Scans the chrome service worker caches and prints url and size for each directory

sw_cache

#!/usr/bin/env python3

# Scans the chrome service worker caches and prints url  and size for each directory
# Add -v to the command line for more details

import struct, os, functools, sys

def uvarint(data,pos):
    x = s = 0
    while True:
        b = data[pos]
        pos += 1
        x = x | ((b&0x7f)<<s)
        if b < 0x80: return x,pos
        s += 7

def readbytes(data,pos):
    l,pos = uvarint(data,pos)
    return data[pos:pos+l], pos+l

def readstruct(fmt,l):
    return lambda data,pos: (struct.unpack_from(fmt,data,pos)[0],pos+l)

readers = [ uvarint, readstruct('<d',8), readbytes, None, None, readstruct('<f',4) ]

def parse(data, schema):
    "parses a protobuf"
    obj = {}
    pos = 0
    while pos < len(data):
        val,pos = uvarint(data,pos)
        typ = val & 7
        key = val >> 3
        val, pos = readers[typ](data,pos)
        if key not in schema: 
            continue
        name, repeated, typ = schema[key]
        if isinstance(typ, dict):
            val = parse(val, typ)
        if typ == 'string':
            val = val.decode('utf8')
        if repeated:
            val = obj.get(name,[]) + [val]
        obj[name] = val
    return obj

dname = os.path.expanduser('~/Library/Application Support/Google/Chrome/Default/Service Worker/CacheStorage')
schema = {
    1: ["workers",1,{
        1: ["name",0,"string"],
        2: ["path", 0,"string"],
        3: ["size",0,0],
        5: [".5",0,0],
        6: [".6",0,0],
    }],
    2: ["urla", 0, "string"],
    3: ["urlb", 0, "string"],
    4: [".4",0,0],
    5: [".5",0,0],
}

records = []
for entry in os.scandir(dname):
    fn = os.path.join(dname, entry, 'index.txt')
    data = open(fn,'rb').read()
    obj = parse(data,schema)
    obj['name'] = entry.name
    obj['size'] = sum(a['size'] for a in obj.get('workers',[]))
    records.append(obj)

records.sort(key=lambda x: x['size'])
verbose = '-v' in sys.argv

for obj in records:
    total = obj['size']
    print(obj['name'], obj['urla'], total/1024./1024.)
    if not verbose: continue
    for w in obj.get('workers',[]):
        size = w['size']
        print('-',w['name'], w['path'],w['size'])
    print('total',total/1024./1024., 'MB')
    print('---')

# Print directory so I can easily cd to it
print(dname)