FAQs

On linux or OSX:

for f in *.txt; do pandoc "$f" -s -o "${f%.txt}.rtf"; done

In Windows Powershell:

gci -r -i *.txt
|foreach{$rtf=$_.directoryname+"\"+$_.basename+".rtf";pandoc -f markdown -s $_.fullname -o $rtf}

Be sure to use the -s or --standalone flag, or you just get a fragment, not a full document with the required header:

pandoc -s -f markdown -t icml -o my.icml my.md

By default, pandoc uses pdflatex to generate the PDF, and pdflatex doesn’t handle Chinese characters. But you can change the default to use xelatex instead. You should also make sure you’re using a font with Chinese glyphs. For example:

pandoc -o c.pdf --pdf-engine=xelatex -V mainfont='Adobe Ming Std'

Run the following command as admin:

msiexec /i pandoc-VERSION.msi ALLUSERS=1

This will put pandoc in C:\Program Files\Pandoc. You can install Pandoc to a different directory by setting APPLICATIONFOLDER parameter, for example:

msiexec /i pandoc-1.11.1.msi ALLUSERS=1 APPLICATIONFOLDER="C:\Pandoc"

The option

-V geometry:margin=1in

will set the margins to one inch on each side. If you don’t want uniform margins, you can do something like

-V geometry:"top=2cm, bottom=1.5cm, left=1cm, right=1cm"

Or

-V geometry:"left=3cm, width=10cm"

For more options, see the documentation for the LaTeX geometry package.

Here is a wiki page comparing the two.

For example, if you convert an image with {width="50%"}, the LaTeX produced will be \includegraphics[width=0.5\textwidth,height=\textheight].

This output presupposes the following code in pandoc’s default latex template:

% Scale images if necessary, so that they will not overflow the page
% margins by default, and it is still possible to overwrite the defaults
% using explicit options in \includegraphics[width, height, ...]{}
\setkeys{Gin}{width=\maxwidth,height=\maxheight,keepaspectratio}

If you don’t have this in your custom template, you should add it. If we didn’t set the height explicitly in this way, the image would not be resized correctly unless it was being resized to smaller than its original size.

pandoc +RTS -M30m -RTS will limit heap memory to 30MB. When converting a document requires more than this, an out of memory error will be issued.

For various reasons, the $header-includes$ are not at the very end of the LaTeX preamble. This poses a problem when the code you are inserting depends on declarations in the preamble coming after the $header-includes$ location. For example, you might want to reference the \author and \title metadata values (set at the very bottom of the preamble) and present them in margins. In that case you can wrap your code in etoolbox’s \AtEndPreamble. The technique is demonstrated in this gist. When using \AtEndPreamble, keep any makeatletter or makeatother outside of the \AtEndPreamble, as shown in the example.

You can’t. You can try opening the PDF in Word or Google Docs and saving in a format from which pandoc can convert directly.

No. You can get by with a relatively small TeX installation, for example, by starting with MacTeX’s Basic TeX distribution and using the tlmgr tool to install a few packages required by pandoc (see the manual).

Or, you can produce PDFs via HTML and wkhtmltopdf, or via groff ms and pdfroff. (These don’t produce as nice typography as TeX, particularly when it comes to math, but they may be fine for many purposes.)

We are not sure why this happens, but we have found that fully specifying the pdflatex path avoids the error. For example,

pandoc -o my.pdf --pdf-engine=/Library/TeX/texbin/pdflatex

First, unless your target is a binary format (docx, odt, epub), you must use either --extract-media or (for HTML only) --embed-resources to make the images in the ipynb container available to your output file.

Second, some Jupyter extensions, especially those that use JavaScript for visualizations, assume the presence of require.js. To ensure that this script is available in your HTML output, you can use:

pandoc -s -o output.html input.ipynb \
-V header-includes='<script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js"></script>'

Use the --citeproc option. If it still doesn’t work, you may need to tell pandoc where your bibliography file is using --bibliography. Your references may not be formatted the same as they are when you use latex and bibtex; you can change the format of the citations by specifying an appropriate CSL bibliography style using --csl (see the manual).

The simplest approach is via ConTeXt:

pandoc --pdf-engine=context -V pdfa

Alternatively, --pdf-engine=pdflatex can be used with the following in header-includes in metadata (or included from a file using --include-in-header):

\usepackage[a-2u,mathxmp]{pdfx}
\usepackage[pdfa]{hyperref}

Or --pdf-engine=lualatex can be used with the following:

\usepackage{hyperxmp}
\hypersetup{pdfapart=3,pdfaconformance=B}
\immediate\pdfobj stream attr{/N 3} file{sRGB.icc}
\pdfcatalog{/OutputIntents [<<
/Type /OutputIntent /S /GTS_PDFA1
/DestOutputProfile \the\pdflastobj\space 0 R
/OutputConditionIdentifier (sRGB) /Info (sRGB)
>>]}

Save this filter as nowidths.lua and then pass --lua-filter nowidths.lua as an additional option to pandoc. (See issue 8139.)

-- Unset the width attribute of HTML colspecs in tables
-- See https://github.com/jgm/pandoc/issues/8139
function Table (tbl)
  if PANDOC_VERSION[1] >= 2 and PANDOC_VERSION[2] >= 10 then
    tbl.colspecs = tbl.colspecs:map(function (colspec)
        local align = colspec[1]
        local width = nil  -- default width
        return {align, width}
    end)
  else
    for i, w in ipairs(tbl.widths) do
      tbl.widths[i] = 0
    end
  end
  return tbl
end

Install antiword and use it to convert the doc to DocBook, which can be read by pandoc.

antiword -x db input.doc | pandoc -f docbook