Fixed bug that results in a false positive error under certain circumstances when passing an overloaded function as an argument to a call. This addresses #8632. (#8644)

packages/pyright-internal/src/analyzer/typeEvaluator.ts

@@ -23979,6 +23979,15 @@ export function createTypeEvaluator(
             }
 
             if (isOverloadedFunction(concreteSrcType)) {
+                // If this is the first pass of an argument assignment, skip
+                // all attempts to assign an overloaded function to a function
+                // because we probably don't have enough information to properly
+                // filter the overloads at this time. We will do this work on
+                // subsequent passes.
+                if ((flags & AssignTypeFlags.ArgAssignmentFirstPass) !== 0) {
+                    return true;
+                }
+
                 // Find all of the overloaded functions that match the parameters.
                 const overloads = OverloadedFunctionType.getOverloads(concreteSrcType);
                 const filteredOverloads: FunctionType[] = [];

packages/pyright-internal/src/tests/samples/solver39.py

@@ -0,0 +1,11 @@
+# This sample tests the case where an overloaded function is passed as
+# an argument and the overloads cannot be filtered during the first
+# pass through the arguments.
+
+from functools import reduce
+from operator import getitem
+from typing import Any
+
+
+def deep_getitem(data: dict[str, Any], attr: str) -> Any:
+    return reduce(getitem, attr.split("."), data)

packages/pyright-internal/src/tests/typeEvaluator2.test.ts

@@ -795,6 +795,12 @@ test('Solver38', () => {
     TestUtils.validateResults(analysisResults, 0);
 });
 
+test('Solver39', () => {
+    const analysisResults = TestUtils.typeAnalyzeSampleFiles(['solver39.py']);
+
+    TestUtils.validateResults(analysisResults, 0);
+});
+
 test('SolverScoring1', () => {
     const analysisResults = TestUtils.typeAnalyzeSampleFiles(['solverScoring1.py']);