fix(iterator): stop silently truncating spread/Array.from at 100,000 elements (#7562)

[proggeramlug]

Fixes #7562.

The bug

[...m.values()] on a 250,000-entry Map returned 100,000 elements. No error, no warning — a plausible-looking wrong answer.

                 before          node
map size:        250000          250000
spread length:   100000    <-    250000
Array.from:      100000    <-    250000

Three drain loops in array/iterator.rs carried a hardcoded for _ in 0..100_000 "safety limit" and simply fell out of the loop when it was hit, returning whatever had accumulated. This affected every spread, Array.from, and iterator-protocol drain — Maps, Sets, generators, user iterables alike.

Same severity class as the JSON key-loss corruption fixed earlier (#7546): silent wrong output is worse than a crash, because nothing signals it.

The fix, and why not simply remove the limit

• The bound becomes MAX_ITERATOR_DRAIN = JavaScript's own maximum array length (u32::MAX - 1). Every realistic workload is unaffected.
• Exhausting it now throws a RangeError instead of truncating.

Node applies no limit at all: [...it] runs until the iterator finishes or the process runs out of memory. Matching that exactly would trade silent truncation for an unbounded loop on a runaway iterator — so I kept a bound but made it report itself. A visible, recoverable error is the right third option; returning short data is strictly worse than either hanging or throwing.

Verification

• test_gap_7562_iterator_no_truncation.ts — all three drain paths past the old bound (Map values/keys/entries and [...m], a 130k Set, a 120k generator), plus a small-iterator case to prove the common path is untouched. Byte-identical to node 26.5.1.
• Sabotage-verified: restoring the 100,000 bound makes that test exit 1. A green run is therefore evidence, not decoration.
• cargo test -p perry-runtime --no-fail-fast: 1808 passed, 0 failed.
• raw_handle_debt.py 998 (baseline), check_file_size.sh, check_test_registration.py, cargo fmt --all --check all clean.

Pre-existing — confirmed at 969b447cc. Found by the map_1m investigation (#7561) while building its semantics matrix.

CI has a deep repo-wide runner backlog and may not report; the above is local validation and I am not claiming CI green.

Summary by CodeRabbit

• Bug Fixes

  • Iterator conversions no longer silently truncate results after 100,000 elements.
  • Large Map, Set, and generator iterators now preserve all values.
  • Exceeding the supported array size now reports a RangeError.

• Tests

  • Added coverage for large synchronous and asynchronous iterator operations.

• Chores

  • Updated the application version to 0.5.1319.

[coderabbitai]

Caution

Review failed

The pull request is closed.

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: c4e0442d-fe5a-4555-b8d0-44a6d6ef1466

📥 Commits

Reviewing files that changed from the base of the PR and between dc0778c and bf4a3b9.

⛔ Files ignored due to path filters (1)

• Cargo.lock is excluded by !**/*.lock

📒 Files selected for processing (5)

• CLAUDE.md
• Cargo.toml
• changelog.d/7565-iterator-no-silent-truncation.md
• crates/perry-runtime/src/array/iterator.rs
• test-files/test_gap_7562_iterator_no_truncation.ts

---

📝 Walkthrough

Walkthrough

Iterator drains now use a shared u32::MAX - 1 limit. Drains throw RangeError instead of silently truncating results. Regression tests cover large Map, Set, and generator inputs. Project version references are updated to 0.5.1319.

Changes

Iterator drain truncation fix

Layer / File(s)	Summary
Runtime limit and drain behavior crates/perry-runtime/src/array/iterator.rs	The runtime adds a shared maximum and RangeError helper. Synchronous and asynchronous iterator drains throw when the maximum is reached.
Regression coverage and release metadata test-files/test_gap_7562_iterator_no_truncation.ts, changelog.d/7565-iterator-no-silent-truncation.md, CLAUDE.md, Cargo.toml	Tests validate complete large-iterator results and unchanged small-iterator behavior. The changelog and version references record the update.

Estimated code review effort: 3 (Moderate) | ~20 minutes

Possibly related PRs

• PerryTS/perry#7495: Both changes modify iterator-to-array handling, but address different runtime issues.

Suggested labels: bug, parity

Suggested reviewers: thehypnoo, andrewtdiz

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/7562-iterator-truncation

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}