From caf7182c6f8d74c9018cbf8675beb4da10a2d0bd Mon Sep 17 00:00:00 2001
From: MrEx3cut0r <legendary.killtell@gmail.com>
Date: Mon, 9 Mar 2026 16:28:48 +0300
Subject: [PATCH] gh-145668: Optimize FOR_ITER virtual iterators for bytes,
 bytearray, and str

---
 Python/bytecodes.c |  6 +++--
 Python/ceval.c     | 55 +++++++++++++++++++++++++++++++++++++++++-----
 2 files changed, 54 insertions(+), 7 deletions(-)

diff --git a/Python/bytecodes.c b/Python/bytecodes.c
index 8a748fec9e4201..d0fc0ca0928c08 100644
--- a/Python/bytecodes.c
+++ b/Python/bytecodes.c
@@ -3277,8 +3277,10 @@ dummy_func(
             _Py_GatherStats_GetIter(iterable);
             #endif
             PyTypeObject *tp = PyStackRef_TYPE(iterable);
-            if (tp == &PyTuple_Type || tp == &PyList_Type) {
-                /* Leave iterable on stack and pushed tagged 0 */
+            if (tp == &PyTuple_Type || tp == &PyList_Type ||
+                tp == &PyBytes_Type || tp == &PyByteArray_Type ||
+                tp == &PyUnicode_Type) {
+                /* Leave iterable on stack and push tagged 0 */
                 iter = iterable;
                 DEAD(iterable);
                 index_or_null = PyStackRef_TagInt(0);
diff --git a/Python/ceval.c b/Python/ceval.c
index 1e5142f4b456a1..e10c1d8a883949 100644
--- a/Python/ceval.c
+++ b/Python/ceval.c
@@ -3704,8 +3704,8 @@ static _PyStackRef
 foriter_next(PyObject *seq, _PyStackRef index)
 {
     assert(PyStackRef_IsTaggedInt(index));
-    assert(PyTuple_CheckExact(seq) || PyList_CheckExact(seq));
     intptr_t i = PyStackRef_UntagInt(index);
+
     if (PyTuple_CheckExact(seq)) {
         size_t size = PyTuple_GET_SIZE(seq);
         if ((size_t)i >= size) {
@@ -3713,11 +3713,56 @@ foriter_next(PyObject *seq, _PyStackRef index)
         }
         return PyStackRef_FromPyObjectNew(PyTuple_GET_ITEM(seq, i));
     }
-    PyObject *item = _PyList_GetItemRef((PyListObject *)seq, i);
-    if (item == NULL) {
-        return PyStackRef_NULL;
+
+    if (PyList_CheckExact(seq)) {
+        PyObject *item = _PyList_GetItemRef((PyListObject *)seq, i);
+        if (item == NULL) {
+            return PyStackRef_NULL;
+        }
+        return PyStackRef_FromPyObjectSteal(item);
     }
-    return PyStackRef_FromPyObjectSteal(item);
+
+    if (PyBytes_CheckExact(seq)) {
+        Py_ssize_t size = PyBytes_GET_SIZE(seq);
+        if (i < 0 || i >= size) {
+            return PyStackRef_NULL;
+        }
+        unsigned char ch = (unsigned char)PyBytes_AS_STRING(seq)[i];
+        PyObject *item = PyLong_FromUnsignedLong((unsigned long)ch);
+        if (item == NULL) {
+            return PyStackRef_ERROR;
+        }
+        return PyStackRef_FromPyObjectSteal(item);
+    }
+
+    if (PyByteArray_CheckExact(seq)) {
+        Py_ssize_t size = PyByteArray_GET_SIZE(seq);
+        if (i < 0 || i >= size) {
+            return PyStackRef_NULL;
+        }
+        unsigned char ch = (unsigned char)PyByteArray_AS_STRING(seq)[i];
+        PyObject *item = PyLong_FromUnsignedLong((unsigned long)ch);
+        if (item == NULL) {
+            return PyStackRef_ERROR;
+        }
+        return PyStackRef_FromPyObjectSteal(item);
+    }
+
+    if (PyUnicode_CheckExact(seq)) {
+        Py_ssize_t size = PyUnicode_GET_LENGTH(seq);
+        if (i < 0 || i >= size) {
+            return PyStackRef_NULL;
+        }
+        // Iteration over str yields 1-character substrings.
+        PyObject *item = PyUnicode_Substring(seq, i, i + 1);
+        if (item == NULL) {
+            return PyStackRef_ERROR;
+        }
+        return PyStackRef_FromPyObjectSteal(item);
+    }
+
+    // Fallback: use the iterator protocol for unsupported types.
+    return PyStackRef_ERROR;
 }
 
 _PyStackRef _PyForIter_VirtualIteratorNext(PyThreadState* tstate, _PyInterpreterFrame* frame, _PyStackRef iter, _PyStackRef* index_ptr)