Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

extra_tests/snippets/stdlib_select.py (1)

48-65: ⚠️ Potential issue | 🟠 Major | ⚡ Quick win

Keep the high-FD socket allocation inside the non-Windows guard.

Line 59-65 are dedented, so Windows still tries to allocate thousands of sockets before reaching the ValueError assertion. That makes the regression test flaky or failing on Windows for the wrong reason.

Suggested fix

 # Too many descriptors for select.select()
 if sys.platform != "win32":
     import resource

     # 100 additional fds should be enough for interpreter needs
     need_fds = TOO_MANY_SELECT_FDS + 100

     soft_max_fds, hard_max_fds = resource.getrlimit(resource.RLIMIT_NOFILE)
     soft_max_fds = max(soft_max_fds, need_fds)
     assert hard_max_fds >= soft_max_fds, "Not enough file descriptors for this test"
     resource.setrlimit(resource.RLIMIT_NOFILE, (soft_max_fds, hard_max_fds))
-sockets = [s for _ in range(TOO_MANY_SELECT_FDS // 2) for s in socket.socketpair()]
-assert_raises(ValueError, select.select, sockets, [], [], 0)
-del sockets
-a, b = socket.socketpair()
-# CPython disallows this on *nix systems too.
-assert_raises(ValueError, select.select, [a] * TOO_MANY_SELECT_FDS, [], [], 0)
-del a, b
+    sockets = [s for _ in range(TOO_MANY_SELECT_FDS // 2) for s in socket.socketpair()]
+    assert_raises(ValueError, select.select, sockets, [], [], 0)
+    del sockets
+    a, b = socket.socketpair()
+    # CPython disallows this on *nix systems too.
+    assert_raises(ValueError, select.select, [a] * TOO_MANY_SELECT_FDS, [], [], 0)
+    del a, b

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@extra_tests/snippets/stdlib_select.py` around lines 48 - 65, The high-FD
socket creation and subsequent assertions are currently outside the Windows
guard and must be moved inside the sys.platform != "win32" block: wrap or indent
the block that creates sockets (the list comprehension using socket.socketpair()
assigned to sockets), the assert_raises call that calls select.select with
sockets, the del sockets, and the CPython-specific test using variables a, b
(socket.socketpair() and the assert_raises on [a] * TOO_MANY_SELECT_FDS) so they
only run when not on Windows; update the code around the symbols sockets,
socket.socketpair, assert_raises(select.select, ...), and variables a,b
accordingly to keep all high-FD allocation and CPython-only assertions inside
the non-Windows guard.

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

extra_tests/snippets/stdlib_select.py (1)

49-69: ⚠️ Potential issue | 🟠 Major | ⚡ Quick win

Keep the high-FD regression block inside the non-Windows guard.

Lines 63-69 are still outside the sys.platform != "win32" block, so Windows will try to allocate thousands of sockets before it ever reaches the Rust-side validation.

🐛 Suggested fix

 if sys.platform != "win32":
     import resource

     soft_max_fds, hard_max_fds = resource.getrlimit(resource.RLIMIT_NOFILE)
     if soft_max_fds != resource.RLIM_INFINITY:
         # 100 additional fds should be enough for interpreter needs
         need_fds = TOO_MANY_SELECT_FDS + 100

         soft_max_fds = max(soft_max_fds, need_fds)
         if hard_max_fds != resource.RLIM_INFINITY:
             assert hard_max_fds >= soft_max_fds, (
                 "Not enough file descriptors for this test"
             )
         resource.setrlimit(resource.RLIMIT_NOFILE, (soft_max_fds, hard_max_fds))
-sockets = [s for _ in range(TOO_MANY_SELECT_FDS // 2) for s in socket.socketpair()]
-assert_raises(ValueError, select.select, sockets, [], [], 0)
-del sockets
-a, b = socket.socketpair()
-# CPython disallows this on *nix systems too.
-assert_raises(ValueError, select.select, [a] * TOO_MANY_SELECT_FDS, [], [], 0)
-del a, b
+    sockets = [s for _ in range(TOO_MANY_SELECT_FDS // 2) for s in socket.socketpair()]
+    assert_raises(ValueError, select.select, sockets, [], [], 0)
+    del sockets
+    a, b = socket.socketpair()
+    # CPython disallows this on *nix systems too.
+    assert_raises(ValueError, select.select, [a] * TOO_MANY_SELECT_FDS, [], [], 0)
+    del a, b

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@extra_tests/snippets/stdlib_select.py` around lines 49 - 69, The high-FD
regression test that allocates many sockets is not guarded by the sys.platform
!= "win32" check; move the block that creates sockets, asserts select.select
raises, and the subsequent single-socket repeated-list test (the code using
TOO_MANY_SELECT_FDS, sockets = [...], assert_raises(ValueError, select.select,
sockets, [], [], 0), del sockets, and the a,b socketpair test) inside the
non-Windows guard so Windows won't try to create thousands of socket pairs and
will rely on the Rust-side validation instead.

crates/stdlib/src/select.rs (1)

85-88: ⚡ Quick win

Extract the platform-specific FD limit before the shared comparison.

Only the limit value differs here. Pulling that out first keeps the branch local and avoids duplicating the comparison logic.

♻️ Proposed refactor

-            let too_many_fds = cfg_select! {
-                windows => v.len() > FD_SETSIZE as usize,
-                _ => v.len() > FD_SETSIZE,
-            };
+            let fd_limit = cfg_select! {
+                windows => FD_SETSIZE as usize,
+                _ => FD_SETSIZE,
+            };
+            let too_many_fds = v.len() > fd_limit;

As per coding guidelines, "When branches differ only in a value but share common logic, extract the differing value first, then call the common logic once to avoid duplicate code".

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@crates/stdlib/src/select.rs` around lines 85 - 88, The comparison duplicates
only the platform-specific FD_SETSIZE value; extract that value first (e.g.,
define a fd_limit via the existing cfg_select! macro) and then perform the
shared comparison once (compute too_many_fds = v.len() > fd_limit). Update the
code that sets too_many_fds to use the new fd_limit variable and keep the
cfg_select! usage limited to producing the platform-specific limit (FD_SETSIZE)
so the comparison logic is not duplicated.