Restore MRO calls to inherited constructors

Sergey Vasilyev · Sergey Vasilyev · commit 46e87b7a4836 · 2023-10-02T20:32:00.000+02:00
Always, always, always call the inherited constructor — even if it does nothing. If not called, it breaks the MRO chain of inheritance and complicates the mixin/class building.

This must be a linting rule.
diff --git a/data_diff/databases/_connect.py b/data_diff/databases/_connect.py
@@ -98,6 +98,7 @@ class Connect:
     conn_cache: MutableMapping[Hashable, Database]
 
     def __init__(self, database_by_scheme: Dict[str, Database] = DATABASE_BY_SCHEME):
+        super().__init__()
         self.database_by_scheme = database_by_scheme
         self.match_uri_path = {name: MatchUriPath(cls) for name, cls in database_by_scheme.items()}
         self.conn_cache = weakref.WeakValueDictionary()
diff --git a/data_diff/databases/base.py b/data_diff/databases/base.py
@@ -180,6 +180,7 @@ class ThreadLocalInterpreter:
     """
 
     def __init__(self, compiler: Compiler, gen: Generator):
+        super().__init__()
         self.gen = gen
         self.compiler = compiler
 
@@ -1109,6 +1110,7 @@ class ThreadedDatabase(Database):
     """
 
     def __init__(self, thread_count=1):
+        super().__init__()
         self._init_error = None
         self._queue = ThreadPoolExecutor(thread_count, initializer=self.set_conn)
         self.thread_local = threading.local()
diff --git a/data_diff/databases/bigquery.py b/data_diff/databases/bigquery.py
@@ -224,6 +224,7 @@ class BigQuery(Database):
     dialect = Dialect()
 
     def __init__(self, project, *, dataset, bigquery_credentials=None, **kw):
+        super().__init__()
         credentials = bigquery_credentials
         bigquery = import_bigquery()
 
diff --git a/data_diff/databases/databricks.py b/data_diff/databases/databricks.py
@@ -104,12 +104,12 @@ class Databricks(ThreadedDatabase):
     CONNECT_URI_PARAMS = ["catalog", "schema"]
 
     def __init__(self, *, thread_count, **kw):
+        super().__init__(thread_count=thread_count)
         logging.getLogger("databricks.sql").setLevel(logging.WARNING)
 
         self._args = kw
         self.default_schema = kw.get("schema", "default")
         self.catalog = self._args.get("catalog", "hive_metastore")
-        super().__init__(thread_count=thread_count)
 
     def create_connection(self):
         databricks = import_databricks()
diff --git a/data_diff/databases/duckdb.py b/data_diff/databases/duckdb.py
@@ -139,6 +139,7 @@ class DuckDB(Database):
     CONNECT_URI_PARAMS = ["database", "dbpath"]
 
     def __init__(self, **kw):
+        super().__init__()
         self._args = kw
         self._conn = self.create_connection()
 
diff --git a/data_diff/databases/mysql.py b/data_diff/databases/mysql.py
@@ -138,9 +138,8 @@ class MySQL(ThreadedDatabase):
     CONNECT_URI_PARAMS = ["database?"]
 
     def __init__(self, *, thread_count, **kw):
-        self._args = kw
-
         super().__init__(thread_count=thread_count)
+        self._args = kw
 
         # In MySQL schema and database are synonymous
         try:
diff --git a/data_diff/databases/oracle.py b/data_diff/databases/oracle.py
@@ -182,12 +182,10 @@ class Oracle(ThreadedDatabase):
     CONNECT_URI_PARAMS = ["database?"]
 
     def __init__(self, *, host, database, thread_count, **kw):
+        super().__init__(thread_count=thread_count)
         self.kwargs = dict(dsn=f"{host}/{database}" if database else host, **kw)
-
         self.default_schema = kw.get("user").upper()
 
-        super().__init__(thread_count=thread_count)
-
     def create_connection(self):
         self._oracle = import_oracle()
         try:
diff --git a/data_diff/databases/postgresql.py b/data_diff/databases/postgresql.py
@@ -128,9 +128,8 @@ class PostgreSQL(ThreadedDatabase):
     default_schema = "public"
 
     def __init__(self, *, thread_count, **kw):
-        self._args = kw
-
         super().__init__(thread_count=thread_count)
+        self._args = kw
 
     def create_connection(self):
         if not self._args:
diff --git a/data_diff/databases/presto.py b/data_diff/databases/presto.py
@@ -161,6 +161,7 @@ class Presto(Database):
     default_schema = "public"
 
     def __init__(self, **kw):
+        super().__init__()
         prestodb = import_presto()
 
         if kw.get("schema"):
diff --git a/data_diff/databases/snowflake.py b/data_diff/databases/snowflake.py
@@ -155,6 +155,7 @@ class Snowflake(Database):
     CONNECT_URI_KWPARAMS = ["warehouse"]
 
     def __init__(self, *, schema: str, **kw):
+        super().__init__()
         snowflake, serialization, default_backend = import_snowflake()
         logging.getLogger("snowflake.connector").setLevel(logging.WARNING)
 
diff --git a/data_diff/databases/trino.py b/data_diff/databases/trino.py
@@ -40,6 +40,7 @@ class Trino(presto.Presto):
     CONNECT_URI_PARAMS = ["catalog", "schema"]
 
     def __init__(self, **kw):
+        super().__init__()
         trino = import_trino()
 
         if kw.get("schema"):
diff --git a/data_diff/databases/vertica.py b/data_diff/databases/vertica.py
@@ -159,11 +159,10 @@ class Vertica(ThreadedDatabase):
     default_schema = "public"
 
     def __init__(self, *, thread_count, **kw):
+        super().__init__(thread_count=thread_count)
         self._args = kw
         self._args["AUTOCOMMIT"] = False
 
-        super().__init__(thread_count=thread_count)
-
     def create_connection(self):
         vertica = import_vertica()
         try:
diff --git a/data_diff/dbt_parser.py b/data_diff/dbt_parser.py
@@ -101,6 +101,8 @@ def __init__(
         project_dir_override: Optional[str] = None,
         state: Optional[str] = None,
     ) -> None:
+        super().__init__()
+
         try_set_dbt_flags()
         self.dbt_runner = try_get_dbt_runner()
         self.project_dir = Path(project_dir_override or default_project_dir())
diff --git a/data_diff/lexicographic_space.py b/data_diff/lexicographic_space.py
@@ -63,6 +63,7 @@ class LexicographicSpace:
     """
 
     def __init__(self, dims: Vector):
+        super().__init__()
         self.dims = dims
 
     def __contains__(self, v: Vector):
@@ -120,6 +121,8 @@ class BoundedLexicographicSpace:
     """
 
     def __init__(self, min_bound: Vector, max_bound: Vector):
+        super().__init__()
+
         dims = tuple(mx - mn for mn, mx in safezip(min_bound, max_bound))
         if not all(d >= 0 for d in dims):
             raise ValueError("Error: Negative dimension!")
diff --git a/data_diff/thread_utils.py b/data_diff/thread_utils.py
@@ -46,6 +46,7 @@ class ThreadedYielder(Iterable):
     """
 
     def __init__(self, max_workers: Optional[int] = None):
+        super().__init__()
         self._pool = PriorityThreadPoolExecutor(max_workers)
         self._futures = deque()
         self._yield = deque()
diff --git a/data_diff/utils.py b/data_diff/utils.py
@@ -73,6 +73,7 @@ def new(self, initial=()) -> Self:
 
 class CaseInsensitiveDict(CaseAwareMapping):
     def __init__(self, initial):
+        super().__init__()
         self._dict = {k.lower(): (k, v) for k, v in dict(initial).items()}
 
     def __getitem__(self, key: str) -> V:
@@ -175,6 +176,8 @@ def alphanums_to_numbers(s1: str, s2: str):
 
 class ArithAlphanumeric(ArithString):
     def __init__(self, s: str, max_len=None):
+        super().__init__()
+
         if s is None:
             raise ValueError("Alphanum string cannot be None")
         if max_len and len(s) > max_len:
diff --git a/tests/test_database_types.py b/tests/test_database_types.py
@@ -367,6 +367,7 @@ class PaginatedTable:
     RECORDS_PER_BATCH = 1000000
 
     def __init__(self, table_path, conn):
+        super().__init__()
         self.table_path = table_path
         self.conn = conn
 
@@ -398,6 +399,7 @@ class DateTimeFaker:
     ]
 
     def __init__(self, max):
+        super().__init__()
         self.max = max
 
     def __iter__(self):
@@ -413,6 +415,7 @@ class IntFaker:
     MANUAL_FAKES = [127, -3, -9, 37, 15, 0]
 
     def __init__(self, max):
+        super().__init__()
         self.max = max
 
     def __iter__(self):
@@ -428,6 +431,7 @@ class BooleanFaker:
     MANUAL_FAKES = [False, True, True, False]
 
     def __init__(self, max):
+        super().__init__()
         self.max = max
 
     def __iter__(self):
@@ -458,6 +462,7 @@ class FloatFaker:
     ]
 
     def __init__(self, max):
+        super().__init__()
         self.max = max
 
     def __iter__(self):
@@ -471,6 +476,7 @@ def __len__(self):
 
 class UUID_Faker:
     def __init__(self, max):
+        super().__init__()
         self.max = max
 
     def __len__(self):
@@ -486,6 +492,7 @@ class JsonFaker:
     ]
 
     def __init__(self, max):
+        super().__init__()
         self.max = max
 
     def __iter__(self):
