binary_tags.py

import json
import math
import sys
import os


infile = os.path.abspath(sys.argv[1])
outdir = os.path.abspath(sys.argv[2] if len(sys.argv) > 2 else '.')
factor = int(sys.argv[3]) if len(sys.argv) > 3 else 2

print('Reading ids from file:', infile)

data = json.load(open(infile))
ids = list(data.keys())

nids = len(ids)
nlayers = math.ceil(math.log(nids, factor))

print(f'Processing {len(ids)} ids, across {nlayers} layers with {factor} ids per layer...')

tags = [{'values': [item for sublist in [ids[sliceidx : sliceidx + factor ** layeridx] for sliceidx in range(0, nids, factor ** (layeridx + 1))] for item in sublist]} for layeridx in range(0, nlayers)]

print(f'Writing {len(tags)} tags to directory:', outdir)

for i, tag in enumerate(tags):
    json.dump(tag, open(os.path.join(outdir, str(i) + '.json'), 'w'), indent=2)