Thanks for offering to contribute to xd!
This is a repo of helper scripts for converting puzzles to .xd and importing them into gxd repo.
In a terminal window, type (or some variant according to your system):
pip3 install -r requirements.txt
Download a puzzle, for example from the eminent Brendan Emmett Quigley:
xdfile/puz2xd.py test.puz > test.xd
The crossword module's from_ipuz needed to be updated, due to changes in the ipuz Python module.
It has been incorporated into this repo with those updates.
First, ensure that Python can find it
export PYTHONPATH=.
and then
xdfile/ipuz2xd.py test.ipuz > test.xd
sh convert_puz.sh <srcdir> <destdir>
or
sh convert_ipuz.sh <srcdir> <destdir>
where:
<srcdir>is the directory holding the .puz<destdir>is the the directory which will hold the .xd
The input .puz currently must adhere to the standard filename format (idYYYY-MM-DD.puz or id-NNN.puz) or the xdfile.py script won't convert it.
To be merged into gxd, the output file must be named correctly, and the file must be in the correct directory. For example:
latimes/2019/lat2019-10-11.xd
(Non-daily publications generally omit the year directory.)
- Create a new empty repo on gitlab.
- Commit and push the puzzles to a branch.
- Invite me to the repo on gitlab.
- You'll be added as a collaborator to the gxd repo.
- Submit a PR from your repo to gxd.
Don't give these puzzles to anyone else. These puzzles are to be used only for research, and not for consumption.
If filename has an unknown pubid, the xd pipeline won't analyze it properly.
If the publication is legitimate but unknown, add a row to publications.tsv
in the root of the gxd repo.
If you publish any research or results using the xd crossword corpus, please cite xd accordingly:
Pwanson, Saul. (2019). "xd Crossword Corpus." Retrieved from http://xd.saul.pw
The year of publication should be the year that the corpus was accessed.