[sandugood]

Added a regression test as well (with the string generated by fuzzer in the issue).

• Issue: MemoryError during parse can escalate to SystemError and abort on small input #150001

[bedevere-app]

Most changes to Python require a NEWS entry. Add one using the blurb_it web app or the blurb command-line tool.

If this change has little impact on Python users, wait for a maintainer to apply the skip news label instead.

[python-cla-bot]

All commit authors signed the Contributor License Agreement.

[bedevere-app]

Most changes to Python require a NEWS entry. Add one using the blurb_it web app or the blurb command-line tool.

If this change has little impact on Python users, wait for a maintainer to apply the skip news label instead.

[sandugood]

I think that this fix doesn't require a NEWS entry. But I am opened to suggestions

[pablogsal]

We need a news entry can you add one please?

[sandugood]

Yes, sure. Thanks for the review

[sandugood]

Seems that the CI error is not related to the PR itself.

[JelleZijlstra]

I retried it

[sandugood]

I retried it

Looks like it is a github issue? The run itself doesnt get to the building stage

[sandugood]

Thank you for the CI rerun. All of the tests have passed

[sandugood]

Could you, please, review this PR if you have time @lysnikolaou?
Thank you in advance

[sandugood]

Sorry for a late response @StanFromIreland
Thank you for your review

[stestagg]

Worth considering if this would be neater? If we assume that parse failure paths always return NULL, then you can

diff --git a/Tools/peg_generator/pegen/c_generator.py b/Tools/peg_generator/pegen/c_generator.py
index d9236dfb22835bd..bb0cc34c68a75e9 100644
--- a/Tools/peg_generator/pegen/c_generator.py
+++ b/Tools/peg_generator/pegen/c_generator.py
@@ -658,10 +658,12 @@ def _handle_loop_rule_body(self, node: Rule, rhs: Rhs) -> None:
             )
             if is_repeat1:
                 self.print("if (_n == 0 || p->error_indicator) {")
-                with self.indent():
-                    self.print("PyMem_Free(_children);")
-                    self.add_return("NULL")
-                self.print("}")
+            else:
+                self.print("if (p->error_indicator) {")
+            with self.indent():
+                self.print("PyMem_Free(_children);")
+                self.add_return("NULL")
+            self.print("}")
             self.print("asdl_seq *_seq = (asdl_seq*)_Py_asdl_generic_seq_new(_n, p->arena);")
             self.out_of_memory_return("!_seq", cleanup_code="PyMem_Free(_children);")
             self.print("for (Py_ssize_t i = 0; i < _n; i++) asdl_seq_SET_UNTYPED(_seq, i, _children[i]);")

This way we have one check shared between loop0 and loop1 without duplicating it on the loop1?

BTW: this code is new to me, so I probably made some assumptions here!

[sandugood]

I think we can merge this ver and then reiterate with a follow up?
Wdyt @stestagg?

[StanFromIreland]

Can you please fix the conflicts?

[sandugood]

Can you please fix the conflicts?

Sure, working on it now