Developement

README.md

@@ -35,14 +35,13 @@ For the former, run the anticheat.py file with some or all of the following flag
 | -h, --help | displays the messages below |
 | -m, --map | checks the leaderboard on the given beatmap id against each other |
 | -u, --user | checks only the given user against the other leaderboard replays. Must be set with -m |
-| -l, --local | compare scores under the user/ directory to a beatmap leaderboard (if set with just -m), a score set by a user on a beatmap (if set with -m and -u) or other locally saved replays (default behavior) |
+| -l, --local | compare scores under the replays/ directory to a beatmap leaderboard (if set with -m), a score set by a user on a beatmap (if set with -m and -u) or the other scores in the folder (default behavior) |
 | -t, --threshold | sets the similarity threshold to print comparisons that score under it. Defaults to 20 |
 | -a, --auto-threshold | sets the number of standard deviations from the average similarity the threshold will automatically be set to. Overrides -t  **Note: If more than ![formula](https://latex.codecogs.com/gif.latex?\frac{1}{2}&space;-&space;\frac{1}{2}&space;\mathbf{erf}\frac{a}{\sqrt{2}}) of the input is stolen this may cause false negatives** |
 | -n, --number | how many replays to get from a beatmap. No effect if not set with -m. Defaults to 50. **Note: the time complexity of the comparisons scales with O(n^2)** |
 | -c, --cache | if set, locally caches replays so they don't have to be redownloaded when checking the same map multiple times |
-| --single | compare all replays under user/ with all other replays under user/. No effect if not set with -l |
 | -s, --silent | if set, you will not be prompted for a visualization of comparisons under the threshold. Results will still be printed |
-
+| -v, --verify | Takes 3 positional arguments - map id, user1 id and user2 id. Verifies that the scores are steals of each other |
 
 #### Some Examples
 

STYLE.md

@@ -29,7 +29,7 @@ def method(arg1, arg2):
         [arg type] [arg name]: [description of arg]
 
     Returns:
-        [Description of return value] 
+        [Description of return value]
 
     Raises:
         [error name]: [description of when this error is raised]
@@ -56,8 +56,8 @@ Classes documentation is a little different. Classes follow all the same guideli
 ```python
 class Comparer:
     """
-    A class for managing a set of replay comparisons. 
-    
+    A class for managing a set of replay comparisons.
+
     Attributes:
         List replays1: A list of Replay instances to compare against replays2.
         List replays2: A list of Replay instances to be compared against.
@@ -116,4 +116,4 @@ Follow general git conventions when committing:
 
 Pull Requests with messy history may be converted to a squash merge at a mantainer's discretion.
 
-Finally, thou shalt not commit directly to master.
+Finally, thou shalt not commit directly to master.

osu-ac/anticheat.py

@@ -18,7 +18,7 @@
 from comparer import Comparer
 from investigator import Investigator
 from cacher import Cacher
-from config import PATH_REPLAYS_USER, PATH_REPLAYS_CHECK, WHITELIST
+from config import PATH_REPLAYS, WHITELIST, VERSION
 
 class Anticheat:
 
@@ -36,27 +36,42 @@ def __init__(self, args):
         if(args.map_id):
             self.users_info = Loader.users_info(args.map_id, args.number)
         if(args.user_id and args.map_id):
-            user_info = Loader.user_info(args.map_id, args.user_id)
-            self.replays_check = [OnlineReplay.from_map(self.cacher, args.map_id, args.user_id, user_info[args.user_id][0], user_info[args.user_id][1])]
+            user_info = Loader.user_info(args.map_id, args.user_id)[args.user_id] # should be guaranteed to only be a single mapping of user_id to a list
+            self.replays_check = [OnlineReplay.from_map(self.cacher, args.map_id, args.user_id, user_info[0], user_info[1], user_info[2])]
 
     def run(self):
         """
         Starts loading and detecting replays based on the args passed through the command line.
         """
-
-        if(self.args.local):
+        if(self.args.verify):
+            self._run_verify()
+        elif(self.args.local):
             self._run_local()
         elif(self.args.map_id):
             self._run_map()
         else:
-            print("Please set either --local (-l) or --map (-m)! ")
-            sys.exit(1)
+            print("Please set either --local (-l), --map (-m), or --verify (-v)! ")
+
+    def _run_verify(self):
+        args = self.args
+
+        map_id = self.args.verify[0]
+        user1_id = self.args.verify[1]
+        user2_id = self.args.verify[2]
+
+        user1_info = Loader.user_info(map_id, user1_id)
+        user2_info = Loader.user_info(map_id, user2_id)
+        replay1 = OnlineReplay.from_user_info(self.cacher, map_id, user1_info)
+        replay2 = OnlineReplay.from_user_info(self.cacher, map_id, user2_info)
+
+        comparer = Comparer(args.threshold, args.silent, replay1, replays2=replay2, stddevs=args.stddevs)
+        comparer.compare(mode="double")
 
     def _run_local(self):
 
         args = self.args
-        # get all local user replays (used in every --local case)
-        replays1 = [LocalReplay.from_path(osr_path) for osr_path in PATH_REPLAYS_USER]
+        # get all local replays (used in every --local case)
+        replays1 = [LocalReplay.from_path(osr_path) for osr_path in PATH_REPLAYS]
 
         threshold = args.threshold
         stddevs = args.stddevs
@@ -72,18 +87,9 @@ def _run_local(self):
             comparer = Comparer(threshold, args.silent, replays1, replays2=replays2, stddevs=stddevs)
             comparer.compare(mode="double")
             return
-
-        if(args.single):
-            # checks every replay listed in PATH_REPLAYS_USER against every other replay there
-            comparer = Comparer(threshold, stddevs, args.silent, replays1)
-            comparer.compare(mode="single")
-            return
         else:
-            # checks every replay listed in PATH_REPLAYS_USER against every replay listed in PATH_REPLAYS_CHECK
-            replays2 = [LocalReplay.from_path(osr_path) for osr_path in PATH_REPLAYS_CHECK]
-            comparer = Comparer(threshold, args.silent, replays1, replays2=replays2, stddevs=stddevs)
-            comparer.compare(mode="double")
-            return
+            comparer = Comparer(threshold, args.silent, replays1, stddevs=stddevs)
+            comparer.compare(mode="single")
 
     def _run_map(self):
 
@@ -109,5 +115,9 @@ def _run_map(self):
             return
 
 if __name__ == '__main__':
-    anticheat = Anticheat(argparser.parse_args())
+    args = argparser.parse_args()
+    if(args.version):
+        print("osu!anticheat {}".format(VERSION))
+        sys.exit(0)
+    anticheat = Anticheat(args)
     anticheat.run()

osu-ac/argparser.py

@@ -7,22 +7,23 @@
 argparser.add_argument("-u", "--user", dest="user_id",
                     help="checks only the given user against the other leaderboard replays. Must be set with -m")
 
-argparser.add_argument("-l", "--local", help=("compare scores under the user/ directory to a beatmap leaderboard (if set with -m), "
-                                             "a score set by a user on a beatmap (if set with -m and -u) or other locally "
-                                            "saved replays (default behavior)"), action="store_true")
+argparser.add_argument("-l", "--local", help=("compare scores under the replays/ directory to a beatmap leaderboard (if set with -m), "
+                                             "a score set by a user on a beatmap (if set with -m and -u) or the other scores in the folder "
+                                            "(default behavior)"), action="store_true")
 
-argparser.add_argument("-t", "--threshold", help="sets the similarity threshold to print results that score under it. Defaults to 20", type=int, default=20)
+argparser.add_argument("-t", "--threshold", help="sets the similarity threshold to print results that score under it. Defaults to 20", type=int, default=18)
 
 argparser.add_argument("-a", "--auto", help="Sets the threshold to a number of standard deviations below the average similarity", type=float, dest="stddevs")
 
-argparser.add_argument("-n", "--number", help="how many replays to get from a beatmap. No effect if not set with -m. Must be between 1 and 100 inclusive,"
+argparser.add_argument("-n", "--number", help="how many replays to get from a beatmap. No effect if not set with -m. Must be between 2 and 100 inclusive,"
                                               "defaults to 50. NOTE: THE TIME COMPLEXITY OF THE COMPARISONS WILL SCALE WITH O(n^2).", type=int, default=50)
 
 argparser.add_argument("-c", "--cache", help="If set, locally caches replays so they don't have to be redownloaded when checking the same map multiple times.",
                                         action="store_true")
 
-argparser.add_argument("--single", help="Compare all replays under user/ with all other replays under user/. No effect if not set with -l",
-                                        action="store_true")
-
 argparser.add_argument("-s", "--silent", help="If set, you will not be prompted for a visualization of comparisons under the threshold",
                                          action="store_true")
+
+argparser.add_argument("-v", "--verify", help="Takes 3 positional arguments - map id, user1 id and user2 id. Verifies that the scores are steals of each other", nargs=3)
+
+argparser.add_argument("--version", help="Prints the program version", action="store_true")

osu-ac/cacher.py

@@ -52,18 +52,22 @@ def cache(self, map_id, user_id, lzma_bytes, replay_id):
         else: # else just insert
             self.write("INSERT INTO replays VALUES(?, ?, ?, ?)", [map_id, user_id, compressed_bytes, replay_id])
 
-    def revalidate(self, map_id, user_to_replay):
+    def revalidate(self, map_id, user_info):
         """
         Re-caches a stored replay if one of the given users has overwritten their score on the given map since it was cached.
 
         Args:
             String map_id: The map to revalidate.
-            Dictionary user_to_replay: The up tp date mapping of user_id to replay_id to revalidate.
+            Dictionary user_info: The up to date mapping of user_id to [username, replay_id, enabled_mods] to revalidate.
+                                       Only contains information for a single map.
         """
 
         result = self.cursor.execute("SELECT user_id, replay_id FROM replays WHERE map_id=?", [map_id]).fetchall()
+
+        # filter result to only contain entries also in user_info
+        result = [info for info in result if info[0] in user_info.keys()]
         for user_id, local_replay_id in result:
-            online_replay_id = user_to_replay[user_id]
+            online_replay_id = user_info[user_id][1]
             if(local_replay_id != online_replay_id): # local (outdated) id does not match online (updated) id
                 print("replay outdated, redownloading...", end="")
                 # this **could** conceivable be the source of a logic error by Loader.replay_data returning None and the cache storing None,
@@ -85,7 +89,11 @@ def check_cache(self, map_id, user_id):
         """
 
         result = self.cursor.execute("SELECT replay_data FROM replays WHERE map_id=? AND user_id=?", [map_id, user_id]).fetchone()
-        return wtc.decompress(result[0]) if result else None
+        if(result):
+            print("Loading replay by {} from cache".format(user_id))
+            return wtc.decompress(result[0])
+
+        return None
 
     def write(self, statement, args):
         """

osu-ac/comparer.py

@@ -5,7 +5,7 @@
 from draw import Draw
 from replay import Replay
 from config import WHITELIST
-
+from exceptions import InvalidArgumentsException
 class Comparer:
     """
     A class for managing a set of replay comparisons.
@@ -42,9 +42,11 @@ def __init__(self, threshold, silent, replays1, replays2=None, stddevs=None):
         self.stddevs = stddevs
         self.silent = silent
 
-        self.replays1 = replays1
-        self.replays2 = replays2
+        # filter beatmaps we had no data for - see Loader.replay_data and OnlineReplay.from_map
+        self.replays1 = [replay for replay in replays1 if replay is not None]
 
+        if(replays2):
+            self.replays2 = [replay for replay in replays2 if replay is not None]
 
     def compare(self, mode):
         """
@@ -56,17 +58,18 @@ def compare(self, mode):
             String mode: One of either "double" or "single", determining how to choose which replays to compare.
         """
 
+        if(not self.replays1): # if this is empty, bad things
+            print("No comparisons could be made. Make sure replay data is available for your args")
+            return
+
         if(mode == "double"):
-            print("comparing first set of replays to second set of replays")
             iterator = itertools.product(self.replays1, self.replays2)
         elif (mode == "single"):
-            print("comparing first set of replays to itself")
             iterator = itertools.combinations(self.replays1, 2)
         else:
-            raise Exception("`mode` must be one of 'double' or 'single'")
-
-
+            raise InvalidArgumentsException("`mode` must be one of 'double' or 'single'")
 
+        print("Starting to compare replays")
         # automatically determine threshold based on standard deviations of similarities if stddevs is set
         if(self.stddevs):
             results = {}
@@ -126,13 +129,13 @@ def _print_result(self, result, replay1, replay2):
         if(mean > self.threshold):
             return
 
-        # if they were both set online, we don't get dates from
-        first_score = None
+        # if they were both set locally, we don't get replay ids to compare
+        last_score = None
         if(replay1.replay_id and replay2.replay_id):
-            first_score = replay1.player_name if(replay1.replay_id < replay2.replay_id) else replay2.player_name
+            last_score = replay1.player_name if(replay1.replay_id > replay2.replay_id) else replay2.player_name
 
         print("{:.1f} similarity, {:.1f} std deviation ({} vs {}{})"
-              .format(mean, sigma, replay1.player_name, replay2.player_name, " - {} set first".format(first_score) if first_score else ""))
+              .format(mean, sigma, replay1.player_name, replay2.player_name, " - {} set later".format(last_score) if last_score else ""))
 
         if(self.silent):
             return

osu-ac/config.py

@@ -5,16 +5,10 @@
 from secret import API_KEY
 
 PATH_ROOT = pathlib.Path(__file__).parent
-PATH_REPLAYS = PATH_ROOT / "replays"
+PATH_REPLAYS_STUB = PATH_ROOT / "replays"
 
-# names of replays to check
-PATH_REPLAYS_USER_STUB = join(PATH_REPLAYS, "user")
-PATH_REPLAYS_CHECK_STUB = join(PATH_REPLAYS, "compare") # path of replays to check against
-
-PATH_REPLAYS_USER = [join(PATH_REPLAYS_USER_STUB, f) for f in os.listdir(PATH_REPLAYS_USER_STUB) if isfile(join(PATH_REPLAYS_USER_STUB, f)) and f != ".DS_Store"]
 # get all replays in path to check against
-PATH_REPLAYS_CHECK = [join(PATH_REPLAYS_CHECK_STUB, f) for f in os.listdir(PATH_REPLAYS_CHECK_STUB) if isfile(join(PATH_REPLAYS_CHECK_STUB, f)) and f != ".DS_Store"]
-
+PATH_REPLAYS = [join(PATH_REPLAYS_STUB, f) for f in os.listdir(PATH_REPLAYS_STUB) if isfile(join(PATH_REPLAYS_STUB, f)) and f != ".DS_Store"]
 
 API_BASE = "https://osu.ppy.sh/api/"
 API_REPLAY = API_BASE + "get_replay?k=" + API_KEY + "&m=0&b={}&u={}"
@@ -25,3 +19,5 @@
 WHITELIST = ["124493", "6304246", "2558286", "2562987", "2757689"]
 
 PATH_DB = PATH_ROOT / "db" / "cache.db" # /absolute/path/db/cache.db
+
+VERSION = "1.0d"

osu-ac/draw.py

@@ -68,6 +68,8 @@ def run(self):
         plot1 = plt.plot('x', 'y', "red", animated=True, label=self.replay1.player_name)[0]
         plot2 = plt.plot('', '', "blue", animated=True, label=self.replay2.player_name)[0]
 
+        fig.legend()
+
         def init():
             ax.set_xlim(0, 512)
             ax.set_ylim(0, 384)

osu-ac/enums.py

@@ -2,9 +2,10 @@
 
 # strings taken from osu api error responses
 class Error(Enum):
-    NO_REPLAY   = "Replay not available."
-    RATELIMITED = "Requesting too fast! Slow your operation, cap'n!"
-    UNKOWN = "Unkown error."
+    NO_REPLAY         = "Replay not available."
+    RATELIMITED       = "Requesting too fast! Slow your operation, cap'n!"
+    RETRIEVAL_FAILED  = "Replay retrieval failed."
+    UNKOWN            = "Unkown error."
 
 class Mod(Enum):
     NoMod          = 0

osu-ac/exceptions.py

@@ -0,0 +1,8 @@
+class AnticheatException(Exception):
+    """Base class for exceptions in the anticheat program."""
+
+class InvalidArgumentsException(AnticheatException):
+    """Indicates an invalid argument was passed to one of the flags."""
+
+class APIException(AnticheatException):
+    """Indicates some error on the API's end that we were not prepared to handle."""

osu-ac/gui.py

@@ -19,13 +19,12 @@ def run():
 
     _number = num.get()
     _cache = cache.get()
-    _single = single.get()
-    _silent = silent.get()
-
+    _silent = True # Visualizations do very very bad things when not called from the main thread, so when using gui, we just...force ignore them
+    _verify = verify.get()
 
     def run_anticheat():
         anticheat = Anticheat(SimpleNamespace(map_id=_map_id, user_id=_user_id, local=_local, threshold=_threshold, stddevs=_stddevs,
-                                              number=_number, cache=_cache, single=_single, silent=_silent))
+                                              number=_number, cache=_cache, silent=_silent, verify=_verify))
         anticheat.run()
 
     thread = threading.Thread(target=run_anticheat)
@@ -56,8 +55,7 @@ def run_anticheat():
 cache = tkinter.BooleanVar(value=False)
 
 # unimplemented
-single = tkinter.BooleanVar(value=False)
-silent = tkinter.BooleanVar(value=False)
+verify = tkinter.BooleanVar(value=False)
 
 # Make visual elements for main frame
 map_label = ttk.Label(main, text="Map id:")
@@ -98,7 +96,7 @@ def run_anticheat():
 top_plays_label1.grid(row=0, column=1)
 top_plays_entry = ttk.Entry(top_x_plays, width=4, textvariable=num)
 top_plays_entry.grid(row=0, column=2)
-top_plays_label2 = ttk.Label(top_x_plays, text="leaderboard plays?\n(Between 1 and 100 inclusive)")
+top_plays_label2 = ttk.Label(top_x_plays, text="leaderboard plays?\n(Between 2 and 100 inclusive)")
 top_plays_label2.grid(row=0, column=3)
 
 auto_threshold = ttk.Frame(options)