Remarks on Pandoc & DocBook

remarks on pandoc

There are plenty docbook converter out. Most of them have common, that they implement there own (less XML) parsers and they always implement only a subset of the DocBook markup.

The coverage of the pandoc DocBook reader is the most advanced, but it is merely a subset of the DocBook markup. If the pandoc reader does not implement a markup, it will not be in the AST. This is, why it can’t be handled in a pandoc filter. The coverage of the pandoc DocBook reader is documented in the sources:

The pandoc reStructuredText writer is also incomplete and has several bugs, for details take a look at the sources:

On my Ubuntu 15.10 sandbox is the pandoc version 1.13.2 via package manager available. In this version, the DocBook reader missed some common used markups like <xref>. Therefore the development version of pandoc is needed (see pandoc developer installation). Remind, building the Haskell compiler and compiling pandoc does massive time, memory and CPU consumption.

The pandoc DocBook reader does not understand XML nor SGML, therefore it will not do all these XML stuff, like reading external entities (this is where the dbxml2rst library comes into account).

DocBook remarks

DocBook documentation is available from: