From 46eaf4ddfb32d167fbc59855fa2d769e6477bea6 Mon Sep 17 00:00:00 2001
From: Michael Weiser <michael.weiser@gmx.de>
Date: Thu, 17 Mar 2022 08:30:19 +0000
Subject: [PATCH] expressions: Ignore non-iterability of None in membership
 tests

Python makes a distinction between something being contained in
something else (or not) and something not being able to be contained in
nothing (i.e. None) and raises a 'NoneType not iterable' exception in
the latter case. We can see no application of this distinction in our
use-case as of now. In fact it causes problems when writing tests
against object properties which are None. Therefore we handle that case
in our containment tests and interpret it as something (obviously) not
being contained in nothing.

Closes #176.
---
 peekaboo/ruleset/expressions.py | 15 +++++++++------
 tests/test.py                   | 11 +++++++++++
 2 files changed, 20 insertions(+), 6 deletions(-)

diff --git a/peekaboo/ruleset/expressions.py b/peekaboo/ruleset/expressions.py
index 329ce7bc..e755084c 100644
--- a/peekaboo/ruleset/expressions.py
+++ b/peekaboo/ruleset/expressions.py
@@ -448,16 +448,19 @@ def __init__(self, tokens):
         }
 
     @staticmethod
-    def in_(a, b):
+    def in_(op1, op2):
         """ Literally implement membership test. Make it a static method so we
         can do identity checks. Do not use operator.contains because it needs
-        operands swapped. """
-        return a in b
+        operands swapped. Also, there's no foreseeable use of the distinction
+        that something cannot be present in nothing in our application. So
+        prevent 'NoneType not iterable' exceptions by checking that op2 is not
+        None. """
+        return op2 is not None and op1 in op2
 
     @staticmethod
-    def not_in(a, b):
-        """ Naively implement non-membership test. """
-        return a not in b
+    def not_in(op1, op2):
+        """ Implement non-membership test. """
+        return op2 is not None and op1 not in op2
 
     @staticmethod
     def and_(op1, op2):
diff --git a/tests/test.py b/tests/test.py
index b0e982ca..45cfb223 100755
--- a/tests/test.py
+++ b/tests/test.py
@@ -1297,6 +1297,7 @@ def test_rule_expression_filetools(self):
                 == "AppleDouble encoded Macintosh file" -> ignore
             expression.1  : sample.file_extension in {"doc", "docx"}
                 and filereport.type_by_content != /application\/.*word/ -> bad
+            expression.2: "text" in filereport.type_by_name -> ignore
         '''
 
         sample_kwargs = {
@@ -1324,6 +1325,13 @@ def test_rule_expression_filetools(self):
         result = rule.evaluate(sample)
         self.assertEqual(result.result, Result.ignored)
 
+        # expression.2 should not raise a 'NoneType not iterable' exception due
+        # to type_by_name being None (and it should not match). This happens if
+        # the file name is empty, for example.
+        sample = Sample(b'dummy')
+        result = rule.evaluate(sample)
+        self.assertEqual(result.result, Result.unknown)
+
     @asynctest
     async def test_rule_expression_knowntools(self):
         """ Test generic rule on knowntoolsreport. """
@@ -1944,6 +1952,9 @@ def test_basic_expressions(self):
             ["'foo' == 'bar'", False],
             ["'foo' == 'foo'", True],
             ["'foo' in 'bar'", False],
+            # we swallow None being non-iterable
+            ["'foo' in None", False],
+            ["'foo' not in None", False],
             # re.search() for "match anywhere in operand"
             ["'foo' in 'foobar'", True],
             ["/foo/ in 'afoobar'", True],