Remove True from Characteristic rules and output

Get rid of `True` in characteristic (rules, output and json) as it is
implicit. This way, the same syntax is used for characteristic as for
the rest of the features.

Co-authored-by: William Ballenthin <william.ballenthin@fireeye.com>

capa/features/__init__.py

@@ -66,26 +66,22 @@ class MatchedRule(Feature):
 
 
 class Characteristic(Feature):
-    def __init__(self, attribute, value=None, description=None):
+    def __init__(self, value, description=None):
-        '''
+        super(Characteristic, self).__init__([value], description)
-        when `value` is not provided, this serves as descriptor for a class of characteristics.
-        this is only used internally, such as in `rules.py` when checking if a statement is
-          supported by a given scope.
-        '''
-        super(Characteristic, self).__init__([attribute, value], description)
-        self.attribute = attribute
         self.value = value
 
-    def evaluate(self, ctx):
+    def freeze_serialize(self):
-        if self.value is None:
+        # in an older version of capa, characteristics could theoretically match non-existence (value=False).
-            raise ValueError('cannot evaluate characteristc %s with empty value' % (str(self)))
+        # but we found this was never used (and better expressed with `not: characteristic: ...`).
-        return super(Characteristic, self).evaluate(ctx)
+        # this was represented using an additional parameter for Characteristic.
+        # its been removed, but we keep it around in the freeze format to maintain backwards compatibility.
+        # this value is ignored, however.
+        return (self.__class__.__name__, [self.value, True])
 
-    def get_args_str(self):
+    @classmethod
-        if self.value is None:
+    def freeze_deserialize(cls, args):
-            return self.attribute
+        # see above. we ignore the second element in the 2-tuple here.
-        else:
+        return cls(args[0])
-            return '%s(%s)' % (self.attribute, self.value)
 
 
 class String(Feature):

capa/features/extractors/__init__.py

@@ -184,22 +184,22 @@ class NullFeatureExtractor(FeatureExtractor):
 
         extractor = NullFeatureExtractor({
             'file features': [
-                (0x402345, capa.features.Characteristic('embedded pe', True)),
+                (0x402345, capa.features.Characteristic('embedded pe')),
             ],
             'functions': {
                 0x401000: {
                     'features': [
-                        (0x401000, capa.features.Characteristic('switch', True)),
+                        (0x401000, capa.features.Characteristic('switch')),
                     ],
                     'basic blocks': {
                         0x401000: {
                             'features': [
-                                (0x401000, capa.features.Characteristic('tight-loop', True)),
+                                (0x401000, capa.features.Characteristic('tight-loop')),
                             ],
                             'instructions': {
                                 0x401000: {
                                     'features': [
-                                        (0x401000, capa.features.Characteristic('nzxor', True)),
+                                        (0x401000, capa.features.Characteristic('nzxor')),
                                     ],
                                 },
                                 0x401002: ...

capa/features/extractors/ida/basicblock.py

@@ -103,7 +103,7 @@ def extract_bb_stackstring(f, bb):
             bb (IDA BasicBlock)
     '''
     if _ida_bb_contains_stackstring(f, bb):
-        yield Characteristic('stack string', True), bb.start_ea
+        yield Characteristic('stack string'), bb.start_ea
 
 
 def _ida_bb_contains_tight_loop(f, bb):
@@ -133,7 +133,7 @@ def extract_bb_tight_loop(f, bb):
             bb (IDA BasicBlock)
     '''
     if _ida_bb_contains_tight_loop(f, bb):
-        yield Characteristic('tight loop', True), bb.start_ea
+        yield Characteristic('tight loop'), bb.start_ea
 
 
 def extract_features(f, bb):

capa/features/extractors/ida/file.py

@@ -68,7 +68,7 @@ def extract_file_embedded_pe():
             continue
 
         for ea, _ in _ida_check_segment_for_pe(seg):
-            yield Characteristic('embedded pe', True), ea
+            yield Characteristic('embedded pe'), ea
 
 
 def extract_file_export_names():

capa/features/extractors/ida/function.py

@@ -29,7 +29,7 @@ def extract_function_switch(f):
             f (IDA func_t)
     '''
     if _ida_function_contains_switch(f):
-        yield Characteristic('switch', True), f.start_ea
+        yield Characteristic('switch'), f.start_ea
 
 
 def extract_function_calls_to(f):
@@ -39,7 +39,7 @@ def extract_function_calls_to(f):
             f (IDA func_t)
     '''
     for ea in idautils.CodeRefsTo(f.start_ea, True):
-        yield Characteristic('calls to', True), ea
+        yield Characteristic('calls to'), ea
 
 
 def extract_function_loop(f):
@@ -53,7 +53,7 @@ def extract_function_loop(f):
         map(lambda s: edges.append((bb.start_ea, s.start_ea)), bb.succs())
 
     if edges and loops.has_loop(edges):
-        yield Characteristic('loop', True), f.start_ea
+        yield Characteristic('loop'), f.start_ea
 
 
 def extract_recursive_call(f):
@@ -64,7 +64,7 @@ def extract_recursive_call(f):
     '''
     for ref in idautils.CodeRefsTo(f.start_ea, True):
         if f.contains(ref):
-            yield Characteristic('recursive call', True), f.start_ea
+            yield Characteristic('recursive call'), f.start_ea
             break
 
 

capa/features/extractors/ida/insn.py

@@ -259,7 +259,7 @@ def extract_insn_nzxor_characteristic_features(f, bb, insn):
     if _is_nzxor_stack_cookie(f, bb, insn):
         return
 
-    yield Characteristic('nzxor', True), insn.ea
+    yield Characteristic('nzxor'), insn.ea
 
 
 def extract_insn_mnemonic_features(f, bb, insn):
@@ -292,7 +292,7 @@ def extract_insn_peb_access_characteristic_features(f, bb, insn):
 
     if ' fs:30h' in disasm or ' gs:60h' in disasm:
         # TODO: replace above with proper IDA
-        yield Characteristic('peb access', True), insn.ea
+        yield Characteristic('peb access'), insn.ea
 
 
 def extract_insn_segment_access_features(f, bb, insn):
@@ -309,11 +309,11 @@ def extract_insn_segment_access_features(f, bb, insn):
 
     if ' fs:' in disasm:
         # TODO: replace above with proper IDA
-        yield Characteristic('fs access', True), insn.ea
+        yield Characteristic('fs access'), insn.ea
 
     if ' gs:' in disasm:
         # TODO: replace above with proper IDA
-        yield Characteristic('gs access', True), insn.ea
+        yield Characteristic('gs access'), insn.ea
 
 
 def extract_insn_cross_section_cflow(f, bb, insn):
@@ -336,7 +336,7 @@ def extract_insn_cross_section_cflow(f, bb, insn):
         if idaapi.getseg(ref) == idaapi.getseg(insn.ea):
             continue
 
-        yield Characteristic('cross section flow', True), insn.ea
+        yield Characteristic('cross section flow'), insn.ea
 
 
 def extract_function_calls_from(f, bb, insn):
@@ -354,7 +354,7 @@ def extract_function_calls_from(f, bb, insn):
         return
 
     for ref in idautils.CodeRefsFrom(insn.ea, False):
-        yield Characteristic('calls from', True), ref
+        yield Characteristic('calls from'), ref
 
 
 def extract_function_indirect_call_characteristic_features(f, bb, insn):
@@ -373,7 +373,7 @@ def extract_function_indirect_call_characteristic_features(f, bb, insn):
         return
 
     if idc.get_operand_type(insn.ea, 0) in (idc.o_reg, idc.o_phrase, idc.o_displ):
-        yield Characteristic('indirect call', True), insn.ea
+        yield Characteristic('indirect call'), insn.ea
 
 
 def extract_features(f, bb, insn):

capa/features/extractors/viv/basicblock.py

@@ -39,7 +39,7 @@ def _bb_has_tight_loop(f, bb):
 def extract_bb_tight_loop(f, bb):
     ''' check basic block for tight loop indicators '''
     if _bb_has_tight_loop(f, bb):
-        yield Characteristic('tight loop', True), bb.va
+        yield Characteristic('tight loop'), bb.va
 
 
 def _bb_has_stackstring(f, bb):
@@ -62,7 +62,7 @@ def _bb_has_stackstring(f, bb):
 def extract_stackstring(f, bb):
     ''' check basic block for stackstring indicators '''
     if _bb_has_stackstring(f, bb):
-        yield Characteristic('stack string', True), bb.va
+        yield Characteristic('stack string'), bb.va
 
 
 def is_mov_imm_to_stack(instr):

capa/features/extractors/viv/file.py

@@ -13,7 +13,7 @@ def extract_file_embedded_pe(vw, file_path):
         fbytes = f.read()
 
     for offset, i in pe_carve.carve(fbytes, 1):
-        yield Characteristic('embedded pe', True), offset
+        yield Characteristic('embedded pe'), offset
 
 
 def extract_file_export_names(vw, file_path):

capa/features/extractors/viv/function.py

@@ -53,12 +53,12 @@ def extract_function_switch(f):
     method can be optimized
     '''
     if f.va in get_functions_with_switch(f.vw):
-        yield Characteristic('switch', True), f.va
+        yield Characteristic('switch'), f.va
 
 
 def extract_function_calls_to(f):
     for src, _, _, _ in f.vw.getXrefsTo(f.va, rtype=vivisect.const.REF_CODE):
-        yield Characteristic('calls to', True), src
+        yield Characteristic('calls to'), src
 
 
 def extract_function_loop(f):
@@ -74,7 +74,7 @@ def extract_function_loop(f):
                     edges.append((bb.va, bva))
 
     if edges and loops.has_loop(edges):
-        yield Characteristic('loop', True), f.va
+        yield Characteristic('loop'), f.va
 
 
 def extract_features(f):

capa/features/extractors/viv/insn.py

@@ -286,7 +286,7 @@ def extract_insn_nzxor_characteristic_features(f, bb, insn):
     if is_security_cookie(f, bb, insn):
         return
 
-    yield Characteristic('nzxor', True), insn.va
+    yield Characteristic('nzxor'), insn.va
 
 
 def extract_insn_mnemonic_features(f, bb, insn):
@@ -314,12 +314,12 @@ def extract_insn_peb_access_characteristic_features(f, bb, insn):
             #     fs: push dword [eax + 0x30]  ; i386RegMemOper, with eax = 0
             if (isinstance(oper, envi.archs.i386.disasm.i386RegMemOper) and oper.disp == 0x30) or \
                     (isinstance(oper, envi.archs.i386.disasm.i386ImmMemOper) and oper.imm == 0x30):
-                yield Characteristic('peb access', True), insn.va
+                yield Characteristic('peb access'), insn.va
     elif 'gs' in insn.getPrefixName():
         for oper in insn.opers:
             if (isinstance(oper, envi.archs.amd64.disasm.i386RegMemOper) and oper.disp == 0x60) or \
                     (isinstance(oper, envi.archs.amd64.disasm.i386ImmMemOper) and oper.imm == 0x60):
-                yield Characteristic('peb access', True), insn.va
+                yield Characteristic('peb access'), insn.va
     else:
         pass
 
@@ -329,10 +329,10 @@ def extract_insn_segment_access_features(f, bb, insn):
     prefix = insn.getPrefixName()
 
     if prefix == 'fs':
-        yield Characteristic('fs access', True), insn.va
+        yield Characteristic('fs access'), insn.va
 
     if prefix == 'gs':
-        yield Characteristic('gs access', True), insn.va
+        yield Characteristic('gs access'), insn.va
 
 
 def get_section(vw, va):
@@ -369,7 +369,7 @@ def extract_insn_cross_section_cflow(f, bb, insn):
                     continue
 
             if get_section(f.vw, insn.va) != get_section(f.vw, va):
-                yield Characteristic('cross section flow', True), insn.va
+                yield Characteristic('cross section flow'), insn.va
 
         except KeyError:
             continue
@@ -387,7 +387,7 @@ def extract_function_calls_from(f, bb, insn):
     if isinstance(insn.opers[0], envi.archs.i386.disasm.i386ImmMemOper):
         oper = insn.opers[0]
         target = oper.getOperAddr(insn)
-        yield Characteristic('calls from', True), target
+        yield Characteristic('calls from'), target
 
     # call via thunk on x86,
     # see 9324d1a8ae37a36ae560c37448c9705a at 0x407985
@@ -396,18 +396,18 @@ def extract_function_calls_from(f, bb, insn):
     # see Lab21-01.exe_:0x140001178
     elif isinstance(insn.opers[0], envi.archs.i386.disasm.i386PcRelOper):
         target = insn.opers[0].getOperValue(insn)
-        yield Characteristic('calls from', True), target
+        yield Characteristic('calls from'), target
 
     # call via IAT, x64
     elif isinstance(insn.opers[0], envi.archs.amd64.disasm.Amd64RipRelOper):
         op = insn.opers[0]
         target = op.getOperAddr(insn)
-        yield Characteristic('calls from', True), target
+        yield Characteristic('calls from'), target
 
     if target and target == f.va:
         # if we found a jump target and it's the function address
         # mark as recursive
-        yield Characteristic('recursive call', True), target
+        yield Characteristic('recursive call'), target
 
 
 # this is a feature that's most relevant at the function or basic block scope,
@@ -423,13 +423,13 @@ def extract_function_indirect_call_characteristic_features(f, bb, insn):
     # Checks below work for x86 and x64
     if isinstance(insn.opers[0], envi.archs.i386.disasm.i386RegOper):
         # call edx
-        yield Characteristic('indirect call', True), insn.va
+        yield Characteristic('indirect call'), insn.va
     elif isinstance(insn.opers[0], envi.archs.i386.disasm.i386RegMemOper):
         # call dword ptr [eax+50h]
-        yield Characteristic('indirect call', True), insn.va
+        yield Characteristic('indirect call'), insn.va
     elif isinstance(insn.opers[0], envi.archs.i386.disasm.i386SibOper):
         # call qword ptr [rsp+78h]
-        yield Characteristic('indirect call', True), insn.va
+        yield Characteristic('indirect call'), insn.va
 
 
 def extract_features(f, bb, insn):

capa/render/vverbose.py

@@ -41,14 +41,12 @@ def render_statement(ostream, match, statement, indent=0):
         # so, we have to inline some of the feature rendering here.
 
         child = statement['child']
-        if child['type'] in ('string', 'api', 'mnemonic', 'basic block', 'export', 'import', 'section', 'match'):
+        if child['type'] in ('string', 'api', 'mnemonic', 'basic block', 'export', 'import', 'section', 'match', 'characteristic'):
             feature = '%s(%s)' % (child['type'], rutils.bold2(child[child['type']]))
         elif child['type'] in ('number', 'offset'):
             feature = '%s(%s)' % (child['type'], rutils.bold2(rutils.hex(child[child['type']])))
         elif child['type'] == 'bytes':
             feature = '%s(%s)' % (child['type'], rutils.bold2(rutils.hex_string(child[child['type']])))
-        elif child['type'] == 'characteristic':
-            feature = 'characteristic(%s)' % (rutils.bold2(child['characteristic']))
         else:
             raise RuntimeError('unexpected feature type: ' + str(child))
 
@@ -80,7 +78,7 @@ def render_statement(ostream, match, statement, indent=0):
 def render_feature(ostream, match, feature, indent=0):
     ostream.write('  ' * indent)
 
-    if feature['type'] in ('string', 'api', 'mnemonic', 'basic block', 'export', 'import', 'section', 'match'):
+    if feature['type'] in ('string', 'api', 'mnemonic', 'basic block', 'export', 'import', 'section', 'match', 'characteristic'):
         ostream.write(feature['type'])
         ostream.write(': ')
         ostream.write(rutils.bold2(feature[feature['type']]))
@@ -93,8 +91,6 @@ def render_feature(ostream, match, feature, indent=0):
         # bytes is the uppercase, hex-encoded string.
         # it should always be an even number of characters (its hex).
         ostream.write(rutils.bold2(rutils.hex_string(feature[feature['type']])))
-    elif feature['type'] == 'characteristic':
-        ostream.write('characteristic(%s)' % (rutils.bold2(feature['characteristic'])))
     # note that regex is found in `render_statement`
     else:
         raise RuntimeError('unexpected feature type: ' + str(feature))

capa/rules.py

@@ -138,7 +138,7 @@ class InvalidRuleSet(ValueError):
 
 def ensure_feature_valid_for_scope(scope, feature):
     if isinstance(feature, capa.features.Characteristic):
-        if capa.features.Characteristic(feature.attribute) not in SUPPORTED_FEATURES[scope]:
+        if capa.features.Characteristic(feature.value) not in SUPPORTED_FEATURES[scope]:
             raise InvalidRule('feature %s not support for scope %s' % (feature, scope))
     elif not isinstance(feature, tuple(filter(lambda t: isinstance(t, type), SUPPORTED_FEATURES[scope]))):
         raise InvalidRule('feature %s not support for scope %s' % (feature, scope))
@@ -205,9 +205,8 @@ def parse_feature(key):
         return capa.features.insn.Mnemonic
     elif key == 'basic blocks':
         return capa.features.basicblock.BasicBlock
-    elif key.startswith('characteristic(') and key.endswith(')'):
+    elif key == 'characteristic':
-        characteristic = key[len('characteristic('):-len(')')]
+        return capa.features.Characteristic
-        return lambda v, description=None: capa.features.Characteristic(characteristic, v, description)
     elif key == 'export':
         return capa.features.file.Export
     elif key == 'import':
@@ -302,48 +301,34 @@ def build_statements(d, scope):
 
         term = key[len('count('):-len(')')]
 
-        if term.startswith('characteristic('):
+        # when looking for the existence of such a feature, our rule might look like:
-            # characteristic features are specified a bit specially:
+        #     - mnemonic: mov
-            # they simply indicate the presence of something unusual/interesting,
+        #
-            # and we embed the name in the feature name, like `characteristic(nzxor)`.
+        # but here we deal with the form: `mnemonic(mov)`.
-            #
+        term, _, arg = term.partition('(')
-            # when we're dealing with counts, like `count(characteristic(nzxor))`,
+        Feature = parse_feature(term)
-            # we can simply extract the feature and assume we're looking for `True` values.
-            Feature = parse_feature(term)
-            feature = Feature(True)
-            ensure_feature_valid_for_scope(scope, feature)
-        else:
-            # however, for remaining counted features, like `count(mnemonic(mov))`,
-            # we have to jump through hoops.
-            #
-            # when looking for the existance of such a feature, our rule might look like:
-            #     - mnemonic: mov
-            #
-            # but here we deal with the form: `mnemonic(mov)`.
-            term, _, arg = term.partition('(')
-            Feature = parse_feature(term)
 
-            if arg:
+        if arg:
-                arg = arg[:-len(')')]
+            arg = arg[:-len(')')]
-                # can't rely on yaml parsing ints embedded within strings
+            # can't rely on yaml parsing ints embedded within strings
-                # like:
+            # like:
-                #
+            #
-                #     count(offset(0xC))
+            #     count(offset(0xC))
-                #     count(number(0x11223344))
+            #     count(number(0x11223344))
-                #     count(number(0x100 = description))
+            #     count(number(0x100 = description))
-                if term in ('number', 'offset', 'bytes'):
+            if term in ('number', 'offset', 'bytes'):
-                    value, description = parse_description(arg, term)
+                value, description = parse_description(arg, term)
-                    feature = Feature(value, description)
+                feature = Feature(value, description)
-                else:
-                    # arg is string, like:
-                    #
-                    #     count(mnemonic(mov))
-                    #     count(string(error))
-                    # TODO: what about embedded newlines?
-                    feature = Feature(arg)
             else:
-                feature = Feature()
+                # arg is string, like:
-            ensure_feature_valid_for_scope(scope, feature)
+                #
+                #     count(mnemonic(mov))
+                #     count(string(error))
+                # TODO: what about embedded newlines?
+                feature = Feature(arg)
+        else:
+            feature = Feature()
+        ensure_feature_valid_for_scope(scope, feature)
 
         count = d[key]
         if isinstance(count, int):

tests/test_freeze.py

@@ -12,23 +12,23 @@ from fixtures import *
 
 EXTRACTOR = capa.features.extractors.NullFeatureExtractor({
     'file features': [
-        (0x402345, capa.features.Characteristic('embedded pe', True)),
+        (0x402345, capa.features.Characteristic('embedded pe')),
     ],
     'functions': {
         0x401000: {
             'features': [
-                (0x401000, capa.features.Characteristic('switch', True)),
+                (0x401000, capa.features.Characteristic('switch')),
             ],
             'basic blocks': {
                 0x401000: {
                     'features': [
-                        (0x401000, capa.features.Characteristic('tight loop', True)),
+                        (0x401000, capa.features.Characteristic('tight loop')),
                     ],
                     'instructions': {
                         0x401000: {
                             'features': [
                                 (0x401000, capa.features.insn.Mnemonic('xor')),
-                                (0x401000, capa.features.Characteristic('nzxor', True)),
+                                (0x401000, capa.features.Characteristic('nzxor')),
                             ],
                         },
                         0x401002: {
@@ -57,9 +57,9 @@ def test_null_feature_extractor():
                     scope: basic block
                 features:
                     - and:
-                        - characteristic(tight loop): true
+                        - characteristic: tight loop
                         - mnemonic: xor
-                        - characteristic(nzxor): true
+                        - characteristic: nzxor
         ''')),
     ])
     capabilities = capa.main.find_capabilities(rules, EXTRACTOR)
@@ -150,7 +150,7 @@ def test_serialize_features():
     roundtrip_feature(capa.features.insn.Offset(0x0))
     roundtrip_feature(capa.features.insn.Mnemonic('push'))
     roundtrip_feature(capa.features.file.Section('.rsrc'))
-    roundtrip_feature(capa.features.Characteristic('tight loop', True))
+    roundtrip_feature(capa.features.Characteristic('tight loop'))
     roundtrip_feature(capa.features.basicblock.BasicBlock())
     roundtrip_feature(capa.features.file.Export('BaseThreadInitThunk'))
     roundtrip_feature(capa.features.file.Import('kernel32.IsWow64Process'))

tests/test_main.py

@@ -42,7 +42,7 @@ def test_ruleset():
                     name: file rule
                     scope: file
                 features:
-                  - characteristic(embedded pe): y
+                  - characteristic: embedded pe
         ''')),
         capa.rules.Rule.from_yaml(textwrap.dedent('''
             rule:
@@ -50,7 +50,7 @@ def test_ruleset():
                     name: function rule
                     scope: function
                 features:
-                  - characteristic(switch): y
+                  - characteristic: switch
         ''')),
          capa.rules.Rule.from_yaml(textwrap.dedent('''
             rule:
@@ -58,7 +58,7 @@ def test_ruleset():
                     name: basic block rule
                     scope: basic block
                 features:
-                  - characteristic(nzxor): y
+                  - characteristic: nzxor
         ''')),
 
     ])
@@ -128,7 +128,7 @@ def test_match_across_scopes(sample_9324d1a8ae37a36ae560c37448c9705a):
                     examples:
                       - 9324d1a8ae37a36ae560c37448c9705a:0x403685
                 features:
-                  - characteristic(tight loop): true
+                  - characteristic: tight loop
         ''')),
         # this rule should match on a function (0x403660)
         # based on API, as well as prior basic block rule match
@@ -176,7 +176,7 @@ def test_subscope_bb_rules(sample_9324d1a8ae37a36ae560c37448c9705a):
                  features:
                      - and:
                          - basic block:
-                             - characteristic(tight loop): true
+                             - characteristic: tight loop
          '''))
     ])
     # tight loop at 0x403685

tests/test_rules.py

@@ -118,7 +118,7 @@ def test_invalid_rule_feature():
                     name: test rule
                     scope: file
                 features:
-                    - characteristic(nzxor): true
+                    - characteristic: nzxor
         '''))
 
     with pytest.raises(capa.rules.InvalidRule):
@@ -128,7 +128,7 @@ def test_invalid_rule_feature():
                     name: test rule
                     scope: function
                 features:
-                    - characteristic(embedded pe): true
+                    - characteristic: embedded pe
         '''))
 
     with pytest.raises(capa.rules.InvalidRule):
@@ -138,7 +138,7 @@ def test_invalid_rule_feature():
                     name: test rule
                     scope: basic block
                 features:
-                    - characteristic(embedded pe): true
+                    - characteristic: embedded pe
         '''))
 
 
@@ -173,11 +173,11 @@ def test_subscope_rules():
                     scope: file
                 features:
                     - and:
-                        - characteristic(embedded pe): true
+                        - characteristic: embedded pe
                         - function:
                             - and:
-                                - characteristic(nzxor): true
+                                - characteristic: nzxor
-                                - characteristic(switch): true
+                                - characteristic: switch
         '''))
     ])
     # the file rule scope will have one rules:
@@ -229,7 +229,7 @@ def test_invalid_rules():
                     meta:
                         name: test rule
                     features:
-                        - characteristic(number(1)): True
+                        - characteristic: number(1)
             '''))
 
     with pytest.raises(capa.rules.InvalidRule):
@@ -238,7 +238,7 @@ def test_invalid_rules():
                     meta:
                         name: test rule
                     features:
-                        - characteristic(count(number(100))): True
+                        - characteristic: count(number(100))
             '''))
 
 

tests/test_viv_features.py

@@ -116,7 +116,7 @@ def test_offset_features(mimikatz):
 
 def test_nzxor_features(mimikatz):
     features = extract_function_features(viv_utils.Function(mimikatz.vw, 0x410DFC))
-    assert capa.features.Characteristic('nzxor', True) in features  # 0x0410F0B
+    assert capa.features.Characteristic('nzxor') in features  # 0x0410F0B
 
 
 def get_bb_insn(f, va):
@@ -154,7 +154,7 @@ def test_mnemonic_features(mimikatz):
 
 def test_peb_access_features(sample_a933a1a402775cfa94b6bee0963f4b46):
     features = extract_function_features(viv_utils.Function(sample_a933a1a402775cfa94b6bee0963f4b46.vw, 0xABA6FEC))
-    assert capa.features.Characteristic('peb access', True) in features
+    assert capa.features.Characteristic('peb access') in features
 
 
 def test_file_section_name_features(mimikatz):
@@ -170,7 +170,7 @@ def test_tight_loop_features(mimikatz):
         if bb.va != 0x402F8E:
             continue
         features = extract_basic_block_features(f, bb)
-        assert capa.features.Characteristic('tight loop', True) in features
+        assert capa.features.Characteristic('tight loop') in features
         assert capa.features.basicblock.BasicBlock() in features
 
 
@@ -180,7 +180,7 @@ def test_tight_loop_bb_features(mimikatz):
         if bb.va != 0x402F8E:
             continue
         features = extract_basic_block_features(f, bb)
-        assert capa.features.Characteristic('tight loop', True) in features
+        assert capa.features.Characteristic('tight loop') in features
         assert capa.features.basicblock.BasicBlock() in features
 
 
@@ -202,17 +202,17 @@ def test_file_import_name_features(mimikatz):
 
 def test_cross_section_flow_features(sample_a198216798ca38f280dc413f8c57f2c2):
     features = extract_function_features(viv_utils.Function(sample_a198216798ca38f280dc413f8c57f2c2.vw, 0x4014D0))
-    assert capa.features.Characteristic('cross section flow', True) in features
+    assert capa.features.Characteristic('cross section flow') in features
 
     # this function has calls to some imports,
     # which should not trigger cross-section flow characteristic
     features = extract_function_features(viv_utils.Function(sample_a198216798ca38f280dc413f8c57f2c2.vw, 0x401563))
-    assert capa.features.Characteristic('cross section flow', True) not in features
+    assert capa.features.Characteristic('cross section flow') not in features
 
 
 def test_segment_access_features(sample_a933a1a402775cfa94b6bee0963f4b46):
     features = extract_function_features(viv_utils.Function(sample_a933a1a402775cfa94b6bee0963f4b46.vw, 0xABA6FEC))
-    assert capa.features.Characteristic('fs access', True) in features
+    assert capa.features.Characteristic('fs access') in features
 
 
 def test_thunk_features(sample_9324d1a8ae37a36ae560c37448c9705a):
@@ -223,36 +223,36 @@ def test_thunk_features(sample_9324d1a8ae37a36ae560c37448c9705a):
 
 def test_file_embedded_pe(pma_lab_12_04):
     features = extract_file_features(pma_lab_12_04.vw, pma_lab_12_04.path)
-    assert capa.features.Characteristic('embedded pe', True) in features
+    assert capa.features.Characteristic('embedded pe') in features
 
 
 def test_stackstring_features(mimikatz):
     features = extract_function_features(viv_utils.Function(mimikatz.vw, 0x4556E5))
-    assert capa.features.Characteristic('stack string', True) in features
+    assert capa.features.Characteristic('stack string') in features
 
 
 def test_switch_features(mimikatz):
     features = extract_function_features(viv_utils.Function(mimikatz.vw, 0x409411))
-    assert capa.features.Characteristic('switch', True) in features
+    assert capa.features.Characteristic('switch') in features
 
     features = extract_function_features(viv_utils.Function(mimikatz.vw, 0x409393))
-    assert capa.features.Characteristic('switch', True) not in features
+    assert capa.features.Characteristic('switch') not in features
 
 
 def test_recursive_call_feature(sample_39c05b15e9834ac93f206bc114d0a00c357c888db567ba8f5345da0529cbed41):
     features = extract_function_features(viv_utils.Function(sample_39c05b15e9834ac93f206bc114d0a00c357c888db567ba8f5345da0529cbed41.vw, 0x10003100))
-    assert capa.features.Characteristic('recursive call', True) in features
+    assert capa.features.Characteristic('recursive call') in features
 
     features = extract_function_features(viv_utils.Function(sample_39c05b15e9834ac93f206bc114d0a00c357c888db567ba8f5345da0529cbed41.vw, 0x10007B00))
-    assert capa.features.Characteristic('recursive call', True) not in features
+    assert capa.features.Characteristic('recursive call') not in features
 
 
 def test_loop_feature(sample_39c05b15e9834ac93f206bc114d0a00c357c888db567ba8f5345da0529cbed41):
     features = extract_function_features(viv_utils.Function(sample_39c05b15e9834ac93f206bc114d0a00c357c888db567ba8f5345da0529cbed41.vw, 0x10003D30))
-    assert capa.features.Characteristic('loop', True) in features
+    assert capa.features.Characteristic('loop') in features
 
     features = extract_function_features(viv_utils.Function(sample_39c05b15e9834ac93f206bc114d0a00c357c888db567ba8f5345da0529cbed41.vw, 0x10007250))
-    assert capa.features.Characteristic('loop', True) not in features
+    assert capa.features.Characteristic('loop') not in features
 
 
 def test_file_string_features(sample_bfb9b5391a13d0afd787e87ab90f14f5):
@@ -263,20 +263,20 @@ def test_file_string_features(sample_bfb9b5391a13d0afd787e87ab90f14f5):
 
 def test_function_calls_to(sample_9324d1a8ae37a36ae560c37448c9705a):
     features = extract_function_features(viv_utils.Function(sample_9324d1a8ae37a36ae560c37448c9705a.vw, 0x406F60))
-    assert capa.features.Characteristic('calls to', True) in features
+    assert capa.features.Characteristic('calls to') in features
-    assert len(features[capa.features.Characteristic('calls to', True)]) == 1
+    assert len(features[capa.features.Characteristic('calls to')]) == 1
 
 
 def test_function_calls_to64(sample_lab21_01):
     features = extract_function_features(viv_utils.Function(sample_lab21_01.vw, 0x1400052D0))  # memcpy
-    assert capa.features.Characteristic('calls to', True) in features
+    assert capa.features.Characteristic('calls to') in features
-    assert len(features[capa.features.Characteristic('calls to', True)]) == 8
+    assert len(features[capa.features.Characteristic('calls to')]) == 8
 
 
 def test_function_calls_from(sample_9324d1a8ae37a36ae560c37448c9705a):
     features = extract_function_features(viv_utils.Function(sample_9324d1a8ae37a36ae560c37448c9705a.vw, 0x406F60))
-    assert capa.features.Characteristic('calls from', True) in features
+    assert capa.features.Characteristic('calls from') in features
-    assert len(features[capa.features.Characteristic('calls from', True)]) == 23
+    assert len(features[capa.features.Characteristic('calls from')]) == 23
 
 
 def test_basic_block_count(sample_9324d1a8ae37a36ae560c37448c9705a):
@@ -286,8 +286,8 @@ def test_basic_block_count(sample_9324d1a8ae37a36ae560c37448c9705a):
 
 def test_indirect_call_features(sample_a933a1a402775cfa94b6bee0963f4b46):
     features = extract_function_features(viv_utils.Function(sample_a933a1a402775cfa94b6bee0963f4b46.vw, 0xABA68A0))
-    assert capa.features.Characteristic('indirect call', True) in features
+    assert capa.features.Characteristic('indirect call') in features
-    assert len(features[capa.features.Characteristic('indirect call', True)]) == 3
+    assert len(features[capa.features.Characteristic('indirect call')]) == 3
 
 
 def test_indirect_calls_resolved(sample_c91887d861d9bd4a5872249b641bc9f9):