foliantcontrib.pandoc 1.1.2

Pandoc backend for Foliant doc maker. Produces PDF and DOCX.

Pandoc Backend for Foliant

Pandoc is a Swiss-army knife document converter. It converts almost any format to any other format: md to pdf, rst to html, adoc to docx, and so on and so on.

Pandoc backend for Foliant adds pdf, docx, odt, epub and tex targets.

Installation

$ pip install foliantcontrib.pandoc

You also need to install Pandoc and TeXLive distribution for your platform.

Usage

Build pdf:

$ foliant make pdf -p my-project
Parsing config... Done
Applying preprocessor flatten... Done
Making pdf with Pandoc... Done
─────────────────────
Result: My_Project-2020-12-04.pdf

Build docx:

$ foliant make docx -p my-project
Parsing config... Done
Applying preprocessor flatten... Done
Making docx with Pandoc... Done
─────────────────────
Result: My_Project-2020-12-04.docx

Build odt:

$ foliant make odt -p my-project
Parsing config... Done
Applying preprocessor flatten... Done
Making odt with Pandoc... Done
─────────────────────
Result: My_Project-2020-12-04.odt

Build docx:

$ foliant make epub -p my-project
Parsing config... Done
Applying preprocessor flatten... Done
Making epub with Pandoc... Done
─────────────────────
Result: My_Project-2020-12-04.epub

Build tex (mostly for pdf debugging):

$ foliant make tex -p my-project
Parsing config... Done
Applying preprocessor flatten... Done
Making docx with Pandoc... Done
─────────────────────
Result: My_Project-2020-12-04.tex

Config

You don't have to put anything in the config to use Pandoc backend. If it's installed, Foliant will detect it.

You can however customize the backend with options in backend_config.pandoc section:

backend_config:
  pandoc:
    pandoc_path: pandoc
    build_whole_project: true
    template: !path template.tex
    vars:
      ...
    meta:
      ...
    reference_docx: !path reference.docx
    reference_odt: !path reference.odt
    css: !path epub.css
    params:
      ...
    filters:
      ...
    markdown_flavor: markdown
    markdown_extensions:
      ...
    slug: My_Awesome_Custom_Slug

pandoc_path : is the path to pandoc executable. By default, it's assumed to be in the PATH.

build_whole_project : added in 1.1.0 If true, whole project will be built into a single flat document. Default: true.

template : is the path to the TeX template to use when building pdf and tex (see “Templates” in the Pandoc documentation).

> **Tip**
>
> Use `!path` tag to ensure the value is converted into a valid path relative to the project directory.

vars : is a mapping of template variables and their values. They will be added to pandoc command as --variable key[=val].

meta : is a mapping of document metadata properties and their values. They will be added to pandoc command as --metadata key[=val].

reference_docx : is the path to the reference document to be used when building docx (see --reference-doc option info in the Pandoc Options documentation).

reference_odt : is the path to the reference document to be used when building odt (see --reference-doc option info in the Pandoc Options documentation).

css : is the path to the stylesheets to be used when building epub (see --css option info in the Pandoc Options documentation).

params : are passed to the pandoc command. Params should be defined by their long names, with dashes replaced with underscores (e.g. --pdf-engine is defined as pdf_engine).

filters : is a list of Pandoc filters to be applied during build.

markdown_flavor : is the Markdown flavor assumed in the source. Default is markdown, Pandoc's extended Markdown. See “Markdown Variants” in the Pandoc documentation.

markdown_extensions : is a list of Markdown extensions applied to the Markdown source. See Pandoc’s Markdown in the Pandoc documentation.

slug : is the result file name without suffix (e.g. .pdf). Overrides top-level config option slug.

Example config:

backend_config:
  pandoc:
    template: !path templates/basic.tex
    vars:
      toc: true
      title: This Is a Title
      second_title: This Is a Subtitle
      logo: !path templates/logo.png
      year: 2020
    params:
      pdf_engine: xelatex
      listings: true
      number_sections: true
    markdown_extensions:
      - simple_tables
      - fenced_code_blocks
      - strikeout

Build modes

Since 1.1.0 you can build parts of your project into separate PDFs, along with the main PDF of the whole project.

If the build_whole_project parameter of Pandoc backend config is true, the whole project will be built in to a flat document as usual. You can disable it by switching build_whole_project to false.

You can also build parts of your project into separate documents. To configure such behavior we will be adding Metadata to chapters or even smaller sections.

To build a chapter into a separate document, add the following meta tag to your chapter's source:

<meta
  pandoc="
    vars:
      toc: true
      title: Our Awesome Product
      second_title: Specifications
      logo: templates/logo.png
      year: 2020
  "></meta>

# Specifications

size: 15
weight: 59
lifespan: 9

In the example above we have added a meta tag with pandoc field, in which we have overriden the vars mapping. The pandoc field is essential in this case. This is how backend determines that we want this chapter built separately. If you don't want to override any parameters, you can just add pandoc="true" field.

All parameters which are not overriden in the meta tag will be taken from main config foliant.yml.

Now, as the pandoc field is present in one of the meta tags in the project, Pandoc backend should build not one but two documents. Let's check if it's true:

$ foliant make pdf
Parsing config... Done
Applying preprocessor flatten... Done
Making pdf with Pandoc... Done
─────────────────────
Result:
My_Project-2020-12-04.pdf
Specifications-2020-12-04.pdf

That's right, we've got the main PDF with whole project and another pdf, with just the Specifications chapter.

If you wish to build even smaller piece of the project into separate file, add meta tag under the heading which you want to build:

# Specifications

size: 15
weight: 59
lifespan: 9

## Additional info

<meta
  pandoc="
    slug: additional
    vars:
      toc: true
      title: Our Awesome Product
      second_title: Additional info
      logo: templates/logo.png
      year: 2020
  "></meta>

Lorem ipsum dolor sit amet consectetur adipisicing elit. Deleniti quos provident dolores eligendi nam quia sequi et tempore enim blanditiis, consequatur nostrum nulla dolor laborum quasi molestiae perspiciatis magni error consectetur nesciunt eaque veritatis voluptates! Cupiditate illum enim id recusandae assumenda excepturi odit tempore incidunt, amet soluta necessitatibus corrupti, aliquam.

In this example only the Additional info section will be built into a separate document. Notice that we've also given it its own slug.

Let's build again and look at the results:

$ foliant make pdf
Parsing config... Done
Applying preprocessor flatten... Done
Making pdf with Pandoc... Done
─────────────────────
Result:
My_Project-2020-12-04.pdf
additional.pdf

Troubleshooting

Could not convert image ...: check that rsvg2pdf is in path

In order to use svg images in pdf, you need to have rsvg-convert executable in PATH.

On macOS, brew install librsvg does the trick. On Ubuntu, apt install librsvg2-bin. On Windows, download rsvg-convert.7z (without fontconfig support), unpack rsvg-convert.exe, and put it anywhere in PATH. For example, you can put it in the same directory where you run foliant make.

LaTeX listings package does not work correctly with non-ASCII characters, e.g. Cyrillic letters

If you use non-ASCII characters inside backticks in your document, you can see an error like this:

Error producing PDF.
! Undefined control sequence.
\lst@arg ->git clone [SSH-к
                           люч]
l.627 ...through{\lstinline!git clone [SSH-ключ]!}

To fix it, set listings parameter to false:

backend_config:
  pandoc:
    ...
    params:
      pdf_engine: xelatex
      listings: false
      number_sections: true
    ...