E.2. OpenOffice.org to DocBook

E.2.1. Open Office 1.0.x

OOo has been tested by LDP volunteers with mostly positive results. Thanks to Charles Curley (charlescurley.com) for the following notes on using OOo version 1.0.x:

	Check the version of your OpenOffice
	These notes may not apply to the version of OOo you are using.

• To be able to export to DocBook, you must have a Java runtime environment (JRE) installed and registered with OOo--a minimum of version 4.2.x is recommended. The configuration instructions will depend on how you installed your JRE. Visit the OOo web site for help with your setup.

  Contrary to the OOo documentation, the Linux OOo did not come with a JRE. I got one from Sun.

• The exported file has lots of empty lines. My 54 line exported file had 5 lines of actual XML code.

• There was no effort at pretty printing.

• The header is: <?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE article PUBLIC "-//OASIS//DTD DocBook XML V4.1.2//EN" "http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd">

• The pull-down menu in the File->Save As dialog box for file format indicates that the export format is "DocBook (simplified)." There is no explanation of what that "simplified" indicates. Does OOo export a subset of DocBook? If so, which elements are ignored? Is there any way to enter any of them manually?

• There is NO documentation on the DocBook export filter or whether OOo will import it again.

Conclusions: OOo 1.1RC is worth looking at if you want a word processor for preparing DocBook documents.

However, I hope they cure the lack of documentation. For one thing, it would be nice to know which native OOo styles map to which DocBook elements. It would also be nice to know how to map one's own OOo styles to DocBook elements.

E.2.2. Open Office 1.1

Tabatha Marshall offers the following additional information for OOo 1.1.

The first problem was when I tried to do everything on version 1.0.1. That was obviously a problem. I have RH8, and it was installed via rpm packages, so I ripped it out and did a full, new install of OpenOffice 1.1. It took a while to find out 1.1 was a requirement for XML to work.

During the install process I believe I was offered the choice to install the XML features. I have a tendency to do full installs of my office programs, so I selected everything.

I can't offer any advice to those trying to update their current OO 1.1. Their "3 ways" aren't documented very well at the site (xml.openoffice.org) and as of this writing, I can't even find THAT on their site anymore. I think more current documentation is needed there to walk people through the process. Most of this was unclear and I had to pretty much experiment to get things working.

Well, after I installed everything I had some configuration to do. I opened the application, and got started by opening a new file, choosing templates, then selecting the DocBook template. A nice menu of Paragraph Styles popped up for me, which are the names for all those tags, I noticed (you can see I don't use WYSIWYG often).

With a blank doc before me (couldn't get to the XML Filter Settings menu unless some type of doc was opened), I went into Tools->XML Filter Settings, and edited the entry for DocBook file. I configured mine as follows:

• Doctype -//OASIS//DTD DocBook XML V4.2//EN

• DTD http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd

• XSLT for export /usr/local/OpenOffice.org1.1.0/share/xslt/docbook/ldp-html.xsl

• XSLT for import /usr/local/OpenOffice.org1.1.0/share/xslt/docbook/docbooktosoffheadings.xsl (this is the default)

• Template for import /home/tabatha/OpenOffice/user/template/DocBook File/DocBookTemplate.stw

At first, if I opened an XML file that had even one parsing error, it would just open the file anyway and display the markup in OO. I have many XML files that use © and other types of entities which show up as parse errors (depending on the encoding) even though they can be processed through. But today I was unable to open any of those files. I got input/output errors instead. Still investigating that one.

However when you do successfully open a document (one parsing with no errors), it puts it automatically into WYSIWYG based on the markup, and you can then work from the paragraph styles menu like any other such editor.

To validate the document, I used Tools->XML Filter Settings, then clicked the Test XSLTs button. On my screen, I set up the XSLT for export to be ldp-html.xsl. If you test and there are errors, a new window pops up with error messages at the bottom, and the lines that need to be changed up at the top. You can change them there and progress through the errors until they're all gone, and keep testing until they're gone.

If you want to open a file to see the source instead of the processed results, go to Tools->XML Filter Settings->Test XSLTs, and then under the Import section, check the Display Source box. My import XSLT is currently docbooktosoffheadings.xsl (the default) and the template for import is DocBookTemplate.stw (also default).

I think this might work for some people, but unfortunately not for me. I've never used WYSIWYG to edit markup. Emacs with PSGML can tell me what my next tag is no matter where I am, validate by moving through the trouble spots, and I can parse and process from command line.

With OpenOffice, you have to visit http://xml.openoffice.org/filters.html to find conversion tools.