torsdag 6. april 2023 kl. 16:22:17 UTC+2 skrev Julien Dutant:

Oops, remove the "print(id)" line in the filter script above, it was meant for debugging.

On Thursday, April 6, 2023 at 3:20:48 PM UTC+1 Julien Dutant wrote:
This filter will remove all empty Spans with id starting with x. Save as "removeXSpans.lua':

function Span(el)
return #el.content == 0 and el.identifier:match('^x%d') and pandoc.Space()
or el
end

And run pandoc with the `-L removeXSpans.lua` option, e.g.
pandoc -f dockbook sourcefile -t markdown -o outfile.md -L removeXSpans.lua

Result:
# 1 Introduction

However, this will break any link to #x1-10001. If there are internal links in the doc (e.g. from the table of
content) that you need to preserve, you need a filter that produces instead:
# 1 Introduction {#x1-10001}

Perhaps this will work (it'd help to have a sample docbook source), saved as removeXSpans.lua and used as above

function Header(hd)
local id = ''
hd.content = hd.content:walk {
Span = function(el)
if #el.content == 0 and el.identifier:match('^x%d') then
id = el.identifier
return pandoc.Space()
end
end
}
print(id)
if id ~= '' then
hd.identifier = id
return hd
end
end

On Wednesday, April 5, 2023 at 11:38:07 PM UTC+1 hcf wrote:
I'm converting from DocBook to Markdown.

In DocBook there are xml:id tags. When I convert to markdown these are rendered as
[]{#x1-10001}.

A markdown heading look like this when converting from DocBook.

# 1[]{#x1-10001}Introduction

Is there a way to turn this off?

best regards
hcf