dropbox-large-files.py

import re
from dataclasses import dataclass

# Usage:
# dbxcli ls -lR > dropbox.txt
# python3 dropbox-large-files.py

with open('dropbox.txt') as fp:
    lines = fp.readlines()

for line in lines:
    if match := re.fullmatch(r'(((- +)- +)- +)/.*\n', line):
        end_offsets = [len(g) for g in reversed(match.groups())]
        offsets = list(zip([0] + end_offsets, end_offsets + [None]))
        break
else:
    raise Exception('Could not find offsets')

@dataclass
class Node:
    size: int
    file_count: int

def parse_size(size_str: str) -> int:
    if size_str.endswith(' GiB'):
        return int(float(size_str[:-4]) * 1024 * 1024 * 1024)
    elif size_str.endswith(' MiB'):
        return int(float(size_str[:-4]) * 1024 * 1024)
    elif size_str.endswith(' KiB'):
        return int(float(size_str[:-4]) * 1024)
    else:
        assert size_str.endswith(' B')
        return int(size_str[:-2])


tree = dict[str, Node]()
for line in lines:
    revision, size_str, updated, path = (line[x:y].strip() for x, y in offsets)

    if size_str == '-':
        continue

    size = parse_size(size_str)
    path_partial = ''
    for segment in path.split('/')[1:]:
        path_partial += '/' + segment

        if path_partial not in tree:
            tree[path_partial] = Node(0, 0)

        tree[path_partial].size += size
        tree[path_partial].file_count += 1

for path in sorted(tree, key=lambda x: tree[x].size, reverse=True):
    print(path, tree[path].size, tree[path].file_count, sep='\t')