hpyproject / top4000-pypi-packages Goto Github PK

Somehow these two projects do not show up in the top downloads. They are ways to wrap C++ and rust code via the C-API, so they are very relevant for the goals of HPy

It would be nice to see what cython constructs are typically used in the wild. Here is a start at an attempt. It does not clean up everything, and does not handle ZIP files:

# Download the sdists as per the README
for f in $(ls *.tar.gz *.tar.bz2); do 
    name=$(tar -tf $f | head -1); 
    name=${name/\/*/};      # Get the name of the directory
    echo $name; 
    tar -xf  $f; 
    for suffix in py txt rst in cfg yaml yml bat sh jar data vec avi xml jpg dic md pdf def so tiff; do 
        find $name -name "*.$suffix" -delete;
    done;
    find $name -type f ! -name "*.*" -delete;  # delete files with no suffix
    find $name -empty -type d -delete;
    rm $f;
done

Unfortunately I cannot make a PR since I don't have the exact json file originally used.

Edit: add jar, data, vec, avi, xml, jpg, dic, md, pdf, def tiff so. That brings the size down to 9.1G