jan/rrt
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Summarize engine type lane families

Browse source
This commit is contained in:
Jan Petykiewicz 2026-04-21 23:07:34 -07:00
commit 2d5f5dec10
3 changed files with 128 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

3
docs/rehost-queue.md
View file

@ -13,7 +13,8 @@ This file is the short active queue for the current runtime and reverse-engineer
- The active static parser head is now the `engine_types` semantics frontier.
The repo now has structural inspectors for `.car`, `.lco`, `.cgo`, and `.cct`, but the binary side is still only partially semantic: the checked 1.05 corpus grounds `.car` fixed strings at `0x0c / 0x48 / 0x84` plus a second fixed stem slot at `0xa2` and a side-view resource name at `0xc0`, while `.lco` carries a stable primary stem at `0x04` and only conditional companion/body slots at `0x0c` and `0x12` when the leading stem slot is padded.
The checked 1.05 corpus now also splits `.car` auxiliary stems into `126` direct matches, `14` role-neutral roots, and only `5` truly distinct cases, while `.cgo` collapses into five stable scalar ladders instead of arbitrary floats.
The next honest static work is to keep promoting those fixed lanes into stable parser fields, explain the five remaining distinct auxiliary-stem cases, and decide how far the `.cgo` ladders and guarded `.lco` companion lanes can be grounded without overclaiming semantics.
The early `.lco` lane block is now partially partitioned too: only offsets `0x20`, `0x34`, `0x38`, `0x3c`, `0x44`, `0x48`, and `0x54` behave like low-cardinality buckets, while the other early lanes still look high-variance.
The next honest static work is to keep promoting those fixed lanes into stable parser fields, explain the five remaining distinct auxiliary-stem cases, and decide how far the `.cgo` ladders plus the low-cardinality `.lco` lanes can be grounded without overclaiming semantics.
Preserved checked parser detail now lives in [EngineTypes parser semantics](rehost-queue/engine-types-parser-semantics-2026-04-21.md).
Preserved checked format inventory detail now lives in [RT3 format inventory](rehost-queue/format-inventory-2026-04-21.md).

17
docs/rehost-queue/engine-types-parser-semantics-2026-04-21.md
View file

@ -30,6 +30,16 @@ first `.car` / `.lco` / `.cgo` / `.cct` inspector pass landed.
- The checked 1.05 corpus (`66` `.lco` files) shows why the guard matters: long primary stems
such as `AtlanticL` naturally spill across `0x0c`, so `0x0c` and `0x12` are not independent
fixed fields unless the earlier slot is actually zero-padded.
- The checked 1.05 corpus now also narrows the early `.lco` lane block into two categories:
- low-cardinality lanes at `0x20`, `0x34`, `0x38`, `0x3c`, `0x44`, `0x48`, and `0x54`
- high-variance lanes at `0x24`, `0x28`, `0x2c`, `0x30`, `0x40`, `0x4c`, and `0x50`
- The low-cardinality side is now preserved directly in the inspector:
- `0x20`: only `0x00000000` or `0x0000cdcd`
- `0x34`: only `0x00000000 / 0x01000000 / 0x02000000 / 0x03000000`
- `0x38`: only `0x02000000 .. 0x0a000000`
- `0x3c` and `0x44`: always `0x00000000`
- `0x48`: only ten values in the `0x003f8000 .. 0x00412000` range
- `0x54`: only `0x00000000 .. 0x09000000`
- `.cgo` looks structurally narrow right now: the checked 1.05 corpus has `37` files, all exactly
`25` bytes long, each carrying one leading scalar lane plus an inline content stem at `0x04`.
- The `.cgo` leading scalar is no longer just a loose raw count. The checked 1.05 corpus now
@ -56,6 +66,7 @@ first `.car` / `.lco` / `.cgo` / `.cct` inspector pass landed.
- conditional companion stem slot
- conditional body-type label
- early raw numeric lane block `0x20..0x54`
- low-cardinality lane counts for the enumerated subset of that block
- `.cgo`
- leading scalar lane
- content stem
@ -73,8 +84,10 @@ first `.car` / `.lco` / `.cgo` / `.cct` inspector pass landed.
- `.lco`
- whether the guarded companion-stem slot is a tender/fallback display family, a foreign reuse
key, or only a subset authoring convenience
- how much of the early numeric lane block can be promoted from raw `u32/f32` views into stable
typed semantics without dynamic evidence
- whether the newly isolated low-cardinality lane subset can be promoted from raw `u32` buckets
into named enums without overclaiming semantics
- how much of the remaining high-variance lane subset can be promoted from raw `u32/f32` views
into stable typed semantics without dynamic evidence
- `.cgo`
- whether the leading scalar ladders are enough to justify a named typed field, or whether they
should stay conservative report-only ladders until more binary/code correlation exists