jgm/pandoc 1.12

pandoc 1.12

on GitHub

Complete changelog

The focus of this release is customizability. Several features
have been added that should make it easier to fit pandoc to your
needs:

• Flexible metadata: metadata is no longer limited to title,
  authors, and date. You can add whatever fields you like. Metadata
  values are also structured: they can be lists, maps, booleans,
  strings, or lists of pandoc Inline or Block elements. A new
  YAML metadata block is provided so structured metadata can be
  included in pandoc markdown documents. (Of course, the old title
  blocks will still work, and can be combined with the YAML blocks.)
  Template variables will automatically be set from metadata, so
  if you want to create a template with variables like subject,
  abstract, or author.institution, you can.

• A new --filter option is provided, to make it easier to use
  JSON filters that transform the pandoc AST. Instead of

  pandoc -t json -s | ./myfilter | pandoc -f json -t html5 -s
  

  you can now do

  pandoc -t html5 -s --filter ./myfilter
  

  Pandoc passes the output format as the first argument to the
  filter, so filters can easily be made sensitive to the output
  format. (For example, a graphviz filter might produce a PNG
  if the output is HTML and a PDF if the output is LaTeX.)

• For those who don't know Haskell, a python library,
  pandocfilters, has been created to make it easy to write pandoc
  filters using python. There are some examples in the
  pandocfilters repository.

• You can now script your own custom pandoc writer using lua. See
  example 33 on the
  demos page. While
  JSON filters are a better approach for most customizations, this
  approach gives you complete control over the writer's behavior,
  and allows you to support completely new output formats.

• Generic Inline (Span) and Block (Div) containers that take
  attributes have been added to the underlying Pandoc types.
  These provide the customization possibilites for formatted
  text that the Code and CodeBlock elements provide for unformatted
  text.

Other significant changes include support for OPML (as both an input
and an output format), support for writing reveal.js slide shows,
and support for reading Haddock markup.

Citation support has been extracted from core pandoc and put into
a new pandoc-citeproc package. (Users who install pandoc using
cabal should cabal install pandoc-citeproc if they want to use
citations.) Bibliographic information can now be included right in
a pandoc markdown document, inside a YAML metadata block, making
the document self-contained. (A tool biblio2yaml is provided to
convert existing bibliographies, for those who want to do this.)