GameCult/CultLib
Watch
9
0
Fork
You've already forked CultLib
 0
Code Issues Pull requests 1 Projects Releases Packages Wiki Activity Actions

Rust: a payload is one MessagePack value (bytes-after-rs) #131

Merged
libby merged 6 commits from hands/bytes-after-rs into main 2026-11-12 05:43:56 +00:00
Conversation 0 Commits 6 Files changed 8 +346 -30
libby commented 2026-11-12 04:43:59 +00:00
Member
Copy link

cultlib-gaps:follow_up:bytes-after-leaves, spec cultlib-gaps:cut_spec:cut-bytes-after-rs.r1, ruling trailing-bytes-refuse: cultcache-rs refuses a record payload with bytes after its first MessagePack value, by count, naming type and key, never echoing the tail, through one owner check_record_payload (depth first, then one value). Single-file, redb and load_soa run it at load for every record (as C# and Python refuse at open), and so does every raw-envelope/batch put. Vectors: bytes-after-outcomes.msgpack (4 rows) walked.

Moved with it, stated: depth is now refused at load too (a store holding a past-bound payload no longer loads), and a cache whose last pull was refused refuses every write and removal (CultCacheLoadRefused, one require_loaded gate) until a pull succeeds. Before, a refused pull left an empty writable cache and the next put_prepared_batch replaced the store; this closes that for the duplicate-index and retype refusals too. Open Low: a never-pulled cache still has that hole (follow-up cache-must-pull-before-write).

Semver: Unreleased/Breaking in the cultcache-rs CHANGELOG; implies a minor bump while 0.x (0.3.1 to 0.4.0), cultnet-rs 0.3.0 to 0.4.0.

Soul: verdicts cut-bytes-after-rs.s1 to .s3 (s1 load forked on index; s2 refused-load batch wiped the store; s3 holds). Reports h1 to h3. Full cultcache-rs, cultnet-rs, cultmesh-rs suites pass (914 tests). Expect a rebase on #127 (slot-readonly-rs) at decode_payload and the changelog head.

🤖 Generated with Claude Code

`cultlib-gaps:follow_up:bytes-after-leaves`, spec `cultlib-gaps:cut_spec:cut-bytes-after-rs.r1`, ruling `trailing-bytes-refuse`: cultcache-rs refuses a record payload with bytes after its first MessagePack value, by count, naming type and key, never echoing the tail, through one owner `check_record_payload` (depth first, then one value). Single-file, redb and `load_soa` run it at load for every record (as C# and Python refuse at open), and so does every raw-envelope/batch put. Vectors: `bytes-after-outcomes.msgpack` (4 rows) walked. Moved with it, stated: depth is now refused at load too (a store holding a past-bound payload no longer loads), and a cache whose last pull was refused refuses every write and removal (`CultCacheLoadRefused`, one `require_loaded` gate) until a pull succeeds. Before, a refused pull left an empty writable cache and the next `put_prepared_batch` replaced the store; this closes that for the duplicate-index and retype refusals too. Open Low: a never-pulled cache still has that hole (follow-up `cache-must-pull-before-write`). Semver: Unreleased/Breaking in the cultcache-rs CHANGELOG; implies a minor bump while 0.x (0.3.1 to 0.4.0), cultnet-rs 0.3.0 to 0.4.0. Soul: verdicts `cut-bytes-after-rs.s1` to `.s3` (s1 load forked on index; s2 refused-load batch wiped the store; s3 holds). Reports h1 to h3. Full cultcache-rs, cultnet-rs, cultmesh-rs suites pass (914 tests). Expect a rebase on #127 (slot-readonly-rs) at `decode_payload` and the changelog head. 🤖 Generated with [Claude Code](https://claude.com/claude-code) 
One private payload check (depth, then one value) behind decode_payload and
raw-envelope/batch admission; refuses a tail by count, never walks or quotes it.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Single-file, redb and load_soa refuse a tail at load through check_record_payload,
indexed or not; depth beats a tail is pinned; the dead end guard is gone.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
A cache whose load was refused refuses every write and removal; depth pinned at load; map names the live owner
Some checks failed
CultMesh portability and performance / ubuntu-latest reactive semantics and leases (pull_request) Has been cancelled
Details
bcc3052aa0 
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Merge origin/main (09b825d6, slot-readonly-rs) into hands/bytes-after-rs
All checks were successful
CultMesh portability and performance / ubuntu-latest reactive semantics and leases (pull_request) Successful in 20m27s
Details
acad89c713 
Both sides kept: the payload check (depth, one value) runs before the mark scope opens
and before the held read-only check; the load gate covers slot-held stores.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
libby merged commit 363a8d6fe5 into main 2026-11-12 05:43:56 +00:00
Sign in to join this conversation.
Reviewers
No reviewers
Labels
Clear labels
No items
No labels
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference
GameCult/CultLib!131
No description provided.