bpo-46765: Replace Locally Cached Strings with Statically Initialized Objects (gh-31366)

https://bugs.python.org/issue46765

6 files changed, 42 insertions, 162 deletions

diff --git a/Python/_warnings.c b/Python/_warnings.c
index 03e6ffc..be962e7 100644
--- a/Python/_warnings.c
+++ b/Python/_warnings.c
@@ -768,25 +768,6 @@ warn_explicit(PyThreadState *tstate, PyObject *category, PyObject *message,
 static int
 is_internal_frame(PyFrameObject *frame)
 {
-    static PyObject *importlib_string = NULL;
-    static PyObject *bootstrap_string = NULL;
-    int contains;
-
-    if (importlib_string == NULL) {
-        importlib_string = PyUnicode_FromString("importlib");
-        if (importlib_string == NULL) {
-            return 0;
-        }
-
-        bootstrap_string = PyUnicode_FromString("_bootstrap");
-        if (bootstrap_string == NULL) {
-            Py_DECREF(importlib_string);
-            return 0;
-        }
-        Py_INCREF(importlib_string);
-        Py_INCREF(bootstrap_string);
-    }
-
     if (frame == NULL) {
         return 0;
     }
@@ -802,12 +783,12 @@ is_internal_frame(PyFrameObject *frame)
         return 0;
     }
 
-    contains = PyUnicode_Contains(filename, importlib_string);
+    int contains = PyUnicode_Contains(filename, &_Py_ID(importlib));
     if (contains < 0) {
         return 0;
     }
     else if (contains > 0) {
-        contains = PyUnicode_Contains(filename, bootstrap_string);
+        contains = PyUnicode_Contains(filename, &_Py_ID(_bootstrap));
         if (contains < 0) {
             return 0;
         }
diff --git a/Python/ast_unparse.c b/Python/ast_unparse.c
index 126e904..f148e99 100644
--- a/Python/ast_unparse.c
+++ b/Python/ast_unparse.c
@@ -1,5 +1,6 @@
 #include "Python.h"
 #include "pycore_ast.h"           // expr_ty
+#include "pycore_runtime.h"       // _Py_ID()
 #include <float.h>                // DBL_MAX_10_EXP
 #include <stdbool.h>
 
@@ -8,11 +9,10 @@
  * See ast.unparse for a full unparser (written in Python)
  */
 
-static PyObject *_str_open_br;
-static PyObject *_str_dbl_open_br;
-static PyObject *_str_close_br;
-static PyObject *_str_dbl_close_br;
-static PyObject *_str_inf;
+_Py_DECLARE_STR(open_br, "{");
+_Py_DECLARE_STR(dbl_open_br, "{{");
+_Py_DECLARE_STR(close_br, "}");
+_Py_DECLARE_STR(dbl_close_br, "}}");
 static PyObject *_str_replace_inf;
 
 /* Forward declarations for recursion via helper functions. */
@@ -80,7 +80,7 @@ append_repr(_PyUnicodeWriter *writer, PyObject *obj)
     {
         PyObject *new_repr = PyUnicode_Replace(
             repr,
-            _str_inf,
+            &_Py_ID(inf),
             _str_replace_inf,
             -1
         );
@@ -575,11 +575,11 @@ escape_braces(PyObject *orig)
 {
     PyObject *temp;
     PyObject *result;
-    temp = PyUnicode_Replace(orig, _str_open_br, _str_dbl_open_br, -1);
+    temp = PyUnicode_Replace(orig, &_Py_STR(open_br), &_Py_STR(dbl_open_br), -1);
     if (!temp) {
         return NULL;
     }
-    result = PyUnicode_Replace(temp, _str_close_br, _str_dbl_close_br, -1);
+    result = PyUnicode_Replace(temp, &_Py_STR(close_br), &_Py_STR(dbl_close_br), -1);
     Py_DECREF(temp);
     return result;
 }
@@ -673,7 +673,7 @@ append_formattedvalue(_PyUnicodeWriter *writer, expr_ty e)
     if (!temp_fv_str) {
         return -1;
     }
-    if (PyUnicode_Find(temp_fv_str, _str_open_br, 0, 1, 1) == 0) {
+    if (PyUnicode_Find(temp_fv_str, &_Py_STR(open_br), 0, 1, 1) == 0) {
         /* Expression starts with a brace, split it with a space from the outer
            one. */
         outer_brace = "{ ";
@@ -927,26 +927,6 @@ append_ast_expr(_PyUnicodeWriter *writer, expr_ty e, int level)
 static int
 maybe_init_static_strings(void)
 {
-    if (!_str_open_br &&
-        !(_str_open_br = PyUnicode_InternFromString("{"))) {
-        return -1;
-    }
-    if (!_str_dbl_open_br &&
-        !(_str_dbl_open_br = PyUnicode_InternFromString("{{"))) {
-        return -1;
-    }
-    if (!_str_close_br &&
-        !(_str_close_br = PyUnicode_InternFromString("}"))) {
-        return -1;
-    }
-    if (!_str_dbl_close_br &&
-        !(_str_dbl_close_br = PyUnicode_InternFromString("}}"))) {
-        return -1;
-    }
-    if (!_str_inf &&
-        !(_str_inf = PyUnicode_FromString("inf"))) {
-        return -1;
-    }
     if (!_str_replace_inf &&
         !(_str_replace_inf = PyUnicode_FromFormat("1e%d", 1 + DBL_MAX_10_EXP))) {
         return -1;
diff --git a/Python/compile.c b/Python/compile.c
index 7f0a6f0..cb6e5e3 100644
--- a/Python/compile.c
+++ b/Python/compile.c
@@ -2522,7 +2522,6 @@ static int
 compiler_class(struct compiler *c, stmt_ty s)
 {
     PyCodeObject *co;
-    PyObject *str;
     int i, firstlineno;
     asdl_expr_seq *decos = s->v.ClassDef.decorator_list;
 
@@ -2557,30 +2556,21 @@ compiler_class(struct compiler *c, stmt_ty s)
         Py_INCREF(s->v.ClassDef.name);
         Py_XSETREF(c->u->u_private, s->v.ClassDef.name);
         /* load (global) __name__ ... */
-        str = PyUnicode_InternFromString("__name__");
-        if (!str || !compiler_nameop(c, str, Load)) {
-            Py_XDECREF(str);
+        if (!compiler_nameop(c, &_Py_ID(__name__), Load)) {
             compiler_exit_scope(c);
             return 0;
         }
-        Py_DECREF(str);
         /* ... and store it as __module__ */
-        str = PyUnicode_InternFromString("__module__");
-        if (!str || !compiler_nameop(c, str, Store)) {
-            Py_XDECREF(str);
+        if (!compiler_nameop(c, &_Py_ID(__module__), Store)) {
             compiler_exit_scope(c);
             return 0;
         }
-        Py_DECREF(str);
         assert(c->u->u_qualname);
         ADDOP_LOAD_CONST(c, c->u->u_qualname);
-        str = PyUnicode_InternFromString("__qualname__");
-        if (!str || !compiler_nameop(c, str, Store)) {
-            Py_XDECREF(str);
+        if (!compiler_nameop(c, &_Py_ID(__qualname__), Store)) {
             compiler_exit_scope(c);
             return 0;
         }
-        Py_DECREF(str);
         /* compile the body proper */
         if (!compiler_body(c, s->v.ClassDef.body)) {
             compiler_exit_scope(c);
@@ -2591,13 +2581,7 @@ compiler_class(struct compiler *c, stmt_ty s)
         /* Return __classcell__ if it is referenced, otherwise return None */
         if (c->u->u_ste->ste_needs_class_closure) {
             /* Store __classcell__ into class namespace & return it */
-            str = PyUnicode_InternFromString("__class__");
-            if (str == NULL) {
-                compiler_exit_scope(c);
-                return 0;
-            }
-            i = compiler_lookup_arg(c->u->u_cellvars, str);
-            Py_DECREF(str);
+            i = compiler_lookup_arg(c->u->u_cellvars, &_Py_ID(__class__));
             if (i < 0) {
                 compiler_exit_scope(c);
                 return 0;
@@ -2606,13 +2590,10 @@ compiler_class(struct compiler *c, stmt_ty s)
 
             ADDOP_I(c, LOAD_CLOSURE, i);
             ADDOP_I(c, COPY, 1);
-            str = PyUnicode_InternFromString("__classcell__");
-            if (!str || !compiler_nameop(c, str, Store)) {
-                Py_XDECREF(str);
+            if (!compiler_nameop(c, &_Py_ID(__classcell__), Store)) {
                 compiler_exit_scope(c);
                 return 0;
             }
-            Py_DECREF(str);
         }
         else {
             /* No methods referenced __class__, so just return None */
@@ -4741,13 +4722,8 @@ compiler_joined_str(struct compiler *c, expr_ty e)
 
     Py_ssize_t value_count = asdl_seq_LEN(e->v.JoinedStr.values);
     if (value_count > STACK_USE_GUIDELINE) {
-        ADDOP_LOAD_CONST_NEW(c, _PyUnicode_FromASCII("", 0));
-        PyObject *join = _PyUnicode_FromASCII("join", 4);
-        if (join == NULL) {
-            return 0;
-        }
-        ADDOP_NAME(c, LOAD_METHOD, join, names);
-        Py_DECREF(join);
+        ADDOP_LOAD_CONST_NEW(c, &_Py_STR(empty));
+        ADDOP_NAME(c, LOAD_METHOD, &_Py_ID(join), names);
         ADDOP_I(c, BUILD_LIST, 0);
         for (Py_ssize_t i = 0; i < asdl_seq_LEN(e->v.JoinedStr.values); i++) {
             VISIT(c, expr, asdl_seq_GET(e->v.JoinedStr.values, i));
diff --git a/Python/pylifecycle.c b/Python/pylifecycle.c
index 4a3a1ab..b6310c9 100644
--- a/Python/pylifecycle.c
+++ b/Python/pylifecycle.c
@@ -24,7 +24,6 @@
 #include "pycore_runtime.h"       // _Py_ID()
 #include "pycore_runtime_init.h"  // _PyRuntimeState_INIT
 #include "pycore_sliceobject.h"   // _PySlice_Fini()
-#include "pycore_symtable.h"      // _PySymtable_Fini()
 #include "pycore_sysmodule.h"     // _PySys_ClearAuditHooks()
 #include "pycore_traceback.h"     // _Py_DumpTracebackThreads()
 #include "pycore_tuple.h"         // _PyTuple_InitTypes()
@@ -1690,9 +1689,6 @@ finalize_interp_clear(PyThreadState *tstate)
     int is_main_interp = _Py_IsMainInterpreter(tstate->interp);
 
     _PyExc_ClearExceptionGroupType(tstate->interp);
-    if (is_main_interp) {
-        _PySymtable_Fini();
-    }
 
     /* Clear interpreter state and all thread states */
     _PyInterpreterState_Clear(tstate);
diff --git a/Python/symtable.c b/Python/symtable.c
index e9bdff3..0b259b0 100644
--- a/Python/symtable.c
+++ b/Python/symtable.c
@@ -230,13 +230,6 @@ static int symtable_raise_if_annotation_block(struct symtable *st, const char *,
 static int symtable_raise_if_comprehension_block(struct symtable *st, expr_ty);
 
 
-static identifier top = NULL, lambda = NULL, genexpr = NULL,
-    listcomp = NULL, setcomp = NULL, dictcomp = NULL,
-    __class__ = NULL, _annotation = NULL;
-
-#define GET_IDENTIFIER(VAR) \
-    ((VAR) ? (VAR) : ((VAR) = PyUnicode_InternFromString(# VAR)))
-
 #define DUPLICATE_ARGUMENT \
 "duplicate argument '%U' in function definition"
 
@@ -313,8 +306,7 @@ _PySymtable_Build(mod_ty mod, PyObject *filename, PyFutureFeatures *future)
         recursion_limit * COMPILER_STACK_FRAME_SCALE : recursion_limit;
 
     /* Make the initial symbol information gathering pass */
-    if (!GET_IDENTIFIER(top) ||
-        !symtable_enter_block(st, top, ModuleBlock, (void *)mod, 0, 0, 0, 0)) {
+    if (!symtable_enter_block(st, &_Py_ID(top), ModuleBlock, (void *)mod, 0, 0, 0, 0)) {
         _PySymtable_Free(st);
         return NULL;
     }
@@ -619,9 +611,7 @@ static int
 drop_class_free(PySTEntryObject *ste, PyObject *free)
 {
     int res;
-    if (!GET_IDENTIFIER(__class__))
-        return 0;
-    res = PySet_Discard(free, __class__);
+    res = PySet_Discard(free, &_Py_ID(__class__));
     if (res < 0)
         return 0;
     if (res)
@@ -834,9 +824,7 @@ analyze_block(PySTEntryObject *ste, PyObject *bound, PyObject *free,
     }
     else {
         /* Special-case __class__ */
-        if (!GET_IDENTIFIER(__class__))
-            goto error;
-        if (PySet_Add(newbound, __class__) < 0)
+        if (PySet_Add(newbound, &_Py_ID(__class__)) < 0)
             goto error;
     }
 
@@ -1610,13 +1598,11 @@ symtable_visit_expr(struct symtable *st, expr_ty e)
         VISIT(st, expr, e->v.UnaryOp.operand);
         break;
     case Lambda_kind: {
-        if (!GET_IDENTIFIER(lambda))
-            VISIT_QUIT(st, 0);
         if (e->v.Lambda.args->defaults)
             VISIT_SEQ(st, expr, e->v.Lambda.args->defaults);
         if (e->v.Lambda.args->kw_defaults)
             VISIT_SEQ_WITH_NULL(st, expr, e->v.Lambda.args->kw_defaults);
-        if (!symtable_enter_block(st, lambda,
+        if (!symtable_enter_block(st, &_Py_ID(lambda),
                                   FunctionBlock, (void *)e,
                                   e->lineno, e->col_offset,
                                   e->end_lineno, e->end_col_offset))
@@ -1730,8 +1716,7 @@ symtable_visit_expr(struct symtable *st, expr_ty e)
         if (e->v.Name.ctx == Load &&
             st->st_cur->ste_type == FunctionBlock &&
             _PyUnicode_EqualToASCIIString(e->v.Name.id, "super")) {
-            if (!GET_IDENTIFIER(__class__) ||
-                !symtable_add_def(st, __class__, USE, LOCATION(e)))
+            if (!symtable_add_def(st, &_Py_ID(__class__), USE, LOCATION(e)))
                 VISIT_QUIT(st, 0);
         }
         break;
@@ -1832,7 +1817,7 @@ symtable_visit_annotation(struct symtable *st, expr_ty annotation)
 {
     int future_annotations = st->st_future->ff_features & CO_FUTURE_ANNOTATIONS;
     if (future_annotations &&
-        !symtable_enter_block(st, GET_IDENTIFIER(_annotation), AnnotationBlock,
+        !symtable_enter_block(st, &_Py_ID(_annotation), AnnotationBlock,
                               (void *)annotation, annotation->lineno,
                               annotation->col_offset, annotation->end_lineno,
                               annotation->end_col_offset)) {
@@ -1867,7 +1852,7 @@ symtable_visit_annotations(struct symtable *st, stmt_ty o, arguments_ty a, expr_
 {
     int future_annotations = st->st_future->ff_features & CO_FUTURE_ANNOTATIONS;
     if (future_annotations &&
-        !symtable_enter_block(st, GET_IDENTIFIER(_annotation), AnnotationBlock,
+        !symtable_enter_block(st, &_Py_ID(_annotation), AnnotationBlock,
                               (void *)o, o->lineno, o->col_offset, o->end_lineno,
                               o->end_col_offset)) {
         VISIT_QUIT(st, 0);
@@ -2085,7 +2070,7 @@ symtable_handle_comprehension(struct symtable *st, expr_ty e,
 static int
 symtable_visit_genexp(struct symtable *st, expr_ty e)
 {
-    return symtable_handle_comprehension(st, e, GET_IDENTIFIER(genexpr),
+    return symtable_handle_comprehension(st, e, &_Py_ID(genexpr),
                                          e->v.GeneratorExp.generators,
                                          e->v.GeneratorExp.elt, NULL);
 }
@@ -2093,7 +2078,7 @@ symtable_visit_genexp(struct symtable *st, expr_ty e)
 static int
 symtable_visit_listcomp(struct symtable *st, expr_ty e)
 {
-    return symtable_handle_comprehension(st, e, GET_IDENTIFIER(listcomp),
+    return symtable_handle_comprehension(st, e, &_Py_ID(listcomp),
                                          e->v.ListComp.generators,
                                          e->v.ListComp.elt, NULL);
 }
@@ -2101,7 +2086,7 @@ symtable_visit_listcomp(struct symtable *st, expr_ty e)
 static int
 symtable_visit_setcomp(struct symtable *st, expr_ty e)
 {
-    return symtable_handle_comprehension(st, e, GET_IDENTIFIER(setcomp),
+    return symtable_handle_comprehension(st, e, &_Py_ID(setcomp),
                                          e->v.SetComp.generators,
                                          e->v.SetComp.elt, NULL);
 }
@@ -2109,7 +2094,7 @@ symtable_visit_setcomp(struct symtable *st, expr_ty e)
 static int
 symtable_visit_dictcomp(struct symtable *st, expr_ty e)
 {
-    return symtable_handle_comprehension(st, e, GET_IDENTIFIER(dictcomp),
+    return symtable_handle_comprehension(st, e, &_Py_ID(dictcomp),
                                          e->v.DictComp.generators,
                                          e->v.DictComp.key,
                                          e->v.DictComp.value);
@@ -2173,16 +2158,3 @@ _Py_SymtableStringObjectFlags(const char *str, PyObject *filename,
     _PyArena_Free(arena);
     return st;
 }
-
-void
-_PySymtable_Fini(void)
-{
-    Py_CLEAR(top);
-    Py_CLEAR(lambda);
-    Py_CLEAR(genexpr);
-    Py_CLEAR(listcomp);
-    Py_CLEAR(setcomp);
-    Py_CLEAR(dictcomp);
-    Py_CLEAR(__class__);
-    Py_CLEAR(_annotation);
-}
diff --git a/Python/sysmodule.c b/Python/sysmodule.c
index 57bf04d..e23b879 100644
--- a/Python/sysmodule.c
+++ b/Python/sysmodule.c
@@ -706,7 +706,6 @@ sys_displayhook(PyObject *module, PyObject *o)
 {
     PyObject *outf;
     PyObject *builtins;
-    static PyObject *newline = NULL;
     PyThreadState *tstate = _PyThreadState_GET();
 
     builtins = PyImport_GetModule(&_Py_ID(builtins));
@@ -747,12 +746,8 @@ sys_displayhook(PyObject *module, PyObject *o)
             return NULL;
         }
     }
-    if (newline == NULL) {
-        newline = PyUnicode_FromString("\n");
-        if (newline == NULL)
-            return NULL;
-    }
-    if (PyFile_WriteObject(newline, outf, Py_PRINT_RAW) != 0)
+    _Py_DECLARE_STR(newline, "\n");
+    if (PyFile_WriteObject(&_Py_STR(newline), outf, Py_PRINT_RAW) != 0)
         return NULL;
     if (PyObject_SetAttr(builtins, &_Py_ID(_), o) != 0)
         return NULL;
@@ -946,30 +941,18 @@ sys_intern_impl(PyObject *module, PyObject *s)
 
 /*
  * Cached interned string objects used for calling the profile and
- * trace functions.  Initialized by trace_init().
+ * trace functions.
  */
-static PyObject *whatstrings[8] = {NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL};
-
-static int
-trace_init(void)
-{
-    static const char * const whatnames[8] = {
-        "call", "exception", "line", "return",
-        "c_call", "c_exception", "c_return",
-        "opcode"
-    };
-    PyObject *name;
-    int i;
-    for (i = 0; i < 8; ++i) {
-        if (whatstrings[i] == NULL) {
-            name = PyUnicode_InternFromString(whatnames[i]);
-            if (name == NULL)
-                return -1;
-            whatstrings[i] = name;
-        }
-    }
-    return 0;
-}
+static PyObject *whatstrings[8] = {
+   &_Py_ID(call),
+   &_Py_ID(exception),
+   &_Py_ID(line),
+   &_Py_ID(return),
+   &_Py_ID(c_call),
+   &_Py_ID(c_exception),
+   &_Py_ID(c_return),
+   &_Py_ID(opcode),
+};
 
 
 static PyObject *
@@ -1050,10 +1033,6 @@ trace_trampoline(PyObject *self, PyFrameObject *frame,
 static PyObject *
 sys_settrace(PyObject *self, PyObject *args)
 {
-    if (trace_init() == -1) {
-        return NULL;
-    }
-
     PyThreadState *tstate = _PyThreadState_GET();
     if (args == Py_None) {
         if (_PyEval_SetTrace(tstate, NULL, NULL) < 0) {
@@ -1099,10 +1078,6 @@ sys_gettrace_impl(PyObject *module)
 static PyObject *
 sys_setprofile(PyObject *self, PyObject *args)
 {
-    if (trace_init() == -1) {
-        return NULL;
-    }
-
     PyThreadState *tstate = _PyThreadState_GET();
     if (args == Py_None) {
         if (_PyEval_SetProfile(tstate, NULL, NULL) < 0) {