Skip to content


Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time


Note This project is no longer maintained: Some features explored here found their way into pandoc proper, while many others have been adopted by Quarto. If you are looking for a replacement, give Quarto a try! For the remaining pandoc-scholar features, as well as additional goodies, check the pandoc-ext organization.

release shield DOI license shield Build status

Create beautiful, semantically enriched articles with pandoc. This package provides utilities to make publishing of scientific articles as simple and pleasant as possible. It simplifies setting authors' metadata in YAML blocks, allows to add semantic annotation to citations, and only requires the programs pandoc and make.


Plain pandoc is already excellent at document conversion, but it lacks in metadata handling. Pandoc-scholar offers simple ways to include metadata on authors, affiliations, contact details, and citations. The data is included into the final output as document headers. Additionally all entries can be exported as JSON-LD, a standardized format for the semantic web.

The background leading to the development of pandoc-scholar is described in the paper published in PeerJ Computer Science.

Note that since version 2.0, most of the functionality of pandoc-scholar is now provided via pandoc Lua filters. If you prefer to mix-and-match selected functionalities provided by pandoc-scholar, you can now use the respective Lua filters directly. Integration with tools like RMarkdown is possible this way.


An example document plus bibliography is provided in the example folder. Running make in the example folder will process the example article, generating output like below:

example article screenshot

Get the full output as pdf, docx, or epub, or take a look at the metadata in JSON-LD format.

Usage via Docker

A very easy way to use pandoc-scholar is via Docker. The ready-made images contain all necessary software to generate a paper in multiple formats. This avoids any compatibility concerns; only Docker is required.

The official images are in the [pandocscholar/ubuntu] and [pandocscholar/alpine] images. The Alpine image is a bit smaller, while the Ubuntu image may be more familiar for people looking to extend the image. Both images come with pandoc, pandoc-citeproc, pandoc-crossref, and LaTeX.

Example call

Docker commands are often unwieldly due to the additional arguments. We recommend to define an alias or short script to simplify its use.

Given an article in file and a simple Makefile like

include $(PANDOC_SCHOLAR_PATH)/Makefile

the conversion can be performed by running

docker run --rm -v "$(pwd):/data" -u "$(id -u)" pandocscholar/alpine

This will generate a set of files whose names all start with out.. Please be aware that existing files of the same name will be overwritten. The pandoc-scholar container calls make internally; additional commands and options can be passed by appending them the above command.

The images are based upon the official pandoc images; for more info and usage examples, see the pandoc/dockerfiles GitHub repo. The Docker images can easily be used in automatic document conversion pipelines; pandoc-actions-example gives a good overview.

A major difference between pandoc and pandoc-scholar images is that pandoc-scholar doesn't use pandoc but make as entrypoint. A basic Makefile must be present in the article directory when running pandoc-scholar.


This package builds on pandoc, the universal document converter. See the pandoc website for installation instructions and suggestions for LaTeX packages, which we use for PDF generation.

Starting with pandoc-scholar 3.0.0, the minimum required pandoc version is 2.11. If you have to use an older pandoc, please combine it with the last 2.* release of pandoc-scholar.

Also note that pandoc's JATS support, especially citation handling, was buggy prior to pandoc v2.11.4. Please use that or a newer version when producing JATS XML.


Archives containing all required files are provided for each release. Use the release button above (or directly go to the latest release) and download a pandoc-scholar archive; both archive files, .zip and .tar.gz, contain the same files. Choose the filetype that is the easiest to unpack on you system.

A pandoc-scholar folder will be created on unpacking. The folder contains all required scripts and templates.



Run make to convert the example article into all supported output formats. The markdown file used to create the output files can be configured via the ARTICLE_FILE variable, either directly in the Makefile or by specifying the value on the command line.


Includable Makefile

The Makefile, which does most of the work, is written in a style that makes it simple to include it from within other Makefiles. This method allows to keep pandoc-scholar installed in a central location and to use the same instance for multiple projects. The ARTICLE_FILE and PANDOC_SCHOLAR_PATH variables must be defined in the including Makefile:

PANDOC_SCHOLAR_PATH = ../path-to-pandoc-scholar-folder
include $(PANDOC_SCHOLAR_PATH)/Makefile

Calling make as usual will create all configured output formats. Per default, this creates pdf, latex, docx, odt, epub, html, and jats output. The set of output files can be reduced by setting the DEFAULT_EXTENSIONS variable to a subset of the aforementioned formats. For example DEFAULT_EXTENSIONS = pdf odt docx

Alternative template files can be set using TEMPLATE_FILE_<FORMAT> variables, where <FORMAT> is one of HTML, EPUB, JATS, or LATEX. The reference files for ODT and DOCX output can be changed using ODT_REFERENCE_FILE and DOCX_REFERENCE_FILE, respectively.

Additional pandoc options can be given on a per-format basis using PANDOC_<FORMAT>_OPTIONS variables. The following uses an actual Makefile as an example to demonstrate usage of those options.


PANDOC_LATEX_OPTIONS  = --latex-engine=xelatex
PANDOC_LATEX_OPTIONS += --csl=peerj.csl
PANDOC_LATEX_OPTIONS += --filter=pandoc-citeproc
PANDOC_LATEX_OPTIONS += -M fontsize=10pt
PANDOC_LATEX_OPTIONS += -M classoption=fleqn


DOCX_REFERENCE_FILE   = pandoc-manuscript.docx
ODT_REFERENCE_FILE    = pandoc-manuscript.odt
TEMPLATE_FILE_LATEX   = pandoc-peerj.latex

PANDOC_SCHOLAR_PATH = pandoc-scholar
include $(PANDOC_SCHOLAR_PATH)/Makefile

Metadata Features

Pandoc-scholar supports additional functionality via metadata fields. Most notably, the augmentation of articles with author and affiliation data, which is essential for academic publishing, is greatly simplified when using pandoc-scholar.

Authors and affiliations

Most metadata should be specified in the YAML block at the top of the article. Author data and affiliations are taken from the author and institute field, respectively. Institutes can be given via user-defined abbreviations, saving unnecessary repetitions while preserving readability.


  - James Dewey Watson:
      institute: cavendish
  - Francis Harry Compton Crick:
      institute: cavendish
  - cavendish: Cavendish Laboratory, Cambridge

Authors are given in the order in which they are listed, while institute order follows from author order.

The separate institute field may add unwanted complexity in some cases. It is hence possible to omit it and to give the affiliations name directly in the author entry:

  - John MacFarlane:
      institute: University of California, Berkeley

Institute address

Often it is not enough to give just a name for institutes. It is hence possible to add arbitrary fields. The name must then explicitly be set via the name field of the institute entry:

  - Robert Winkler:
      institute: cinvestav
  - cinvestav:
      name: 'CINVESTAV Unidad Irapuato, Department of Biochemistry and Biotechnology'
      address: 'Km. 9.6 Libramiento Norte Carr. Irapuato-León, 36821 Irapuato Gto. México'
      phone: +52 (462) 623 9635

Currently only the institute's address is used in the default template, but future extensions will be based on this convention.

Semantic citations

Understanding the reason a citations is included in scholarly articles usually requires natural language processing of the article. However, navigating the current literature landscape can be improved and by having that information accesible and in a machine-readable form. Pandoc-scholar supports the CiTO ontology, allowing authors to specify important meta-information on the citation directly while writing the text. The property is simply prepended to the citation key, separated by a colon: @<property>:citationKey.

The following table contains all supported keywords and the respective CiTO properties. Authors are free to use the short-form, the full-length property, or any of the alternatives listed below (i.e., all word in a row denote the property and have the same effect).

CiTO property Keyword alternatives
agrees_with agrees_with agree_with
cites_as_authority authority as_authority
cites_as_data_source data_source as_data_source
cites_as_evidence evidence as_evidence
cites_as_metadata_document metadata as_metadata_document
cites_as_recommended_reading recommended_reading as_recommended_reading
includes_excerpt_from excerpt excerpt_from
includes_quotation_from quotation quotation_from
obtaines_background_from background background_from
uses_data_from data_from data
uses_method_in method method_in


DNA strands form a double-helix [@evidence:watson_crick_1953].


Copyright © 2016–2021 Albert Krewinkel and Robert Winkler except for the following components:

  • HTML template: © 2016 Andrew G. York and Diana Mounter
  • dkjson: © 2010-2013 David Heiko Kolf
  • lua-filters: © 2017-2021 Albert Krewinkel, John MacFarlane, and contributors.

This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.