PythonPowerTools

These tools are ports of the BSD utilities from the PerlPowerTools (PPT) project, started by Tom Christiansen in February, 1999, and currently maintained by brian d foy.

The Perl sources are drawn from the PPT GitHub repository . Default license for each tool is Perl's Artistic License 2.0, which can be found in LICENSE.md.

The steward for this license is the Perl and Raku Foundation. I've copied the license from this URL, adding markdown to the title and subheads by hand to make it look a little nicer.

At brian's suggestion, I've adopted this license for the project as a whole, but some individual tools have a different license, which is specified in the source of that tool. For example, fortune is GPL-licensed, so that license should be retained for the converted version as well. Even if a lawyer could argue that the original license need not be applied to the converted version, re-imposing it seems like good manners.

TODO

• I'm going to try to create a first draft of each tool through a direct conversion by an AI assistant. Will that work? Who knows.

  If you try to use a Python version of one of the tools, assume it's completely untested.

• I'll have to add tests.

The PPT repository testing directory, PPT/t only lists individual tests for 25 of the 125 tools in PPT/bin, so test-driven-development (TDD) is impractical.

I have no plan to add tests until I have created drafts of all the tools.

Caveat emptor.

• pyproject.toml needs to be augmented to add package dependencies for a few utilities.

• I need to enhance bin/port and bin/writeport to work on Linux. Right now, they just work on my Mac.

• Starting with spell, I ran out of quota for Gemini Pro 2.5 and had to fall back to Gemini 2.5 Flash. Perhaps these won't be as good. The conversions are much faster, and lack both the preamble, about what the conversion will do, and the explanation afterwards of how the converted code works.

• Even with Flash, tools with huge amounts of internal data, like factor or banner convert slowly.

Notes

• First cuts were all done with Google Gemini 2.5 Pro, starting with the tool with the fewest lines, rot13, and proceding to the one with the most, ching.

• The tools used for the conversion, port and writeport can be found in bin/.

I've only put the code into python/<toolname>.py, but Gemini typically generated a summary before the code and discussion of how the code worked afterwards. To see either, just convert the code yourself!

• Many utilities are no longer very useful.

The utility asa interprets ASA/FORTRAN carriage-controls, which is only useful to folks who work in an institution that still uses FORTRAN, such as the National Center for Atmospheric Research (NCAR).

For the rest of us, asa is historically interesting. It offers a glimpse into the history of computing: what used to be important enough to write utilities around.

• Some PPT utilities are really Perl-specific. Some have been converted into Python analogues. For example, par generates a self-extracting Python script instead of a self-extracting Perl script.

Others like PPT/bin/awk and PPT/bin/find don't convert at all, because they depend on Perl. awk is a front-end for a2p, and find is a front end for find2perl. Python implementations of either are possible, but need to be done as an independent effort.

• Gemini Pro 2.5 thinks PPT/bin/dc is impossible to implement in Python. The dc language is too complex and un-Pythonic. When Gemini Flash had no problem with PPT/bin/bc, I tried giving it dc, and it converted that, too. This makes me suspicious of Gemini Flash.