Python: detecting header splitting in synthetic app

python/ql/lib/change-notes/2025-11-26-model-ParseResult.md

@@ -0,0 +1,4 @@
+---
+category: minorAnalysis
+---
+* Added taint flow model and type model for `urllib.parseurl`.

python/ql/lib/semmle/python/frameworks/Stdlib.model.yml

@@ -142,6 +142,8 @@ extensions:
       - ["typing", "Member[cast]", "Argument[1,val:]", "ReturnValue", "value"]
       # See https://docs.python.org/3/library/urllib.parse.html#urllib.parse.parse_qs
       - ["urllib", "Member[parse].Member[parse_qs]", "Argument[0,qs:]", "ReturnValue", "taint"]
+      # See https://docs.python.org/3/library/urllib.parse.html#urllib.parse.urlparse
+      - ["urllib", "Member[parse].Member[urlparse]", "Argument[0,urlstring:]", "ReturnValue", "taint"]
       # See https://docs.python.org/3/library/urllib.parse.html#urllib.parse.quote
       - ["urllib", "Member[parse].Member[quote]", "Argument[0,string:]", "ReturnValue", "taint"]
       # See https://docs.python.org/3/library/urllib.parse.html#urllib.parse.quote_plus
@@ -181,7 +183,9 @@ extensions:
   - addsTo:
       pack: codeql/python-all
       extensible: typeModel
-    data: []
+    data:
+      # See https://docs.python.org/3/library/urllib.parse.html#urllib.parse.urlparse
+      - ["urllib.parse.ParseResult~Subclass", 'urllib', 'Member[parse].Member[urlparse]']
 
   - addsTo:
       pack: codeql/python-all

python/ql/lib/semmle/python/frameworks/Stdlib.qll

@@ -245,6 +245,67 @@ module Stdlib {
     }
   }
 
+  /**
+   * Provides models for the `urllib.parse.ParseResult` class
+   *
+   * See https://docs.python.org/3.9/library/urllib.parse.html#urllib.parse.ParseResult.
+   */
+  module ParseResult {
+    /** Gets a reference to the `urllib.parse.ParseResult` class. */
+    API::Node classRef() {
+      result = API::moduleImport("urllib").getMember("parse").getMember("ParseResult")
+      or
+      result = ModelOutput::getATypeNode("urllib.parse.ParseResult~Subclass").getASubclass*()
+    }
+
+    /**
+     * A source of instances of `urllib.parse.ParseResult`, extend this class to model new instances.
+     *
+     * This can include instantiations of the class, return values from function
+     * calls, or a special parameter that will be set when functions are called by an external
+     * library.
+     *
+     * Use the predicate `ParseResult::instance()` to get references to instances of `urllib.parse.ParseResult`.
+     */
+    abstract class InstanceSource extends DataFlow::LocalSourceNode { }
+
+    /** A direct instantiation of `urllib.parse.ParseResult`. */
+    private class ClassInstantiation extends InstanceSource, DataFlow::CallCfgNode {
+      ClassInstantiation() { this = classRef().getACall() }
+    }
+
+    /** Gets a reference to an instance of `urllib.parse.ParseResult`. */
+    private DataFlow::TypeTrackingNode instance(DataFlow::TypeTracker t) {
+      t.start() and
+      result instanceof InstanceSource
+      or
+      exists(DataFlow::TypeTracker t2 | result = instance(t2).track(t2, t))
+    }
+
+    /** Gets a reference to an instance of `urllib.parse.ParseResult`. */
+    DataFlow::Node instance() { instance(DataFlow::TypeTracker::end()).flowsTo(result) }
+
+    /**
+     * Taint propagation for `urllib.parse.ParseResult`.
+     */
+    private class InstanceTaintSteps extends InstanceTaintStepsHelper {
+      InstanceTaintSteps() { this = "urllib.parse.ParseResult" }
+
+      override DataFlow::Node getInstance() { result = instance() }
+
+      override string getAttributeName() {
+        result in [
+            "netloc", "path", "params", "query", "fragment", "username", "password", "hostname",
+            "port"
+          ]
+      }
+
+      override string getMethodName() { none() }
+
+      override string getAsyncMethodName() { none() }
+    }
+  }
+
   // ---------------------------------------------------------------------------
   // logging
   // ---------------------------------------------------------------------------

python/ql/test/experimental/query-tests/Security/CWE-409/DecompressionBombs.expected

@@ -1,23 +1,23 @@
 edges
 | test.py:10:16:10:24 | ControlFlowNode for file_path | test.py:11:21:11:29 | ControlFlowNode for file_path | provenance |  |
 | test.py:11:5:11:35 | ControlFlowNode for Attribute() | test.py:11:5:11:52 | ControlFlowNode for Attribute() | provenance | Config |
-| test.py:11:21:11:29 | ControlFlowNode for file_path | test.py:11:5:11:35 | ControlFlowNode for Attribute() | provenance | MaD:85 |
+| test.py:11:21:11:29 | ControlFlowNode for file_path | test.py:11:5:11:35 | ControlFlowNode for Attribute() | provenance | MaD:86 |
 | test.py:11:21:11:29 | ControlFlowNode for file_path | test.py:11:5:11:52 | ControlFlowNode for Attribute() | provenance | Config |
 | test.py:11:21:11:29 | ControlFlowNode for file_path | test.py:12:21:12:29 | ControlFlowNode for file_path | provenance |  |
 | test.py:12:5:12:35 | ControlFlowNode for Attribute() | test.py:12:5:12:48 | ControlFlowNode for Attribute() | provenance | Config |
-| test.py:12:21:12:29 | ControlFlowNode for file_path | test.py:12:5:12:35 | ControlFlowNode for Attribute() | provenance | MaD:85 |
+| test.py:12:21:12:29 | ControlFlowNode for file_path | test.py:12:5:12:35 | ControlFlowNode for Attribute() | provenance | MaD:86 |
 | test.py:12:21:12:29 | ControlFlowNode for file_path | test.py:12:5:12:48 | ControlFlowNode for Attribute() | provenance | Config |
 | test.py:12:21:12:29 | ControlFlowNode for file_path | test.py:14:26:14:34 | ControlFlowNode for file_path | provenance |  |
 | test.py:14:10:14:35 | ControlFlowNode for Attribute() | test.py:15:14:15:29 | ControlFlowNode for Attribute() | provenance | Config |
-| test.py:14:26:14:34 | ControlFlowNode for file_path | test.py:14:10:14:35 | ControlFlowNode for Attribute() | provenance | MaD:85 |
+| test.py:14:26:14:34 | ControlFlowNode for file_path | test.py:14:10:14:35 | ControlFlowNode for Attribute() | provenance | MaD:86 |
 | test.py:14:26:14:34 | ControlFlowNode for file_path | test.py:15:14:15:29 | ControlFlowNode for Attribute() | provenance | Config |
 | test.py:14:26:14:34 | ControlFlowNode for file_path | test.py:18:26:18:34 | ControlFlowNode for file_path | provenance |  |
 | test.py:18:10:18:35 | ControlFlowNode for Attribute() | test.py:19:14:19:39 | ControlFlowNode for Attribute() | provenance | Config |
-| test.py:18:26:18:34 | ControlFlowNode for file_path | test.py:18:10:18:35 | ControlFlowNode for Attribute() | provenance | MaD:85 |
+| test.py:18:26:18:34 | ControlFlowNode for file_path | test.py:18:10:18:35 | ControlFlowNode for Attribute() | provenance | MaD:86 |
 | test.py:18:26:18:34 | ControlFlowNode for file_path | test.py:19:14:19:39 | ControlFlowNode for Attribute() | provenance | Config |
 | test.py:18:26:18:34 | ControlFlowNode for file_path | test.py:22:21:22:29 | ControlFlowNode for file_path | provenance |  |
 | test.py:22:5:22:30 | ControlFlowNode for Attribute() | test.py:22:5:22:60 | ControlFlowNode for Attribute() | provenance | Config |
-| test.py:22:21:22:29 | ControlFlowNode for file_path | test.py:22:5:22:30 | ControlFlowNode for Attribute() | provenance | MaD:85 |
+| test.py:22:21:22:29 | ControlFlowNode for file_path | test.py:22:5:22:30 | ControlFlowNode for Attribute() | provenance | MaD:86 |
 | test.py:22:21:22:29 | ControlFlowNode for file_path | test.py:22:5:22:60 | ControlFlowNode for Attribute() | provenance | Config |
 | test.py:22:21:22:29 | ControlFlowNode for file_path | test.py:24:18:24:26 | ControlFlowNode for file_path | provenance |  |
 | test.py:24:18:24:26 | ControlFlowNode for file_path | test.py:24:5:24:52 | ControlFlowNode for Attribute() | provenance | Config |

python/ql/test/query-tests/Security/CWE-113-HeaderInjection/Tests1/HeaderInjection.expected

@@ -11,6 +11,16 @@ edges
 | flask_tests.py:31:5:31:14 | ControlFlowNode for rfs_header | flask_tests.py:33:11:33:20 | ControlFlowNode for rfs_header | provenance |  |
 | flask_tests.py:31:5:31:14 | ControlFlowNode for rfs_header | flask_tests.py:35:12:35:21 | ControlFlowNode for rfs_header | provenance |  |
 | flask_tests.py:31:18:31:24 | ControlFlowNode for request | flask_tests.py:31:5:31:14 | ControlFlowNode for rfs_header | provenance | AdditionalTaintStep |
+| http_test.py:5:16:5:19 | ControlFlowNode for self | http_test.py:6:45:6:53 | ControlFlowNode for Attribute | provenance | AdditionalTaintStep |
+| http_test.py:6:9:6:19 | ControlFlowNode for parsed_path | http_test.py:7:40:7:56 | ControlFlowNode for Attribute | provenance | AdditionalTaintStep |
+| http_test.py:6:23:6:54 | ControlFlowNode for Attribute() | http_test.py:6:9:6:19 | ControlFlowNode for parsed_path | provenance |  |
+| http_test.py:6:45:6:53 | ControlFlowNode for Attribute | http_test.py:6:23:6:54 | ControlFlowNode for Attribute() | provenance | MaD:77 |
+| http_test.py:7:9:7:14 | ControlFlowNode for params | http_test.py:8:23:8:28 | ControlFlowNode for params | provenance |  |
+| http_test.py:7:18:7:57 | ControlFlowNode for Attribute() | http_test.py:7:9:7:14 | ControlFlowNode for params | provenance |  |
+| http_test.py:7:40:7:56 | ControlFlowNode for Attribute | http_test.py:7:18:7:57 | ControlFlowNode for Attribute() | provenance | MaD:76 |
+| http_test.py:8:9:8:19 | ControlFlowNode for input_value | http_test.py:12:40:12:50 | ControlFlowNode for input_value | provenance |  |
+| http_test.py:8:23:8:28 | ControlFlowNode for params | http_test.py:8:23:8:47 | ControlFlowNode for Attribute() | provenance | dict.get |
+| http_test.py:8:23:8:47 | ControlFlowNode for Attribute() | http_test.py:8:9:8:19 | ControlFlowNode for input_value | provenance |  |
 | wsgiref_tests.py:4:14:4:20 | ControlFlowNode for environ | wsgiref_tests.py:6:5:6:10 | ControlFlowNode for h_name | provenance |  |
 | wsgiref_tests.py:4:14:4:20 | ControlFlowNode for environ | wsgiref_tests.py:7:5:7:9 | ControlFlowNode for h_val | provenance |  |
 | wsgiref_tests.py:6:5:6:10 | ControlFlowNode for h_name | wsgiref_tests.py:8:17:8:22 | ControlFlowNode for h_name | provenance |  |
@@ -28,6 +38,17 @@ nodes
 | flask_tests.py:31:18:31:24 | ControlFlowNode for request | semmle.label | ControlFlowNode for request |
 | flask_tests.py:33:11:33:20 | ControlFlowNode for rfs_header | semmle.label | ControlFlowNode for rfs_header |
 | flask_tests.py:35:12:35:21 | ControlFlowNode for rfs_header | semmle.label | ControlFlowNode for rfs_header |
+| http_test.py:5:16:5:19 | ControlFlowNode for self | semmle.label | ControlFlowNode for self |
+| http_test.py:6:9:6:19 | ControlFlowNode for parsed_path | semmle.label | ControlFlowNode for parsed_path |
+| http_test.py:6:23:6:54 | ControlFlowNode for Attribute() | semmle.label | ControlFlowNode for Attribute() |
+| http_test.py:6:45:6:53 | ControlFlowNode for Attribute | semmle.label | ControlFlowNode for Attribute |
+| http_test.py:7:9:7:14 | ControlFlowNode for params | semmle.label | ControlFlowNode for params |
+| http_test.py:7:18:7:57 | ControlFlowNode for Attribute() | semmle.label | ControlFlowNode for Attribute() |
+| http_test.py:7:40:7:56 | ControlFlowNode for Attribute | semmle.label | ControlFlowNode for Attribute |
+| http_test.py:8:9:8:19 | ControlFlowNode for input_value | semmle.label | ControlFlowNode for input_value |
+| http_test.py:8:23:8:28 | ControlFlowNode for params | semmle.label | ControlFlowNode for params |
+| http_test.py:8:23:8:47 | ControlFlowNode for Attribute() | semmle.label | ControlFlowNode for Attribute() |
+| http_test.py:12:40:12:50 | ControlFlowNode for input_value | semmle.label | ControlFlowNode for input_value |
 | wsgiref_tests.py:4:14:4:20 | ControlFlowNode for environ | semmle.label | ControlFlowNode for environ |
 | wsgiref_tests.py:6:5:6:10 | ControlFlowNode for h_name | semmle.label | ControlFlowNode for h_name |
 | wsgiref_tests.py:7:5:7:9 | ControlFlowNode for h_val | semmle.label | ControlFlowNode for h_val |
@@ -39,5 +60,6 @@ subpaths
 | flask_tests.py:20:36:20:61 | ControlFlowNode for Subscript | flask_tests.py:1:29:1:35 | ControlFlowNode for ImportMember | flask_tests.py:20:36:20:61 | ControlFlowNode for Subscript | This HTTP header is constructed from a $@. | flask_tests.py:1:29:1:35 | ControlFlowNode for ImportMember | user-provided value |
 | flask_tests.py:33:11:33:20 | ControlFlowNode for rfs_header | flask_tests.py:1:29:1:35 | ControlFlowNode for ImportMember | flask_tests.py:33:11:33:20 | ControlFlowNode for rfs_header | This HTTP header is constructed from a $@. | flask_tests.py:1:29:1:35 | ControlFlowNode for ImportMember | user-provided value |
 | flask_tests.py:35:12:35:21 | ControlFlowNode for rfs_header | flask_tests.py:1:29:1:35 | ControlFlowNode for ImportMember | flask_tests.py:35:12:35:21 | ControlFlowNode for rfs_header | This HTTP header is constructed from a $@. | flask_tests.py:1:29:1:35 | ControlFlowNode for ImportMember | user-provided value |
+| http_test.py:12:40:12:50 | ControlFlowNode for input_value | http_test.py:5:16:5:19 | ControlFlowNode for self | http_test.py:12:40:12:50 | ControlFlowNode for input_value | This HTTP header is constructed from a $@. | http_test.py:5:16:5:19 | ControlFlowNode for self | user-provided value |
 | wsgiref_tests.py:8:17:8:22 | ControlFlowNode for h_name | wsgiref_tests.py:4:14:4:20 | ControlFlowNode for environ | wsgiref_tests.py:8:17:8:22 | ControlFlowNode for h_name | This HTTP header is constructed from a $@. | wsgiref_tests.py:4:14:4:20 | ControlFlowNode for environ | user-provided value |
 | wsgiref_tests.py:8:42:8:46 | ControlFlowNode for h_val | wsgiref_tests.py:4:14:4:20 | ControlFlowNode for environ | wsgiref_tests.py:8:42:8:46 | ControlFlowNode for h_val | This HTTP header is constructed from a $@. | wsgiref_tests.py:4:14:4:20 | ControlFlowNode for environ | user-provided value |

python/ql/test/query-tests/Security/CWE-113-HeaderInjection/Tests1/http_test.py

@@ -0,0 +1,22 @@
+from http.server import HTTPServer, BaseHTTPRequestHandler
+import urllib.parse
+
+class VulnerableHandler(BaseHTTPRequestHandler):
+    def do_GET(self):
+        parsed_path = urllib.parse.urlparse(self.path)
+        params = urllib.parse.parse_qs(parsed_path.query)
+        input_value = params.get("input", [""])[0]
+        # Unsafe: Directly including user input in headers
+        self.send_response(200)
+        try:
+            self.send_header("X-Info", input_value) # BAD
+        except Exception as e:
+            print(f"[!] Header injection failed: {e}")
+        self.end_headers()
+        self.wfile.write(b"Hello world!")
+
+
+# if __name__ == "__main__":
+#     print("Serving vulnerable app on http://127.0.0.1:8080")
+#     httpd = HTTPServer(("127.0.0.1", 8080), VulnerableHandler)
+#     httpd.serve_forever()