PYTHON-6140 [Build Failure] TestBSON.test_encode_size_limit MemoryError on Windows

[blink1073]

PYTHON-6140

Changes in this PR

Skip TestBSON.test_encode_size_limit on hosts with less than 6GiB of total physical memory. The test needs ~4GiB of peak memory to encode a document that exceeds the BSON size limit, but on 4GB windows-64-vsMulti-small hosts the allocations fail with MemoryError before the expected ValueError is raised.

Test Plan

• Verified locally: test_encode_size_limit now skips on a 4GiB host.
• All Win32 tests pass in CI.

Checklist

Checklist for Author

• Did you update the changelog (if necessary)? — Not necessary, test-only change.
• Is there test coverage? — N/A, this change is test infrastructure.
• Is any followup work tracked in a JIRA ticket? — Backport to v4.18 required (the failure also occurs on that branch).

Checklist for Reviewer

• Does the title of the PR reference a JIRA Ticket?
• Do you fully understand the implementation? (Would you be comfortable explaining how this code works to someone else?)
• Is all relevant documentation (README or docstring) updated?

[Copilot]

Copilot review overview

🟢 Approval recommended

The focused test-only change matches the stated requirements, with no blocking issues identified; Windows CI verification remains pending.

Review effort: Balanced
Findings: None

What changed in this PR

Prevents low-memory hosts from failing the BSON size-limit test with MemoryError instead of the expected ValueError.

Changes:

• Detect total physical memory on Windows and other supported platforms.
• Skip the affected test below 6 GiB; retain execution when memory is unknown.

File	Description
test/​test_bson.py	Adds memory detection and a targeted test skip.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[NoahStapp]

Honest question: what if we just stopped testing on any platforms that can't handle this test? Do we know how many users could even encounter this?

[blink1073]

Honest question: what if we just stopped testing on any platforms that can't handle this test? Do we know how many users could even encounter this?

If we hard-code an OS, it would still fail in a container. We could try to nail down the conditions, but I think it would be more likely that we accidentally stop testing. What if we just unconditionally catch the MemoryError?

[blink1073]

I updated to only do the memory probe on linux.