How to modify the document core property "Keywords"?

[Bart Jenkins <ba...@mac.com>]

All
  MS-Word 2010
  POI-3.9

  I'm trying to pass in a version string to a field code in some boilerplate docx file that I read in at startup.  For example:

XWPFDocument document = new XWPFDocument(new FileInputStream("c:\template.docx");
CoreProperties coreprops = document.getProperties().getCoreProperties();
coreprops.setKeywords("Version 1.0");


And in my "template.docx" document, I've got a field code that looks like the following:

{INFO Keywords xxx \* MERGEFORMAT}    // <-- this is how it looks with the field code toggled ON in the word document.

The 3rd line of code above (coreprops.setKeywords()) does not fail and if I re-read the coreprops variable I see that it gets updated.  But the value is NOT showing up in my document.

Any idea / strategy on how I can pass parameters to a document and have them get dropped into field codes in my document at processing time?

Thanks

Bart

Re: How to modify the document core property "Keywords"?

[Mark Kimsal <me...@gmail.com>]

I had the code to do this, but I guess I didn't save it to git, because I
can't find it now.  I changed my whole strategy so as not to rely on
document properties at all.  This decision mostly came from the fact that
OOo and LibreO have a long standing bug where they cannot save custom
document properties to a docx file, nor can they read them.

The strategy to update the cached values involves looking for w:instrText,
then finding its w:r parent, then finding that w:r's siblings that contain
w:fldChar where w:fldCharType="end".  Then update any w:t tags in any w:r
between those two.  This all involves XPath queries on CTNodes.

Since I had to update the word/document.xml main document body with XPath,
and I was in control of the format of the documents entering the system, I
switched to just using plain old text in the word doc and then XPath swap
on that: {INSERT some variable name} .  Doing this is easier if you can
highlight the text and change the background color, this will make that
text key it's own w:t tag inside a separate w:r tag.  Once the value is
swapped, remove any w:rPr/w:color from the parent w:r.  You just have to
look at the XML and write XPath.


On Thu, Mar 7, 2013 at 4:45 PM, Bart Jenkins <ba...@mac.com> wrote:

> Thanks Nick,
>   I'm going for plan B which is to use the Word merge-record feature. If I
> add mergerec fields into my document and then point it to a CSV file with
> the needed data to be inserted into my document, it seems to work.  Only
> issue now is that Word puts up a popup dialog warning me I am grabbing data
> from an outside source and should I allow that.
>
> Anyway to turn that off?
>
> Thanks
>
> Bart
> On Mar 7, 2013, at 4:42 PM, Nick Burch <ap...@gagravarr.org> wrote:
>
> > On Thu, 7 Mar 2013, Bart Jenkins wrote:
> >> I'm trying to pass in a version string to a field code in some
> boilerplate docx file that I read in at startup.  For example:
> >>
> >> XWPFDocument document = new XWPFDocument(new
> FileInputStream("c:\template.docx");
> >> CoreProperties coreprops = document.getProperties().getCoreProperties();
> >> coreprops.setKeywords("Version 1.0");
> >
> > That looks write to change the OOXML level properties
> >
> >> And in my "template.docx" document, I've got a field code that looks
> like the following:
> >>
> >> {INFO Keywords xxx \* MERGEFORMAT}    // <-- this is how it looks with
> the field code toggled ON in the word document.
> >>
> >> The 3rd line of code above (coreprops.setKeywords()) does not fail and
> if I re-read the coreprops variable I see that it gets updated.  But the
> value is NOT showing up in my document.
> >
> > The office file formats contain caching at various points for
> performance reasons. For example, Excel caches formula values, which often
> catches out people new to POI.
> >
> > Word also caches property values, which is what's biting you here. In
> fact, you can see that by loading a copy of word, going into the document
> properties, and changing them. In most versions of word, with many common
> settings, the property won't change in the text, until you trigger an
> update properties call (right click then update property, if memory serves
> - I don't have word to hand)
> >
> > Sadly, that means you too need to update the cached properties in your
> word document after the change. It's likely a common problem though, so if
> you do manage to write some code to process the word document and update
> the values of all the properties it finds, please do share it back!
> >
> > Nick
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: user-unsubscribe@poi.apache.org
> > For additional commands, e-mail: user-help@poi.apache.org
> >
>
>

Re: How to modify the document core property "Keywords"?

[Bart Jenkins <ba...@mac.com>]

Thanks Nick,
  I'm going for plan B which is to use the Word merge-record feature. If I add mergerec fields into my document and then point it to a CSV file with the needed data to be inserted into my document, it seems to work.  Only issue now is that Word puts up a popup dialog warning me I am grabbing data from an outside source and should I allow that.

Anyway to turn that off?

Thanks

Bart
On Mar 7, 2013, at 4:42 PM, Nick Burch <ap...@gagravarr.org> wrote:

> On Thu, 7 Mar 2013, Bart Jenkins wrote:
>> I'm trying to pass in a version string to a field code in some boilerplate docx file that I read in at startup.  For example:
>> 
>> XWPFDocument document = new XWPFDocument(new FileInputStream("c:\template.docx");
>> CoreProperties coreprops = document.getProperties().getCoreProperties();
>> coreprops.setKeywords("Version 1.0");
> 
> That looks write to change the OOXML level properties
> 
>> And in my "template.docx" document, I've got a field code that looks like the following:
>> 
>> {INFO Keywords xxx \* MERGEFORMAT}    // <-- this is how it looks with the field code toggled ON in the word document.
>> 
>> The 3rd line of code above (coreprops.setKeywords()) does not fail and if I re-read the coreprops variable I see that it gets updated.  But the value is NOT showing up in my document.
> 
> The office file formats contain caching at various points for performance reasons. For example, Excel caches formula values, which often catches out people new to POI.
> 
> Word also caches property values, which is what's biting you here. In fact, you can see that by loading a copy of word, going into the document properties, and changing them. In most versions of word, with many common settings, the property won't change in the text, until you trigger an update properties call (right click then update property, if memory serves - I don't have word to hand)
> 
> Sadly, that means you too need to update the cached properties in your word document after the change. It's likely a common problem though, so if you do manage to write some code to process the word document and update the values of all the properties it finds, please do share it back!
> 
> Nick
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscribe@poi.apache.org
> For additional commands, e-mail: user-help@poi.apache.org
> 

Re: How to modify the document core property "Keywords"?

[Nick Burch <ap...@gagravarr.org>]

On Thu, 7 Mar 2013, Bart Jenkins wrote:
> I'm trying to pass in a version string to a field code in some 
> boilerplate docx file that I read in at startup.  For example:
>
> XWPFDocument document = new XWPFDocument(new FileInputStream("c:\template.docx");
> CoreProperties coreprops = document.getProperties().getCoreProperties();
> coreprops.setKeywords("Version 1.0");

That looks write to change the OOXML level properties

> And in my "template.docx" document, I've got a field code that looks 
> like the following:
>
> {INFO Keywords xxx \* MERGEFORMAT}    // <-- this is how it looks with the field code toggled ON in the word document.
>
> The 3rd line of code above (coreprops.setKeywords()) does not fail and 
> if I re-read the coreprops variable I see that it gets updated.  But the 
> value is NOT showing up in my document.

The office file formats contain caching at various points for performance 
reasons. For example, Excel caches formula values, which often catches out 
people new to POI.

Word also caches property values, which is what's biting you here. In 
fact, you can see that by loading a copy of word, going into the document 
properties, and changing them. In most versions of word, with many common 
settings, the property won't change in the text, until you trigger an 
update properties call (right click then update property, if memory serves 
- I don't have word to hand)

Sadly, that means you too need to update the cached properties in your 
word document after the change. It's likely a common problem though, so if 
you do manage to write some code to process the word document and update 
the values of all the properties it finds, please do share it back!

Nick

---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@poi.apache.org
For additional commands, e-mail: user-help@poi.apache.org