Re: How to insert extra and hidden information for each element of document via LibreOffice API?

On 24/07/18 12:20, MUNOZ JULHO Yupanqui wrote:
> Hi everyone,
>
> I am working with the Eclipse Papyrus team on a UML to odt generator.
> Besides one way generating, it should be able to handle bidirectional
> synchronization between these two "worlds".
> For reaching this goal, I need to map each model element with its
> respective doc element. For example, a package name will be transformed
> to a section in a document or a comment in a paragraph. I am looking for
> a way to reference the UML element from the odt document, for example
> adding the ID of the UML element to an odt "paragraph".
>
> In the users mailing list, Regina Henschel, suggested me to use
> bookmarks. It is a good idea, but, if is possible, I want to hide this
> information.
>
> Does anyone have another suggestion?
Might need someone to suggest a change to the Open Document format, but
is there anything about "preserved but ignored" elements?

WordPerfect had such a thing with their new layout for v6 (crumbs,
that's 1994!). They defined a certain layout for their tags - bit like
an XML element - but also explicitly stated that unrecognised properly
formed tags would be "preserved and ignored".

So if say v8 introduced a whole bunch of new tags, the resulting
document could be round-tripped through v6, and unless the user
*ex*plicitly deleted the tags in v6, they would still be there when the
document was returned back to v8.

If you could guarantee that unrecognised elements would be preserved,
you could simply stick them in the XML that makes up the odt, knowing
that when it's retrieved back in to Papyrus, they would still be there.

Cheers,
Wol