Python: Handle get_collection on pymongo DB

Author: RasmusWL

python/ql/src/experimental/semmle/python/frameworks/NoSQL.qll

@@ -96,6 +96,12 @@ private module NoSql {
       )
       or
       result.(DataFlow::AttrRead).getObject() = mongoDBInstance()
+      or
+      // see https://pymongo.readthedocs.io/en/stable/api/pymongo/database.html#pymongo.database.Database.get_collection
+      // see https://pymongo.readthedocs.io/en/stable/api/pymongo/database.html#pymongo.database.Database.create_collection
+      result
+          .(DataFlow::MethodCallNode)
+          .calls(mongoDBInstance(), ["get_collection", "create_collection"])
     )
     or
     exists(DataFlow::TypeTracker t2 | result = mongoCollection(t2).track(t2, t))

python/ql/test/experimental/query-tests/Security/CWE-943/NoSQLInjection.expected

@@ -53,10 +53,10 @@ edges
 | pymongo_test.py:29:27:29:33 | ControlFlowNode for request | pymongo_test.py:29:27:29:38 | ControlFlowNode for Attribute |
 | pymongo_test.py:29:27:29:38 | ControlFlowNode for Attribute | pymongo_test.py:29:27:29:50 | ControlFlowNode for Subscript |
 | pymongo_test.py:29:27:29:50 | ControlFlowNode for Subscript | pymongo_test.py:29:16:29:51 | ControlFlowNode for Attribute() |
-| pymongo_test.py:38:16:38:51 | ControlFlowNode for Attribute() | pymongo_test.py:42:34:42:73 | ControlFlowNode for Dict |
-| pymongo_test.py:38:27:38:33 | ControlFlowNode for request | pymongo_test.py:38:27:38:38 | ControlFlowNode for Attribute |
-| pymongo_test.py:38:27:38:38 | ControlFlowNode for Attribute | pymongo_test.py:38:27:38:50 | ControlFlowNode for Subscript |
-| pymongo_test.py:38:27:38:50 | ControlFlowNode for Subscript | pymongo_test.py:38:16:38:51 | ControlFlowNode for Attribute() |
+| pymongo_test.py:39:16:39:51 | ControlFlowNode for Attribute() | pymongo_test.py:43:34:43:73 | ControlFlowNode for Dict |
+| pymongo_test.py:39:27:39:33 | ControlFlowNode for request | pymongo_test.py:39:27:39:38 | ControlFlowNode for Attribute |
+| pymongo_test.py:39:27:39:38 | ControlFlowNode for Attribute | pymongo_test.py:39:27:39:50 | ControlFlowNode for Subscript |
+| pymongo_test.py:39:27:39:50 | ControlFlowNode for Subscript | pymongo_test.py:39:16:39:51 | ControlFlowNode for Attribute() |
 nodes
 | flask_mongoengine_bad.py:19:21:19:27 | ControlFlowNode for request | semmle.label | ControlFlowNode for request |
 | flask_mongoengine_bad.py:19:21:19:32 | ControlFlowNode for Attribute | semmle.label | ControlFlowNode for Attribute |
@@ -123,11 +123,11 @@ nodes
 | pymongo_test.py:29:27:29:38 | ControlFlowNode for Attribute | semmle.label | ControlFlowNode for Attribute |
 | pymongo_test.py:29:27:29:50 | ControlFlowNode for Subscript | semmle.label | ControlFlowNode for Subscript |
 | pymongo_test.py:33:34:33:73 | ControlFlowNode for Dict | semmle.label | ControlFlowNode for Dict |
-| pymongo_test.py:38:16:38:51 | ControlFlowNode for Attribute() | semmle.label | ControlFlowNode for Attribute() |
-| pymongo_test.py:38:27:38:33 | ControlFlowNode for request | semmle.label | ControlFlowNode for request |
-| pymongo_test.py:38:27:38:38 | ControlFlowNode for Attribute | semmle.label | ControlFlowNode for Attribute |
-| pymongo_test.py:38:27:38:50 | ControlFlowNode for Subscript | semmle.label | ControlFlowNode for Subscript |
-| pymongo_test.py:42:34:42:73 | ControlFlowNode for Dict | semmle.label | ControlFlowNode for Dict |
+| pymongo_test.py:39:16:39:51 | ControlFlowNode for Attribute() | semmle.label | ControlFlowNode for Attribute() |
+| pymongo_test.py:39:27:39:33 | ControlFlowNode for request | semmle.label | ControlFlowNode for request |
+| pymongo_test.py:39:27:39:38 | ControlFlowNode for Attribute | semmle.label | ControlFlowNode for Attribute |
+| pymongo_test.py:39:27:39:50 | ControlFlowNode for Subscript | semmle.label | ControlFlowNode for Subscript |
+| pymongo_test.py:43:34:43:73 | ControlFlowNode for Dict | semmle.label | ControlFlowNode for Dict |
 subpaths
 #select
 | flask_mongoengine_bad.py:22:34:22:44 | ControlFlowNode for json_search | flask_mongoengine_bad.py:19:21:19:27 | ControlFlowNode for request | flask_mongoengine_bad.py:22:34:22:44 | ControlFlowNode for json_search | $@ NoSQL query contains an unsanitized $@ | flask_mongoengine_bad.py:22:34:22:44 | ControlFlowNode for json_search | This | flask_mongoengine_bad.py:19:21:19:27 | ControlFlowNode for request | user-provided value |
@@ -141,4 +141,4 @@ subpaths
 | mongoengine_bad.py:61:29:61:49 | ControlFlowNode for Dict | mongoengine_bad.py:57:21:57:27 | ControlFlowNode for request | mongoengine_bad.py:61:29:61:49 | ControlFlowNode for Dict | $@ NoSQL query contains an unsanitized $@ | mongoengine_bad.py:61:29:61:49 | ControlFlowNode for Dict | This | mongoengine_bad.py:57:21:57:27 | ControlFlowNode for request | user-provided value |
 | pymongo_test.py:15:42:15:62 | ControlFlowNode for Dict | pymongo_test.py:12:21:12:27 | ControlFlowNode for request | pymongo_test.py:15:42:15:62 | ControlFlowNode for Dict | $@ NoSQL query contains an unsanitized $@ | pymongo_test.py:15:42:15:62 | ControlFlowNode for Dict | This | pymongo_test.py:12:21:12:27 | ControlFlowNode for request | user-provided value |
 | pymongo_test.py:33:34:33:73 | ControlFlowNode for Dict | pymongo_test.py:29:27:29:33 | ControlFlowNode for request | pymongo_test.py:33:34:33:73 | ControlFlowNode for Dict | $@ NoSQL query contains an unsanitized $@ | pymongo_test.py:33:34:33:73 | ControlFlowNode for Dict | This | pymongo_test.py:29:27:29:33 | ControlFlowNode for request | user-provided value |
-| pymongo_test.py:42:34:42:73 | ControlFlowNode for Dict | pymongo_test.py:38:27:38:33 | ControlFlowNode for request | pymongo_test.py:42:34:42:73 | ControlFlowNode for Dict | $@ NoSQL query contains an unsanitized $@ | pymongo_test.py:42:34:42:73 | ControlFlowNode for Dict | This | pymongo_test.py:38:27:38:33 | ControlFlowNode for request | user-provided value |
+| pymongo_test.py:43:34:43:73 | ControlFlowNode for Dict | pymongo_test.py:39:27:39:33 | ControlFlowNode for request | pymongo_test.py:43:34:43:73 | ControlFlowNode for Dict | $@ NoSQL query contains an unsanitized $@ | pymongo_test.py:43:34:43:73 | ControlFlowNode for Dict | This | pymongo_test.py:39:27:39:33 | ControlFlowNode for request | user-provided value |

python/ql/test/experimental/query-tests/Security/CWE-943/pymongo_test.py

@@ -35,10 +35,11 @@ def bad2():
 
 @app.route("/bad3")
 def bad3():
+    # using `get_` methods instead of subscript/attribute lookups
     event_id = json.loads(request.args['event_id'])
     client = MongoClient("localhost", 27017, maxPoolSize=50)
     db = client.get_database(name="localhost")
-    collection = db['collection']
+    collection = db.get_collection("collection")
     cursor = collection.find_one({"$where": f"this._id == '${event_id}'"})