latex-rubber 1.6.5

an automated system for building LaTeX documents

This is Rubber.

Rubber is a building system for LaTeX documents. It is based on a routine that runs just as many compilations as necessary and runs companion programs like BibTeX and Makeindex when needed. The module system provides a great flexibility that virtually allows support for any package with no user intervention, as well as pre- and post-processing of the document. A good number of standard packages are supported, including graphics/graphicx with automatic conversion between various graphics formats and Metapost compilation. The associated tool "rubber-info" extracts information, such as dependency relations or post-compilation diagnostics.

Installation

Running Rubber just requires Python 3.8. Of course it won't be of much use without a working LaTeX environment (Rubber is known to work on TeXLive and VTeX on various flavors of Unix including Darwin and Cygwin, any feedback is welcome about other systems).

For compilation, you will need the Python Distutils, which are usually included in development packages (in Debian, this is the python3-dev package).

To build the documentation, you need texinfo (Debian package: texinfo). Note that all the documentation formats are disabled by default.

To compile and install Rubber, just follow the usual procedure:

# python3 setup.py --help
# python3 setup.py install
# python3 setup.py clean --all

Some useful options to setup.py include:

Disabling info docs:

# python3 setup.py build --info=False install

and similar for --html, --man, --pdf.

Changing the installation path for manpages:

# python3 setup.py install --mandir=/path/to/man/pages

Installing to a staging directory instead of the root/prefix:

# python3 setup.py install --root=/staging/directory

Note that if you need build and install to be two different steps (for example when building packages for distribution purposes), Python's distutils will forget about any "build" options, and re-build with default options during the "install" stage. This is worrysome if you'd like not to build some of the documentation. It is then best to make options permanent by putting them info a setup.cfg file. For example:

[build]
man = 1
html = 0
pdf = 0
info = 0
txt = 0
[install]
prefix  = /usr

Finally, invoke

# python3 setup.py build
# python3 setup.py install --root=/staging/directory

PEP 517 Build

The newfangled build process for Python processes is specified in PEP 517. There's initial support for this, but the packages will not include the documentation.

The development flow is as follows:

Setup a virtual environment, using a Python version of your choice:

# python3 -m venv .venv

Activate the virtual environment, and install the requirements:

# source .venv/bin/activate
(.venv) # pip install -U pip setuptools wheel
(.venv) # pip install -r requirements-dev.txt

This is the virtual environment setup, it only needs to be done once. If the requirements change, rerun this latest step.

To create a PEP 571-compatible sdist and wheel, run:

# make build-wheel

This uses the "build" module to create the files in the "dist" folder. These are the files to be uploaded to PyPI.

PyPI Release

To publish the PEP 517

See https://twine.readthedocs.io for how to setup twine regarding credentials.

After building the sdist and wheel, upload them to PyPI by using:

# make publish-wheel

Usage

As civility requires, saying rubber --help and rubber-info --help provides a short description of the command line's syntax. The included manual pages (available in English and French) and Texinfo documentation contain more precise usage information.

Known Bugs

Rubber is generally working fine, though there are some known issues.

1. Rubber tries to do too much. Rubber attempts to provide a one-stop solution to compiling a TeX document, and run a lot of related software as needed. This approach is fragile for three reasons: a) It does so by parsing the .tex file itself, and all included files, and discovering any related input files. TeX is a hard language to parse correctly without embedding a full TeX interpreter, which rubber does not do. b) To do its work, Rubber needs to be taught about every version of every package in the TeX ecosystem in order to gauge its impact on the compilation of a TeX document. It needs to know the preferences of any TeX compiler with regards to image formats, search paths etc. All this information is hard to keep up to date. c) In some cases (like image conversion), it needs to outright guess what the user intends to do, and may in fact guess incorrectly. In a future release, some of these features may be taken out in favor of more modern ways to accomplish the same thing, with to goal to render Rubber simpler and more robust. One might want to make use of the -recorder feature (.fls) for example instead of attempting to read the same information from the human-readable log file.

2. The codebase has been cleaned up considerably, has been converted to Python3 and is generally in a sane state. Nevertheless, it has been written over a number of years, and some features would be implemented differently or skipped altogether if a rewrite were attempted (e.g. the onchange mechanism, modules, ...).

3. In some cases, Rubber will trigger a recompile that is arguably unnecessary. Rubber tries to err on the side of caution here.

Author

Rubber was originally written by Emmanuel Beffara [e-mail]. It is currently maintained by Florian Schmaus [e-mail], Sebastian Kapfer [e-mail] and Nicolas Boulenguez [e-mail].

Its homepage can be found at https://gitlab.com/latex-rubber/rubber

Thanks to all those who provided testing, comments and suggestions, and re-thanks to those who wrote patches and bugfixes.

Any kind of feedback is appreciated, in order to make this program as useful and robust as possible.

License

Rubber is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program. If not, see https://www.gnu.org/licenses/.

SPDX-License-Identifier: GPL-3.0-or-later