From 80c6f082d114ce5772dd38330b528868e3914363 Mon Sep 17 00:00:00 2001
From: Owen Mansel-Chan <owen-mc@github.com>
Date: Thu, 28 May 2026 11:34:02 +0100
Subject: [PATCH] Fix TODO in `containerStep`

---
 .../new/internal/TaintTrackingPrivate.qll     | 57 ++++++++++---------
 1 file changed, 29 insertions(+), 28 deletions(-)

diff --git a/python/ql/lib/semmle/python/dataflow/new/internal/TaintTrackingPrivate.qll b/python/ql/lib/semmle/python/dataflow/new/internal/TaintTrackingPrivate.qll
index 2d2cceb73c1..de5a06eaaad 100644
--- a/python/ql/lib/semmle/python/dataflow/new/internal/TaintTrackingPrivate.qll
+++ b/python/ql/lib/semmle/python/dataflow/new/internal/TaintTrackingPrivate.qll
@@ -11,21 +11,35 @@ private import semmle.python.ApiGraphs
  */
 predicate defaultTaintSanitizer(DataFlow::Node node) { none() }
 
+/**
+ * Holds if default taint tracking should read content `contentSet` implicitly and
+ * propagate taint from a container to reads of that content.
+ */
+private predicate defaultTaintReadContent(DataFlow::ContentSet contentSet) {
+  // Tuple and dictionary content is precise, so use wildcard content sets to avoid
+  // blowing up the size of `Stage1::readSetEx` (otherwise this predicate would
+  // expand to one row per (node, distinct key or index) and the framework's
+  // read-set relation grows quadratically). `ContentSet.getAReadContent` expands
+  // these wildcards back to the specific contents when matching against stores.
+  contentSet.isAnyTupleElement()
+  or
+  contentSet.isAnyDictionaryElement()
+  or
+  // List and set element content is already imprecise, so no wildcard expansion is
+  // needed.
+  contentSet.getAStoreContent() instanceof DataFlow::ListElementContent
+  or
+  contentSet.getAStoreContent() instanceof DataFlow::SetElementContent
+}
+
 /**
  * Holds if default `TaintTracking::Configuration`s should allow implicit reads
  * of `c` at sinks and inputs to additional taint steps.
  */
 bindingset[node]
 predicate defaultImplicitTaintRead(DataFlow::Node node, DataFlow::ContentSet c) {
-  // We allow implicit reads of precise content; imprecise content has already
-  // bubbled up. We use the wildcard content sets here rather than the
-  // per-key/per-index ones to avoid blowing up the size of `Stage1::readSetEx`
-  // (otherwise this predicate would expand to one row per (node, distinct key
-  // or index) and the framework's read-set relation grows quadratically).
-  // `ContentSet.getAReadContent` expands these wildcards back to the specific
-  // contents when matching against stores.
   exists(node) and
-  (c.isAnyTupleElement() or c.isAnyDictionaryElement())
+  defaultTaintReadContent(c)
 }
 
 private module Cached {
@@ -171,28 +185,15 @@ predicate stringManipulation(DataFlow::CfgNode nodeFrom, DataFlow::CfgNode nodeT
 }
 
 /**
- * Holds if taint can flow from `nodeFrom` to `nodeTo` with a step related to containers
- * (lists/sets/dictionaries): literals, constructor invocation, methods. Note that this
- * is currently very imprecise, as an example, since we model `dict.get`, we treat any
- * `<tainted object>.get(<arg>)` will be tainted, whether it's true or not.
+ * Holds if taint can flow from `nodeFrom` to `nodeTo` with a step related to reading
+ * content from containers (lists/sets/dictionaries/tuples): subscripts, iteration,
+ * constructor invocation, methods.
  */
 predicate containerStep(DataFlow::Node nodeFrom, DataFlow::Node nodeTo) {
-  // construction by literal
-  //
-  // TODO: once we have proper flow-summary modeling, we might not need this step any
-  // longer -- but there needs to be a matching read-step for the store-step, and we
-  // don't provide that right now.
-  DataFlowPrivate::listStoreStep(nodeFrom, _, nodeTo)
-  or
-  DataFlowPrivate::setStoreStep(nodeFrom, _, nodeTo)
-  or
-  // comprehension, so there is taint-flow from `x` in `[x for x in xs]` to the
-  // resulting list of the list-comprehension.
-  //
-  // TODO: once we have proper flow-summary modeling, we might not need this step any
-  // longer -- but there needs to be a matching read-step for the store-step, and we
-  // don't provide that right now.
-  DataFlowPrivate::yieldStoreStep(nodeFrom, _, nodeTo)
+  exists(DataFlow::ContentSet contentSet |
+    DataFlowPrivate::readStep(nodeFrom, contentSet, nodeTo) and
+    defaultTaintReadContent(contentSet)
+  )
 }
 
 /**