Print known Uniform Type Identifiers as JSON

cuti.py

#!/usr/bin/env python
import json
import re
import subprocess
import sys


if len(sys.argv) > 1:
    lsregister = sys.argv[1]
else:
    lsregister = ('/System/Library/Frameworks/CoreServices.framework'
                  '/Versions/A/Frameworks/LaunchServices.framework'
                  '/Versions/A/Support/lsregister')

p = subprocess.Popen([lsregister, '-dump'], stdout=subprocess.PIPE,
                                            stderr=subprocess.PIPE)
out, err = p.communicate()
chunks = ['']
for line in out.splitlines():
    if line == '-' * 80:
        chunks.append('')
    else:
        chunks[-1] += line + '\n'

kvsplit = re.compile(r'^\s*(?P<key>.*?):\s*(?P<value>.*)$').search
def parse(chunk):
    entry = {}
    for line in chunk.splitlines():
        match = kvsplit(line)
        if match is not None:
            entry[match.group('key')] = match.group('value')
    return entry

entries = {}
csvsplit = re.compile(r'\s*,\s*').split
for entry in map(parse, chunks):
    if 'uti' in entry:
        content_types, extensions = [], []
        for tag in csvsplit(entry['tags']):
            if tag.startswith('.'):
                extensions.append(tag)
            elif '/' in tag:
                content_types.append(tag)
        entries[entry['uti']] = {
            'conforms_to': csvsplit(entry['conforms to']),
            'content_types': content_types,
            'description': entry['description'],
            'extensions': extensions,
        }
print json.dumps(entries)

$ ./cuti.py | python -mjson.tool | grep org.bittorrent.torrent --after-context=13
    "org.bittorrent.torrent": {
        "conforms_to": [
            "public.data", 
            "public.item", 
            "com.bittorrent.torrent"
        ], 
        "content_types": [
            "application/x-bittorrent"
        ], 
        "description": "BitTorrent Document", 
        "extensions": [
            ".torrent"
        ]
    },