Thank you very much, fiddlosopher. I'll try with a lua filter which sets all column width to an equal size:

```lua

function myfunction(el)
    -- divide 1 by number of columns
    -- subtract some spacing
    equalwidth = (1 / #el.colspecs) - 0.01
    -- print(equalwidth)
    -- walk through colspecs
    for idx, _ in ipairs(el.colspecs) do
        -- set width per column
        table.insert(el.colspecs[idx], equalwidth)
    end
    -- return Table element
    return el
end

return {{ Table = myfunction }}

```

fiddlosopher schrieb am Montag, 24. Oktober 2022 um 06:12:47 UTC+2:

Pandoc tries to preserve the relative column widths in the original table. If you have table with a very narrow second column, pandoc will reproduce that in a grid table (since the colmun widths in this format matter).

One possible workaround would be to use a Lua filter to adjsut the column widths of Table elements after they are parsed.

Another would be to alter the column widths of the source table.



> On Oct 23, 2022, at 3:25 AM, Jan Stühler <jan.st...-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
>
> Hello group.
>
> I have been using
> ```
> pandoc --reference-links --extract-media=dir -o file.md file.docx
> ```
> on a set of documents, each of those documents has many tables. Especially those tables which have "only" images in the right column, pandoc will break the image reference after one character. The image references can therefore not be found when converting the document back to docx.
>
> ```
> +------------------------+---+------------------------------------------+
>
> | If the Home Screen is | ! | |
>
> | locked, the | [ | |
>
> | Examination button is | ] | |
>
> | dimmed. Please first | [ | |
>
> | click on Unlock and | 4 | |
>
> | then click on | 8 | |
>
> | Examination, | ] | |
>
> +========================+===+==========================================+
>
> ```
> I do not have that problem when passing `--wrap=none` to markdown, but then all the lines in vim are unmanageably long.
>
> `wrap=preserver|auto|none` with or without `--columns=x` does not resolve the one-character breaking issue.
>
> Also, converting to `-t gfm` first and then converting `-f gfm -t markdown` still creates one-character line breaks.
>
> Is there any way to define a minimum column width?
>
> --
> 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-discus...-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org.
> To view this discussion on the web visit https://groups.google.com/d/msgid/pandoc-discuss/22b620a8-4869-4cd8-b040-a4c55d0ae521n%40googlegroups.com.