Merge pull request #121 from omergunal/patch-2

Whitelist lines ending in # nosec

Author: KevinHock

pyt/__main__.py

@@ -142,6 +142,8 @@ def parse_args(args):
                              '(only JSON-formatted files are accepted)',
                         type=str,
                         default=False)
+    parser.add_argument('--ignore-nosec', dest='ignore_nosec', action='store_true',
+                         help='do not skip lines with # nosec comments')
 
     save_parser = subparsers.add_parser('save', help='Save menu.')
     save_parser.set_defaults(which='save')
@@ -192,7 +194,7 @@ def parse_args(args):
     return parser.parse_args(args)
 
 
-def analyse_repo(args, github_repo, analysis_type, ui_mode):
+def analyse_repo(args, github_repo, analysis_type, ui_mode, nosec_lines):
     cfg_list = list()
     directory = os.path.dirname(github_repo.path)
     project_modules = get_modules(directory)
@@ -215,7 +217,8 @@ def analyse_repo(args, github_repo, analysis_type, ui_mode):
         VulnerabilityFiles(
             args.blackbox_mapping_file,
             args.trigger_word_file
-        )
+        ),
+        nosec_lines
     )
     return vulnerabilities
 
@@ -235,12 +238,23 @@ def main(command_line_args=sys.argv[1:]):
     elif args.trim_reassigned_in:
         ui_mode = UImode.TRIM
 
+    path = os.path.normpath(args.filepath)
     cfg_list = list()
+    if args.ignore_nosec:
+        nosec_lines = set()
+    else:
+        file = open(path, "r")
+        lines = file.readlines()
+        nosec_lines = set(
+                    lineno for
+                    (lineno, line) in enumerate(lines, start=1)
+                    if '#nosec' in line or '# nosec' in line)
+        
     if args.git_repos:
         repos = get_repos(args.git_repos)
         for repo in repos:
             repo.clone()
-            vulnerabilities = analyse_repo(args, repo, analysis, ui_mode)
+            vulnerabilities = analyse_repo(args, repo, analysis, ui_mode, nosec_lines)
             if args.json:
                 json.report(vulnerabilities, sys.stdout)
             else:
@@ -263,8 +277,6 @@ def main(command_line_args=sys.argv[1:]):
         )
         exit()
 
-    path = os.path.normpath(args.filepath)
-
     directory = None
     if args.project_root:
         directory = os.path.normpath(args.project_root)
@@ -305,8 +317,10 @@ def main(command_line_args=sys.argv[1:]):
         VulnerabilityFiles(
             args.blackbox_mapping_file,
             args.trigger_word_file
-        )
+        ),
+        nosec_lines
     )
+    
     if args.baseline:
             vulnerabilities = get_vulnerabilities_not_in_baseline(vulnerabilities, args.baseline)
 

pyt/vulnerabilities.py

@@ -73,7 +73,8 @@ def identify_triggers(
     cfg,
     sources,
     sinks,
-    lattice
+    lattice,
+    nosec_lines
 ):
     """Identify sources, sinks and sanitisers in a CFG.
 
@@ -89,12 +90,12 @@ def identify_triggers(
     tainted_nodes = filter_cfg_nodes(cfg, TaintedNode)
     tainted_trigger_nodes = [TriggerNode('Framework function URL parameter', None,
                                          node) for node in tainted_nodes]
-    sources_in_file = find_triggers(assignment_nodes, sources)
+    sources_in_file = find_triggers(assignment_nodes, sources, nosec_lines)
     sources_in_file.extend(tainted_trigger_nodes)
 
     find_secondary_sources(assignment_nodes, sources_in_file, lattice)
 
-    sinks_in_file = find_triggers(cfg.nodes, sinks)
+    sinks_in_file = find_triggers(cfg.nodes, sinks, nosec_lines)
 
     sanitiser_node_dict = build_sanitiser_node_dict(cfg, sinks_in_file)
 
@@ -170,7 +171,8 @@ def append_node_if_reassigned(
 
 def find_triggers(
     nodes,
-    trigger_words
+    trigger_words,
+    nosec_lines=set()
 ):
     """Find triggers from the trigger_word_list in the nodes.
 
@@ -183,7 +185,8 @@ def find_triggers(
     """
     trigger_nodes = list()
     for node in nodes:
-        trigger_nodes.extend(iter(label_contains(node, trigger_words)))
+        if node.line_number not in nosec_lines:
+            trigger_nodes.extend(iter(label_contains(node, trigger_words)))
     return trigger_nodes
 
 
@@ -466,7 +469,8 @@ def find_vulnerabilities_in_cfg(
     lattice,
     ui_mode,
     blackbox_mapping,
-    vulnerabilities_list
+    vulnerabilities_list,
+    nosec_lines
 ):
     """Find vulnerabilities in a cfg.
 
@@ -482,7 +486,8 @@ def find_vulnerabilities_in_cfg(
         cfg,
         definitions.sources,
         definitions.sinks,
-        lattice
+        lattice,
+        nosec_lines
     )
     for sink in triggers.sinks:
         for source in triggers.sources:
@@ -503,7 +508,8 @@ def find_vulnerabilities(
     cfg_list,
     analysis_type,
     ui_mode,
-    vulnerability_files
+    vulnerability_files,
+    nosec_lines=set()
 ):
     """Find vulnerabilities in a list of CFGs from a trigger_word_file.
 
@@ -518,7 +524,6 @@ def find_vulnerabilities(
     """
     vulnerabilities = list()
     definitions = parse(vulnerability_files.triggers)
-
     with open(vulnerability_files.blackbox_mapping) as infile:
         blackbox_mapping = json.load(infile)
     for cfg in cfg_list:
@@ -528,9 +533,9 @@ def find_vulnerabilities(
             Lattice(cfg.nodes, analysis_type),
             ui_mode,
             blackbox_mapping,
-            vulnerabilities
+            vulnerabilities,
+            nosec_lines
         )
     with open(vulnerability_files.blackbox_mapping, 'w') as outfile:
         json.dump(blackbox_mapping, outfile, indent=4)
-
     return vulnerabilities

tests/command_line_test.py

@@ -28,6 +28,7 @@ def test_no_args(self):
                      [-m BLACKBOX_MAPPING_FILE] [-py2] [-l LOG_LEVEL]
                      [-a ADAPTOR] [-db] [-dl DRAW_LATTICE [DRAW_LATTICE ...]]
                      [-j] [-li | -re | -rt] [-ppm] [-b BASELINE]
+                     [--ignore-nosec]
                      {save,github_search} ...\n""" + \
                      "python -m pyt: error: one of the arguments " + \
                      "-f/--filepath -gr/--git-repos is required\n"