One way is to make several several passes. In the first pass you wrap all
Cite elements in Span elements with a class which does not otherwise occur.
In the second pass you visit Note elements and use `walk_block` to again
unwrap those Span elements inside footnotes. In the third pass you use
Inlines to find those Span elements elsewhere, unwrap them and adjust
punctuation around the Cite.

You make a multi-pass filter script by returning an array table with
several filter map tables from the script.

``````lua
local function wrap_cite (cite)
  return pandoc.Span({cite}, {class = 'is-cite'})
end

local function unwrap_cite (span)
  if span.classes:includes('is-cite') then
    return span.content[1]
  return nil
end

local unwrap_filter = { Span = unwrap_cite }

local function unwrap_in_notes (note)
  local old_blocks = pandoc.Div(note.content)
  local new_blocks = pandoc.walk_block(old_blocks, unwrap_filter)
  return pandoc.Note(new_blocks.content)
end

end

local function adjust_punct (inlines)
  for i=1, #inlines do
    if 'Span' == inlines[i].tag then
      local cite = unwrap_cite(inlines[i])
      if cite then
        -- move stuff around!
      end
    end
  end
  return inlines
end

return {
  { Cite = wrap_cite },
  { Note = unwrap_in_notes },
  { Inlines = adjust_punct },
}
``````



Den tors 25 nov. 2021 20:02FI Apps <ffi.appdev-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> skrev:

> The particular case I’m dealing with is fixing a long and complex text
> that followed an Italian style in placing punctuation after footnotes, even
> though the text is in English. The script is a one-off, but since there are
> over 1000 footnotes, fixing it with a script is the easiest solution. Since
> I’m using BibLaTeX, I could tell it to move punctuation for footnotes
> generated with Cite, but it just swaps the footnote and the punctuation: it
> doesn’t move periods or commas into quotes that may precede the footnote.
>
> The first time I wrote a script that needed to know if it was in a
> footnote, I resorted to the solution you suggest: I wrote a filter function
> for Note and used walk_block. But since this is the second script that
> wants to know whether or not it’s in a footnote, I thought I should mention
> this as a desideratum for the future.
>
> > On 24 Nov 2021, at 18:59, John MacFarlane <jgm-TVLZxgkOlNX2fBVCVOL8/A@public.gmane.org> wrote:
> >
> >
> > This is a limitation of the current architecture -- there's no
> > way to determine the "parent" context.  Sometimes you can work
> > around this by using walk_block to do a transformation inside
> > a particular kind of block (e.g. a footnote) -- but in this
> > case you want to do the transformation OUTSIDE of the block,
> > and that's more difficult.
> >
> > Doesn't pandoc's --citeproc do this punctuation moving for you
> > (in the case of citations automatically added as footnotes)?
> > If not, try setting `notes-after-punctuation` as described in
> > the manual.
> >
> > (If you are talking about footnotoes you insert explicitly,
> > instead of citations that become footnotes, then this doesn't
> > apply, but in that case why would you need to adjust the
> > punctuation?)
> >
> > jcr <ffi.appdev-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> writes:
> >
> >> I find in Lua filters that I sometimes would like to know whether or
> not
> >> I'm in a footnote. Currently, I'm trying to move punctuation before
> >> footnotes. Given my citation style, I know that a Cite in body text
> will
> >> produce a footnote, while a Cite in a footnote will not. So I want to
> move
> >> punctuation before a Cite when it's not in a footnote. Since a filter
> >> function for Inlines will descend into footnotes as well, there doesn't
> >> seem to be any way to tell when the Cite is in a footnote.
> >>
> >> In this particular case, I can work around the limitation because any
> Cite
> >> in a footnote will either be the first element or will have a Space
> before
> >> it. So with that assumption, I can look for the last innermost element
> >> before the Cite and check its type: if it's a Str,  can append the
> >> punctuation to it and delete the punctuation from where it was. if it's
> a
> >> Space, I do nothing, because I must be in a footnote. However, at least
> in
> >> the long term, I'd like to be able to tell whether or not I'm in a
> footnote.
> >>
> >> --
> >> You received this message because you are subscribed to the Google
> Groups "pandoc-discuss" group.
> >> To unsubscribe from this group and stop receiving emails from it, send
> an email to pandoc-discuss+unsubscribe-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
> >> To view this discussion on the web visit
> https://groups.google.com/d/msgid/pandoc-discuss/111b665a-1b7a-4856-bf37-d96780a07c24n%40googlegroups.com
> .
>
> --
> You received this message because you are subscribed to the Google Groups
> "pandoc-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to pandoc-discuss+unsubscribe-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/pandoc-discuss/F6A8DF67-F34E-4FF9-A7A2-CF451E96D683%40gmail.com
> .
>

-- 
You received this message because you are subscribed to the Google Groups "pandoc-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pandoc-discuss+unsubscribe-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
To view this discussion on the web visit https://groups.google.com/d/msgid/pandoc-discuss/CADAJKhByRxPwZOqgkfkV6ZoSFfBZhHHX52mpyB36_ShLR40Gyg%40mail.gmail.com.