Philippe Coq wrote:
> I am using xmlmind for few weeks and I am really satisfied with it.
> I am try now to process a DocBook book composed of several .xml files
> tahat are linked together via <olink> elements.
> (I have attached a very simple example of what we are doing).
> I have some problem to make the same with the convertdoc tool
> I do the following:
> convertdoc -p collect.xref.targets 'only' -p targets.filename docs/A.db \
> docb.toHTML A.xml
> -> /Writing docs/A.db for section/
>
> convertdoc -p collect.xref.targets 'only' -p targets.filename docs/B.db \
> docb.toHTML B.xml \
> -> /Writing docs/ABdb for section
>
> / convertdoc -p target.database.document 'targetdb.xml' -p
> current.docid A \
> docb.toHTML A.xml \
> -u docs/
>
> /Olink error: could not open target database 'targetdb.xml'.
> Olink debug: root element of target.database '' is ''.
> Error: unresolved olink: targetdoc/targetptr = 'B/Config_Rules'/
>
> Is there anybody that can tell me :
> - how must I give the name of the master database document (in my
> example targetdb.xml is on the working directory) ?
I've no experience with <olink>s and therefore I cannot help you.
> - is docbtoHTML the correct process command to use?
No. docb.toHTML has been designed to convert a DocBook document to
multi-page HTML.
Here's a real world example (doc.xml contains a modular DocBook <book>):
---
convertdoc -p callout.graphics 1 \
-p toc.section.depth 4 -p chunk.section.depth 2 \
-p use.id.as.filename 1 \
docb.toHTML doc.xml -u ../../doc/configure
---
Arguments "-u" "../../doc/configure" specify the directory which is to
contain the converted HTML pages.
> - is there another way to define the targetdb.xml in order to avoid
> entities and hard coded paths?
Don't know.
> - when these commands will be running, will it be possible to fire them
> via a button in the xmlmind gui?
Yes. You'll have to create your own elaborate process command and add a
button for it in the DocBook menu.
--> My recommendations (because I'm afraid you are loosing your time):
[1] Please consider XXE as a *pure* *authoring* *tool*. Use it in a way
that maximizes the interoperability with xsltproc (I've explained the
rules in a previous email and you have my word that it will be easier in
next version).
[2] In these conditions, converting a set of linked DocBook documents
created with XXE is not an issue related to XXE. Please solve this
problem separately by posting to mailing lists related to DocBook and to
its XSL style sheets.
