How latexpp works

The latexpp preprocessor relies on pylatexenc 2.0 to parse the latex document into an internal node structure. For instance, the chunk of latex code:

Hello, \textit{world}! % show a greeting

will be parsed into a list of four nodes, a ‘normal characters node’ "Hello, ", a ‘macro node’ \textit with argument a ‘group node’ {world} which itself contains a ‘normal characters node’ world, a ‘normal characters node’ "! ", and a ‘latex comment node’ % show a greeting. The structure is recursive, with e.g. macro arguments and environment contents themselves represented as nodes which can contain further macros and environments. See pylatexenc.latexwalker for more information. The pylatexenc library has a list of some known macros and environments, and knows how to parse their arguments. Some fixes in latexpp add their own macro and environment definitions.

Once the latex document is parsed into the node structure, then the fix classes are given a chance one by one to traverse the node structure and make the necessary fixes. After all the fixes have been applied, the resulting structure is written to the output file as latex code.