NBNF Tools

Converting NBNF grammars embedded in Markdown documents to other formats

What are NBNF grammars?

NBNF (eNriched BNF) is an extension of BNF which adds a bit syntactic sugar to the latter. In particular this means that every grammar which can be expressed in NBNF can also be expressed in BNF. Hence NBNF does not leave the scope of context-free grammars.

For a description of the syntax and semantics of NBNF, have a look at the file docs/nbnf-description.md. A fully formal specification of the syntax of NBNF as it is accepted by this package's parser can be found in the file docs/nbnf-syntax.md.

Features

This package provides a tool called nbnf-exe which can be used to convert Pandoc Markdown (in the following just called Markdown) files containing NBNF rules to the following formats:

• Plain text
• XML
• HTML

When converted to plain text or XML the resulting file will only contain the NBNF rules of the Markdown file, whereas when converting to HTML it will also contain any sourrounding text.

Thus for nbnf-exe to be able to distinguish between NBNF rules and other content of the Markdown file, it is necessary to enclose every single rule within ```nbnf and ```, e.g.

```nbnf
<foo> = <bar>
      | "baz"
```

Build

To build the executable nbnf-exe, run the following command from within the root directory of this repository:

stack build

Usage

To convert a Markdown document which contains NBNF rules to plain text, XML or HTML run the following command:

stack exec nbnf-exe -- -i INFILE -o OUTFILE [-h (standalone|jekyll)]

The -i option specifies an input file. If several instances of it occur, the specified files will be concatenated in the order they appear and treated as one big file.

The option -o specifies the output file. Its file name extension determines outptu format. Supported are:

• .txt (plain text)
• .html (HTML)
• .xml (XML)
• .log (debug log)

Warning: If a file with the same name as the output file already exists it will be overridden without further inquiry!

If the output format is HTML, it must be specify whether the output file should be a stand-alone document (via -h standalone) or a HTML file which is adapted to be used with Jekyll (via -h jekyll).

Moreover, the option --help prints a little help message.