Merge pull request #14120 from asgerf/dynamic/typemodel-istypeused

Dynamic: add TypeModel.isTypeUsed

python/ql/lib/semmle/python/dataflow/new/internal/DataFlowPublic.qll

@@ -9,6 +9,7 @@ import Attributes
 import LocalSources
 private import semmle.python.essa.SsaCompute
 private import semmle.python.dataflow.new.internal.ImportStar
+private import semmle.python.frameworks.data.ModelsAsData
 private import FlowSummaryImpl as FlowSummaryImpl
 private import semmle.python.frameworks.data.ModelsAsData
 
@@ -125,6 +126,13 @@ newtype TNode =
     f = any(VariableCapture::CapturedVariable v).getACapturingScope() and
     // TODO: Remove this restriction when adding proper support for captured variables in the body of the function we generate for comprehensions
     exists(TFunction(f))
+  } or
+  /** An empty, unused node type that exists to prevent unwanted dependencies on data flow nodes. */
+  TForbiddenRecursionGuard() {
+    none() and
+    // We want to prune irrelevant models before materialising data flow nodes, so types contributed
+    // directly from CodeQL must expose their pruning info without depending on data flow nodes.
+    (any(ModelInput::TypeModel tm).isTypeUsed("") implies any())
   }
 
 private import semmle.python.internal.CachedStages

python/ql/lib/semmle/python/frameworks/data/internal/ApiGraphModels.qll

@@ -168,9 +168,20 @@ module ModelInput {
    * A unit class for adding additional type model rows from CodeQL models.
    */
   class TypeModel extends Unit {
+    /**
+     * Holds if any of the other predicates in this class might have a result
+     * for the given `type`.
+     *
+     * The implementation of this predicate should not depend on `DataFlow::Node`.
+     */
+    bindingset[type]
+    predicate isTypeUsed(string type) { none() }
+
     /**
      * Gets a data-flow node that is a source of the given `type`.
      *
+     * Note that `type` should also be included in `isTypeUsed`.
+     *
      * This must not depend on API graphs, but ensures that an API node is generated for
      * the source.
      */
@@ -180,6 +191,8 @@ module ModelInput {
      * Gets a data-flow node that is a sink of the given `type`,
      * usually because it is an argument passed to a parameter of that type.
      *
+     * Note that `type` should also be included in `isTypeUsed`.
+     *
      * This must not depend on API graphs, but ensures that an API node is generated for
      * the sink.
      */
@@ -188,6 +201,8 @@ module ModelInput {
     /**
      * Gets an API node that is a source or sink of the given `type`.
      *
+     * Note that `type` should also be included in `isTypeUsed`.
+     *
      * Unlike `getASource` and `getASink`, this may depend on API graphs.
      */
     API::Node getAnApiNode(string type) { none() }
@@ -367,6 +382,8 @@ predicate isRelevantType(string type) {
   (
     Specific::isTypeUsed(type)
     or
+    any(TypeModel model).isTypeUsed(type)
+    or
     exists(TestAllModels t)
   )
   or