C++: Port implementation to CPP.

cpp/ql/lib/semmle/code/cpp/dataflow/ExternalFlow.qll

@@ -13,8 +13,8 @@
  *
  * The interpretation of a row is similar to API-graphs with a left-to-right
  * reading.
- * 1. The `namespace` column selects a package.
- * 2. The `type` column selects a type within that package.
+ * 1. The `namespace` column selects a namespace.
+ * 2. The `type` column selects a type within that namespace.
  * 3. The `subtypes` is a boolean that indicates whether to jump to an
  *    arbitrary subtype of that type. Set this to `false` if leaving the `type`
  *    blank (for example, a free function).
@@ -65,16 +65,14 @@
  *    globally applicable value-preserving step.
  */
 
-import swift
-private import internal.DataFlowDispatch
-private import internal.DataFlowPrivate
-private import internal.DataFlowPublic
+import cpp
+private import new.DataFlow
 private import internal.FlowSummaryImpl
 private import internal.FlowSummaryImpl::Public
 private import internal.FlowSummaryImpl::Private
 private import internal.FlowSummaryImpl::Private::External
-private import FlowSummary as FlowSummary
 private import codeql.mad.ModelValidation as SharedModelVal
+private import codeql.util.Unit
 
 /**
  * A unit class for adding additional source model rows.
@@ -354,13 +352,13 @@ private predicate elementSpec(
 private string paramsStringPart(Function c, int i) {
   i = -1 and result = "(" and exists(c)
   or
-  exists(int n, string p | c.getParam(n).getType().toString() = p |
+  exists(int n, string p | c.getParameter(n).getType().toString() = p |
     i = 2 * n and result = p
     or
     i = 2 * n - 1 and result = "," and n != 0
   )
   or
-  i = 2 * c.getNumberOfParams() and result = ")"
+  i = 2 * c.getNumberOfParameters() and result = ")"
 }
 
 /**
@@ -401,42 +399,42 @@ private Element interpretElement0(
       type = "" and
       matchesSignature(func, signature) and
       subtypes = false and
-      not result instanceof Method and
+      not exists(func.getDeclaringType()) and
       result = func
     )
     or
     // Member functions
-    exists(NominalTypeDecl namedTypeDecl, Decl declWithMethod, Method method |
+    exists(Class namedClass, Class classWithMethod, Function method |
       method.getName() = name and
-      method = declWithMethod.getAMember() and
-      namedTypeDecl.getFullName() = type and
+      method = classWithMethod.getAMember() and
+      namedClass.getName() = type and
       matchesSignature(method, signature) and
       result = method
     |
-      // member declared in the named type or a subtype of it (or an extension of any)
+      // member declared in the named type or a subtype of it
       subtypes = true and
-      declWithMethod.asNominalTypeDecl() = namedTypeDecl.getADerivedTypeDecl*()
+      classWithMethod = namedClass.getADerivedClass*()
       or
-      // member declared directly in the named type (or an extension of it)
+      // member declared directly in the named type
       subtypes = false and
-      declWithMethod.asNominalTypeDecl() = namedTypeDecl
+      classWithMethod = namedClass
     )
     or
-    // Fields
+    // Member variables
     signature = "" and
-    exists(NominalTypeDecl namedTypeDecl, Decl declWithField, FieldDecl field |
-      field.getName() = name and
-      field = declWithField.getAMember() and
-      namedTypeDecl.getFullName() = type and
-      result = field
+    exists(Class namedClass, Class classWithMember, MemberVariable member |
+      member.getName() = name and
+      member = classWithMember.getAMember() and
+      namedClass.getName() = type and
+      result = member
     |
       // field declared in the named type or a subtype of it (or an extension of any)
       subtypes = true and
-      declWithField.asNominalTypeDecl() = namedTypeDecl.getADerivedTypeDecl*()
+      classWithMember = namedClass.getADerivedClass*()
       or
       // field declared directly in the named type (or an extension of it)
       subtypes = false and
-      declWithField.asNominalTypeDecl() = namedTypeDecl
+      classWithMember = namedClass
     )
   )
 }
@@ -451,44 +449,6 @@ Element interpretElement(
   )
 }
 
-deprecated private predicate parseField(AccessPathToken c, Content::FieldContent f) {
-  exists(string fieldRegex, string name |
-    c.getName() = "Field" and
-    fieldRegex = "^([^.]+)$" and
-    name = c.getAnArgument().regexpCapture(fieldRegex, 1) and
-    f.getField().getName() = name
-  )
-}
-
-deprecated private predicate parseTuple(AccessPathToken c, Content::TupleContent t) {
-  c.getName() = "TupleElement" and
-  t.getIndex() = c.getAnArgument().toInt()
-}
-
-deprecated private predicate parseEnum(AccessPathToken c, Content::EnumContent e) {
-  c.getName() = "EnumElement" and
-  c.getAnArgument() = e.getSignature()
-  or
-  c.getName() = "OptionalSome" and
-  e.getSignature() = "some:0"
-}
-
-/** Holds if the specification component parses as a `Content`. */
-deprecated predicate parseContent(AccessPathToken component, Content content) {
-  parseField(component, content)
-  or
-  parseTuple(component, content)
-  or
-  parseEnum(component, content)
-  or
-  // map legacy "ArrayElement" specification components to `CollectionContent`
-  component.getName() = "ArrayElement" and
-  content instanceof Content::CollectionContent
-  or
-  component.getName() = "CollectionElement" and
-  content instanceof Content::CollectionContent
-}
-
 cached
 private module Cached {
   /**
@@ -496,7 +456,7 @@ private module Cached {
    * model.
    */
   cached
-  predicate sourceNode(Node node, string kind) {
+  predicate sourceNode(DataFlow::Node node, string kind) {
     exists(SourceSinkInterpretationInput::InterpretNode n |
       isSourceNode(n, kind) and n.asNode() = node
     )
@@ -507,7 +467,7 @@ private module Cached {
    * model.
    */
   cached
-  predicate sinkNode(Node node, string kind) {
+  predicate sinkNode(DataFlow::Node node, string kind) {
     exists(SourceSinkInterpretationInput::InterpretNode n |
       isSinkNode(n, kind) and n.asNode() = node
     )
@@ -567,7 +527,7 @@ private class NeutralCallableAdapter extends NeutralCallable {
   string provenance_;
 
   NeutralCallableAdapter() {
-    // Neutral models have not been implemented for Swift.
+    // Neutral models have not been implemented for CPP.
     none() and
     exists(this) and
     exists(kind) and

cpp/ql/lib/semmle/code/cpp/dataflow/internal/FlowSummaryImpl.qll

@@ -2,21 +2,20 @@
  * Provides classes and predicates for defining flow summaries.
  */
 
-private import swift
+private import cpp
 private import codeql.dataflow.internal.FlowSummaryImpl
 private import codeql.dataflow.internal.AccessPathSyntax as AccessPath
-private import DataFlowImplSpecific as DataFlowImplSpecific
-private import DataFlowImplSpecific::Private
-private import DataFlowImplSpecific::Public
-private import DataFlowImplCommon
-private import codeql.swift.dataflow.ExternalFlow
+private import semmle.code.cpp.ir.dataflow.internal.DataFlowPrivate
+private import semmle.code.cpp.ir.dataflow.internal.DataFlowUtil
+private import semmle.code.cpp.ir.dataflow.internal.DataFlowImplSpecific as DataFlowImplSpecific
+private import semmle.code.cpp.dataflow.ExternalFlow
 
-module Input implements InputSig<DataFlowImplSpecific::SwiftDataFlow> {
+module Input implements InputSig<DataFlowImplSpecific::CppDataFlow> {
   class SummarizedCallableBase = Function;
 
-  ArgumentPosition callbackSelfParameterPosition() { result instanceof ThisArgumentPosition }
+  ArgumentPosition callbackSelfParameterPosition() { result = TDirectPosition(-1) }
 
-  ReturnKind getStandardReturnValueKind() { result instanceof NormalReturnKind }
+  ReturnKind getStandardReturnValueKind() { result.(NormalReturnKind).getIndirectionIndex() = 0 }
 
   string encodeParameterPosition(ParameterPosition pos) { result = pos.toString() }
 
@@ -29,37 +28,40 @@ module Input implements InputSig<DataFlowImplSpecific::SwiftDataFlow> {
   }
 
   string encodeContent(ContentSet cs, string arg) {
-    exists(Content::FieldContent c |
+    exists(FieldContent c |
       cs.isSingleton(c) and
       result = "Field" and
       arg = c.getField().getName()
     )
-    or
-    exists(Content::TupleContent c |
-      cs.isSingleton(c) and
-      result = "TupleElement" and
-      arg = c.getIndex().toString()
-    )
-    or
-    exists(Content::EnumContent c, string sig |
-      cs.isSingleton(c) and
-      sig = c.getSignature()
-    |
-      if sig = "some:0"
-      then
-        result = "OptionalSome" and
-        arg = ""
-      else (
-        result = "EnumElement" and
-        arg = sig
-      )
-    )
-    or
-    exists(Content::CollectionContent c |
-      cs.isSingleton(c) and
-      result = "CollectionElement" and
-      arg = ""
-    )
+    /*
+     * or
+     *    exists(Content::TupleContent c |
+     *      cs.isSingleton(c) and
+     *      result = "TupleElement" and
+     *      arg = c.getIndex().toString()
+     *    )
+     *    or
+     *    exists(Content::EnumContent c, string sig |
+     *      cs.isSingleton(c) and
+     *      sig = c.getSignature()
+     *    |
+     *      if sig = "some:0"
+     *      then
+     *        result = "OptionalSome" and
+     *        arg = ""
+     *      else (
+     *        result = "EnumElement" and
+     *        arg = sig
+     *      )
+     *    )
+     *    or
+     *    exists(Content::CollectionContent c |
+     *      cs.isSingleton(c) and
+     *      result = "CollectionElement" and
+     *      arg = ""
+     *    )
+     */
+
     }
 
   string encodeWithoutContent(ContentSet c, string arg) {
@@ -68,24 +70,12 @@ module Input implements InputSig<DataFlowImplSpecific::SwiftDataFlow> {
 
   string encodeWithContent(ContentSet c, string arg) { result = "WithContent" + c and arg = "" }
 
-  bindingset[token]
-  ContentSet decodeUnknownContent(AccessPath::AccessPathTokenBase token) {
-    // map legacy "ArrayElement" specification components to `CollectionContent`
-    token.getName() = "ArrayElement" and
-    result.isSingleton(any(Content::CollectionContent c))
-    or
-    token.getName() = "CollectionElement" and
-    result.isSingleton(any(Content::CollectionContent c))
-  }
-
   bindingset[token]
   ParameterPosition decodeUnknownParameterPosition(AccessPath::AccessPathTokenBase token) {
     // needed to support `Argument[x..y]` ranges and `Argument[-1]`
     token.getName() = "Argument" and
     exists(int pos | pos = AccessPath::parseInt(token.getAnArgument()) |
-      result.(PositionalParameterPosition).getIndex() = pos
-      or
-      pos = -1 and result instanceof ThisParameterPosition
+      result = TDirectPosition(pos)
     )
   }
 
@@ -94,24 +84,21 @@ module Input implements InputSig<DataFlowImplSpecific::SwiftDataFlow> {
     // needed to support `Parameter[x..y]` ranges and `Parameter[-1]`
     token.getName() = "Parameter" and
     exists(int pos | pos = AccessPath::parseInt(token.getAnArgument()) |
-      result.(PositionalArgumentPosition).getIndex() = pos
-      or
-      pos = -1 and
-      result instanceof ThisArgumentPosition
+      result = TDirectPosition(pos)
     )
   }
 }
 
-private import Make<DataFlowImplSpecific::SwiftDataFlow, Input> as Impl
+/*private*/ import Make<DataFlowImplSpecific::CppDataFlow, Input> as Impl
 
 private module StepsInput implements Impl::Private::StepsInputSig {
-  DataFlowCall getACall(Public::SummarizedCallable sc) { result.asCall().getStaticTarget() = sc }
+  DataFlowCall getACall(Public::SummarizedCallable sc) { result.getStaticCallTarget() = sc }
 }
 
 module SourceSinkInterpretationInput implements
   Impl::Private::External::SourceSinkInterpretationInputSig<Location>
 {
-  class Element = AstNode;
+  class Element = Element;
 
   class SourceOrSinkElement = Element;
 
@@ -158,10 +145,12 @@ module SourceSinkInterpretationInput implements
     DataFlowCall asCall() { this = TDataFlowCall_(result) }
 
     /** Gets the callable that this node corresponds to, if any. */
-    DataFlowCallable asCallable() { result.getUnderlyingCallable() = this.asElement() }
+    DataFlowCallable asCallable() { result.(Function) = this.asElement() }
 
     /** Gets the target of this call, if any. */
-    Element getCallTarget() { result = this.asCall().asCall().getStaticTarget() }
+    Element getCallTarget() {
+      result = this.asNode().asExpr().(Call).getTarget()
+    }
 
     /** Gets a textual representation of this node. */
     string toString() {
@@ -186,31 +175,31 @@ module SourceSinkInterpretationInput implements
   bindingset[c]
   predicate interpretOutput(string c, InterpretNode mid, InterpretNode node) {
     // Allow fields to be picked as output nodes.
-    exists(Node n, AstNode ast |
+    exists(Node n, Element ast |
       n = node.asNode() and
       ast = mid.asElement()
     |
       c = "" and
-      n.asExpr().(MemberRefExpr).getMember() = ast
+      n.asExpr().(VariableAccess).getTarget() = ast
     )
   }
 
   /** Provides additional source specification logic. */
   bindingset[c]
   predicate interpretInput(string c, InterpretNode mid, InterpretNode node) {
-    exists(Node n, AstNode ast, MemberRefExpr e |
+    exists(Node n, Element ast, VariableAccess e |
       n = node.asNode() and
       ast = mid.asElement() and
-      e.getMember() = ast
+      e.getTarget() = ast
     |
       // Allow fields to be picked as input nodes.
       c = "" and
-      e.getBase() = n.asExpr()
+      e.getQualifier() = n.asExpr()
       or
       // Allow post update nodes to be picked as input nodes when the `input` column
       // of the row is `PostUpdate`.
       c = "PostUpdate" and
-      e.getBase() = n.(PostUpdateNode).getPreUpdateNode().asExpr()
+      e.getQualifier() = n.(PostUpdateNode).getPreUpdateNode().asExpr()
     )
   }
 }

cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowPrivate.qll

@@ -458,11 +458,13 @@ class ReturnKind extends TReturnKind {
   abstract string toString();
 }
 
-private class NormalReturnKind extends ReturnKind, TNormalReturnKind {
+class NormalReturnKind extends ReturnKind, TNormalReturnKind {
   int index;
 
   NormalReturnKind() { this = TNormalReturnKind(index) }
 
+  int getIndirectionIndex() { result = index }
+
   override string toString() { result = "indirect return" }
 }
 

cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowUtil.qll

@@ -2330,6 +2330,12 @@ class UnionContent extends Content, TUnionContent {
  * stored into (`getAStoreContent`) or read from (`getAReadContent`).
  */
 class ContentSet instanceof Content {
+  /**
+   * Holds if this content set is the singleton `{c}`. At present, this is
+   * the only kind of content set supported in C/C++.
+   */
+  predicate isSingleton(Content c) { this = c }
+
   /** Gets a content that may be stored into when storing into this set. */
   Content getAStoreContent() { result = this }
 

cpp/ql/lib/semmle/code/cpp/models/interfaces/FlowSource.qll

@@ -9,6 +9,7 @@
 import cpp
 import FunctionInputsAndOutputs
 import semmle.code.cpp.models.Models
+import semmle.code.cpp.dataflow.ExternalFlow
 
 /**
  * A library function that returns data that may be read from a network connection.

cpp/ql/lib/semmle/code/cpp/security/FlowSources.qll

@@ -20,6 +20,9 @@ abstract class RemoteFlowSource extends FlowSource { }
 /** A data flow source of local user input. */
 abstract class LocalFlowSource extends FlowSource { }
 
+/**
+ * A remote data flow source that is defined through a `RemoteFlowSourceFunction` model.
+ */
 private class RemoteModelSource extends RemoteFlowSource {
   string sourceType;
 
@@ -34,6 +37,9 @@ private class RemoteModelSource extends RemoteFlowSource {
   override string getSourceType() { result = sourceType }
 }
 
+/**
+ * A local data flow source that is defined through a `LocalFlowSourceFunction` model.
+ */
 private class LocalModelSource extends LocalFlowSource {
   string sourceType;
 
@@ -48,6 +54,9 @@ private class LocalModelSource extends LocalFlowSource {
   override string getSourceType() { result = sourceType }
 }
 
+/**
+ * A local data flow source that the `argv` parameter to `main`.
+ */
 private class ArgvSource extends LocalFlowSource {
   ArgvSource() {
     exists(Function main, Parameter argv |
@@ -60,6 +69,24 @@ private class ArgvSource extends LocalFlowSource {
   override string getSourceType() { result = "a command-line argument" }
 }
 
+/**
+ * A remote data flow source that is defined through 'models as data'.
+ */
+private class ExternalRemoteFlowSource extends RemoteFlowSource {
+  ExternalRemoteFlowSource() { sourceNode(this, "remote") }
+
+  override string getSourceType() { result = "external" }
+}
+
+/**
+ * A local data flow source that is defined through 'models as data'.
+ */
+private class ExternalLocalFlowSource extends LocalFlowSource {
+  ExternalLocalFlowSource() { sourceNode(this, "local") }
+
+  override string getSourceType() { result = "external" }
+}
+
 /** A remote data flow sink. */
 abstract class RemoteFlowSink extends DataFlow::Node {
   /** Gets a string that describes the type of this flow sink. */

shared/dataflow/codeql/dataflow/internal/FlowSummaryImpl.qll

@@ -1588,7 +1588,8 @@ module Make<DF::InputSig DataFlowLang, InputSig<DataFlowLang> Input> {
           SourceSinkAccessPath output, int n, InterpretNode ref, InterpretNode node
         ) {
           sourceElementRef(ref, output, _) and
-          n = 0 and
+          //n = 0 and
+          n = [0,1] and // TODO: fix this, there's no good reason for it.
           (
             if output = ""
             then