ccl_bplist is a small Python module for dealing with binary property lists (bplists). It also contains special helper functionality to simplify the processing of NSKeyedArchiver structures often found in binary property lists.
It converts the binary property list file contents into a native Python object.
| Property List Type | Returned Python Type |
|---|---|
| null | None |
| true | True |
| false | False |
| integer | int |
| real | float |
| uid | int |
| date | datetime.datetime |
| string | str |
| data | bytes |
| array | list |
| dict | dict |
| set | list |
Here's a quick overview of the module's features:
At its most basic, just import the module, create a file-like object around your plist (here I'm opening a file, but often I find that I'm dealing with embedded plists, in which case io.BytesIO is very useful).
>>> import ccl_bplist
>>> f = open("plist.plist", "rb")
>>> plist = ccl_bplist.load(f)
>>>
>>> plist
{'$objects': ['$null', {'apiKey': UID: 10, 'eventLog': UID: 7, 'locale': UID: 12, '$class': UID: 14, 'pauseTime': UID: 4, 'totalErrorCount': 0, 'internetAvailable': True, 'errors': UID: 9, 'userID': UID: 0, 'appVersion': UID: 11, 'latitude': 0.0, 'eventCounts': UID: 5, 'eventLogComplete': True, 'crashed': False, 'pageViewCount': 0, 'startTime': UID: 2, 'pauseIntervalMillis': 0, 'gender': -1, 'longitude': 0.0, 'serializationVersion': 1, 'timeZone': UID: 13, 'accuracy': 0.0}, {'$class': UID: 3, 'NS.time': 378032076.116256}, {'$classes': ['NSDate', 'NSObject'], '$classname': 'NSDate'}, {'$class': UID: 3, 'NS.time': 378032077.674404}, {'$class': UID: 6, 'NS.keys': [], 'NS.objects': []}, {'$classes': ['NSMutableDictionary', 'NSDictionary', 'NSObject'], '$classname': 'NSMutableDictionary'}, {'$class': UID: 8, 'NS.objects': []}, {'$classes': ['NSMutableArray', 'NSArray', 'NSObject'], '$classname': 'NSMutableArray'}, {'$class': UID: 8, 'NS.objects': []}, 'BMPNB5HT2H63KM42BB83', '4.0.1', 'en_GB', 'Europe/London', {'$classes': ['FlurrySession', 'NSObject'], '$classname': 'FlurrySession'}], '$top': {'root': UID: 1}, '$version': 100000, '$archiver': 'NSKeyedArchiver'}The plist is converted to a native Python object (the translations of plist data-type to Python data-types are listed above).
In this case the plist is one that has been created by Apple's NSKeyedArchiver class which means it's not a whole lot of fun to work with. Luckily, ccl_bplist can deserialise the object and give you the actual structure of the data.
>>> ns_keyed_archiver_obj = ccl_bplist.deserialise_NsKeyedArchiver(plist)
>>> ns_keyed_archiver_obj
{'apiKey': UID: 10, 'eventLog': UID: 7, 'locale': UID: 12, '$class': UID: 14, 'eventLogComplete': True, 'crashed': False, 'pageViewCount': 0, 'startTime': UID: 2, 'pauseTime': UID: 4, 'totalErrorCount': 0, 'internetAvailable': True, 'errors': UID: 9, 'pauseIntervalMillis': 0, 'gender': -1, 'userID': UID: 0, 'appVersion': UID: 11, 'longitude': 0.0, 'serializationVersion': 1, 'latitude': 0.0, 'timeZone': UID: 13, 'accuracy': 0.0, 'eventCounts': UID: 5}The module assumes that if the top-level dictionary structure is $top/root that the object that you want returning sits beneath top. You can override this behaviour is you prefer:
>>> ns_keyed_archiver_obj = ccl_bplist.deserialise_NsKeyedArchiver(plist, parse_whole_structure=True)
>>> ns_keyed_archiver_obj
{'root': UID: 1}You'll notice in both of these cases that we have UID objects. This is because NSKeyedArchiver maintains an object-table, and in the structure the objects are referred by their position in the object-table (more details here: http://www.cclgroupltd.com/digital-forensics/rd/geek-post-nskeyedarchiver-files-%E2%80%93-what-are-they-and-how-can-i-use-them.html. The module doesn't look-up the objects until they are requested, both for speed and (mostly) because you can have infinitely recursing structures. If you access a UID object in a list or dictionary it will be looked up and returned:
>>> ns_keyed_archiver_obj["root"]
{'apiKey': UID: 10, 'eventLog': UID: 7, 'locale': UID: 12, '$class': UID: 14, 'eventLogComplete': True, 'crashed': False, 'pageViewCount': 0, 'startTime': UID: 2, 'pauseTime': UID: 4, 'totalErrorCount': 0, 'internetAvailable': True, 'errors': UID: 9, 'pauseIntervalMillis': 0, 'gender': -1, 'userID': UID: 0, 'appVersion': UID: 11, 'longitude': 0.0, 'serializationVersion': 1, 'latitude': 0.0, 'timeZone': UID: 13, 'accuracy': 0.0, 'eventCounts': UID: 5}NSKeyed Archiver contains serialisations of ObjectiveC objects which would usually require extra code to deal with, but in version 0.13 of the module we've added a set of convenience features to make dealing with these data types a lot simpler.
>>> ns_keyed_archiver_obj["root"]["eventLog"]
{'$class': UID: 8, 'NS.objects': []}Here we've accessed an object in this plist, we can inspect the type of the object by looking at the "$class" key:
>>> ns_keyed_archiver_obj["root"]["eventLog"]["$class"]
{'$classes': ['NSMutableArray', 'NSArray', 'NSObject'], '$classname': 'NSMutableArray'}We can see that this is a serialisation of an NSMutableArray. We could access the contents of the array in the NS.objects key (which in this case contains an empty list), alternatively we can set a converter function for the module to use, you can write your own, but the module comes with one built in which you can enable like this:
>>> ccl_bplist.set_object_converter(ccl_bplist.NSKeyedArchiver_common_objects_convertor)(NB this enables the converter module-wide, not just for objects that you're currently working on)
If we access the same object now, it automatically gets converted into a nice, native Python list:
>>> ns_keyed_archiver_obj["root"]["eventLog"]
[]If you want to turn this functionality off, just set the converter function to a pass-through function:
>>> ccl_bplist.set_object_converter(lambda x: x)This built-in converter also works for NSDictionary:
>>> ns_keyed_archiver_obj["root"]["eventCounts"]
{'$class': UID: 6, 'NS.keys': [], 'NS.objects': []}
>>> ns_keyed_archiver_obj["root"]["eventCounts"]["$class"]
{'$classes': ['NSMutableDictionary', 'NSDictionary', 'NSObject'], '$classname': 'NSMutableDictionary'}
>>> ccl_bplist.set_object_converter(ccl_bplist.NSKeyedArchiver_common_objects_convertor)
>>> ns_keyed_archiver_obj["root"]["eventCounts"]
{}
>>> ccl_bplist.set_object_converter(lambda x: x)NSDate:
>>> ns_keyed_archiver_obj["root"]["startTime"]
{'$class': UID: 3, 'NS.time': 378032076.116256}
>>> ns_keyed_archiver_obj["root"]["startTime"]["$class"]
{'$classes': ['NSDate', 'NSObject'], '$classname': 'NSDate'}
>>> ccl_bplist.set_object_converter(ccl_bplist.NSKeyedArchiver_common_objects_convertor)
>>> ns_keyed_archiver_obj["root"]["startTime"]
datetime.datetime(2012, 12, 24, 8, 54, 36, 116256)
>>> ccl_bplist.set_object_converter(lambda x: x)and the way that NSKeyedArchiver serialises Null as a string containing "$null":
>>> ns_keyed_archiver_obj["root"]["userID"]
'$null'
>>> ccl_bplist.set_object_converter(ccl_bplist.NSKeyedArchiver_common_objects_convertor)
>>> print(ns_keyed_archiver_obj["root"]["userID"])
None