Python quote counting

counts.csv

naive-quote-counting.py

import json
import subprocess
import tarfile
import tempfile
from pathlib import Path
from zipfile import BadZipFile, ZipFile


def count_package(package):
    with tempfile.TemporaryDirectory() as tmpdir:
        try:
            subprocess.run(
                [
                    'pip', 'download', '--no-deps', '--dest', tmpdir,
                    '--no-binary', ':all:', package
                ],
                check=True,
                stdout=subprocess.PIPE,
                stderr=subprocess.PIPE)
        except:
            # not checking all of these individually
            return (None, None)

        archive = next(Path(tmpdir).iterdir())

        try:
            with ZipFile(str(archive), 'r') as f:
                pyfiles = [
                    path for path in f.namelist() if path.endswith('.py')
                ]
                pyfiles = list(
                    map(lambda x: f.open(x).read().decode('utf-8', 'ignore'),
                        pyfiles))
        except BadZipFile:
            # probably a tarball then
            with tarfile.open(str(archive), 'r:gz') as f:
                pyfiles = [p for p in f.getmembers() if p.name.endswith('.py')]
                pyfiles = list(map(lambda x: f.extractfile(x).read().decode('utf-8', 'ignore'), pyfiles))

        counts = list(map(lambda x: (x.count('\''), x.count('"')), pyfiles))
        single = sum([pair[0] for pair in counts])
        double = sum([pair[1] for pair in counts])

    return single, double
  
  
if __name__ == '__main__':

    # Sourced here:
    # https://github.com/hugovk/top-pypi-packages
    with open('top-pypi-packages-365-days.json') as f:
        top = json.load(f)

    packages = [p['project'] for p in top['rows'][:100]]

    for p in packages:
        counts = count_package(p)
        print('{p},{single},{double}'.format(
            p=p, single=counts[0], double=counts[1]))

Density plot of single / (single + double).

    ++----------+----------+----------+-----------+----------+-+
    |                                 **********               |
    |                               **         **              |
    |                             ***            **            |
1.5 +                            **               **           +
    |                           **                 **          |
    |                          **                   *          |
    |                         **                     *         |
  1 +                        **                      **        +
    |                       **                        **       |
    |                     ***                          **      |
    |     ******         **                             **     |
0.5 +  ****    ***** *****                               ***   +
    |  *           ***                                     **  |
    ++----------+----------+----------+-----------+----------+-+
     0         0.2        0.4        0.6         0.8         1