puzpy 0.6.1

python crossword puzzle library

puz.py: Python parser for .puz files

Python library to read and write crossword puzzle files in Across Lite .puz and .txt formats.

Installing

From pypi (recommended)

pip install puzpy

From source (for development)

uv sync --extra dev

Or with pip:

pip install -e .[dev]

Example Usage

import puz

#
# Load a puzzle file:
#
p = puz.read('testfiles/washpost.puz')

#
# Print all clues and their answers
#
clues = p.clue_numbering()

print('Across')
for clue in clues.across:
    print(f'{clue.number}. {clue.text} - {clue.solution}')

print('Down')
for clue in clues.down:
    print(f'{clue.number}. {clue.text} - {clue.solution}')

#
# Print the puzzle grid
#
for row in p.grid():
    print(' '.join(row))

#
# Unlock a puzzle that has a locked solution
#
p.unlock_solution(7844)

# Now print the unscrambed solution grid:
for row in p.solution_grid():
    print(' '.join(row))

#
# Save a puzzle with modifications:
#
p.fill = 'LAMB' + p.fill[:4]
p.save('example.puz')

#
# New! Convert from Across Lite text format to .puz:
#
p2 = puz.read_text('testfiles/text_format_v1.txt')
p2.save('example.puz')

Notes

The parser is as strict as Across Lite, enforcing internal checksums and magic strings. The parser is designed to round-trip all data in the file, even fields whose utility is unknown. This makes testing easier. It is resilient to garbage at the beginning and end of the file (for example some publishers put the filename on the first line and some files have a rn at the end).

In addition to the handful of tests checked in here, the library has been tested on over 9700 crossword puzzles in .puz format drawn from the archives of several publications including The New York Times, The Washington Post, The Onion, and, the Wall Street Journal. As of writing, it can round-trip 100% of them with full fidelity.

Running tests

Unit tests

uv run pytest

Lint

uv run ruff check

Type check

uv run pyright

Viewer

This repo includes a puz file viewer. The script renders a .puz or .txt file as a print-ready stand-alone HTML file. Its only dependency is puz.py. Examples: 15x15, 21x21, ...

$ python puz_viewer.py --help
usage: puz_viewer.py [-h] [-o OUTFILE] [--outdir OUTDIR] [-f {auto,puz,txt}] [--index] [puzzles ...]

Generate an HTML viewer for a crossword puzzle or puzzles

positional arguments:
  puzzles               Paths to .puz or .txt files (default: stdin)

options:
  -h, --help            show this help message and exit
  -o, --outfile OUTFILE
                        Output HTML filename (default: stdout, or auto-generated from input filename in batch mode)
  --outdir OUTDIR       Output directory for HTML files (default: .)
  -f, --format {auto,puz,txt}
                        Input format (default: auto-detect)
  --index               Generate index.html in output directory (batch mode)

Python version support

Python >=3.9 required since 0.3.1. For older python versions including 2.x use puzpy==0.2.6

Resources

• AcrossLite .puz file format (archived from http://code.google.com/p/puz/wiki/FileFormat)
• Across text format (archived from https://www.litsoft.com/across/docs/AcrossTextFormat.pdf)

License

MIT License.