latextree 0.2

A document object model and related tools for Latex

The latextree Package

Creates a document object model for LaTeX and provides some useful functions. The parser understands the syntax of most common macros but is NOT a replacement for a full LaTeX engine. The parser should only be invoked only after the input has been successfully compiled under LaTeX.

Example:

>>> from latextree.parser import LatexParser
>>> pa = LatexParser()
>>> filename = "/path/to/main.tex""
>>> doc = pa.parse_latex_file(filename)
>>> doc.make_website()			# create website
>>> bbqs = doc.bbq()			# extract mc/ma questions in Blackboard format (plain text)
>>> docx = doc.root.get_xml()	# create xml document
>>> from lxml import etree
>>> print(etree.tostring(docx, pretty_print=True))

If latextree is installed as a package:

>>> import latextree
>>> filename = "/path/to/main.tex""
>>> doc = load(filename)
>>> doc.make_website()	# create website
>>> bbqs = doc.bbq()	# extract mc/ma questions in Blackboard format (plain text)
>>> exit()

---

Standards

The following are yet to be implemented properly.

• Exception handling
• Testing
• Logging
• Documentation

---

make_website

>>> doc.make_website(copy_static=True, copy_figures=True, WEB_ROOT=None)

By default, the website is created in a folder called web-latexmainfilename in the directory containing the main latex source file. . The output directory can be changed by setting the WEB_ROOT variable.

• html files are written to WEB_ROOT
• style files are copied to WEB_ROOT/static/css. These can be modified post-production.
• figures are copied to WEB_ROOT/static/img
• urls are constructed from chapter and/or section numbers (if any)

Templates

The LatexDocument.make_website() function uses jinja2 templates.

• scripts.html loads the MathJax and jquery CDNs from cloudflare.
• node.html does most of the work. This needs to be broken up.

Static files

• css/: The stylesheets are rough-and-ready and need a lot of work (especially for phones and tablets).
• img/: Image files are copied into here.
• js/: Currently contains only a simple show/hide function.

---

Modules

• bibliography: wrapper for bibtexparser
• content: content nodes
• document: LatexDocument class and output functions
• factory: creates node classes and objects
• ltree: command line tools
• macrosdef: macro definitions for pylatexenc.latexwalker
• node: base class for LatexTreeNode objects.
• parser: LatexParser class. Creates LatexDocument objects.
• preprocessor: prepare latex source for pylatexenc.latexwalker(replace $$...$$ etc.)
• reader: input functions and utilities
• settings: website settings (e.g. mathjax cdn)
• tabular: parse tabular environments.
• taxonomy: register of species (e.g. list) and genera (e.g. itemize).
• walker: wrapper for pylatexenc.latexwalker.

Dependencies

• bibtexparser
• jinja2
• lxml
• pylatexenc

Test files

The tex directory contains:

• LatexTreeTestBook for testing the book document class
• LatexTreeTestArticle for testing the article document class
• LatexTreeTestExam for testing the exam document class

Existing software for Latex processing

• pylatexenc

  • LaTeX -> Unicode
  • python package

• plasTeX

  • LaTeX -> XML and XHTML
  • python package

• tex4ht

  • LaTeX -> XML, HTML and Braille!
  • java and c libraries
  • provides a htlatex cmd to replace latex.

• pandoc

  • LaTeX -> HTML
  • haskell library

• tralics (INRIA)

  • LaTeX -> XML