OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

docbook-apps message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [docbook-apps] DcBk assembly- contentonly=true - is including the meta-info of discarded subsuming root element - Is this correct behavior?


On Thu, Sep 21, 2017 at 1:08 PM, Simon Dew <simonjabadaw@gmail.com> wrote:
>
> Hi Otto,
>
> My understanding is that contentonly strips off the root element; omittitles strips the titles from the resource. See e.g.
>
> http://tdg.docbook.org/tdg/5.1/module.html
>
> However, when you're using contentonly and omittitles to aggregate several resources into a single module, there is still a problem with spurious info elements appearing in the output. I raised an issue on this back in March:
>
> https://github.com/docbook/xslt10-stylesheets/issues/26
>

Simon,

Yes, this is the same related behavior. And likewise, inserting both
contentonly=1 and omitttitles=1, the result is info-element, without
titles.

Yet fundamentally, contentonly=1 should be a superset of omittitles=1.

A title(s) is a portion of what may be in an info-element.

It is logically inconsistent that contentonly=1, which is stripping
off the subsuming root-element, then leave the the meta-information
for the missing element.

It seems to me that contentonly=1 would strip the outer root-element
AND any meta-information associated with that stripped element.

As a made up example, looking at all of the elements that can be
within the info-element, (http://tdg.docbook.org/tdg/5.1/info.html),
one may want to have something like, contentonly=0 and omittitles=1,
where something like biblioid-element is retained but not the titles.

However if the content is the only thing that is desired,
contentonly=1, then there is nothing to which the meta-information
associated with the stripped root-element may attach. The author
simply wants just content, not the enspousing wrapper and its related
information.

So I can picture, in a very general manner, that these two attributes
are partially orthogonal in use.
* contentonly means enclosing root-element is stripped along with all
of its meta-info.
* omittitles means that the enclosing root-element is retained,
however some of the meta-info is not utilized. It could be generalized
with something like omit={list of elements to omit}, obviously one
does not want an omittitles, omitauthor, omitbiblioid, and so on,
which would be a nightmare.
* Or one could have the dual (negative?) where it would be
retain={list of elements to retain} and hence dropping all other
elements, along with constraining that the author use either
attributes "omit" or "retain" but not both for the same
module-element.

Just some additional thoughts.

..Otto


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]