A text processing and literate programming tool inspired by antiweb, noweb

## Project description

pnw is a text processing tool with a wide range of applications. It can be used to compile chunks of text from various sources into a new text file. The advantage over cut-and-paste techniques is that changes in the source file are automatically followed. Secondly, we provide a thin wrapper around the pandoc document converter. Therefore, chunks of text can be converted from and to a wide range of lightweight markup languages as well as html and tex/latex. Although literate programming has become a little out of fashion, pnw can be used as a literate programming tool, too. Not surprisingly, pnw itself and its documentation is written using pnw. pnw is remotely inspired by noweb and heavily inspired by (even borrowing part of the syntax from) antiweb. An important feature of pnw, which is not present in antiweb, is the support of namespaces. This allows to create reusable block libraries which can be imported and addressed in a dot-notation syntax similar to python modules.

A pnw file is a text file with tags defining named blocks of text. In a way this is similar to an xml file. Tags may be hidden behind comment characters, such that the file remains processable by another program. The blocks of text from various pnw files can be processed and rendered by another pnw file.

## Documentation

The documentation is written in latex. To compile the pdf pnw-doc.pdf run:

make doc

from the command line. You can also find pnw on the python package index. There pnw-doc.pdf is provided. I don’t put it under version control because it will unnecessarily inflate the repo.

## Example

Assuming that pnw is installed, from the main directory of the pnw package open an interactive python session, I recommend to use ipython, and write

import pnw
pnw.pnwimport('pnw.pnw','')

Now you can explore the pnw document tree of the pnw.pnw source file by inspecting the dictionary

pnw.chunks

To see the structure of a block, just print its repr

pnw.chunks['moduledoc']

To render it, try

print pnw.chunks['moduledoc'].render(indent=0,depth=2,toformat='rst')

The pnw source consists of one file pnw.pnw. It is a working python file containing pnw tags hidden behind the # comment character. To produce a clean python file we invoke (compare the makefile in the main directory)

python ./pnw.pnw  -F pnw.pnw > pnw.py

The -F directive means that the whole file is treated as one block named u''. The previous command is used when installing pnw for the first time.

The main driver file for the documentation is pnw-doc.pnw in the doc directory. Here is the head of this file

#@chunk(main,format=latex)
#@path(../..)
#@path(..)
#@import(latex-blocks.tex)
#@import(md-blocks)
#@import(pnw.pnw)
#@include(ltxheader)

First the whole file is declared to be a block named main which is latex formatted. Then we add the first two parent directories to the search path for imports. Then the files latex-blocks.tex, md-blocks, pnw.pnw are included into the main namespace u''. The block ltxheader, which contains what its name says, and which can be found in latex-blocks.tex is included.

pnw-doc.pnw contains a little of the documentation written in latex. Mostly, it includes blocks from other files. The bulk of the documentation, written in markdown, can be found in md-blocks.

Finally, an executable latex file is produced by invoking

pnw -Rmain pnw-doc.pnw > pnw-doc.tex

pnw is on Bitbucket and on PyPi.

## Project details

Uploaded source