Introduce skiptitlepage for docx output

[tdewin]

Describe your proposed improvement and the problem it solves.
Currently there is a forced title page generated based on the meta data. Since generating a custom "whitepaper", the first page is some custom graphics, But I would like to keep the meta data set in the document properties. I tried looking at lua filters but they only modify the AST tree not the docx write.

If I would know/understand haskel, I would try to edit but as far as I can tell the fix should be something like

Some switch called skiptitlepage
pandoc/src/Text/Pandoc/Writers/Docx.hs Line : 764

  let notIncludeMetaPage = lookupMetaBool "skiptitlepage" meta

Then where the doc is generated, make meta only the toc if skipped

pandoc/src/Text/Pandoc/Writers/Docx.hs Line : 812

  meta' <- if notIncludeMetaPage
    then return map Elem toc
    else return title ++ subtitle ++ authors ++ date ++ abstract ++ map Elem toc

  return (meta' ++ doc', notes', comments')

Describe alternatives you've considered.

• Zip unzip the docx after generation and sed remove it
• changing the style but it would still render the data
• Lua filter on block, but it starts at the very first item (AST)

[jgm]

Relevant: #7256, #2928.
In general, it would be nice if we could make the reference.docx behave more like a template, so that you could remove the title and author.

[tdewin]

Thanks for creating pandoc and responding so quickly

I have worked around this by not supplying any meta data at all, then unzip/zip and overwriting core.xml manually. It is pretty crude, eg you need to escape & with & etc.

OUTFILE="pandoc-output.docx"
REMASTER="${OUTFILE%.*}-remaster"
unzip -o -qq $OUTFILE -d $REMASTER
PREVPATH=$(pwd)
cd $REMASTER

cat <<'EOF' > docProps/core.xml
<?xml version="1.0" encoding="UTF-8"?>
<cp:coreProperties xmlns:cp="http://schemas.openxmlformats.org/package/2006/metadata/core-properties"
    xmlns:dc="http://purl.org/dc/elements/1.1/"
    xmlns:dcterms="http://purl.org/dc/terms/"
    xmlns:dcmitype="http://purl.org/dc/dcmitype/"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
    <dc:title>title</dc:title>
    <dc:creator>author1;author2</dc:creator>
    <dc:description>description</dc:description>
    <dc:language>en-us</dc:language>
    <cp:keywords>some, key, words</cp:keywords>
    <dcterms:created xsi:type="dcterms:W3CDTF">2023-01-01T00:00:00Z</dcterms:created>
    <dcterms:modified xsi:type="dcterms:W3CDTF">2023-01-01T00:00:00Z</dcterms:modified>
</cp:coreProperties>
EOF

zip -r "$REMASTER.docx" *
cd $PREVPATH
  
#for safety
#rm -rf $REMASTER

Similar, I've removed the toc completely but include it with include lua filter https://github.com/pandoc-ext/include-files/blob/main/include-files.lua and then this segment https://gist.github.com/tdewin/e8cb8bc35a0fbabb630ff3f4a033ab77

Hopefully this is useful for somebody out there

[ghost]

One option for this specific issue would be for the Docx Writer to use "title-meta", "author-meta" fields like HTML and LaTeX do. That is, have the Docx writer set the document metadata from the "title-meta" etc. fields, and set the title block from "title", "author" fields, setting "title-meta" etc. automatically from "title" if present. That way, if we want to set the metadata without generating the title block, we can set the "title-meta" fields and omit "title". This doesn't look like it would be a huge change (though I can barely read Haskell, so I could be missing something), and would make the behavior consistent with other writers.

[jgm]

Still, you'd have to set title-meta, etc. manually, which doesn't seem ideal.
Btw, the reason for title-meta in HTML and other formats is that sometimes there are restrictions on what formatting can go in these fields.

[aditivin]

Thanks for creating pandoc and responding so quickly

I have worked around this by not supplying any meta data at all, then unzip/zip and overwriting core.xml manually. It is pretty crude, eg you need to escape & with & etc.

OUTFILE="pandoc-output.docx"
REMASTER="${OUTFILE%.*}-remaster"
unzip -o -qq $OUTFILE -d $REMASTER
PREVPATH=$(pwd)
cd $REMASTER

cat <<'EOF' > docProps/core.xml
<?xml version="1.0" encoding="UTF-8"?>
<cp:coreProperties xmlns:cp="http://schemas.openxmlformats.org/package/2006/metadata/core-properties"
    xmlns:dc="http://purl.org/dc/elements/1.1/"
    xmlns:dcterms="http://purl.org/dc/terms/"
    xmlns:dcmitype="http://purl.org/dc/dcmitype/"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
    <dc:title>title</dc:title>
    <dc:creator>author1;author2</dc:creator>
    <dc:description>description</dc:description>
    <dc:language>en-us</dc:language>
    <cp:keywords>some, key, words</cp:keywords>
    <dcterms:created xsi:type="dcterms:W3CDTF">2023-01-01T00:00:00Z</dcterms:created>
    <dcterms:modified xsi:type="dcterms:W3CDTF">2023-01-01T00:00:00Z</dcterms:modified>
</cp:coreProperties>
EOF

zip -r "$REMASTER.docx" *
cd $PREVPATH
  
#for safety
#rm -rf $REMASTER

Similar, I've removed the toc completely but include it with include lua filter https://github.com/pandoc-ext/include-files/blob/main/include-files.lua and then this segment https://gist.github.com/tdewin/e8cb8bc35a0fbabb630ff3f4a033ab77

Hopefully this is useful for somebody out there

Thank you, very helpful! The openxml TOC format worked for me