--- Comment #57 from Dimitri Goloborodko <dv...@yahoo.com> 2009-04-14 03:33:19
(In reply to comment #50)
> (In reply to comment #49)
> > (In reply to comment #48)
> > >
> > > AFAIU the reason why FootnoteBodyLM is re-parented is that it put its
> > > areas at
> > > the right place (as children of the footnote-reference-area, instead of
> > > the
> > > block containing the footnote). Simply moving the setParent call after
> > > the call
> > > to getNextKnuthElements makes the warnings disappear, and doesn't break
> > > any
> > > test.
> > > Confirmation from specialists of this part of the code would be
> > > appreciated ;-)
> > Seems reasonable to me. Cleaner than overriding getParent() anyway.
> > In the meantime, I've also been playing with adding an interface
> > FootnoteCitationHolder. Such an interface could then be implemented by
> > KnuthBlockBox and ActiveCell. The interface methods can be used by
> > LineLayoutManager, PageBreaker, ListItemLayoutManager, TableStepper...
> > The methods would roughly be:
> > hasAnchors()
> > getFootnoteBodyLMs()
> > addFootnotes(List<KnuthElement>)
> > addFootnotes(FootnoteCitationHolder)
> > addFootnote(FootnoteCitationHolder.Citation)
> > expandFootnotes(LayoutManager, LayoutContext, int)
> > While this would still leave the related portions of code distributed over
> > those classes, the interface makes it a bit easier to locate them in an IDE,
> > and makes those pieces of code a bit more uniform.
> > Most of the loops we see now, would move to KnuthBlockBox, as the only
> > complete
> > implementation. ActiveCell would only implement what is needed to make the
> > citations accessible to the box created by TableStepper. Slight compromise
> > in
> > comparison to the last patch is that, in the iteration over the active
> > cells,
> > we would only create a temporary list with those having citations. If the
> > list
> > is empty, we create a regular box. If not, then we iterate over that
> > temporary
> > list of cells, and instruct the created KnuthBlockBox to add the citations
> > from
> > those cells. The same pattern can be used by ListItemLayoutManager:
> > - create a temporary list of FootnoteCitationHolders for which hasAnchors()
> > returns true
> > - if non-empty, iterate over that temporary list
> > - for each element, ask the higher level FootnoteCitationHolder
> > (KnuthBlockBox)
> > to extract the citations, and add them to its own list.
> > I'll see if I can attach a patch to demonstrate one of these days.
> I've been asked to look into this issue, so I committed a partial and
> fix based on the latest patch:
> Footnotes in table headers and footers are not handled yet, and anyway I think
> it's best to wait for clarification from xsl-editors before implementing
> anything (which gives us a couple of months ;-) ).
> That doesn't prevent you from exploring your ideas above, though. I await your
I try to use the trunk 660979 and find a case, when a footnote defined in
table-body disappears. An example of this is attached. I look for solution
since some days, but don’t get much with my very modest knowledge of fop. Do
you have any ideas about that?
Configure bugmail: https://issues.apache.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.