Re: Inquiry Regarding SUMPRODUCT Support in Apache POI

[Nick Burch]

On Tue, 21 Nov 2023, Reddy, Sri Rama wrote:
I've recently encountered an issue with SUMPRODUCT advanced use cases in 
Apache POI


What specifically didn't work?

Detailed steps to reproduce the problem, and/or a failing junit unit test 
showing the issue, would be most helpful



as outlined in the official GitHub documentation.


Our official documentation is on https://poi.apache.org/ , the github 
mirror is aimed at contributors to the project


I'm reaching out to inquire about the roadmap for SUMPRODUCT support. 
Could you provide information on when we can expect a full 
implementation of SUMPRODUCT in Apache POI?


Apache POI is a volunter led project.

If you'd like to contribute additional support, that would be great! 
Alternately, if Amazon would like to sponsor some work on the specific 
formula support you need, I'm sure one of the volunteer contributors would 
be able to take on that work


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: POI PMC roll call

[Nick Burch]

On Fri, 3 Mar 2023, PJ Fanning wrote:
Apologies if anyone thinks I'm going in the wrong direction here. The 
TLDR is: I am asking if PMC members could respond with whether they are 
still interested in working on POI or if they are no longer in a 
position to help.


Still here, still watching, but I don't currently have any ringfenced time 
for open source work in my $DAYJOB so it's quite variable how much time 
I've got free to give at any point.


Happy to help mentor some new people towards committership, if there's any 
candidates, as that (long term) gives more benefit for the time invested 
:)


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

[jira] [Commented] (XMLBEANS-629) org.apache.xmlbeans.XmlObject.selectPath(String arg0) API returns wrong values after org.apache.xmlbeans.XmlCursor.moveXml(XmlCursor arg0) API call

[Nick Burch (Jira)]

[ 
https://issues.apache.org/jira/browse/XMLBEANS-629?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17616342#comment-17616342
 ] 

Nick Burch commented on XMLBEANS-629:
-

As was already stated, XmlBeans is in maintenance mode. Your options are:
 * Use a different API
 * Fix the bug and submit a patch
 * Pay your favourite open source support vendor to fix it for you

There is no unpaid volunteer energy to fix this code behaviour that has been in 
place for 10+ years

> org.apache.xmlbeans.XmlObject.selectPath(String arg0) API returns wrong 
> values after org.apache.xmlbeans.XmlCursor.moveXml(XmlCursor arg0) API call
> ---
>
> Key: XMLBEANS-629
> URL: https://issues.apache.org/jira/browse/XMLBEANS-629
> Project: XMLBeans
>  Issue Type: Bug
>  Components: XPath
>Affects Versions:  Version 2.3, Version 3.1.0, Version 5.1.1
> Environment: Windows/Linux
>Reporter: Shinith PP
>Priority: Major
> Attachments: XMLBeanBugTest.zip
>
>
> When I search for an XmlObject within the root xml document using 
> org.apache.xmlbeans.XmlObject.selectPath(String arg0) API with following 
> query expression selectPath() is returning the matching XmlObject.
> selectPath("declare namespace wf='http://www.example.org/SimplifiedSchema' 
> //wf:data[@id='data.85']")
> *{color:#de350b}but calling the same method with same query expression 
> returns no matching XmlObjects after moving one child xml object from one 
> location to another location within using 
> org.apache.xmlbeans.XmlCursor.moveXml(XmlCursor arg0) API{color}.*
> As per my xml schema my xml have the following structure.
> 
>     
>         
>             
>         
>     
> 
> root document is session. Inside session we can have task. Inside task we can 
> have task and data.
> *{color:#de350b}The issue arises when I search for a data using it's id after 
> moving one child task to new location within a parent task.{color}*
> I have attached the schema file and sample code to reproduce the issue for 
> reference.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

[jira] [Commented] (XMLBEANS-616) is there a dead lock in loading the SOAP Service Concurrently

[Nick Burch (Jira)]

[ 
https://issues.apache.org/jira/browse/XMLBEANS-616?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17562666#comment-17562666
 ] 

Nick Burch commented on XMLBEANS-616:
-

What version of Apache XMLBeans are you using? If not the latest, what happens 
when you upgrade?

> is there a dead lock in loading the SOAP Service Concurrently
> -
>
> Key: XMLBEANS-616
> URL: https://issues.apache.org/jira/browse/XMLBEANS-616
> Project: XMLBeans
>  Issue Type: Bug
>Reporter: yanglinfu
>Priority: Major
> Attachments: 007337a1-7e50-40bc-ac1f-d2ebd469f095.png, 
> 2a6c2727-50e1-4ccc-a2f5-33abd746117a.png, 
> 40f87ce3-cbcd-495e-9ce7-0ae861a3bdb7-1.png, 
> 40f87ce3-cbcd-495e-9ce7-0ae861a3bdb7-2.png, 
> 40f87ce3-cbcd-495e-9ce7-0ae861a3bdb7.png, 
> b3d30bc4-a060-419a-b990-503b188fe262.png
>
>
> *i have 10 threads to loading the SOAP Service Concurrently,and i find a dead 
> lock in xmlBeans's some  class loading process.*
> !40f87ce3-cbcd-495e-9ce7-0ae861a3bdb7.png!
> !2.png!!1.png!
> *xmlBeans code analyze*
>  # the object XmlBeans init is waiting for BuiltinSchemaTypeSystem  inited
> !007337a1-7e50-40bc-ac1f-d2ebd469f095.png!
>  # BuiltinSchemaTypeSystem  is the sub class of SchemaTypeLoaderBase ,so  
> BuiltinSchemaTypeSystem is inited after SchemaTypeLoaderBase ;
> !b3d30bc4-a060-419a-b990-503b188fe262.png!
>  # SchemaTypeLoaderBase init is need XmlBeans init in line 38;
> !2a6c2727-50e1-4ccc-a2f5-33abd746117a.png!
>  
>  



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: poi site - missing trademark declaration

[Nick Burch]

On Thu, 9 Jun 2022, PJ Fanning wrote:
The src/documentation/index.xml that we have for Forrest build seems to 
try to add it as a footer but it doesn't get picked up. I don't much 
about Forrest. Anyone know how this can be fixed?


I've a few ideas, but currently I can't seem to build POI from a svn 
checkout using ant...


Gradle builds the source just fine, but we need Ant to use with Forrest


If I try "ant clean jar" it fails on compile-ooxml with

poi-ooxml/src/main/java/org/apache/poi/xslf/draw/SVGUserAgent.java:66: 
error: cannot access ViewCSS
[javac] String viewBoxStr = el.getAttributeNS(null, 
SVGConstants.SVG_VIEW_BOX_ATTRIBUTE);

[javac]   ^
[javac]   class file for org.w3c.dom.css.ViewCSS not found


If I try with "ant docs" then it fails with
 [exec] load-project-props:
 [exec]
 [exec] BUILD FAILED
 [exec] java.lang.NoClassDefFoundError: org/w3c/dom/ls/DocumentLS
 [exec] 	at java.base/java.lang.ClassLoader.defineClass1(Native 
Method)
 [exec] 	at 
java.base/java.lang.ClassLoader.defineClass(ClassLoader.java:1017)



Is there a trick to getting POI to build happily with Ant, and Java 11, 
that anyone knows?


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: want to provide patch

[Nick Burch]

On Thu, 28 Apr 2022, stanton fisque wrote:
i'd like to provide a patch for some missing functionality, specifically 
the support handler for AverageIf (the generalized version is present 
but not the singular version).  what is the best way to provide that? 
is there a repo where a pull request can be provided?


Great! You'll want:
 * https://poi.apache.org/devel/guidelines.html
 * https://github.com/apache/poi

For formula functions, see also
 * https://poi.apache.org/components/spreadsheet/eval.html
 * 
http://home.apache.org/~yegor/apachecon_us2010/Evaluation_Of_Excel_Formulas_In_POI.pptx

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Duplicate classes in classpath from poi-ooxml-lite and ooxml-schemas

[Nick Burch]

On Fri, 21 Jan 2022, Валерий Зайченко wrote:

I used fr.opensagres.xdocreport:org.apache.poi.xwpf.converter.pdf:1.0.5 for
convert *.docx files to *.pdf.


That's a totally different project, which merely depends on Apache POI

Can you please put in order dependencies for 
fr.opensagres.xdocreport:fr.opensagres.poi.xwpf.converter.pdf ?


You would need to ask the authors of that project for help with their 
versions and dependencies, we have no control over their project


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: XWPF - worth adding helpers for creating styles?

[Nick Burch]

On Wed, 22 Dec 2021, Nick Burch wrote:
I was wondering about adding a couple of friendly helper methods on 
XWPFStyles, which would either set you up with some sensible default styles 
(inspired by the ones Word pops in a new file), or let you quickly add a 
bunch of header or paragraph styles by font size/colour/family.


Anyone have any objections to me doing that? Anyone have strong preferences 
for what to call the methods / what to put on XWPFStyles vs XWPFStyle?


My current idea, if no objections, would be
 * XWPFStyle.createHeaderStyle(String name, int headingLevel,
 String fontFamily, String fontColor, int fontSizeHalfPoints)
 * XWPFStyle.createParagraphStyle(String name, String fontFamily,
 String fontColor, int fontSizeHalfPoints)
 * XWPFStyle.createDefaults()
- calls the XWPFStyle.create* methods with sensible defaults

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

XWPF - worth adding helpers for creating styles?

[Nick Burch]

Hi All

I found myself wanting to create a DOCX file from scratch today, and in 
order to have it look nice that meant setting up a bunch of styles. It's a 
bit fiddly, and would've been pretty tough if I hadn't already been happy 
with reading the XML from word and re-doing that in CT xmlbeans classes. 
Based on StackOverflow posts, it's a fairly common thing for our users to 
want to do.


I was wondering about adding a couple of friendly helper methods on 
XWPFStyles, which would either set you up with some sensible default 
styles (inspired by the ones Word pops in a new file), or let you quickly 
add a bunch of header or paragraph styles by font size/colour/family.


Anyone have any objections to me doing that? Anyone have strong 
preferences for what to call the methods / what to put on XWPFStyles vs 
XWPFStyle?


(I wouldn't be doing a full set of XWPFStyle get/set/etc methods, just 
enough to help you easily create new ones. The alternative to all this is 
to start with a template .docx file, having set the styles up by hand in 
Word, but I wanted to avoid that.)


Cheers
Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: flaky tests on CI/CD

[Nick Burch]

On Thu, 25 Nov 2021, PJ Fanning wrote:
There are other issues on the CI/CD tests. For 
example, https://ci-builds.apache.org/job/POI/job/POI-DSL-1.11/ just 
keeps failing with svn issues. It's like the machine(s) that support 
Java 11 have an issue. Is there somewhere we can report this?


Based on https://ci.apache.org/ I think the folks who could help hang out 
on builds@ - see https://mail-archives.apache.org/mod_mbox/www-builds/


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: svn commit: r1889815 - in /poi/trunk: build.gradle gradle/wrapper/gradle-wrapper.jar gradle/wrapper/gradle-wrapper.properties gradlew gradlew.bat

[Nick Burch]

On Wed, 12 May 2021, fannin...@apache.org wrote:

Modified:
   poi/trunk/build.gradle
   poi/trunk/gradle/wrapper/gradle-wrapper.jar
   poi/trunk/gradle/wrapper/gradle-wrapper.properties
   poi/trunk/gradlew
   poi/trunk/gradlew.bat


Need to be careful with the Gradle wrapper, see
https://issues.apache.org/jira/browse/LEGAL-570

If I have followed that issue properly, you can have the gradle wrapper in 
source control to help new developers, but it shouldn't include it in 
source releases


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Commons IO dependency

[Nick Burch]

On Tue, 4 May 2021, Andreas Beeker wrote:

Any objections?
(... I always wonder, why we haven't had commons-io as an dependency ...)


Historically, we tried to avoid having too many dependencies. Partly that 
was to help downstream users with jar-hell, partly because many of our new 
users are also new to Java programing and struggle with getting everything 
on the classpath.


Now that most people are using Maven or Gradle, rather than ant, adding + 
consuming dependencies is less of an issue for many. However, if you check 
stackoverflow, you'll see that a lot of our new users really struggle with 
the dependencies we already have...


A couple of the classes in poi/src/main/java/org/apache/poi/util/ are 
lifted straight from Commons IO. If it is just one more class we need, I'd 
suggest pulling that over too. If it is a bunch, may be worth adding the 
Commons IO dependency + removing the copied classes we already have


Thanks
Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: TRIM function issue

[Nick Burch]

On Tue, 16 Mar 2021, Emil Elmarsson wrote:
I'm not sure where to contact you or if you have an issue tracker 
somewhere, so tell me if I should ask this question somewhere else.


Apache POI uses Bugzilla for issue tracking:
https://bz.apache.org/bugzilla/describecomponents.cgi?product=POI


I was wondering why your implementation of the Excel TRIM function doesn't
remove extra whitespace, but rather only removes leading and trailing
whitespace like Java's trim function.


Probably because it's implemented with the Java trim function! :)
https://github.com/apache/poi/blob/trunk/src/java/org/apache/poi/ss/formula/functions/TextFunction.java#L153


Excel: TRIM(" a   b ") -> "a b" (3 characters)
Java: String a = " a   b ".trim(); -> "a   b" (5 characters)


If you're able to work up a fix for this, ideally with a unit test that 
shows the issue (possibly by adding a few more cells into the TRIM rows in 
test-data/spreadsheet/FormulaEvalTestData.xls ?), that'd be great!


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Apache POI for Modular JavaFX Projects

[Nick Burch]

On Tue, 23 Feb 2021, Julio Lopez wrote:

My team and I have been using POI for quite some time now on our Java
applications. We already developed a wide variety of applications that are
being used in our company, thanks to your excellent Java-Excel tools.
However, we came across this on a particular problem in integrating Apache
POI to a JavaFX modular project.


What is the problem you are facing with using Apache POI with JavaFX?

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: poi-ooxml-lite 5.0.0

[Nick Burch]

On Tue, 2 Feb 2021, Feiger, Dieter wrote:
I am locking for the source code of the poi-ooxml-lite of the Version 
5.0.0. I am planning to use this library in one of my projects, but 
before I can do this the legal department has to check the terms and 
conditions of usage. This includes also the source code.


See https://poi.apache.org/help/faq.html#faq-N10192 - these classes are 
auto-generated. If you really need the source for them, do a full rebuild 
of Apache POI from the source release and the build system will regenerate 
the source files from the OOXML XSD definitions


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: License file

[Nick Burch]

On Tue, 12 Jan 2021, Andreas Beeker wrote:

do we really need that long LICENSE file?

Junit / Jacoco and few others aren't necessary to run POI and only used to 
build it.


Do utility libraries need to be included in the official LICENSE file?


If we ship them we need to detail them


May we create an additional LICENSE file for development with POI.


I seem to recall that some projects do have different license files for 
their source and binary releases, to account for the different components 
at build and run time


The best advice on license files at the ASF can be normally found in the 
Incubator, as they're always helping new incoming projects get it right. 
If you can't get a clear answer from reading the stuff at 
 and 
 I'd suggest ping'ing 
general@incubator and asking one of the experts there for a hand!


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Website update

[Nick Burch]

On Wed, 11 Nov 2020, Marius Volkhart wrote:
I wanted to submit a change for the website (poi.apache.org) but I can't 
seem to find the files for the documentation. I found them at the last 
2.x branch tag.


The website source should be at 

I don't think it is mirrored to git like main source tree, I think the 
site currently is svn-only


Thanks
Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: XLSX wrapped in an OLE2 CompObj/Package - should WorkbookFactory handle it?

[Nick Burch]

On Sun, 11 Oct 2020, Andreas Beeker wrote:
Should we have WorkbookFactory spot this case, grab the OOXML out of the 

POIFS and try to load that?

Actually I've updated the factories to handle that case - it might not work 
...
We should have an example in our test corpus - Dominik/Tim, can you provide a 
sample file for .ppt(x) / .xls(x)?


Looks like you're right, I'd missed those commits! Support is all there in 
XSSFWorkbookFactory and friends.


I've added a unit test for this based on the sample file from Apache Tika

Thanks
Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

XLSX wrapped in an OLE2 CompObj/Package - should WorkbookFactory handle it?

[Nick Burch]

Hi All

Over on Stackoverflow  
there's a user who was getting what they thought was an embedded XSLX file 
out of a PPT, but finding it was an OLE2 wrapper with CompObj and Package 
entries. The real XLSX was in the Package part. Passing the outer OLE2 
stream to WorkbookFactory didn't work


What do people think here? Should we have WorkbookFactory spot this case, 
grab the OOXML out of the POIFS and try to load that? Update HSLF to 
optionally extract the OOXML out of the OLE2? Record the gotcha in the 
docs somewhere? Something else?


Cheers
Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Apache POI Example page is broken

[Nick Burch]

On Wed, 29 Jul 2020, Ivan Mladenov wrote:
it seems like Apache POI Examples web page is broken a lot (probably 
all) of links render 404HSSF and XSSF Examples


Hmm, looks like there's a problem, eg
http://poi.apache.org/components/spreadsheet/examples.html#business-plan
links to
http://svn.apache.org/repos/asf/poi/trunk/src/examples/src/org/apache/poi/ss/examples/BusinessPlan.java
which needs to be
http://svn.apache.org/repos/asf/poi/trunk/src/examples/src/org/apache/poi/examples/ss/BusinessPlan.java

Need to swap eg ss.example for example.ss in the package names

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: XSSF doesn't work - but HSSF works fine

[Nick Burch]

On Fri, 10 Jul 2020, Jan Fenten wrote:
for school I need to analyse some xlsm files. I thought about using your 
API. But sadly the XSSF Workbook doesn't work properly. The HSSF does 
it's job fine, but I have to handle newer Office Versions and large 
tables. The error I get is the NoClassDefFoundError.


Make sure you have all the required jars and their dependencies, see
http://poi.apache.org/components/index.html#components

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: New Bugzilla version tag

[Nick Burch]

On Sat, 15 Feb 2020, Andreas Beeker wrote:
please do the Bugzilla magic (i.e. adding the version, disabling the old 
if necessary) for the released POI 4.1.2 version.


4.1.2-FINAL added, 4.1.0-FINAL disabled for new bugs, 4.1.1-FINAL and 
4.1.x-dev unchanged + open


Thanks
Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Feedback download.html

[Nick Burch]

On Tue, 11 Feb 2020, Joris Aerts wrote:
The changelog (https://poi.apache.org/changes.html) states there must be 
a version 4.1.2 of POI, released on 02-20-2020, but I can’t find it 
nowhere.


Note that that date is in the future! It's a planned future release


Is this version actually released, and can I obtain it somewhere?


Not yet, the release candidate is being voted on
https://lists.apache.org/thread.html/r55c9e1473f454cb75aa6a94720113746306431fc569d867b3389fb67%40%3Cdev.poi.apache.org%3E

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Snapshot repository

[Nick Burch]

On Fri, 10 Jan 2020, Ghazi TRIKI wrote:
I know that the latest snapshot can be downloaded from the official 
Jenkins. However I did not find it in any of the online repositories 
because I would like to use it on a project. Did I miss a repository 
where it could be in? Or is there a policy not deploying snapshots to 
artifacts repositories?


General ASF policy is that snapshots / nightly builds / release candidates 
etc are only for testing use by involved community members, and shouldn't 
be a substitute for a proper release. So, they are not generally widely 
published, as per http://www.apache.org/legal/release-policy.html#host-rc


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Need help getting Specs for One Note versions older than 2012

[Nick Burch]

On Mon, 23 Dec 2019, Nicholas DiPiazza wrote:
Unfortunately, the formats are *drastically* different. Any attempt to 
reverse engineer would be a massive project... mostly impossible.


Pesky. I have heard of that with products that get bought-in then redone, 
unusual for Microsoft fully in-house



I can use a java program based on the linux "strings" program which will
make two passes through the older OneNote files.

Does tika or POI already come equipped with a utility that can detect 
the likelihood of useful text versus garbage?


Tika has some stuff for that, part of Tika-eval. Tim Allison is the expert 
on that bit of the code


I'd suggest you try with that, then see if you can identify some common 
header/block patterns for the bits before the real text


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Need help getting Specs for One Note versions older than 2012

[Nick Burch]

On Mon, 23 Dec 2019, Nicholas DiPiazza wrote:
I'm needing some help gathering the older Spec documentation for 
Microsoft OneNote prior to year 2012. See 
https://docs.microsoft.com/en-us/openspecs/office_file_formats/ms-onestore/ae670cd2-4b38-4b24-82d1-87cfb2cc3725 
and you'll notice that the specifications are available beginning 1.5 
which was released in 2012.


Unfortunately, you probably can't, at least not very quickly.

Microsoft do have internal docs on many (but not all) file formats coverin 
much (but not all) of the details. I believe that the public docs are 
largely written from scratch. Assuming nothing has changed in the last few 
years, the doc writers read the internal stuff, go spend quite a bit of 
time chatting with the devs, read bits of the code, then write.


You can certainly ask Microsoft if they'd mind releasing the earlier ones, 
and they'll add it to the wish-list. Unless there's minimal changes from 
the published stuff, as old formats are a bit niche, you may well be 
waiting rather a long time for them to get round to it. (Most of the focus 
is on the new stuff)



That said, the file formats from Microsoft tend to evolve, so it'll 
probably be similar to what you already have. Possibly the header block 
will be smaller (fewer flags / shorter lengths), possibly missing some 
other parts, maybe with some bits joined or split out, but likely similar. 
Good luck!


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Add 4.1.1 to Bugzilla

[Nick Burch]

On Mon, 21 Oct 2019, Andreas Beeker wrote:

On 20.10.19 22:47, Nick Burch wrote:
Done! Open versions are now 4.0.1, 4.1.0, 4.1.1 and 4.1.x-dev. Shall I 
close off 4.0.1, or leave it open for now?


Thank you Nick - please close 4.0.1.


Disabled for new bugs (already open ones tagged with this unchanged)

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Add 4.1.1 to Bugzilla

[Nick Burch]

On Sun, 20 Oct 2019, Andreas Beeker wrote:

please add 4.1.1 to our bugzilla.


Done! Open versions are now 4.0.1, 4.1.0, 4.1.1 and 4.1.x-dev. Shall I 
close off 4.0.1, or leave it open for now?


Cheers
Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Add support for LocalDate/LocalDateTime

[Nick Burch]

On Fri, 27 Sep 2019, Axel Howind wrote:
since version 4.0.1, POI requires at least Java 8. I suggest adding 
support for the new Java Date/Time API introduced with that version of 
the JDK, i.e. `Cell.getLocalDateValue()` `Cell.getLocalDateTimeValue()` 
and two matching overloads of `Cell.setCellValue()`. What do you think? 
If I get positive feedback, I'll probably find the time to submit a 
patch after mid october.


Seems like a useful update to me!

DateUtil could also potentially use similar methods too

Thanks
Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: POIFS int size - limits size to 2G

[Nick Burch]

On Mon, 15 Apr 2019, pj.fanning wrote:

I'm not very familiar with the OLE specification that POIFS implements.
Right now, we have the size parameter implemented as an int which means that
we are limited to a file size of approx 2G.


That's correct. That's a published restriction on OLE2 (CFB) files with 
the small block size

https://docs.microsoft.com/en-us/openspecs/windows_protocols/ms-cfb/d4c8084a-e583-473f-b2f4-4af13b3101d9


I had a look at changing the int to a long but this has a knock-on effect
for lots of classes.


And breaks most file format readers!


Is there a reason for us to stick with int sizes or should we look at
supporting long sizes (and bigger files)?


The "correct" (aka Microsoft documented) fix is to switch to the larger 
block size (4kb sectors) when your file gets close to 2gb. However, not 
all software copes with the switch...


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Feedback download.html

[Nick Burch]

On Thu, 8 Nov 2018, Gerardi Francesco (CAP-SST/ESM1) wrote:
We are using poi 3.9 already for excel file generation and plan to 
migrate to 4.0 using enhanced excel graph features.


There's a lot of bug fixes since 3.9! See everything above
http://poi.apache.org/changes.html#3.9 

A few things have been deprecated over the years, so you might need to 
upgrade partway, fix the deprecation warnings with the recommended fixes, 
then upgrade more.


This one has still poi v3.9. : 
http://download.eclipse.org/tools/orbit/downloads/drops/S20181031145145/


We don't control that download site, we only control the official 
apache.org download site [1] and mirrors [2], and the org.apache.poi maven 
co-ordinates [3]. You'd need to talk to the folks over at Eclipse about 
updating their mirrors and/or builds


Nick

[1] http://poi.apache.org/download.html
[2] https://www.apache.org/dyn/closer.lua/poi/release/
[3] https://repo1.maven.org/maven2/org/apache/poi/

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Proposal - host javadocs of recent versions too

[Nick Burch]

On Fri, 19 Oct 2018, Nick Burch wrote:

The immediate change would be:
* Everything at https://poi.apache.org/apidocs/ gets pushed to
  https://poi.apache.org/apidocs/dev/
* A new page at https://poi.apache.org/apidocs/ which says which
  versions you can see javadocs for online, and instructions on getting
  older ones from the released files
* Redirects for as much as possible from https://poi.apache.org/apidocs/
  to https://poi.apache.org/apidocs/dev/ to avoid breaking links
* Tweaks to the site build ant file to support the changes


These are now hopefully all done and sorted!


Then, and this is the bit which could use some more feedback:
* /apidocs/4.0/ for the latest 4.0.x release (eg when we do 4.0.1 we
  replace them)
* /apidocs/3.17/ for 3.17 final


These two are done as well


I think the main thing that remains now, if anyone feels keen is that 
building the latest Javadocs shows a bunch of broken references to other 
classes / methods / fields. If someone has maybe an hour to spare, it'd be 
great to run "ant javadocs", see the warnings for the broken references, 
and fix as many of them as possible!


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Proposal - host javadocs of recent versions too

[Nick Burch]

Hi All

As many of you will know, we have javadocs on the website, but these 
correspond to the latest (typically unreleased) code. Javadocs for a 
specific release are shipped in the binary package, and as a javadoc jar 
via Maven central (eg for IDE use). Based on stackoverflow posts though, 
this approach can sometimes cause new user confusion.


What I'd propose is that we host additionally the javadocs of the last few 
main releases, so people can easily look up the javadocs of the version 
they're using, without needing to re-find a past download / download 
something else.



The immediate change would be:
 * Everything at https://poi.apache.org/apidocs/ gets pushed to
   https://poi.apache.org/apidocs/dev/
 * A new page at https://poi.apache.org/apidocs/ which says which
   versions you can see javadocs for online, and instructions on getting
   older ones from the released files
 * Redirects for as much as possible from https://poi.apache.org/apidocs/
   to https://poi.apache.org/apidocs/dev/ to avoid breaking links
 * Tweaks to the site build ant file to support the changes


Then, and this is the bit which could use some more feedback:
 * /apidocs/40/ for the latest 4.0.x release (eg when we do 4.0.1 we
   replace them)
 * /apidocs/317/ for 3.17 final
 * /apidocs/316/ for 3.16 final
 * No betas, nothing older than 3.16
 * Using 40 not 4.0 nor 4.0.x
 * Using 317 not 3.17

On these last parts especially, what do people think?

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: POI 4.0.0 issues with new commons-compress library "InputStream of class [..] is not implementing InputStreamStatistics"

[Nick Burch]

On Sat, 29 Sep 2018, Jörn Franke wrote:
as part of the HadoopOffice library ( 
https://github.com/zuinnote/hadoopoffice/wiki) we provide the 
functionality to read office documents, such as MS Excel, on Big Data 
platforms, such as Hadoop/Hive/Spark/Flink.


We should probably list that on the website! Do you have a few paragraph 
blurb we can use?



I want to release a new version supporting POI 4.0.0, but I have one
remaining blocking issue: The Big Data platforms use an old version of
commons-compress (between 1.4.x and 1.9.x). This means I am always running
into the exception in ZipArchiveThresholdInputStream "InputStream of class
[..] is not implementing InputStreamStatistics" (
https://svn.apache.org/viewvc/poi/trunk/src/ooxml/java/org/apache/poi/openxml4j/util/ZipArchiveThresholdInputStream.java?view=markup=1832789
).


We need that for security reasons - newer Java versions won't let us 
protect against zip bomb attacks as they inconveniently hide the expansion 
stats, so we had to switch to commons to guard against it.



Unfortunately, updating these platforms to the latest commons-compress is
very intrusive and for many organizations not possible.


Wave some CVEs at them and see if you can tempt an upgrade?

If not, you'd probably need to work with the commons folks to backport the 
zip stats stuff to your old version, so you can keep the security stuff we 
need? dev@commons is moderately quiet and fairly friendly :)


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Worth doing a 4.0.1 release soon?

[Nick Burch]

Hi All

It looks like we've had a few issues with 4.0.0, which while they're not a 
problem for most people, are a significant issue for a few of our user's 
platforms.


Good news - they've all been fixed, and fixed super speedily, yey! :)

Bad news - they're only fixed in svn...

What do people think about doing a 4.0.1 release shortly? Any other bugs 
we want to try to get fixed before we roll it? Any volunteers to release 
manager it?


Cheers
Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Speaking on POI at China Open Source Conference in October

[Nick Burch]

On Sat, 15 Sep 2018, Dave Fisher wrote:

I’ll be at Apachecon Montreal, anyone else?


I'll be there! Happy to look at draft slides then, and offer advice :)

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Provide alpha releases via Apache repo

[Nick Burch]

On Fri, 7 Sep 2018, Alain FAGOT BÉAREZ wrote:
This was also the way I started to contribute, first testing my changes 
against my pet project by local publishing Maven artifacts to my .m2 for 
a first try.


However, there had been a steep step until I managed to gather all 
necessary info and… the will to solve an issue that was already several 
years on BugZilla.


IIRC it's something like "ant clean mvn-install", which should do a build 
of the jars then the poms then prod maven to install it all locally


If someone with a build that currently passes the unit tests wants to try 
that, confirm/fix, then update the website, that'd be great :)


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Provide alpha releases via Apache repo

[Nick Burch]

On Fri, 7 Sep 2018, Dave Fisher wrote:
Release Policy is such that publishing a snapshot to a public, open 
repository is not allowed.


See http://www.apache.org/legal/release-policy.html#release-definition 



Just what I was about to post!

So, nope, not allowed, sorry. We can make it available for testing, just 
not to people who aren't actively invovled and largely aware of the risks 
/ lack of foundation sign-off etc.



If we want to produce Alpha and Beta then we will need to VOTE on these.


That's what we have done in the past in the run up to big releases / 
changes. I don't know how that'd fit with our new semver-style world 
though?


Alternatively are we using GitBox so that the repos is on GitHub? If so 
then they can fork on GitHub to test.


We don't use the read/write gitbox, but we do have a github mirror, and 
IIRC some docs on the website about forking it there, opening pull 
requests etc


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: POI 4.0.0 in bugzilla

[Nick Burch]

On Thu, 6 Sep 2018, Andreas Beeker wrote:
please check the bugzilla version tags and decide how to proceed with 
4.0.0 and the upcoming - in a few months - 4.0.1 version.


I've gone for
 * 4.0.0-FINAL
 * 4.0.x-dev

I've also disabled 3.17-dev and a couple of the older 3.1x-FINAL versions 
from accepting new bugs (old bugs remain on them)


Hopefully that makes sense, sorts well, and ought to be fairly easy to 
maintain?


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Help on POI JVM Languages Kotlin

[Nick Burch]

On Wed, 6 Jun 2018, Luiz Oliveira wrote:

I just finish some code in Kotlin using Apache POI and I see that you need
a want some of this. The repository here https://github.com/
luizleiteoliveira/kotlin-poi-example

How can I add my code on POI website?


Do you mean adding a Kotlin section to 
https://poi.apache.org/poi-jvm-languages.html ?


If so, is this the right file to add to the page?
https://github.com/luizleiteoliveira/kotlin-poi-example/blob/master/src/main/java/com/example/kotlinpoi/poi/entity/ExcelKotlinExample.kt
And any text / build guides / etc to go with it?

Thanks
Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

3.18? Or push for 4?

[Nick Burch]

Hi All

We released 3.17 back in September, which is over 6 months ago. The 
changelog  says we've done quite a bit

since then

What do we think about another release? Do a 3.18, then breaking changes 
for 4.0 afterwards to release that in the summer? Or just finish stuff up 
and push 4.0 out soon?


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Single Java 9 module jar?

[Nick Burch]

On Thu, 19 Apr 2018, Michal Schielmann wrote:

Is there any way to run poi with ooxml at the moment on Java 9?


Apache POI works just fine with Java 9 on the classpath

Java 9 / 10 module path is not supported at the moment

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: git repo is not in synch with svn

[Nick Burch]

On Wed, 21 Mar 2018, pj.fanning wrote:
Anyone know who to contact about https://github.com/apache/poi being out 
of sync with svn?


What happens is that SVN syncs with http://git.apache.org/ , then Github 
syncs with that


Drop an email to us...@infra.apache.org with details of the last SVN 
revision that sync'd over OK, the git commit hash of that commit, and the 
SVN revisions that have failed to sync. They'll then investigate / prod 
the sync system / etc


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Single Java 9 module jar?

[Nick Burch]

On Fri, 2 Mar 2018, pj.fanning wrote:
Could we consider keeping all the jars but ensure that no package spans 
jars?


We'd need to re-package all the jars by component / area. You couldn't 
have a HSSF only one for example, it'd need to be HSSF+XSSF+SS if there 
was to be no package spanning.


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Single Java 9 module jar?

[Nick Burch]

Hi All

From my understanding of Java 9 modules, you basically can't have one 
package split across multiple jars. This is an issue for us, as we do 
split things betwen core, scratchpad and ooxml, to permit people to 
minimise their jar sizes by excluding components they don't want.


Longer term, we could maybe look at having per-component jars. In the 
short term... I think if we just bundled the 3 jars together and dropped 
in a module info file, that'd fix things for Java 9 module users at the 
expense of them having to get all of POI (no exclusions).


Build wise, it'd be really simple - just generate an extra jar with the 
contents of the 3 base jars included, along with the module info file, and 
call it something like poi-all-java9.


What do people think of this idea? Any downsides? Comments? Suggestions?

Also, would David North like to share his Java 9 module slides from his 
talk a few weeks back? :)


Cheers
Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

RE: publishing poi xmlbeans jars

[Nick Burch]

On Wed, 28 Feb 2018, pj.fanning wrote:

http://mail-archives.apache.org/mod_mbox/attic-general/201802.mbox/%3C1519744891.257410.1285160920.6801732D%40webmail.messagingengine.com%3E

Looks like we're not going to make much progress on resurrecting Apache
XMLBeans.


The big problem is with Maven users:
 * Anyone using Maven to depend on xmlbeans explicitly won't know to
   switch to org.apache.poi:xmlbeans to get the fixes
 * Maven won't know that the the old 2.6.0 release and our new fix are
   the same thing, so will happy pull both in

The smaller problem is by-hand-dependency users - the number of people on 
stackoverflow or the mailing list who try combining old and new POI jars + 
get stuck is rather high - so the chances of those people getting it right 
is near zero


As long as all existing XMLBeans unit tests pass, it should be safe 
(+sensible) for existing xmlbeans users to upgrade to our fixed version.



Could someone help work up a good explanation of the jar-hell + maven 
issues, so we can go back to the attic with a reason why we should be 
allowed to do it? (My reading of the attic wasn't a flat-out no, just a 
"no unless you have a good reason otherwise")


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: XMLBeans in 4.0.0 release

[Nick Burch]

On Wed, 10 Jan 2018, pj.fanning wrote:
I don't fancy trying to merge the xmlbeans and poi ant builds. I've been 
trying for an hour or more and still getting issues.


There's no need to do this!

We'd keep the two codebases separate, and probably release at different 
times. (XMLBeans hardly ever, except when we find bugs) All that would 
happen is that we'd swap the group ID, add POI to the release names, and 
update the readme etc to refer to POI as the PMC providing the oversight


So, assuming we're releasing it as POI rather than getting it out of the 
attic:

 * Copy xmlbeans trunk in svn to
   https://svn.apache.org/repos/asf/poi/xmlbeans/trunk
 * Merge in the fixes
 * Update the bin/src packages to have poi in them, eg
   xmlbeans-poi-2.6.5-bin.zip
 * Update the poms to be under org.apache.poi rather than xmlbeans
 * Test, vote, release!

Or, if we want to take xmlbeans out of the attic, we'd do:
 * Vote to take it on
 * Wait for attic PMC to confirm
 * Get infra to unlock svn for xmlbeans
 * Merge in fixes
 * Test, vote, release!

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: XMLBeans in 4.0.0 release

[Nick Burch]

On Mon, 8 Jan 2018, pj.fanning wrote:
I understand the desire to replace XMLBeans altogether but I don't think 
we have enough developer time available to do this in the 4.0.0 time 
frame.


Given the amount of work that'd take, and a desire for other features and 
bugs to be worked on in the mean time, I'd agree!




1. Keep Apache XMLBeans 2.6.0 dependency - if users run into issues with
XMLBeans, they can choose to configure their own project builds to swap in
https://github.com/pjfanning/xmlbeans instead.


I don't think that's a good new-user experience (nor upgrading user 
experience), the "default" they're pushed towards is known to have 
problems and they'll likely struggle with them for a while before 
discovering the fix



2. Make https://github.com/pjfanning/xmlbeans the default dependency
3. Delay POI 4.0.0 release until we can get an adequate replacement for
XMLBeans.


I'd suggest we either go back to the Attic PMC and do a fixed release of 
XMLBeans there (based on your fixed github fork), or fork XMLBeans within 
POI (based on your fixed fork) and release a fixed release ourselves.


Either way, we then have POI depend on a known-fixed rather than 
known-broken jar by default!


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: svn commit: r1816500 [1/23] - in /poi/site: publish/apidocs/ publish/apidocs/org/apache/poi/class-use/ publish/apidocs/org/apache/poi/common/usermodel/fonts/ publish/apidocs/org/apache/poi/common/

[Nick Burch]

On Mon, 27 Nov 2017, abea...@apache.org wrote:

URL: http://svn.apache.org/viewvc?rev=1816500=rev
Log:
update status and javadocs for closed github-68


What jdk did you use to generate the javadocs? Generally we try to use 
(oldest POI supported + openjdk), so currently openjdk 8. Everyone using 
the same one tends to stop javadoc changing everything for no good reason, 
then creating huge diffs!


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: [DISCUSS] Getting a fixed version of XMLBeans

[Nick Burch]

On Thu, 9 Nov 2017, Dave Fisher wrote:
e) take XMLBeans out of the attic. Fix the bugs and make maintenance 
releases as 2.7, 2.8, etc. This would be an official version. Probably 
needs a board resolution.


I think it might be enough for the Attic PMC to vote with 3 +1s to 
transfer the responsibilities over to us. Probably best to ask them 
though!


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Suspected Spam:Re: Non-maintainer upload of bugfixes for the XMLBeans library in the Attic

[Nick Burch]

On Thu, 9 Nov 2017, Dustin Spicuzza wrote:

I'm curious, does anyone know the rationale for closing down the
XMLBeans product in the first place? It doesn't seem to be stated on the
XMLBeans website.


See https://attic.apache.org/ for details of how the attic works and how 
things get put there. Basically though, the community died, and there 
weren't enough people left to provide proper oversight of the project


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Excel formula saving oddity

[Nick Burch]

On Fri, 9 Jun 2017, Greg Woolsey wrote:

However, for unknown reasons not documented in Excel help as far as I can
see, the named range reference has to be specified as:

'workbook file name.xlsx'!named_range

however, when saving, the XML doesn't store the file name, but rather:

[0]!named_range

which of course is not recognized by POI's formula parsing.


I've come across this before. There's a little bit of support for these in 
XSSF, if you search the XSSF unit tests for "]!" you'll see some



My question is, does it seem reasonable, when the expression fails to parse
otherwise, to check if it starts with

[0]!


It can parse, but you have to supply the referenced workbooks for 1+. It's 
a bit messy though - some things expose the [#] form, some things can work 
with the [name.xlsx] form. Ideally we'd make it possible/easy to replicate 
the Excel behaviour


#56737 and #56752 are two bugs that spring to mind,
TestXSSFFormulaParser#formulaReferencesSameWorkbook() and
TestXSSFFormulaEvaluation#testReferencesToOtherWorkbooks() are two unit 
tests to start with


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

XSSF tables - anyone know much about them?

[Nick Burch]

Hi All

Yesterday I came across a post on stackoverflow with a user having trouble 
with XSSF tables. In the process of fixing the bug they'd found, I also 
had a go at tidying up their code to reduce the need to touch the raw CT 
classes from end-user java code.


As part of that, I've added a couple of methods, but I don't know enough 
about the table stuff to know if they're right for all cases, plus there's 
a few more bits we might want to improve


Anyone know about the XSSF tables stuff? If so, any chance you could 
review my commits from yesterday, as well as looking at the updated 
CreateTable example, and let me know your thoughts / go ahead and fix more 
things?


Cheers
Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: POI @ ApacheCon

[Nick Burch]

On Mon, 15 May 2017, David North wrote:
Who else is in Miami for ApacheCon? Do we have critical mass for an 
in-person chat over breakfast/beer/... ?


I'm here!

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: svn commit: r1795002 [3/3] - in /poi/trunk: ./ src/java/org/apache/poi/ss/usermodel/ src/ooxml/java/org/apache/poi/xssf/model/ src/ooxml/java/org/apache/poi/xssf/usermodel/ src/ooxml/testcases/org

[Nick Burch]

On Fri, 12 May 2017, Greg Woolsey wrote:

Argh, fat fingered it before I could type my comment.


No worries! We all do it eventually... Don't forget you can edit the 
message later:

http://stackoverflow.com/questions/2184317/change-svn-commit-message-retroactively

 My implementation is really ugly imho, but it works.  I couldn't figure 
out a way to get XMLBeans factories on the CT classes to initialize the 
full objects from an XML fragment, only top-level documents, so I fake 
up a styles.xml for each built-in style found, and parse those.  Takes 
about 2.5 seconds on my system, which is why I made it lazy.


I *think* I got that working in the past elsewhere. If you let me know 
which class has the faking in, and the best unit test to use to check 
with, I can try and have a go!


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: svn commit: r1794084 - in /poi/trunk: src/java/org/apache/poi/hssf/usermodel/ src/java/org/apache/poi/ss/formula/ src/java/org/apache/poi/ss/usermodel/ src/ooxml/java/org/apache/poi/xssf/usermodel

[Nick Burch]

On Fri, 5 May 2017, gwool...@apache.org wrote:
Currently only implemented for XSSF, as there is no API available for 
HSSF conditional formatting rule number formats (if it is even in the 
files).


Best way I can think of to check if it's stored - build a xlsx file with 
this in, save, open, save-as xls, open, save-as xlsx, check. If the info 
survives that trip then it's somewhere in the xls format!


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: New Bugzilla version entry

[Nick Burch]

On Wed, 19 Apr 2017, Andreas Beeker wrote:

please add a new version entry to our bugzilla.


I've added two new versions:
 * 3.16-FINAL
 * 3.17-dev

I've also edited two older versions to prevent new bugs being raised 
against them (old bugs are unchanged):

 * 3.15-dev
 * 3.12-FINAL

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

NPOIFS, files and the read-only flag

[Nick Burch]

Hi All

Currently, in NPOIFS (NPOIFSFileSystem + POIFSFileSystem, but not 
OPOIFSFileSystem), you can either load from a File or from a Stream. 
With a Stream, everything gets buffered into memory


When loading from a File, you can load read-write or read-only. In 
read-write mode, we use NIO to mmap the file so we can quickly + low 
memory read + write.


The other case, File + read-only, isn't so ideal. We need to support 
both people doing true read-only cases, eg reading or text extraction, 
plus people doing read-template + write or read-change-saveas cases. To 
support that, we buffer into memory.


Taking another look at the NIO stuff for bug #60670, I was wondering 
about splitting those two cases out. For true read-only cases, we could 
then mmap in read-only mode. For cases where you're loading a file and 
want to change + save elsewhere with the original unchanged, I think we 
could mmap in "private" mode, which looks to be copy-on-write, so only 
changes would be buffered in memory.


Any thoughts? Is it worth making the change? Can anyone spot any 
downsides to the change? Anyone think of any good enum names for the 
different modes if we replace the read-only boolean with a 3-way enum?


(Changing all file-based access to mmap, rather than a mix of mmap + 
in-memory buffers, should make it easier to tweak the logic to mmap 
bigger chunks)


Cheers
Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Build failed in Jenkins: POI-DSL-Maven #93

[Nick Burch]

I think the key errors for this are things like:

[WARNING] 
/home/jenkins/jenkins-slave/workspace/POI-DSL-Maven/sonar/main/src/test/java/org/apache/poi/ss/usermodel/TestDateUtil.java:[113,27] 
unmappable character for encoding US-ASCII
[WARNING] 
/home/jenkins/jenkins-slave/workspace/POI-DSL-Maven/sonar/main/src/test/java/org/apache/poi/ss/usermodel/TestDateUtil.java:[113,28] 
unmappable character for encoding US-ASCII


Does someone need to tweak the Jenkins job to tell it we're now using UTF8 
encoding? Or do we need to use \u escape style encodings in the test 
file for safety?


Nick

On Fri, 10 Feb 2017, Apache Jenkins Server wrote:


See 

Changes:

[nick] Update the big file test to use POIFSFileSystem.create(File), and tweak 
javadocs

--
[...truncated 1055 lines...]
Running org.apache.poi.hssf.record.chart.TestSeriesLabelsRecord
Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.chart.TestSeriesLabelsRecord
Running org.apache.poi.hssf.record.chart.TestLinkedDataRecord
Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.chart.TestLinkedDataRecord
Running org.apache.poi.hssf.record.chart.TestSeriesIndexRecord
Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.chart.TestSeriesIndexRecord
Running org.apache.poi.hssf.record.chart.TestObjectLinkRecord
Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.chart.TestObjectLinkRecord
Running org.apache.poi.hssf.record.chart.TestAreaRecord
Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.chart.TestAreaRecord
Running org.apache.poi.hssf.record.chart.TestDataFormatRecord
Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.chart.TestDataFormatRecord
Running org.apache.poi.hssf.record.chart.TestSeriesRecord
Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.chart.TestSeriesRecord
Running org.apache.poi.hssf.record.chart.TestTextRecord
Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.chart.TestTextRecord
Running org.apache.poi.hssf.record.chart.TestFrameRecord
Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.chart.TestFrameRecord
Running org.apache.poi.hssf.record.chart.TestAxisUsedRecord
Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.chart.TestAxisUsedRecord
Running org.apache.poi.hssf.record.chart.TestFontIndexRecord
Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.chart.TestFontIndexRecord
Running org.apache.poi.hssf.record.TestExtendedFormatRecord
Tests run: 4, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.TestExtendedFormatRecord
Running org.apache.poi.hssf.record.TestFeatRecord
Tests run: 3, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.TestFeatRecord
Running org.apache.poi.hssf.record.TestCommonObjectDataSubRecord
Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.TestCommonObjectDataSubRecord
Running org.apache.poi.hssf.record.TestObjRecord
Tests run: 5, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.TestObjRecord
Running org.apache.poi.hssf.record.TestDrawingGroupRecord
Tests run: 3, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.TestDrawingGroupRecord
Running org.apache.poi.hssf.record.TestNoteStructureSubRecord
Tests run: 3, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.TestNoteStructureSubRecord
Running org.apache.poi.hssf.record.TestCFHeaderRecord
Tests run: 4, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.TestCFHeaderRecord
Running org.apache.poi.hssf.record.TestNameCommentRecord
Tests run: 1, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.TestNameCommentRecord
Running org.apache.poi.hssf.record.TestCFRuleRecord
Tests run: 8, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0.128 sec - in 
org.apache.poi.hssf.record.TestCFRuleRecord
Running org.apache.poi.hssf.record.TestWriteAccessRecord
Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0 sec - in 
org.apache.poi.hssf.record.TestWriteAccessRecord
Running org.apache.poi.hssf.eventusermodel.TestMissingRecordAwareHSSFListener
Tests run: 8, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0.011 sec - in 
org.apache.poi.hssf.eventusermodel.TestMissingRecordAwareHSSFListener

Re: Saving modified Workbook in place

[Nick Burch]

On Wed, 1 Feb 2017, Murphy, Mark wrote:
I thought that there was a way to save a workbook in place when using a 
File type input. But I can't find the secret sauce.


The method will be write() with no arguments:
https://poi.apache.org/apidocs/org/apache/poi/POIDocument.html#write()

Off the top of my head, I can't remember how much more work there is to do 
to finish the HSSF support for it...


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Evaluating expressions outside a cell value context

[Nick Burch]

On Sat, 28 Jan 2017, Greg Woolsey wrote:

As noted in one of the method JavaDocs, we also need to expose and make use
of the ConditionalFormattingRule "priority" attribute.  That's key to
matching the right rule when more than one rule applies to a cell.  Only
the first match in priority order is applied.


Your slight challenge is that not all Conditional Formatting rules have a 
priority... XLSX ones do, and newer XLS ones based on CFRule12Record (sid 
= 0x087A) do, but the older XLS ones (CFRuleRecord / 0x01B1) don't. I'm 
not sure what Excel does for those, but my hunch (based on our API) is 
that it uses their order as a priority.



I've created a fork in GitHub for this, and committed a stab at 
high-level API methods that could be added to the FormulaEvaluator 
interface:

https://github.com/WoozyG/poi/commit/d44fee7bd03ed450af589467ec90e2581b9f2b16$


FormulaEvaluator is an interface, which we have 4 implementations of in 
our codebabse, and I'd guess that other complex users of POI will have 
dozens more. I'm not sure, therefore, that we want to be putting all of 
the CF and DV logic methods on there, especially as it'll be common to all 
implementations


The HSSF classes for CF all use org.apache.poi.ss.formula.Formula which is 
PTG based. The HSSF classes for DV seem to store the raw PTGs.


If we added two new SS usermodel classes, eg 
ConditionalFormattingEvaluator and DataValidationEvaluator, these could be 
classes (not interfaces) with your proposed new methods on. They could 
hold the logic (once) for all formats (as it's basically the same on all) 
for priority, checking etc


Doing that would also mean that "our" new classes could call out to our 
existing low-level ones to evaluate formulas. That would mean we wouldn't 
have to make a breaking change to the FormulaEvaluator interface too


Might that work for you?


No implementations have been done yet, and the Vaadin comments indicate
HSSF doesn't parse conditional formatting properly or something, and can't
be evaluated correctly currently.  I don't know exactly what they found
wrong, and it's rather annoying they didn't file any bugs.


I think that comment is out of date, from before the CF work in 3.13

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Evaluating expressions outside a cell value context

[Nick Burch]

On Fri, 27 Jan 2017, Greg Woolsey wrote:

This can be done today by accessing the evaluate* package-scoped methods in
o.a.p.ss.formula.WorkbookEvaluator from a class defined in that package.
This requires some understanding of formula parsing and the resulting Ptg
array, and uses a method explicitly noted to not be for public use.

A general-purpose public method would look something like the other
evaluate* methods:

public ValueEval evaluate(String formula, CellReference ref, FormulaType
formulaType) {


We have some evaluation related interfaces, maybe we could put one of 
those onto the Conditional Formatting and Data Validation classes, then 
add a evaluate method on WorkbookEvaluator based on that? (Add / expose)



However, we may also want a more semantic way to handle evaluating data
validations and conditional formatting.  Using them has several
complexities and wrinkles that make it hard to do right in all cases


I'd lean towards adding ConditionalFormattingEvaluator and 
DataValidationEvaluator classes, which do the hard work of working out 
what applies, which of two competing things win etc. Put user-facing 
methods on those. Have those delegate down to the new WorkbookEvaluator to 
check if a given formula is true / false / etc



This is definitely a candidate for a branch/discuss/refine/merge workflow,
which I'm willing to pursue.  I haven't done SVN branches and merges in
about 7 years, but how bad can it be? :D


You can always branch it in github if needed!


One wrinkle is that I've seen the Vaadin code that does the conditional
formatting, because I'm a licensed customer, but it is using their
commercial license.  That means I can't do the POI work on the more complex
and complete steps above.


I think it's publically available source, just not under a license we can 
use. I've certainly found the source when tracking down some odd bugs! You 
won't be able to re-use their code, but I don't think that having seen it 
precludes you writing the POI one



I'm also willing to get my boss to contact Vaadin and see if they will
release the relevant bits of conditional formatting code to us for use in
POI with the Apache 2 license.  She's very good at that sort of thing.  I
wouldn't want it directly, but the patterns it follows likely mirror some
of what we need, and I don't want to cause a problem where there doesn't
need to be one.


Maybe the way to sell it is not as "can you give" or "can you relicense", 
but "would you like to make someone else take over the development and 
maintenance for you" :)


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Time for another release?

[Nick Burch]

On Wed, 25 Jan 2017, Greg Woolsey wrote:
I have something I'm going to file an issue about tomorrow, but it could 
wait. Although having it in would help fix some bugs I'm working around 
with Vaadin Spreadsheet.  It would be my first official commit, so I 
don't want to do it near a code freeze, just in case.  It's a helper to 
expose the formula evaluation code more cleanly for evaluating formulas 
not tied to a specific cell, like conditional formatting and data 
validation expressions.


I'd suggest you create a branch of trunk, and commit to that. I'm happy to 
help review, especially the conditional formatting angle. Once we're all 
happy, and the release is done, merge away!


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Time for another release?

[Nick Burch]

Hi All

We seem to be about 2 months after our last beta release now, and about 4 
months since our last final release.


Is it worth doing another beta soon, to let people test possible breaking 
changes from some of the enum and deprecated stuff? Or hold off a bit 
longer, and just go for 3.16 final?


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Using Apache Commons IO

[Nick Burch]

On Tue, 17 Jan 2017, Javen O'Neal wrote:

In the spirit of "the best code is no code", how would you feel about
replacing our endian classes and IOUtils with Apache Commons IO?

The downside is that it adds a dependency.
https://poi.apache.org/overview.html#components


How many classes do we need? Do those classes include all the methods we 
need, or are there gaps?


(Having dealt with yet another StackOverflow question today from someone 
stuck on a really old POI version, and knowing that the ones who make it 
to StackOverflow are actually the better ones, I'd rather inline a few 
classes / do nothing instead of adding a dependency that people can get 
wrong)


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: svn commit: r1771220 - in /poi/site: publish/download.html src/documentation/content/xdocs/download.xml

[Nick Burch]

On Thu, 24 Nov 2016, one...@apache.org wrote:

URL: http://svn.apache.org/viewvc?rev=1771220=rev
Log:
remove snarky disclaimer

The POI nightly builds are run on the https://builds.apache.org/view/POI/;>Jenkins continuous integration server.
- Note that these are not officially verified builds and they are not 
endorsed or even supported by the POI team.
 


We're not supposed to direct "normal" users to nightly builds, see 
http://www.apache.org/dev/release.html#host-rc . So, some sort of 
disclaimer and warning is officially required


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Interesting question on SO

[Nick Burch]

On Mon, 21 Nov 2016, Murphy, Mark wrote:

http://stackoverflow.com/questions/40723370/was-opensagres-allowed-to-use-apache-poi-like-packages

Does ASF defend its package names, or do we just ignore these kinds of 
things and folks use them at their own peril?


I had a quick look at http://apache.org/foundation/marks/ but couldn't see 
anything there about java or maven package names. I have a feeling there 
is something somewhere that I saw a few years ago...


If you have a bit of time, would you mind asking on trademarks@ [1] for 
advice / pointers?


Nick

[1] http://www.apache.org/foundation/marks/contact

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: [Bug 59907] HSSFClientAnchor.setAnchorType API was broken (problem with jasperreports-5.1.0 and poi-3.14)

[Nick Burch]

On Tue, 8 Nov 2016, bugzi...@apache.org wrote:
(HX)SSFClientAnchor.setAnchorType(int) was restored in r1760617, which 
was after the release of POI 3.15 final. This will be available again in 
POI 3.16 beta 1 and 3.16 final.


Is it worth doing a 3.16 beta 1 release quite soon, so that people 
affected by this (and the related) issue can easily upgrade to get the 
fix?


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Document validation hook in POIXMLDocument

[Nick Burch]

On Sun, 30 Oct 2016, Mark Murphy wrote:

Is there a validation hook in POIXMLDocument that will allow a document to
validate itself before it writes itself to disk?


Maybe prepareForCommit() ? (Need to call superclass too!)

Here is the use case. An XWPFDocument header or footer must contain at 
least an empty paragraph. If a header or footer is created without a 
paragraph, this hook would let me determine this, and just put one in 
there rather than generating a corrupt document.


You could do it in prepareForCommit or commit of XWPFHeaderFooter

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: svn commit: r1766520 - /poi/site/src/documentation/content/xdocs/status.xml

[Nick Burch]

On Tue, 25 Oct 2016, Javen O'Neal wrote:

Nick said in r1766520:
Raw Release Notes for a few more releases, needs someone else to re-format
them into the summary form

Would it be better to paste the summary text into status.xml and write
document2changes.xsl to create the list items by splitting on newlines?


Probably needs a bit more than that - there's the bug numbers that need 
tagging, and possibly different sections for major changes / breaking 
changes / security fixes. Probably not too much work, but I ran out of 
time after the work getting the notes out of SVN!


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Apache POI 3.15 released

[Nick Burch]

On Fri, 21 Oct 2016, Javen O'Neal wrote:

In the meantime, what's the best way to retrieve old release notes?
Were they committed to a private or public SVN server?


We've been using svnpubsub for dist for a while (2011 IIRC), so you ought 
to be able to get them from there by asking SVN for the release notes file 
at older revision numbers


(Shout if you can't do that easily, and I can try to script something up)

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Functionality injection technique for XWPF

[Nick Burch]

On Mon, 17 Oct 2016, Mark Murphy wrote:
I found I should be able to do it using reflection to avoid long strings 
of instanceof conditions which is really the same as duplicating the 
code, but all inside one class. Any issue with using Reflection?


If you post some of the code you're working on, with some ideas, we can 
hopefully all take a look and see if we can help improve things


Feel free to post to a github gist, a branch on your own github, a branch 
in svn, patch to jira etc!


(Quite a bit of XWPF has come from user contributions, so may well be 
missing some overall architecture improvements that piece-meal fixes 
didn't require)


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Apache POI 3.15 released

[Nick Burch]

On 29/09/16 15:29, Javen O'Neal wrote:

Do we archive the release notes somewhere so that someone can read a
high-level summary of the changes for each release? Digging through
the dev@ email list archive to find the release notes draft or
searching a web archiver of one of the mirrors at a date in the past
seem suboptimal. There is no release notes file in the subversion tag
or a page on our site that summarizes the changes in a release more
concisely than changes.html.

The current https://poi.apache.org links to both 3.14 and 3.15, but we
overwrote the 3.14 release notes on
https://www.apache.org/dyn/closer.lua/poi/release/RELEASE-NOTES.txt
and all mirrors when we released 3.15.


After a final release, the homepage and download page should only ever 
link to the most current final version. No other versions should live on 
the dist area, otherwise infra get grumpy...


When we do an alpha or beta release, then homepage + download should 
link to both


If we don't already, it might be good to deploy the release notes twice, 
once as "RELEASE-NOTES.txt" (overwritten often), once as 
"RELEASE-NOTES-.txt". The latter would then always be 
available from archive.apache.org even after a newer final release. 
Thoughts?


Nick


-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Minimal Java version requirement

[Nick Burch]

On Sun, 25 Sep 2016, Yaniv Kunda wrote:

What are the plans in regards to bumping the minimal Java runtime version
for POI?


None. Until there's a compelling feature to upgrade for, or enough users 
have upgraded, bumping the version will only make life hard for users. 
(Most people stuck on old JVMs aren't doing it by their own choice!)



Currently set at Java 6, I believe it should be at least Java 7, which is
already 18 months EOL.


POI works just fine on newer Java versions, so you're not affected if 
you've already upgraded!


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: [VOTE] Apache POI 3.15 RC3

[Nick Burch]

On Wed, 21 Sep 2016, David North wrote:

Nick - please can you sort the Bugzilla versions?


Added 3.15-FINAL and 3.16-dev

3.14-dev and 3.11-FINAL disabled for new bugs

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: [VOTE] Apache POI 3.15 (RC1)

[Nick Burch]

On Mon, 22 Aug 2016, Javen O'Neal wrote:

   switch (cell.getCellType()) {
   case Cell.CELL_TYPE_NUMERIC:


I have fixed the backwards compatibility problem in r1757235. See [1]
for more info.


Thanks! Some of this stuff can be a real pain :(

(Personally I figure I've still got another 1-2 years of faff to finish 
the POIFS -> OPOIFS, NPOIFS -> POIFS switch because of that stuff...)


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: [VOTE] Apache POI 3.15 (RC1)

[Nick Burch]

On Mon, 22 Aug 2016, Javen O'Neal wrote:

Cell.CELL_TYPE_NUMERIC -> CellType.NUMERIC


We can't make a backwards incompatible change of this magnitude just like 
that. It would break almost every single user of POI's code, as well as 
breaking almost every example of using POI that's out there!


(Do a google search for "poi switch cell_type_numeric" and see how many 
hits you get!)



Aside from including this in the release notes, can we add some kind
of indicator on https://poi.apache.org/changes.html indicating that a
bug add/fix is a breaking change? This would save time for developers
reviewing the changelog and updating their code without needing to
read through every bug.


Anything that's a breaking change needs to be in a different area of the 
release notes


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Image Blocks Text POI XSLF

[Nick Burch]

On Tue, 9 Aug 2016, Javen O'Neal wrote:

Slightly off topic...

On Aug 9, 2016 3:10 PM, "Andreas Beeker"  wrote:

Then you unzip both files and find out the differences in the xml files.
If you know which xml element to change, than you can modify it via the

xmlbeans object.

Do we have any tools to make this even quicker?


I think OOXMLPrettyPrint might do it, we've also got some format-specific 
dev tools that can sometimes help too


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: POI xlsx indeterminism

[Nick Burch]

On Mon, 1 Aug 2016, armin.winte...@ldbv.bayern.de wrote:
we've been experiencing an indeterminism problem with POI's xlsx format, 
when generating hash values with the following method in testng test 
cases:


XLSX uses Zip files, which contain within them file dates. If you're 
comparing the outer zip file, you would expect an otherwise identical file 
to change every second as the datetimes move on


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: 3.15 beta 3 soon?

[Nick Burch]

On Sun, 31 Jul 2016, Nick Burch wrote:

On Sun, 31 Jul 2016, Mark Murphy wrote:
So does this mean the trunk is locked for beta 3? I have a couple changes 
to add, but I will not do so if we are going to roll a new beta soon.


Could you hold off for a couple of hours? Just about to start...


Actually, the tigris.org server seems to be down so I can't run the 
release-prep1 step. So, unless someone else feels like stepping up to cut 
the release who can get to it (release notes already written up in SVN!), 
looks like you're good to go on trunk for non-breaking changes


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: 3.15 beta 3 soon?

[Nick Burch]

On Sun, 31 Jul 2016, Mark Murphy wrote:
So does this mean the trunk is locked for beta 3? I have a couple 
changes to add, but I will not do so if we are going to roll a new beta 
soon.


Could you hold off for a couple of hours? Just about to start...

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: svn commit: r1754521 - in /poi/trunk: ./ maven/ src/java/org/apache/poi/hssf/usermodel/ src/java/org/apache/poi/ss/usermodel/ src/ooxml/java/org/apache/poi/xssf/streaming/ src/ooxml/java/org/apach

[Nick Burch]

On Fri, 29 Jul 2016, dno...@apache.org wrote:

Add a dependency on commons collections 4, also ASL licensed so no problem.


Can you update http://poi.apache.org/overview.html#components to reflect 
the new dependency?



   poi/trunk/maven/poi-ooxml.pom


Since HSSF uses it too, should it be the main poi pom that gets updated 
with the dependency?


The sonar pom might need updating with it too

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: 3.15 beta 3 soon?

[Nick Burch]

On Fri, 22 Jul 2016, Nick Burch wrote:

On Fri, 22 Jul 2016, Andreas Beeker wrote:
Who will roll the release? As usual, I'll be the fallback (as long as my 
key is valid ...)


Anyone else want to give the docs a try? :)


Actually, thinking about it, it'd be good to try to get POI 3.15 final out 
around the 20th of August, just in time for our 15 year anniversary![1]


I can roll a 3.15 beta 3 release this weekend, if that works for everyone? 
That'd give a couple of weeks for people to test it + do any remaining bug 
fixes, in time for a final release just before 15 years!


Nick

[1] http://poi.apache.org/changes.html#0.1

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Volunteers wanted - Test HSSF + HSLF in-place write + help document

[Nick Burch]

Hi All

About 10 years after someone first asked for it, we've finally gotten 
in-place-write and write-to-file available for HSSF! (And HSLF too)


This means that if you open a HSSFWorkbook or HSLFSlideShowImpl from a 
File, in read-write mode (not read-only, which is the default), you can 
write your changes back to the same open file without issue or error. You 
can also write out to a brand new file, without having to mess with 
OutputStreams (should be faster)


The main two methods for this, available for HSSF and HSLF, are:
https://poi.apache.org/apidocs/org/apache/poi/POIDocument.html#write()
https://poi.apache.org/apidocs/org/apache/poi/POIDocument.html#write(java.io.File)


So, two requests:
 * Please test this for HSLF + HSSF!
 * Please help us document this on the site, have examples etc

Assuming feedback from HSSF + HSLF suggests we've gone about it the right 
way, we can then roll it out to HWPF and the OPC / X??F classes too


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: 3.15 beta 3 soon?

[Nick Burch]

On Fri, 22 Jul 2016, Andreas Beeker wrote:

When are your current tasks completed (e.g. in-place-writing...)?


In-place write + write-to-File is now done for HSSF + HSLF + HPSF. That's 
probably enough for now. Once we've got some feedback, we can add it to 
HWPF, then decide how to do the same for X??F without breaking too much 
backwards compatibility + solving that "close may change things" 
unexpected issue


Who will roll the release? As usual, I'll be the fallback (as long as my 
key is valid ...)


Anyone else want to give the docs a try? :)

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: svn commit: r1752809 - in /poi/site/src/documentation/content/xdocs: book.xml download.xml

[Nick Burch]

On Fri, 15 Jul 2016, Javen O'Neal wrote:
I had the same idea. Would an https license URL need to be blessed by 
legal?


Pass Someone from Creadur ought to know :)

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: poi-ruby.html no longer being built by Forrest

[Nick Burch]

On Thu, 14 Jul 2016, Javen O'Neal wrote:

After some recent changes to the site [1], "ant docs" no longer
creates/updates publish/poi-ruby.html.


I've just done a "svn up" and "ant docs", and poi-ruby.html was re-created 
for me


Does Forrest have some rules about file naming (does 
poi-jvm-languages.xml clobber poi-ruby.xml)? Are there config files that 
specify the mapping between xml and html file?


I think it does stuff with links and stuff from the menu. Shouldn't be any 
name overlap issues


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: svn commit: r1752809 - in /poi/site/src/documentation/content/xdocs: book.xml download.xml

[Nick Burch]

On Fri, 15 Jul 2016, one...@apache.org wrote:

Author: onealj
Date: Fri Jul 15 09:55:48 2016
New Revision: 1752809

URL: http://svn.apache.org/viewvc?rev=1752809=rev
Log:
use http URL in Apache License for rat check


It might be worth getting the Apache Creadur project (who maintain rat) if 
they could tweak the valid patterns the handle https links for this too


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Regenerate and publish javadocs

[Nick Burch]

On Tue, 12 Jul 2016, Dominik Stadler wrote:
Ideally we would also update the Subversion repository as part of the 
config, but I did not want to put my credentials in there, not sure if 
there are some "automation" credentials or accessKeys available for such 
cases.


You'd need to ask infra - there may be a suitable role account already, 
otherwise they may be able to create one


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: DrawingML

[Nick Burch]

On Fri, 8 Jul 2016, Murphy, Mark wrote:
Much of the Chart code and other things are common parts between 
Spreadsheets, Documents, and Presentations. At least for the XML side of 
things. I propose that the DrawingML portions of these three projects be 
consolidated into a separate package or group of packages to deal with 
things like charts and themes. If it is done correctly, this code should 
be able to be shared among all areas of POI (XML format anyway). I 
haven't checked into the older binary formats, but I suspect that there 
could be a similar relationship.


For the binary formats, we have the "DDF" package which contains the 
common drawing records and classes. Sounds like we might be able to create 
a XDDF package for the common DrawingML drawing xml stuff too!


Nick


-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: BitField#setValue is not a setter

[Nick Burch]

On 04/07/16 03:17, Javen O'Neal wrote:

Andi and I noticed a RV_RETURN_VALUE_IGNORED_NO_SIDE_EFFECT warning
from FindBugs in HWPF. BitField#setValue is not a setter. It returns
the value of the holder if a value was set through the BitMask.

The BitField class should probably probably use a name that doesn't
look like a setter and/or make it extremely clear in the Javadoc that
no state is modified.


I think that the BitField classes are only ever used internally. I can't 
spot any code elsewhere on the internet using it from a google, can't 
spot it used in the usermodel, and I don't recall seeing user code on 
the lists using it.


It's also ancient code IIRC, so don't see why we can't change the name 
or javadocs to something clearer


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: [VOTE] Apache POI 3.15-beta2 release (RC1)

[Nick Burch]

On Thu, 30 Jun 2016, Andreas Beeker wrote:

I think we ought to fix that before the release, no?

Actually I'd prefer to push the release and unfreeze the trunk.
How about having a beta3 in a month and you skip beta2 for tika?


As long as it's in the release notes clearly, and as long as we aim ot 
have another beta *out* within a month (so voting within 3 weeks max), 
then I wouldn't object to the release going out without the fix


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: How do those of you who use Git produce patches

[Nick Burch]

On Fri, 24 Jun 2016, Murphy, Mark wrote:
I have a small change to xwpf that I am working on in a local Git 
repository, but I need to know how to create a patch. The documentation 
on the How to contribute page isn't particularly helpful.


The easiest is probably to fork the POI github repo, push the changes to 
your fork, then open a pull request


Otherwise, something like this should give you all the patches pending:
   git diff --stat --cached origin/master

And this for the diffs:
   git diff --cached origin/master

Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: [PATCH] Use of Hashtable in POI codebase

[Nick Burch]

On Thu, 23 Jun 2016, Axel Howind wrote:

Nick Burch  gagravarr.org> writes:

Patches to the list tend to get lost. Any chance you could either open a
bugzilla entry and attach it there, or open it as a new pull request on
github?


I have created https://bz.apache.org/bugzilla/show_bug.cgi?id=59748 with an
updated patch.


Thanks! Looks good at first glance. If someone has time to review it fully 
and apply, great! If not, I'll apply it in a few days, allowing for lazy 
consensus


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: [PATCH] Use of Hashtable in POI codebase

[Nick Burch]

On Wed, 22 Jun 2016, Axel Howind wrote:
When looking at the current source, I noticed Hashtable still being used 
in a number of places. Is there any reason for not replacing all these 
hashtable instances with HashMap (besides TestContentType.java failing, 
see below)?


I suspect just history! The only place I could think of us needing the 
synchronisation of a hashtable is any system-wide constants, but even then 
maybe not.


I attach a patch that replaces Hashtable instantiations with HashMap and 
Hashtable declarations with Map (the common interface). Current git tests 
pass with this patch applied.


Patches to the list tend to get lost. Any chance you could either open a 
bugzilla entry and attach it there, or open it as a new pull request on 
github?




The failing test is this one:
assertEquals(typeResqml+";version=2.0;type=obj_global1dCrs", 
part.getContentTypeDetails().toString());


part is of type org.apache.poi.openxml4j.opc.PackagePart and toString() just 
concatenates all the entries contained in the parameters attribute 
(Hashtable) together by iterating over its entryset.


This is rather fragile and could break anytime since AFAIK the order in which 
hashtable/HashMap entries are returned is not guaranteed and could change 
with any (minor) Java release.


It might be better to replace that with an assertContains or similar?


-leave the hashtable in
-use a LinkedHashMap for org.apache.poi.openxml4j.opc.PackagePart.parameters 
to ensure consistent ordering


Maybe use a linked map, as there might be some odd ooxml compatibility 
stuff we've all forgotten about that needs a certain ordering... I don't 
think for 0-3 params it'll make much performance difference!


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: source code for org.apache.poi:poi-ooxml-schemas - 3.8

[Nick Burch]

On Wed, 22 Jun 2016, Rongli Shi wrote:
I have successfully build the project 'poi-3.8' and get the source jar 
'ooxml-lib/ooxml-schemas-src-1.1.jar¹. Could this file be regarded as 
the original source code for 'poi-ooxml-schemas-3.8.jar¹?


Sort of... See http://poi.apache.org/faq.html#faq-N10025 for how the full 
schema file gets cut down to the smaller poi-ooxml-schemas jar


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: source code for org.apache.poi:poi-ooxml-schemas - 3.8

[Nick Burch]

On Wed, 22 Jun 2016, Rongli Shi wrote:
I could not find the source code of "poi-ooxml-schemas - 3.8" neither in 
the maven repository nor in the svn/git repo.


See http://poi.apache.org/faq.html#faq-N10186 for details of how it is 
generated, and how to get the source


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Finding deprecated methods

[Nick Burch]

On Tue, 14 Jun 2016, Javen O'Neal wrote:
Is there a smarter way to find deprecated classes besides checking out 
the code from 2 final releases ago and removing any deprecated code that 
still exists in trunk@HEAD?


I'd probably attempt it with:
* grep trunk for @deprecated with no date, with a -A 4 or similar, then
  grab the method name from that
* grab a list of old versions from "svn ls ^/poi/tags/"
* use "svn switch ^/poi/tags//" to quickly walk your tree back through
  older versions
* do the same greps, and spot where they vanish

If that all sounds a bit hairy, let me know and I can try it later in the 
week!


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

Re: Git

[Nick Burch]

On Tue, 14 Jun 2016, Mark Murphy wrote:
Trying to find documentation site on Git, but it does not appear to 
exist. Is this true?


IIRC only trunk, branches and tags get mirrored into github. THe docs are 
in a different bit of the tree, in part so that the ASF webservers can 
watch svn for commits to the publish directory and immediately deploy the 
new files


We could ask for a mirror of docs too, if it would help, but I think the 
main copy needs to stay svn so the auto-deploy on publish work


Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org

3.15 beta 2 soon?

[Nick Burch]

Hi All

We're now about 2 months since we released 3.15-beta1, and it looks like 
we've got quite a few fixes and bits of new behaviour in, which is great! 
Based on bugzilla emails, we've also got a few user-contributed patches 
which are almost ready/fully-applied too.


So, two questions:
 - Can we have a volunteer to be Release Manager for 3.15 beta 2?
 - What bugs do people still need to fix / finish fixing (eg the ones
   we're working with contributors on right now) before we can release
   another beta?

Thanks
Nick

-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org