About PDF J001E

Here should go questions about transforming XML with XSLT and FOP.
Eddie
Posts: 106
Joined: Wed Dec 18, 2013 3:07 am

About PDF J001E

Post by Eddie »

Hello,

We have our own tool for PDF output which includes DITA OT 1.6.3.
We are thinking of switching to oXygen's output engine for convenience but noticed the following phenomenon:
Using Author 17.1 (which includes OT 1.8.5), I noticed that we cannot get an index output for any terms that have an index-sort-as entry that starts with a 2-byte character. I get the following error (eg):
[PDFJ001E] Index entry 'dsdsds[利用者]' is dropped, because corresponding group is not found.


Is there a away to get oXygen to output these index entries?

Cheers,
Eddie.
Radu
Posts: 8992
Joined: Fri Jul 09, 2004 5:18 pm

Re: About PDF J001E

Post by Radu »

Hi Eddie,

Usually the DITA Open Toolkit has a specific folder which has for each language configuration files which decide the index categories:

DITA-OT\plugins\org.dita.pdf2\cfg\common\index

so probably in your case that configuration file needs to be edited to fit your needs. Did you generate an index page for the PDF also using the DITA OT 1.6.3? Or do you want to inhibit the index page generation? Are you using a DITA Map or a DITA bookmap?

Regards,
Radu
Radu Coravu
<oXygen/> XML Editor
http://www.oxygenxml.com
Eddie
Posts: 106
Joined: Wed Dec 18, 2013 3:07 am

Re: About PDF J001E

Post by Eddie »

Hello Radu,

Thanks for the reply.
Did you generate an index page for the PDF also using the DITA OT 1.6.3?
Yes.
Or do you want to inhibit the index page generation?
No, we want the index included in the output.
Are you using a DITA Map or a DITA bookmap?
Using bookmaps.

Cheers,
Eddie
Radu
Posts: 8992
Joined: Fri Jul 09, 2004 5:18 pm

Re: About PDF J001E

Post by Radu »

Hi Eddie,

In the Oxygen Preferences->DITA page there is a setting which allows you to configure the default DITA OT to use for publishing (by default it is the DITA OT 1.8 bundled with Oxygen). So another option for you would be to use that older DITA OT for publishing.
Or you could compare the files (Oxygen Tools->Compare Files):

OXYGEN_INSTALL_DIR\frameworks\dita\DITA-OT\plugins\org.dita.pdf2\cfg\common\index\ja.xml

and:

CUSTOM-DITA-OT-1-6\plugins\org.dita.pdf2\cfg\common\index\ja.xml

Maybe you made changes to the older ja.xml to better support your scenario.

Regards,
Radu
Radu Coravu
<oXygen/> XML Editor
http://www.oxygenxml.com
Eddie
Posts: 106
Joined: Wed Dec 18, 2013 3:07 am

Re: About PDF J001E

Post by Eddie »

Thanks again, Radu.

Just to confirm (I know this is a DITA OT phenomenon, not oXygen, but please bear with me):
So, by default, index-sort-as entries that start with a double-byte character are not output in PDF by DITA OT. To ouput them, you have to customize one or more files in the OT.
(I can't readily confirm this myself with our tool because it's server-based and developed/managed by another department.)

Cheers,
Eddie.
Radu
Posts: 8992
Joined: Fri Jul 09, 2004 5:18 pm

Re: About PDF J001E

Post by Radu »

Hi Eddie,

If you can put together a small DITA map project (map + a small topic) to reproduce the problem and send it to support@oxygenxml.com I could try to look into it. For example I could also try to test it with DITA OT 2.x, see if the behavior changed, maybe add an issue for it.

Regards,
Radu
Radu Coravu
<oXygen/> XML Editor
http://www.oxygenxml.com
Eddie
Posts: 106
Joined: Wed Dec 18, 2013 3:07 am

Re: About PDF J001E

Post by Eddie »

Hello Radu,

There's a sample on the way.

I found out that our tool defaults to "ja-JP" if there is no xml:lang attribute, but even so it outputs the index-sort-as correctly whether or not xml:lang is set and whether it is set to "en-US" or "ja-JP"

oXygen outputs correctly if xml:lang is set to "ja-JP", otherwise the indexterm entries are ignored.

I suppose this isn't really a big deal, but it does seem odd that indexterm and index-sort-as are handled differently.

Cheers,
Eddie.
Radu
Posts: 8992
Joined: Fri Jul 09, 2004 5:18 pm

Re: About PDF J001E

Post by Radu »

Hi Eddie,

Thanks for the samples, please see some answers below:
I found out that our tool defaults to "ja-JP" if there is no xml:lang attribute, but even so it outputs the index-sort-as correctly whether or not xml:lang is set and whether it is set to "en-US" or "ja-JP"
Possibly your tool has a customized DITA OT. Or possibly you are using another plugin to generate the PDF on the server. Maybe you are using the Antenna House DITA OT PDF plugin.
oXygen outputs correctly if xml:lang is set to "ja-JP", otherwise the indexterm entries are ignored.
Oxygen comes bundled with a stock DITA OT distribution. We did not make changes to alter the behavior in this part. And the specs states that you should specify xml:lang on any topic or map in order for the publishing to properly take this into account.

Regards,
Radu
Radu Coravu
<oXygen/> XML Editor
http://www.oxygenxml.com
Eddie
Posts: 106
Joined: Wed Dec 18, 2013 3:07 am

Re: About PDF J001E

Post by Eddie »

Understood. Thanks for the replies, Radu.

Cheers,

Eddie.
Post Reply