\documentclass[11pt]{article} \topmargin=0in \oddsidemargin=0.25in \evensidemargin=0.25in \textwidth=6in \textheight=8in \setlength{\vfuzz}{4pt} % to ignore trivial vertical overfull boxes \sloppy \usepackage[noBBpl]{mathpazo} \usepackage{enumerate} \begin{document} \title{DE-MACRO} \author{Copyright 2005-2007 P\'eter G\'acs\\ Licensed under the Academic Free Licence version 2.1} \maketitle Version 1.4 - Luca Citi made it python2.7 and python3 compatible. Peter Gacs improved the parsing of \verb`\input{}`, and made @ a letter in the style files. Version 1.3 - this version is much more conservative about deleting comments and inserting or deleting blank space: tries to leave in all comments, adds space only when necessary, and tries not to delete space in the main text. The motivating comments came from Daniel Webb. Version 1.2 - a syntactical bug corrected, thanks Brian de Alwis! \section*{Purpose} This program can eliminate most private macros from a LaTeX file. Applications: \begin{enumerate}[--] \item your publisher has difficulty dealing with many private macros \item you cooperate with colleagues who do not understand your macros \item preprocessing before a system like latex2html, which is somewhat unpredictable with private macros. \end{enumerate} \section*{Platform} This is a Python program, which I only have tried to run under Unix. But the Unix dependence is minimal (for example all the directory path references are platform-independent). It should be easy to adapt the program to Windows, and also to avoid command-line arguments. In case your Python is not in \verb`/usr/bin`, you should change the top line (the "shebang" line) of the program accordingly. This top line uses the \verb`-O` option for python (stands for ``optimize''). Without it, the program may run too slowly. If you do not care for speed, a number of other complications (the database, the checking for newer versions) could be eliminated. \section*{Usage} Command line: \begin{verbatim} de-macro [--defs ] [.tex] [[.tex] ...] \end{verbatim} {\bfseries Simplest example:} \verb`de-macro testament`\\[0pt] (As you see, the \verb`<>` is used only in the notation of this documentation, you should not type it.) If \verb`` contains a command \verb`\usepackage{-private}` then the file \verb`-private.sty` will be read, and its macros will be replaced in \verb`` with their definitions. The result is in \verb`-clean.tex`. Only \verb`newcommand`, \verb`renewcommand`, \verb`newenvironment` and \verb`renewenvironment` are understood (it does not matter, whether you write \verb`new` or \verb`renew`). These can be nested but do not be too clever, since I do not guarantee the same expansion order as in TeX. \section*{Files} \begin{verbatim} .db -clean.tex -private.sty \end{verbatim} For speed, a macro database file called \verb`.db` is created. If such a file exists already then it is used. If \verb`-private.sty` is older than \verb`.db` then it will not be used. It is possible to specify another database filename via \verb`--defs `. Then \verb`.db` will be used. (Warning: with some Python versions and/or Unix platforms, the database file name conventions may be different from what is said here.) For each \verb``, a file \verb`-clean.tex` will be produced. If \verb`-clean.tex` is newer than \verb`.tex` then it stays. \section*{Input command} If a tex file contains a command \verb`\input{}` or \verb`\input ` then \verb`.tex` is processed recursively, and \verb`-clean.tex` will be inserted into the final output. For speed, if \verb`-clean.tex` is newer than \verb`.tex` then \verb`.tex` will not be reprocessed. The dependency checking is not sophisticated, so if you rewrite some macros then remove all \verb`*-clean.tex` files! \end{document}