.html
> > *
> >
> http://people.apache.org/~centic/poi_regression/reportsAll/index411RC2to412RC1.html
> >
> > I'll do a run with all recent changes for a "RC2" soon to get the actual
> > exceptions.
> >
>
>
>
Reply | Threaded
Open this post in threaded view
|

Re: next release?

kiwiwings
Hi,

although I've fixed two issues *) locally, I'll postpone those commits after the release.

So I would roll the release now.

If there are any late-commits ... not a problem, but please inform us before Dominik and Tim run the integration tests again - not sure if this still makes sense, as there were only marginal changes.

Andi

*) a missing move-to exception from the mass tests and using a indexed palette for duotone bitmaps



signature.asc (499 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: next release?

kiwiwings
I'm running into 4x OOMs when running the integration tests in Open or Oracle Jdk 8, even when raising the heap from 1512 to 2048 mb.
This didn't happen with OpenJDK 12 - which was my default setting before :|

It looks like there's a problem with XSLF (see below, although this is probably not the location where the memory leak is happening)

Andi


Testcase: testAllFiles[845: slideshow/Divino_Revelado.pptx using org.apache.poi.stress.XSLFFileHandler@396a51ab] took 2.339 sec
    Caused an ERROR
Java heap space
java.lang.OutOfMemoryError: Java heap space
    at java.util.Arrays.copyOf(Arrays.java:3332)
    at java.lang.AbstractStringBuilder.ensureCapacityInternal(AbstractStringBuilder.java:124)
    at java.lang.AbstractStringBuilder.append(AbstractStringBuilder.java:649)
    at java.lang.StringBuilder.append(StringBuilder.java:202)
    at org.apache.poi.sl.draw.DrawTextParagraph.tab2space(DrawTextParagraph.java:427)
    at org.apache.poi.sl.draw.DrawTextParagraph.getRenderableText(DrawTextParagraph.java:386)
    at org.apache.poi.sl.draw.DrawTextParagraph.getRenderableText(DrawTextParagraph.java:381)
    at org.apache.poi.sl.draw.DrawTextParagraph.getAttributedString(DrawTextParagraph.java:564)
    at org.apache.poi.sl.draw.DrawTextParagraph.breakText(DrawTextParagraph.java:258)
    at org.apache.poi.sl.draw.DrawTextShape.drawParagraphs(DrawTextShape.java:158)
    at org.apache.poi.sl.draw.DrawTextShape.getTextHeight(DrawTextShape.java:220)
    at org.apache.poi.sl.draw.DrawTextShape.drawContent(DrawTextShape.java:104)
    at org.apache.poi.sl.draw.DrawSimpleShape.draw(DrawSimpleShape.java:107)
    at org.apache.poi.sl.draw.DrawSheet.draw(DrawSheet.java:71)
    at org.apache.poi.sl.draw.DrawSlide.draw(DrawSlide.java:41)
    at org.apache.poi.xslf.usermodel.XSLFSlide.draw(XSLFSlide.java:373)
    at org.apache.poi.stress.SlideShowHandler.renderSlides(SlideShowHandler.java:152)
    at org.apache.poi.stress.SlideShowHandler.handleSlideShow(SlideShowHandler.java:46)
    at org.apache.poi.stress.XSLFFileHandler.handleFile(XSLFFileHandler.java:43)
    at org.apache.poi.TestAllFiles.testAllFiles(TestAllFiles.java:433)



signature.asc (499 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: next release?

Dominik Stadler
Hi,

target "jenkins" works fine for me with

export ANT_OPTS="-Xmx1024m"

and no changes to the build.xml both with JDK 8_201 and OpenJDK 8_232

Dominik.

On Fri, Jan 31, 2020 at 9:01 PM Andreas Beeker <[hidden email]> wrote:

> I'm running into 4x OOMs when running the integration tests in Open or
> Oracle Jdk 8, even when raising the heap from 1512 to 2048 mb.
> This didn't happen with OpenJDK 12 - which was my default setting before :|
>
> It looks like there's a problem with XSLF (see below, although this is
> probably not the location where the memory leak is happening)
>
> Andi
>
>
> Testcase: testAllFiles[845: slideshow/Divino_Revelado.pptx using
> org.apache.poi.stress.XSLFFileHandler@396a51ab] took 2.339 sec
>     Caused an ERROR
> Java heap space
> java.lang.OutOfMemoryError: Java heap space
>     at java.util.Arrays.copyOf(Arrays.java:3332)
>     at
> java.lang.AbstractStringBuilder.ensureCapacityInternal(AbstractStringBuilder.java:124)
>     at
> java.lang.AbstractStringBuilder.append(AbstractStringBuilder.java:649)
>     at java.lang.StringBuilder.append(StringBuilder.java:202)
>     at
> org.apache.poi.sl.draw.DrawTextParagraph.tab2space(DrawTextParagraph.java:427)
>     at
> org.apache.poi.sl.draw.DrawTextParagraph.getRenderableText(DrawTextParagraph.java:386)
>     at
> org.apache.poi.sl.draw.DrawTextParagraph.getRenderableText(DrawTextParagraph.java:381)
>     at
> org.apache.poi.sl.draw.DrawTextParagraph.getAttributedString(DrawTextParagraph.java:564)
>     at
> org.apache.poi.sl.draw.DrawTextParagraph.breakText(DrawTextParagraph.java:258)
>     at
> org.apache.poi.sl.draw.DrawTextShape.drawParagraphs(DrawTextShape.java:158)
>     at
> org.apache.poi.sl.draw.DrawTextShape.getTextHeight(DrawTextShape.java:220)
>     at
> org.apache.poi.sl.draw.DrawTextShape.drawContent(DrawTextShape.java:104)
>     at
> org.apache.poi.sl.draw.DrawSimpleShape.draw(DrawSimpleShape.java:107)
>     at org.apache.poi.sl.draw.DrawSheet.draw(DrawSheet.java:71)
>     at org.apache.poi.sl.draw.DrawSlide.draw(DrawSlide.java:41)
>     at org.apache.poi.xslf.usermodel.XSLFSlide.draw(XSLFSlide.java:373)
>     at
> org.apache.poi.stress.SlideShowHandler.renderSlides(SlideShowHandler.java:152)
>     at
> org.apache.poi.stress.SlideShowHandler.handleSlideShow(SlideShowHandler.java:46)
>     at
> org.apache.poi.stress.XSLFFileHandler.handleFile(XSLFFileHandler.java:43)
>     at org.apache.poi.TestAllFiles.testAllFiles(TestAllFiles.java:433)
>
>
>
Reply | Threaded
Open this post in threaded view
|

Re: next release?

Dominik Stadler
Sorry, I was wrong, it actually only works with JDK 8_201, but fails with
OOM on OpenJKD 8_232, so some change between those two versions of the JDK
might be related.

BTW, CI on Jenkins still uses patch-level 191, which is quite outdated
nowadays...

Dominik.

On Sat, Feb 1, 2020 at 10:34 AM Dominik Stadler <[hidden email]>
wrote:

> Hi,
>
> target "jenkins" works fine for me with
>
> export ANT_OPTS="-Xmx1024m"
>
> and no changes to the build.xml both with JDK 8_201 and OpenJDK 8_232
>
> Dominik.
>
> On Fri, Jan 31, 2020 at 9:01 PM Andreas Beeker <[hidden email]>
> wrote:
>
>> I'm running into 4x OOMs when running the integration tests in Open or
>> Oracle Jdk 8, even when raising the heap from 1512 to 2048 mb.
>> This didn't happen with OpenJDK 12 - which was my default setting before
>> :|
>>
>> It looks like there's a problem with XSLF (see below, although this is
>> probably not the location where the memory leak is happening)
>>
>> Andi
>>
>>
>> Testcase: testAllFiles[845: slideshow/Divino_Revelado.pptx using
>> org.apache.poi.stress.XSLFFileHandler@396a51ab] took 2.339 sec
>>     Caused an ERROR
>> Java heap space
>> java.lang.OutOfMemoryError: Java heap space
>>     at java.util.Arrays.copyOf(Arrays.java:3332)
>>     at
>> java.lang.AbstractStringBuilder.ensureCapacityInternal(AbstractStringBuilder.java:124)
>>     at
>> java.lang.AbstractStringBuilder.append(AbstractStringBuilder.java:649)
>>     at java.lang.StringBuilder.append(StringBuilder.java:202)
>>     at
>> org.apache.poi.sl.draw.DrawTextParagraph.tab2space(DrawTextParagraph.java:427)
>>     at
>> org.apache.poi.sl.draw.DrawTextParagraph.getRenderableText(DrawTextParagraph.java:386)
>>     at
>> org.apache.poi.sl.draw.DrawTextParagraph.getRenderableText(DrawTextParagraph.java:381)
>>     at
>> org.apache.poi.sl.draw.DrawTextParagraph.getAttributedString(DrawTextParagraph.java:564)
>>     at
>> org.apache.poi.sl.draw.DrawTextParagraph.breakText(DrawTextParagraph.java:258)
>>     at
>> org.apache.poi.sl.draw.DrawTextShape.drawParagraphs(DrawTextShape.java:158)
>>     at
>> org.apache.poi.sl.draw.DrawTextShape.getTextHeight(DrawTextShape.java:220)
>>     at
>> org.apache.poi.sl.draw.DrawTextShape.drawContent(DrawTextShape.java:104)
>>     at
>> org.apache.poi.sl.draw.DrawSimpleShape.draw(DrawSimpleShape.java:107)
>>     at org.apache.poi.sl.draw.DrawSheet.draw(DrawSheet.java:71)
>>     at org.apache.poi.sl.draw.DrawSlide.draw(DrawSlide.java:41)
>>     at org.apache.poi.xslf.usermodel.XSLFSlide.draw(XSLFSlide.java:373)
>>     at
>> org.apache.poi.stress.SlideShowHandler.renderSlides(SlideShowHandler.java:152)
>>     at
>> org.apache.poi.stress.SlideShowHandler.handleSlideShow(SlideShowHandler.java:46)
>>     at
>> org.apache.poi.stress.XSLFFileHandler.handleFile(XSLFFileHandler.java:43)
>>     at org.apache.poi.TestAllFiles.testAllFiles(TestAllFiles.java:433)
>>
>>
>>
Reply | Threaded
Open this post in threaded view
|

Re: next release?

Dominik Stadler
Hi,

problem seems to be at DrawTextParagraph.java:417, TextLayout.getAdvance()
sometimes returns 0 for OpenJDK. In the statements "(int)Math.ceil(tabSz /
wspace)" the division by zero leads to "Infinity", which Math.ceil()
converts to Integer.MAX_VALUE, thus we try to create a string with that
many spaces.

Simple fix may be to adjust wspace to non-zero in this case as it probably
is not useful to use "0" for wspace here.

I didn't see any related change in POI for some time, so it may be caused
by changes in recent OpenJDK versions.

Dominik.


On Sat, Feb 1, 2020 at 10:48 AM Dominik Stadler <[hidden email]>
wrote:

>
> Sorry, I was wrong, it actually only works with JDK 8_201, but fails with
> OOM on OpenJKD 8_232, so some change between those two versions of the JDK
> might be related.
>
> BTW, CI on Jenkins still uses patch-level 191, which is quite outdated
> nowadays...
>
> Dominik.
>
> On Sat, Feb 1, 2020 at 10:34 AM Dominik Stadler <[hidden email]>
> wrote:
>
>> Hi,
>>
>> target "jenkins" works fine for me with
>>
>> export ANT_OPTS="-Xmx1024m"
>>
>> and no changes to the build.xml both with JDK 8_201 and OpenJDK 8_232
>>
>> Dominik.
>>
>> On Fri, Jan 31, 2020 at 9:01 PM Andreas Beeker <[hidden email]>
>> wrote:
>>
>>> I'm running into 4x OOMs when running the integration tests in Open or
>>> Oracle Jdk 8, even when raising the heap from 1512 to 2048 mb.
>>> This didn't happen with OpenJDK 12 - which was my default setting before
>>> :|
>>>
>>> It looks like there's a problem with XSLF (see below, although this is
>>> probably not the location where the memory leak is happening)
>>>
>>> Andi
>>>
>>>
>>> Testcase: testAllFiles[845: slideshow/Divino_Revelado.pptx using
>>> org.apache.poi.stress.XSLFFileHandler@396a51ab] took 2.339 sec
>>>     Caused an ERROR
>>> Java heap space
>>> java.lang.OutOfMemoryError: Java heap space
>>>     at java.util.Arrays.copyOf(Arrays.java:3332)
>>>     at
>>> java.lang.AbstractStringBuilder.ensureCapacityInternal(AbstractStringBuilder.java:124)
>>>     at
>>> java.lang.AbstractStringBuilder.append(AbstractStringBuilder.java:649)
>>>     at java.lang.StringBuilder.append(StringBuilder.java:202)
>>>     at
>>> org.apache.poi.sl.draw.DrawTextParagraph.tab2space(DrawTextParagraph.java:427)
>>>     at
>>> org.apache.poi.sl.draw.DrawTextParagraph.getRenderableText(DrawTextParagraph.java:386)
>>>     at
>>> org.apache.poi.sl.draw.DrawTextParagraph.getRenderableText(DrawTextParagraph.java:381)
>>>     at
>>> org.apache.poi.sl.draw.DrawTextParagraph.getAttributedString(DrawTextParagraph.java:564)
>>>     at
>>> org.apache.poi.sl.draw.DrawTextParagraph.breakText(DrawTextParagraph.java:258)
>>>     at
>>> org.apache.poi.sl.draw.DrawTextShape.drawParagraphs(DrawTextShape.java:158)
>>>     at
>>> org.apache.poi.sl.draw.DrawTextShape.getTextHeight(DrawTextShape.java:220)
>>>     at
>>> org.apache.poi.sl.draw.DrawTextShape.drawContent(DrawTextShape.java:104)
>>>     at
>>> org.apache.poi.sl.draw.DrawSimpleShape.draw(DrawSimpleShape.java:107)
>>>     at org.apache.poi.sl.draw.DrawSheet.draw(DrawSheet.java:71)
>>>     at org.apache.poi.sl.draw.DrawSlide.draw(DrawSlide.java:41)
>>>     at org.apache.poi.xslf.usermodel.XSLFSlide.draw(XSLFSlide.java:373)
>>>     at
>>> org.apache.poi.stress.SlideShowHandler.renderSlides(SlideShowHandler.java:152)
>>>     at
>>> org.apache.poi.stress.SlideShowHandler.handleSlideShow(SlideShowHandler.java:46)
>>>     at
>>> org.apache.poi.stress.XSLFFileHandler.handleFile(XSLFFileHandler.java:43)
>>>     at org.apache.poi.TestAllFiles.testAllFiles(TestAllFiles.java:433)
>>>
>>>
>>>