--> XXE v4 will have such command-line tool.
XXE v4 will not generate duplicate IDs (though, in our opinion, its not
a good idea to stop processing a document just because the document
contains duplicate IDs).
--> Meanwhile, with XXE v3:
* Instead of using docb.toHTML1, I would recommend to write a custom,
very simple, process command, just copying a flattened copy of a
document to the location you specify.
* Command:
---
convertdoc -d docb.toHTML1 operator-manual.docbook.xml \
-u build/test.html
---
fails because the docb.toHTML1 expects 3 arguments.
The only way to learn how to use a process command through convertdoc is
to read the XML source of this process command.
Excerpts of XXE_install_dir/addon/config/docbook/xslMenu.incl
---
<command name="docb.toHTML1">
<process>
<subProcess name="docb.convertStep1" />
<copyProcessResources resources="xsl/css/html.css" to="html.css" />
<transform stylesheet="xsl/html/docbook.xsl"
file="__doc.xml" to="__doc.html"
label="Convert to single-page HTML"
documentation="http://docbook.sourceforge.net/release/xsl/current/doc/html/%{parameter.name|html.stylesheet}.html">
<parameter name="use.extensions">1</parameter>
<!-- Cannot work and generates a lot of error messages. -->
<parameter name="graphicsize.extension">0</parameter>
<parameter name="generate.toc">%0</parameter>
<parameter name="toc.section.depth">3</parameter>
<parameter name="section.autolabel">%1</parameter>
<parameter name="callout.graphics">1</parameter>
<parameter name="html.stylesheet">html.css</parameter>
<parameterGroup name="docb.toHTML1.transformParameters" />
</transform>
<upload base="%2">
<copyFile file="__doc.html" to="%2" />
<copyFile file="html.css" to="html.css" />
<copyFiles files="resources/*" toDir="resources" />
<copyFiles files="images/*" toDir="images" />
</upload>
</process>
</command>
---
Therefore, command:
---
convertdoc -d docb.toHTML1 operator-manual.docbook.xml \
-s " " -s 0 -u build/test.html
---
Should work fine.
Jean Jordaan wrote:
>
> Short version: I'd like to use http://www.deltaxml.com/free/docbook/
> to generate a PDF with changebars for the differences between two
> versions of a DocBook document. This requires that xincludes are
> expanded, and it doesn't like duplicate IDs.
>
> It looks like XXE is the only xinclude processor out there which can
> handle the recursive xincludes which it uses for modular documents. Is
> it perhaps possible to make it output the combined document?
>
> It looks like this should work:
>
> $ convertdoc -d docb.toHTML1 operator-manual.docbook.xml -u build/test.html
>
> There's something wrong with the way I call it, as it dies with:
>
> """
> [...]
> Unexpected section.autolabel value: ; using default.
> Unexpected section.autolabel value: ; using default.
> convertdoc: *** error: Command execution has failed:
> java.net.MalformedURLException: no protocol:
> +---------------------------------------
> | java.net.URL.<init>(URL.java:567)
> | java.net.URL.<init>(URL.java:464)
> | java.net.URL.<init>(URL.java:413)
> | com.xmlmind.xmleditapp.process.UploadItem.execute(UploadItem.java:83)
> | com.xmlmind.xmleditapp.process.ProcessCommand$1.run(ProcessCommand.java:176)
> """
>
> Anyway, it leaves a /tmp/xxe56203/__doc.xml file which looks like what
> I'm looking for, except that it has duplicate ids.
>
> If I do this:
>
> $ convertdoc -d docb.toHTML operator-manual.docbook.xml -u build/out/
>
> I don't get any errors, but the /tmp/xxeNNNNN/ directory doesn't
> contain any __doc.xml file, which leads me to suspect that the
> docb.toHTML1 version won't leave a __doc.xml either, if the invocation
> is fixed so that it completes successfully.
>
