bpo-42128: Structural Pattern Matching (PEP 634) (GH-22917)

Co-authored-by: Guido van Rossum <guido@python.org>
Co-authored-by: Talin <viridia@gmail.com>
Co-authored-by: Pablo Galindo <pablogsal@gmail.com>

Python/ast.c

@@ -309,6 +309,14 @@ validate_expr(expr_ty exp, expr_context_ty ctx)
         return validate_exprs(exp->v.Tuple.elts, ctx, 0);
     case NamedExpr_kind:
         return validate_expr(exp->v.NamedExpr.value, Load);
+    case MatchAs_kind:
+        PyErr_SetString(PyExc_ValueError,
+                        "MatchAs is only valid in match_case patterns");
+        return 0;
+    case MatchOr_kind:
+        PyErr_SetString(PyExc_ValueError,
+                        "MatchOr is only valid in match_case patterns");
+        return 0;
     /* This last case doesn't have any checking. */
     case Name_kind:
         return 1;
@@ -317,6 +325,13 @@ validate_expr(expr_ty exp, expr_context_ty ctx)
     return 0;
 }
 
+static int
+validate_pattern(expr_ty p)
+{
+    // Coming soon (thanks Batuhan)!
+    return 1;
+}
+
 static int
 _validate_nonempty_seq(asdl_seq *seq, const char *what, const char *owner)
 {
@@ -415,6 +430,20 @@ validate_stmt(stmt_ty stmt)
                 return 0;
         }
         return validate_body(stmt->v.AsyncWith.body, "AsyncWith");
+    case Match_kind:
+        if (!validate_expr(stmt->v.Match.subject, Load)
+            || !validate_nonempty_seq(stmt->v.Match.cases, "cases", "Match")) {
+            return 0;
+        }
+        for (i = 0; i < asdl_seq_LEN(stmt->v.Match.cases); i++) {
+            match_case_ty m = asdl_seq_GET(stmt->v.Match.cases, i);
+            if (!validate_pattern(m->pattern)
+                || (m->guard && !validate_expr(m->guard, Load))
+                || !validate_body(m->body, "match_case")) {
+                return 0;
+            }
+        }
+        return 1;
     case Raise_kind:
         if (stmt->v.Raise.exc) {
             return validate_expr(stmt->v.Raise.exc, Load) &&

Python/ast_opt.c

@@ -408,6 +408,9 @@ static int astfold_comprehension(comprehension_ty node_, PyArena *ctx_, _PyASTOp
 static int astfold_keyword(keyword_ty node_, PyArena *ctx_, _PyASTOptimizeState *state);
 static int astfold_withitem(withitem_ty node_, PyArena *ctx_, _PyASTOptimizeState *state);
 static int astfold_excepthandler(excepthandler_ty node_, PyArena *ctx_, _PyASTOptimizeState *state);
+static int astfold_match_case(match_case_ty node_, PyArena *ctx_, _PyASTOptimizeState *state);
+static int astfold_pattern(expr_ty node_, PyArena *ctx_, _PyASTOptimizeState *state);
+
 #define CALL(FUNC, TYPE, ARG) \
     if (!FUNC((ARG), ctx_, state)) \
         return 0;
@@ -590,6 +593,10 @@ astfold_expr(expr_ty node_, PyArena *ctx_, _PyASTOptimizeState *state)
     case Constant_kind:
         // Already a constant, nothing further to do
         break;
+    case MatchAs_kind:
+    case MatchOr_kind:
+        // These can't occur outside of patterns.
+        Py_UNREACHABLE();
     // No default case, so the compiler will emit a warning if new expression
     // kinds are added without being handled here
     }
@@ -709,6 +716,10 @@ astfold_stmt(stmt_ty node_, PyArena *ctx_, _PyASTOptimizeState *state)
     case Expr_kind:
         CALL(astfold_expr, expr_ty, node_->v.Expr.value);
         break;
+    case Match_kind:
+        CALL(astfold_expr, expr_ty, node_->v.Match.subject);
+        CALL_SEQ(astfold_match_case, match_case, node_->v.Match.cases);
+        break;
     // The following statements don't contain any subexpressions to be folded
     case Import_kind:
     case ImportFrom_kind:
@@ -746,6 +757,125 @@ astfold_withitem(withitem_ty node_, PyArena *ctx_, _PyASTOptimizeState *state)
     return 1;
 }
 
+static int
+astfold_pattern_negative(expr_ty node_, PyArena *ctx_, _PyASTOptimizeState *state)
+{
+    assert(node_->kind == UnaryOp_kind);
+    assert(node_->v.UnaryOp.op == USub);
+    assert(node_->v.UnaryOp.operand->kind == Constant_kind);
+    PyObject *value = node_->v.UnaryOp.operand->v.Constant.value;
+    assert(PyComplex_CheckExact(value) ||
+           PyFloat_CheckExact(value) ||
+           PyLong_CheckExact(value));
+    PyObject *negated = PyNumber_Negative(value);
+    if (negated == NULL) {
+        return 0;
+    }
+    assert(PyComplex_CheckExact(negated) ||
+           PyFloat_CheckExact(negated) ||
+           PyLong_CheckExact(negated));
+    return make_const(node_, negated, ctx_);
+}
+
+static int
+astfold_pattern_complex(expr_ty node_, PyArena *ctx_, _PyASTOptimizeState *state)
+{
+    expr_ty left = node_->v.BinOp.left;
+    expr_ty right = node_->v.BinOp.right;
+    if (left->kind == UnaryOp_kind) {
+        CALL(astfold_pattern_negative, expr_ty, left);
+    }
+    assert(left->kind = Constant_kind);
+    assert(right->kind = Constant_kind);
+    // LHS must be real, RHS must be imaginary:
+    if (!(PyFloat_CheckExact(left->v.Constant.value) ||
+          PyLong_CheckExact(left->v.Constant.value)) ||
+        !PyComplex_CheckExact(right->v.Constant.value))
+    {
+        // Not actually valid, but it's the compiler's job to complain:
+        return 1;
+    }
+    PyObject *new;
+    if (node_->v.BinOp.op == Add) {
+        new = PyNumber_Add(left->v.Constant.value, right->v.Constant.value);
+    }
+    else {
+        assert(node_->v.BinOp.op == Sub);
+        new = PyNumber_Subtract(left->v.Constant.value, right->v.Constant.value);
+    }
+    if (new == NULL) {
+        return 0;
+    }
+    assert(PyComplex_CheckExact(new));
+    return make_const(node_, new, ctx_);
+}
+
+static int
+astfold_pattern_keyword(keyword_ty node_, PyArena *ctx_, _PyASTOptimizeState *state)
+{
+    CALL(astfold_pattern, expr_ty, node_->value);
+    return 1;
+}
+
+static int
+astfold_pattern(expr_ty node_, PyArena *ctx_, _PyASTOptimizeState *state)
+{
+    // Don't blindly optimize the pattern as an expr; it plays by its own rules!
+    // Currently, this is only used to form complex/negative numeric constants.
+    switch (node_->kind) {
+        case Attribute_kind:
+            break;
+        case BinOp_kind:
+            CALL(astfold_pattern_complex, expr_ty, node_);
+            break;
+        case Call_kind:
+            CALL_SEQ(astfold_pattern, expr, node_->v.Call.args);
+            CALL_SEQ(astfold_pattern_keyword, keyword, node_->v.Call.keywords);
+            break;
+        case Constant_kind:
+            break;
+        case Dict_kind:
+            CALL_SEQ(astfold_pattern, expr, node_->v.Dict.keys);
+            CALL_SEQ(astfold_pattern, expr, node_->v.Dict.values);
+            break;
+        // Not actually valid, but it's the compiler's job to complain:
+        case JoinedStr_kind:
+            break;
+        case List_kind:
+            CALL_SEQ(astfold_pattern, expr, node_->v.List.elts);
+            break;
+        case MatchAs_kind:
+            CALL(astfold_pattern, expr_ty, node_->v.MatchAs.pattern);
+            break;
+        case MatchOr_kind:
+            CALL_SEQ(astfold_pattern, expr, node_->v.MatchOr.patterns);
+            break;
+        case Name_kind:
+            break;
+        case Starred_kind:
+            CALL(astfold_pattern, expr_ty, node_->v.Starred.value);
+            break;
+        case Tuple_kind:
+            CALL_SEQ(astfold_pattern, expr, node_->v.Tuple.elts);
+            break;
+        case UnaryOp_kind:
+            CALL(astfold_pattern_negative, expr_ty, node_);
+            break;
+        default:
+            Py_UNREACHABLE();
+    }
+    return 1;
+}
+
+static int
+astfold_match_case(match_case_ty node_, PyArena *ctx_, _PyASTOptimizeState *state)
+{
+    CALL(astfold_pattern, expr_ty, node_->pattern);
+    CALL_OPT(astfold_expr, expr_ty, node_->guard);
+    CALL_SEQ(astfold_stmt, stmt, node_->body);
+    return 1;
+}
+
 #undef CALL
 #undef CALL_OPT
 #undef CALL_SEQ

Python/ceval.c

@@ -880,6 +880,230 @@ _Py_CheckRecursiveCall(PyThreadState *tstate, const char *where)
     return 0;
 }
 
+
+// PEP 634: Structural Pattern Matching
+
+
+// Return a tuple of values corresponding to keys, with error checks for
+// duplicate/missing keys.
+static PyObject*
+match_keys(PyThreadState *tstate, PyObject *map, PyObject *keys)
+{
+    assert(PyTuple_CheckExact(keys));
+    Py_ssize_t nkeys = PyTuple_GET_SIZE(keys);
+    if (!nkeys) {
+        // No keys means no items.
+        return PyTuple_New(0);
+    }
+    PyObject *seen = NULL;
+    PyObject *dummy = NULL;
+    PyObject *values = NULL;
+    // We use the two argument form of map.get(key, default) for two reasons:
+    // - Atomically check for a key and get its value without error handling.
+    // - Don't cause key creation or resizing in dict subclasses like
+    //   collections.defaultdict that define __missing__ (or similar).
+    _Py_IDENTIFIER(get);
+    PyObject *get = _PyObject_GetAttrId(map, &PyId_get);
+    if (get == NULL) {
+        goto fail;
+    }
+    seen = PySet_New(NULL);
+    if (seen == NULL) {
+        goto fail;
+    }
+    // dummy = object()
+    dummy = _PyObject_CallNoArg((PyObject *)&PyBaseObject_Type);
+    if (dummy == NULL) {
+        goto fail;
+    }
+    values = PyList_New(0);
+    if (values == NULL) {
+        goto fail;
+    }
+    for (Py_ssize_t i = 0; i < nkeys; i++) {
+        PyObject *key = PyTuple_GET_ITEM(keys, i);
+        if (PySet_Contains(seen, key) || PySet_Add(seen, key)) {
+            if (!_PyErr_Occurred(tstate)) {
+                // Seen it before!
+                _PyErr_Format(tstate, PyExc_ValueError,
+                              "mapping pattern checks duplicate key (%R)", key);
+            }
+            goto fail;
+        }
+        PyObject *value = PyObject_CallFunctionObjArgs(get, key, dummy, NULL);
+        if (value == NULL) {
+            goto fail;
+        }
+        if (value == dummy) {
+            // key not in map!
+            Py_DECREF(value);
+            Py_DECREF(values);
+            // Return None:
+            Py_INCREF(Py_None);
+            values = Py_None;
+            goto done;
+        }
+        PyList_Append(values, value);
+        Py_DECREF(value);
+    }
+    Py_SETREF(values, PyList_AsTuple(values));
+    // Success:
+done:
+    Py_DECREF(get);
+    Py_DECREF(seen);
+    Py_DECREF(dummy);
+    return values;
+fail:
+    Py_XDECREF(get);
+    Py_XDECREF(seen);
+    Py_XDECREF(dummy);
+    Py_XDECREF(values);
+    return NULL;
+}
+
+// Extract a named attribute from the subject, with additional bookkeeping to
+// raise TypeErrors for repeated lookups. On failure, return NULL (with no
+// error set). Use _PyErr_Occurred(tstate) to disambiguate.
+static PyObject*
+match_class_attr(PyThreadState *tstate, PyObject *subject, PyObject *type,
+                 PyObject *name, PyObject *seen)
+{
+    assert(PyUnicode_CheckExact(name));
+    assert(PySet_CheckExact(seen));
+    if (PySet_Contains(seen, name) || PySet_Add(seen, name)) {
+        if (!_PyErr_Occurred(tstate)) {
+            // Seen it before!
+            _PyErr_Format(tstate, PyExc_TypeError,
+                          "%s() got multiple sub-patterns for attribute %R",
+                          ((PyTypeObject*)type)->tp_name, name);
+        }
+        return NULL;
+    }
+    PyObject *attr = PyObject_GetAttr(subject, name);
+    if (attr == NULL && _PyErr_ExceptionMatches(tstate, PyExc_AttributeError)) {
+        _PyErr_Clear(tstate);
+    }
+    return attr;
+}
+
+// On success (match), return a tuple of extracted attributes. On failure (no
+// match), return NULL. Use _PyErr_Occurred(tstate) to disambiguate.
+static PyObject*
+match_class(PyThreadState *tstate, PyObject *subject, PyObject *type,
+            Py_ssize_t nargs, PyObject *kwargs)
+{
+    if (!PyType_Check(type)) {
+        const char *e = "called match pattern must be a type";
+        _PyErr_Format(tstate, PyExc_TypeError, e);
+        return NULL;
+    }
+    assert(PyTuple_CheckExact(kwargs));
+    // First, an isinstance check:
+    if (PyObject_IsInstance(subject, type) <= 0) {
+        return NULL;
+    }
+    // So far so good:
+    PyObject *seen = PySet_New(NULL);
+    if (seen == NULL) {
+        return NULL;
+    }
+    PyObject *attrs = PyList_New(0);
+    if (attrs == NULL) {
+        Py_DECREF(seen);
+        return NULL;
+    }
+    // NOTE: From this point on, goto fail on failure:
+    PyObject *match_args = NULL;
+    // First, the positional subpatterns:
+    if (nargs) {
+        int match_self = 0;
+        match_args = PyObject_GetAttrString(type, "__match_args__");
+        if (match_args) {
+            if (PyList_CheckExact(match_args)) {
+                Py_SETREF(match_args, PyList_AsTuple(match_args));
+            }
+            if (match_args == NULL) {
+                goto fail;
+            }
+            if (!PyTuple_CheckExact(match_args)) {
+                const char *e = "%s.__match_args__ must be a list or tuple "
+                                "(got %s)";
+                _PyErr_Format(tstate, PyExc_TypeError, e,
+                              ((PyTypeObject *)type)->tp_name,
+                              Py_TYPE(match_args)->tp_name);
+                goto fail;
+            }
+        }
+        else if (_PyErr_ExceptionMatches(tstate, PyExc_AttributeError)) {
+            _PyErr_Clear(tstate);
+            // _Py_TPFLAGS_MATCH_SELF is only acknowledged if the type does not
+            // define __match_args__. This is natural behavior for subclasses:
+            // it's as if __match_args__ is some "magic" value that is lost as
+            // soon as they redefine it.
+            match_args = PyTuple_New(0);
+            match_self = PyType_HasFeature((PyTypeObject*)type,
+                                            _Py_TPFLAGS_MATCH_SELF);
+        }
+        else {
+            goto fail;
+        }
+        assert(PyTuple_CheckExact(match_args));
+        Py_ssize_t allowed = match_self ? 1 : PyTuple_GET_SIZE(match_args);
+        if (allowed < nargs) {
+            const char *plural = (allowed == 1) ? "" : "s";
+            _PyErr_Format(tstate, PyExc_TypeError,
+                          "%s() accepts %d positional sub-pattern%s (%d given)",
+                          ((PyTypeObject*)type)->tp_name,
+                          allowed, plural, nargs);
+            goto fail;
+        }
+        if (match_self) {
+            // Easy. Copy the subject itself, and move on to kwargs.
+            PyList_Append(attrs, subject);
+        }
+        else {
+            for (Py_ssize_t i = 0; i < nargs; i++) {
+                PyObject *name = PyTuple_GET_ITEM(match_args, i);
+                if (!PyUnicode_CheckExact(name)) {
+                    _PyErr_Format(tstate, PyExc_TypeError,
+                                  "__match_args__ elements must be strings "
+                                  "(got %s)", Py_TYPE(name)->tp_name);
+                    goto fail;
+                }
+                PyObject *attr = match_class_attr(tstate, subject, type, name,
+                                                  seen);
+                if (attr == NULL) {
+                    goto fail;
+                }
+                PyList_Append(attrs, attr);
+                Py_DECREF(attr);
+            }
+        }
+        Py_CLEAR(match_args);
+    }
+    // Finally, the keyword subpatterns:
+    for (Py_ssize_t i = 0; i < PyTuple_GET_SIZE(kwargs); i++) {
+        PyObject *name = PyTuple_GET_ITEM(kwargs, i);
+        PyObject *attr = match_class_attr(tstate, subject, type, name, seen);
+        if (attr == NULL) {
+            goto fail;
+        }
+        PyList_Append(attrs, attr);
+        Py_DECREF(attr);
+    }
+    Py_SETREF(attrs, PyList_AsTuple(attrs));
+    Py_DECREF(seen);
+    return attrs;
+fail:
+    // We really don't care whether an error was raised or not... that's our
+    // caller's problem. All we know is that the match failed.
+    Py_XDECREF(match_args);
+    Py_DECREF(seen);
+    Py_DECREF(attrs);
+    return NULL;
+}
+
+
 static int do_raise(PyThreadState *tstate, PyObject *exc, PyObject *cause);
 static int unpack_iterable(PyThreadState *, PyObject *, int, int, PyObject **);
 
@@ -3618,6 +3842,164 @@ main_loop:
 #endif
         }
 
+        case TARGET(GET_LEN): {
+            // PUSH(len(TOS))
+            Py_ssize_t len_i = PyObject_Length(TOP());
+            if (len_i < 0) {
+                goto error;
+            }
+            PyObject *len_o = PyLong_FromSsize_t(len_i);
+            if (len_o == NULL) {
+                goto error;
+            }
+            PUSH(len_o);
+            DISPATCH();
+        }
+
+        case TARGET(MATCH_CLASS): {
+            // Pop TOS. On success, set TOS to True and TOS1 to a tuple of
+            // attributes. On failure, set TOS to False.
+            PyObject *names = POP();
+            PyObject *type = TOP();
+            PyObject *subject = SECOND();
+            assert(PyTuple_CheckExact(names));
+            PyObject *attrs = match_class(tstate, subject, type, oparg, names);
+            Py_DECREF(names);
+            if (attrs) {
+                // Success!
+                assert(PyTuple_CheckExact(attrs));
+                Py_DECREF(subject);
+                SET_SECOND(attrs);
+            }
+            else if (_PyErr_Occurred(tstate)) {
+                goto error;
+            }
+            Py_DECREF(type);
+            SET_TOP(PyBool_FromLong(!!attrs));
+            DISPATCH();
+        }
+
+        case TARGET(MATCH_MAPPING): {
+            // PUSH(isinstance(TOS, _collections_abc.Mapping))
+            PyObject *subject = TOP();
+            // Fast path for dicts:
+            if (PyDict_Check(subject)) {
+                Py_INCREF(Py_True);
+                PUSH(Py_True);
+                DISPATCH();
+            }
+            // Lazily import _collections_abc.Mapping, and keep it handy on the
+            // PyInterpreterState struct (it gets cleaned up at exit):
+            PyInterpreterState *interp = PyInterpreterState_Get();
+            if (interp->map_abc == NULL) {
+                PyObject *abc = PyImport_ImportModule("_collections_abc");
+                if (abc == NULL) {
+                    goto error;
+                }
+                interp->map_abc = PyObject_GetAttrString(abc, "Mapping");
+                if (interp->map_abc == NULL) {
+                    goto error;
+                }
+            }
+            int match = PyObject_IsInstance(subject, interp->map_abc);
+            if (match < 0) {
+                goto error;
+            }
+            PUSH(PyBool_FromLong(match));
+            DISPATCH();
+        }
+
+        case TARGET(MATCH_SEQUENCE): {
+            // PUSH(not isinstance(TOS, (bytearray, bytes, str))
+            //      and isinstance(TOS, _collections_abc.Sequence))
+            PyObject *subject = TOP();
+            // Fast path for lists and tuples:
+            if (PyType_FastSubclass(Py_TYPE(subject),
+                                    Py_TPFLAGS_LIST_SUBCLASS |
+                                    Py_TPFLAGS_TUPLE_SUBCLASS))
+            {
+                Py_INCREF(Py_True);
+                PUSH(Py_True);
+                DISPATCH();
+            }
+            // Bail on some possible Sequences that we intentionally exclude:
+            if (PyType_FastSubclass(Py_TYPE(subject),
+                                    Py_TPFLAGS_BYTES_SUBCLASS |
+                                    Py_TPFLAGS_UNICODE_SUBCLASS) ||
+                PyByteArray_Check(subject))
+            {
+                Py_INCREF(Py_False);
+                PUSH(Py_False);
+                DISPATCH();
+            }
+            // Lazily import _collections_abc.Sequence, and keep it handy on the
+            // PyInterpreterState struct (it gets cleaned up at exit):
+            PyInterpreterState *interp = PyInterpreterState_Get();
+            if (interp->seq_abc == NULL) {
+                PyObject *abc = PyImport_ImportModule("_collections_abc");
+                if (abc == NULL) {
+                    goto error;
+                }
+                interp->seq_abc = PyObject_GetAttrString(abc, "Sequence");
+                if (interp->seq_abc == NULL) {
+                    goto error;
+                }
+            }
+            int match = PyObject_IsInstance(subject, interp->seq_abc);
+            if (match < 0) {
+                goto error;
+            }
+            PUSH(PyBool_FromLong(match));
+            DISPATCH();
+        }
+
+        case TARGET(MATCH_KEYS): {
+            // On successful match for all keys, PUSH(values) and PUSH(True).
+            // Otherwise, PUSH(None) and PUSH(False).
+            PyObject *keys = TOP();
+            PyObject *subject = SECOND();
+            PyObject *values_or_none = match_keys(tstate, subject, keys);
+            if (values_or_none == NULL) {
+                goto error;
+            }
+            PUSH(values_or_none);
+            if (values_or_none == Py_None) {
+                Py_INCREF(Py_False);
+                PUSH(Py_False);
+                DISPATCH();
+            }
+            assert(PyTuple_CheckExact(values_or_none));
+            Py_INCREF(Py_True);
+            PUSH(Py_True);
+            DISPATCH();
+        }
+
+        case TARGET(COPY_DICT_WITHOUT_KEYS): {
+            // rest = dict(TOS1)
+            // for key in TOS:
+            //     del rest[key]
+            // SET_TOP(rest)
+            PyObject *keys = TOP();
+            PyObject *subject = SECOND();
+            PyObject *rest = PyDict_New();
+            if (rest == NULL || PyDict_Update(rest, subject)) {
+                Py_XDECREF(rest);
+                goto error;
+            }
+            // This may seem a bit inefficient, but keys is rarely big enough to
+            // actually impact runtime.
+            assert(PyTuple_CheckExact(keys));
+            for (Py_ssize_t i = 0; i < PyTuple_GET_SIZE(keys); i++) {
+                if (PyDict_DelItem(rest, PyTuple_GET_ITEM(keys, i))) {
+                    Py_DECREF(rest);
+                    goto error;
+                }
+            }
+            Py_DECREF(keys);
+            SET_TOP(rest);
+            DISPATCH();
+        }
+
         case TARGET(GET_ITER): {
             /* before: [obj]; after [getiter(obj)] */
             PyObject *iterable = TOP();

Python/opcode_targets.h

@@ -29,11 +29,11 @@ static void *opcode_targets[256] = {
     &&TARGET_BINARY_TRUE_DIVIDE,
     &&TARGET_INPLACE_FLOOR_DIVIDE,
     &&TARGET_INPLACE_TRUE_DIVIDE,
-    &&_unknown_opcode,
-    &&_unknown_opcode,
-    &&_unknown_opcode,
-    &&_unknown_opcode,
-    &&_unknown_opcode,
+    &&TARGET_GET_LEN,
+    &&TARGET_MATCH_MAPPING,
+    &&TARGET_MATCH_SEQUENCE,
+    &&TARGET_MATCH_KEYS,
+    &&TARGET_COPY_DICT_WITHOUT_KEYS,
     &&_unknown_opcode,
     &&_unknown_opcode,
     &&_unknown_opcode,
@@ -151,7 +151,7 @@ static void *opcode_targets[256] = {
     &&_unknown_opcode,
     &&_unknown_opcode,
     &&_unknown_opcode,
-    &&_unknown_opcode,
+    &&TARGET_MATCH_CLASS,
     &&_unknown_opcode,
     &&TARGET_SETUP_ASYNC_WITH,
     &&TARGET_FORMAT_VALUE,

Python/pystate.c

@@ -308,6 +308,8 @@ interpreter_clear(PyInterpreterState *interp, PyThreadState *tstate)
     Py_CLEAR(interp->importlib);
     Py_CLEAR(interp->import_func);
     Py_CLEAR(interp->dict);
+    Py_CLEAR(interp->map_abc);
+    Py_CLEAR(interp->seq_abc);
 #ifdef HAVE_FORK
     Py_CLEAR(interp->before_forkers);
     Py_CLEAR(interp->after_forkers_parent);

Python/symtable.c

@@ -207,6 +207,7 @@ static int symtable_visit_argannotations(struct symtable *st, asdl_arg_seq *args
 static int symtable_implicit_arg(struct symtable *st, int pos);
 static int symtable_visit_annotations(struct symtable *st, arguments_ty, expr_ty);
 static int symtable_visit_withitem(struct symtable *st, withitem_ty item);
+static int symtable_visit_match_case(struct symtable *st, match_case_ty m);
 
 
 static identifier top = NULL, lambda = NULL, genexpr = NULL,
@@ -239,6 +240,7 @@ symtable_new(void)
         goto fail;
     st->st_cur = NULL;
     st->st_private = NULL;
+    st->in_pattern = 0;
     return st;
  fail:
     PySymtable_Free(st);
@@ -1294,6 +1296,10 @@ symtable_visit_stmt(struct symtable *st, stmt_ty s)
         if (s->v.If.orelse)
             VISIT_SEQ(st, stmt, s->v.If.orelse);
         break;
+    case Match_kind:
+        VISIT(st, expr, s->v.Match.subject);
+        VISIT_SEQ(st, match_case, s->v.Match.cases);
+        break;
     case Raise_kind:
         if (s->v.Raise.exc) {
             VISIT(st, expr, s->v.Raise.exc);
@@ -1648,6 +1654,13 @@ symtable_visit_expr(struct symtable *st, expr_ty e)
             VISIT(st, expr, e->v.Slice.step)
         break;
     case Name_kind:
+        // Don't make "_" a local when used in a pattern:
+        if (st->in_pattern &&
+            e->v.Name.ctx == Store &&
+            _PyUnicode_EqualToASCIIString(e->v.Name.id, "_"))
+        {
+            break;
+        }
         if (!symtable_add_def(st, e->v.Name.id,
                               e->v.Name.ctx == Load ? USE : DEF_LOCAL))
             VISIT_QUIT(st, 0);
@@ -1667,6 +1680,13 @@ symtable_visit_expr(struct symtable *st, expr_ty e)
     case Tuple_kind:
         VISIT_SEQ(st, expr, e->v.Tuple.elts);
         break;
+    case MatchAs_kind:
+        VISIT(st, expr, e->v.MatchAs.pattern);
+        symtable_add_def(st, e->v.MatchAs.name, DEF_LOCAL);
+        break;
+    case MatchOr_kind:
+        VISIT_SEQ(st, expr, e->v.MatchOr.patterns);
+        break;
     }
     VISIT_QUIT(st, 1);
 }
@@ -1785,6 +1805,20 @@ symtable_visit_withitem(struct symtable *st, withitem_ty item)
     return 1;
 }
 
+static int
+symtable_visit_match_case(struct symtable *st, match_case_ty m)
+{
+    assert(!st->in_pattern);
+    st->in_pattern = 1;
+    VISIT(st, expr, m->pattern);
+    assert(st->in_pattern);
+    st->in_pattern = 0;
+    if (m->guard) {
+        VISIT(st, expr, m->guard);
+    }
+    VISIT_SEQ(st, stmt, m->body);
+    return 1;
+}
 
 static int
 symtable_visit_alias(struct symtable *st, alias_ty a)