gh-150207: Raise MemoryError on tokenizer allocation failure instead of crashing (GH-150275)

(cherry picked from commit 262625f)

Co-authored-by: Grant Herman <grantlouisherman041@gmail.com>

Author: grantlouisherman

Misc/NEWS.d/next/Core_and_Builtins/2026-05-22-21-52-38.gh-issue-150207.l2BUtI.rst

@@ -0,0 +1 @@
+Fix a crash when a memory allocation fails during tokenizer initialization. A proper :exc:`MemoryError` is now raised instead.

Parser/lexer/state.c

@@ -15,8 +15,11 @@ _PyTokenizer_tok_new(void)
     struct tok_state *tok = (struct tok_state *)PyMem_Calloc(
                                             1,
                                             sizeof(struct tok_state));
-    if (tok == NULL)
+    if (tok == NULL) {
+        PyErr_NoMemory();
         return NULL;
+    }
+
     tok->buf = tok->cur = tok->inp = NULL;
     tok->fp_interactive = 0;
     tok->interactive_src_start = NULL;

Parser/tokenizer/file_tokenizer.c

@@ -355,6 +355,7 @@ _PyTokenizer_FromFile(FILE *fp, const char* enc,
         return NULL;
     if ((tok->buf = (char *)PyMem_Malloc(BUFSIZ)) == NULL) {
         _PyTokenizer_Free(tok);
+        PyErr_NoMemory();
         return NULL;
     }
     tok->cur = tok->inp = tok->buf;

Parser/tokenizer/helpers.c

@@ -183,6 +183,7 @@ _PyTokenizer_new_string(const char *s, Py_ssize_t len, struct tok_state *tok)
     char* result = (char *)PyMem_Malloc(len + 1);
     if (!result) {
         tok->done = E_NOMEM;
+        PyErr_NoMemory();
         return NULL;
     }
     memcpy(result, s, len);
@@ -211,6 +212,7 @@ _PyTokenizer_translate_newlines(const char *s, int exec_input, int preserve_crlf
     buf = PyMem_Malloc(needed_length);
     if (buf == NULL) {
         tok->done = E_NOMEM;
+        PyErr_NoMemory();
         return NULL;
     }
     for (current = buf; *s; s++, current++) {

Parser/tokenizer/readline_tokenizer.c

@@ -114,6 +114,7 @@ _PyTokenizer_FromReadline(PyObject* readline, const char* enc,
         return NULL;
     if ((tok->buf = (char *)PyMem_Malloc(BUFSIZ)) == NULL) {
         _PyTokenizer_Free(tok);
+        PyErr_NoMemory();
         return NULL;
     }
     tok->cur = tok->inp = tok->buf;