19 KiB
Project Structure
Layout for the 姫狩りダンジョンマイスター (Himegari) → open AGE-engine reimplementation. The guiding rule is source vs. derived vs. our work: the shipped game is read-only input, the extracted archives and everything our tools generate are disposable/reproducible, and our code + docs live entirely apart from the game install. Nothing we produce is ever written back into the game folder.
Workspace root: S:\Game Hacking\Eushully\Himegari\
S:\Game Hacking\Eushully\Himegari\ ← workspace root (three siblings)
│
├── Himegari_Game/ ← SOURCE — pristine game install (read-only)
│ │ Never edit, move, or add to this folder. It holds ORIGINALS ONLY.
│ ├── AGE.EXE, AGERC.DLL, *.dll shipped engine (packed). Stays intact and
│ │ runnable in place — Frida launches it if needed.
│ ├── DATA1-5.ALF, APPEND01.ALF/.AAI shipped archives (~2.3 GB).
│ ├── *.BIN 49 loose patch-override scripts (v1.03) —
│ │ AUTHORITATIVE over DATA1. Plus two root-only
│ │ engine files (SYS4INI=S4IC, SYS4AB=S4AB).
│ └── *.exe (uninstallers), SAS0099.OGG … other shipped files.
│
├── extracted/ ← DERIVED (game-side) — extracted ALF contents,
│ │ ~3.9 GB, regenerable via age-reimpl/bin/BinExtractALF.
│ └── DATA1/ … DATA5/ DATA1 = 481 .BIN scripts (the corpus we parse)
│ + AGF/BMP/WAV in the others.
│
└── age-reimpl/ ← OUR WORK (everything we made lives here)
│
├── README.md navigation-only repository front door; canonical facts stay in docs/
├── run-godot.ps1 / run-godot.cmd tracked development launcher + Windows wrapper;
│ resolves Godot/game-root from parameters, environment,
│ PATH, and the conventional sibling install
├── tools/ Python tooling (parser/disassembler + extractors + VM)
│ ├── paths.py ★ central path anchor — the ONLY place that knows
│ │ where the game / extracted / build dirs are. All
│ │ tools import it; relocatable with no other edits.
│ ├── validate.py layered core/workspace/runtime/full validation driver
│ ├── test_validate.py pure resolver + validation-plan regressions
│ ├── sys4load.py loader + disassembler (opcode-decoding)
│ ├── age_opcodes.py 548-entry Kelebek AGE opcode/arg-type table (PRISTINE; never edit)
│ ├── opcodes_build.py generator/linter: vm-map/opcodes.toml -> the 4 artifacts below
│ ├── opcodes_model.py load + lint (dangling-ref, confidence-ceiling, vocab) + dependents
│ ├── age_opcodes_himegari.py GENERATED from opcodes.toml (do not hand-edit)
│ ├── globals_build.py generator/linter: vm-map/globals.toml (+auto map) -> build/globals.json, docs/global-reference.md
│ ├── story_flags.py static story-flag miner (branch-condition) -> build/story-flags-candidates.json; --bootstrap
│ ├── scjump_decode.py decode SCJUMP progression logic -> build/scjump-decisions.* ; --verify (VM cross-check)
│ ├── test_globals.py, test_opcodes.py unit tests for the globals / opcode tooling
│ ├── vm0.py headless Python VM (Phase A0); `--test` = RECOVER unit test
│ ├── extract_phase2.py batch: disasm + text + data extraction
│ ├── extract_init.py, extract_message_table.py, global_map.py …
│ │ *INIT / ID-dispatched message parsers,
│ │ global-var map builder
│ ├── validate_opcode_table*.py decode-coverage validators
│ ├── movie-corpus-gate/ C# full-corpus FFmpeg decode/lifecycle acceptance tool
│ └── probe_*.py format reverse-engineering probes (historical)
│
├── bin/ 3rd-party binaries we use (not ours, not the game's)
│ ├── BinExtractALF.exe ALF archive extractor → produces extracted/
│ ├── LzssCpp.dll its LZSS codec dependency
│ └── pe-sieve32.exe (hollowsprocess) unpacked-PE dumper — engine-RE escalation (see docs/engine-re.md)
│
├── vm-map/ VM / reverse-engineering reference artifacts
│ ├── opcodes.toml ★ CANONICAL opcode reference (hand-edited: ABI + semantics
│ │ + provenance + depends_on). Single source of truth for opcodes.
│ ├── globals.toml ★ CANONICAL global-variable registry (hand-edited: name + category
│ │ + value_domain + provenance). Single source of truth for globals/story-flags.
│ ├── kelebek1-age-shared.cpp / -disassembler.cpp upstream opcode-table source
│ └── opcode-leads.json, small-script-listings.md
│
├── docs/ all documentation
│ ├── PROJECT-STRUCTURE.md this file (where things live)
│ ├── tools-reference.md every tool: purpose, usage, I/O (operational companion)
│ ├── asset-resolution-re.md resId→file RE (graphics/audio); asset-index steering
│ ├── global-memory-re.md runtime global observation RE (SHELVED; future starting point)
│ ├── remake-architecture-and-roadmap.md THE direction doc (phases A–E)
│ ├── phase-a-slice-plan.md the current slice (A0/A1/A2)
│ ├── phase-b-framework.md natural boot/menu/session → first-gameplay framework
│ ├── platform-portability.md OS dependencies + future cross-platform readiness
│ ├── vm-mapping-plan.md the phased decode plan
│ ├── himegari-port-reference.md master reference + engine background
│ ├── name-resolution.md call-script + global-var name recovery (+ globals.toml registry)
│ ├── sys4-format-notes.md byte-level container format
│ ├── script-inventory.md what the 481 scripts are
│ ├── opcode-reference.md GENERATED from opcodes.toml (human-readable opcode reference)
│ ├── global-reference.md GENERATED from globals.toml (human-readable global registry)
│ ├── scjump-progression.md SCJUMP progression decode + SCINIT decision→scene registry
│ └── engine-re.md native-engine RE (Ghidra+MCP): dispatch table, decompiled ops
│
├── build/ DERIVED (our-work-side) — generated by tools/; disposable
│ ├── disasm/ <NAME>.asm — human-readable disassembly, one per script
│ ├── text/ extracted text:
│ │ ├── <NAME>.strings.txt all inline strings in a script
│ │ ├── dialogue.jsonl show-text lines only (the translation corpus)
│ │ └── strings.jsonl every string, tagged by source opcode
│ ├── data/ parsed data tables (*INIT → JSON)
│ ├── scripts-json/ machine-readable full dumps (on demand via --json)
│ ├── textures/ AGF-to-BMP stills (convert_agf.py) - diagnostic pixel oracle only
│ ├── engine-dump/ UNPACKED engine dump (frida/dump_engine.py): range_<base>.bin + manifest.json
│ ├── asset-index.json, asset-sections.json asset resolver data (parse_sys4ini / resolve_asset)
│ ├── global-var-map.{json,md} partial global-variable name map (auto shape inference; feeds globals.toml merge)
│ ├── globals.json GENERATED from globals.toml merged over global-var-map (sys4load labels + C# VM)
│ ├── story-flags-candidates.json GENERATED by story_flags.py (ranked story-flag review surface)
│ ├── scjump-decisions.{json,md} GENERATED by scjump_decode.py (progression decision table)
│ ├── opcodes.json GENERATED from opcodes.toml (machine view for the C# VM)
│ └── manifest.json, opcode-coverage.md (opcode-coverage.md GENERATED from opcodes.toml)
│
├── engine/ DELIVERABLE — the .NET VM core (AgeEngine.sln: Age.Engine / Age.Cli / tests)
│ ├── Age.Engine/Model/ retained graphics state plus separately navigable public
│ render, transition, diagnostic, and persistence contracts
│ ├── Age.Engine/Sys4/ runtime game-root selection, catalog parser, loose-first bounded
│ ALF asset store, script provider, AGF/LZSS and Windows CUR decoders,
│ and resource facade
│ ├── Age.Engine/Text/ backend-neutral glyph-mask requests/results, deterministic AGE
│ compositor/layout, retained live/History glyph and wait-atlas
│ publication, and bounded font/glyph cache primitives
│ ├── Age.Engine/Persistence/ native S3SD/S4SD + S3RT codecs, shared payload/ReadTextDB,
│ layout-3 numbered state/history/gfx, DAT/STH pair + BMP codec,
│ and profile-owned state
│ └── Age.Engine.Text.Windows/ optional exact Windows GDI gray-4 reference backend; kept
│ outside the platform-neutral VM/runtime core
├── native/ authored native runtime boundaries
│ └── age_movie_ffmpeg/ project-owned FFmpeg C ABI, immutable Windows/Linux dependency manifests,
│ and target bootstrap/build scripts (outputs stay under disposable build/)
├── tools/frida/ runtime-capture + engine-dump scripts (see tools/frida/README.md)
└── godot/ DELIVERABLE — the Godot/C# ADV front-end (references Age.Engine
plus the optional exact Windows text adapter). Its assembly embeds
generated build/opcodes.json through HimegariRuntimeMetadata, so
exports do not need repository discovery. Also contains the portable
TextServer glyph adapter, per-profile font-substitution configuration,
and TITLE-only F4 debug scene launcher
The Godot deliverable includes Himegari.sln because Godot's .NET exporter requires a solution,
export_presets.cfg for the Linux x86_64 target, and build/.gdignore so disposable capture frames below
the project do not enter its import or export scan. tools/export-linux-x64.ps1 produces and validates the
complete disposable artifact under build/export/linux-x64/.
godot/Main.cs retains the front-end's startup and runtime coordination. Behavior-neutral partial-class
companions keep cohesive surfaces independently navigable without changing the Godot node type or invocation
paths: godot/Main.SelfTest.cs owns the synthetic threaded/headless regression harness, while
godot/Main.Audio.cs owns BGM, voice, sound-effect, mixer-routing/persistence, and audio-bus control, and
godot/Main.Movie.cs owns decoder staging, movie frame/audio publication, completion, and teardown.
godot/Main.Compositor.cs owns retained/GPU/software composition state, texture resolution and caching,
surface-transition drawing, raster helpers, and compositor decision logging. godot/Main.Input.cs owns
Godot input routing, locator/debug hotkeys, debug-scene dispatch, cursor control, native alerts, and
full-width text entry.
godot/GodotAdvHost.cs retains cross-domain host coordination. Its partial-class companion
godot/GodotAdvHost.AdvText.cs owns live/retained ADV text, surface glyph rasterization and caching,
history presentation, message-window alpha, and retained wait-indicator configuration/publication.
godot/GodotAdvHost.PresentationInput.cs owns script/presentation synchronization, waits and timing,
message-skip/input services, cursor and foreground waits, frame/backbuffer publication, transitions, and
scene-context lifecycle coordination. godot/GodotAdvHost.Surfaces.cs owns decoded-image caching, mutable
surface pixels/resources/dimensions, fill/copy/resolve operations, render-target publication, and surface/range
teardown. godot/GodotAdvHost.Movies.cs owns movie surface bindings, ordinary and modal playback, movie-mask
transitions, diagnostic snapshots, frame/completion publication, and mask teardown.
godot/GodotAdvHost.Audio.cs owns BGM, voice and SFX resolution/dispatch, delayed voice state, volume/routing
control, and blocking BGM fades; presentation/input retains the message-skip, reset, and frame-pulse consumers
of that state through the sealed partial class.
engine/Age.Engine/Model/GfxState.cs retains cross-domain retained-graphics coordination.
engine/Age.Engine/Model/GfxState.Contracts.cs owns its public render, transition, diagnostic, persistence,
animation, numeric-glyph, and handle-range contracts. engine/Age.Engine/Model/GfxState.Surfaces.cs owns surface
resource/color-key state, created/reloadable classification, movie stop-time metadata, render-target/tile
configuration, and surface lifecycle operations. engine/Age.Engine/Model/GfxState.RetainedObjects.cs owns the
retained-object record and registry/index, range-transform state, object creation/query/clone/erase and draw
binding, plus numeric-glyph object generation. engine/Age.Engine/Model/GfxState.Animation.cs owns the shared
animation clock, object color/source-cell/matrix/cyclic channels, animation control and forced completion, and
the interpolation helpers consumed by retained-scene sampling. engine/Age.Engine/Model/GfxState.Presentation.cs
owns surface/movie transition queues, presentation activity and click-skip logic, diagnostics and dirty-reason
accounting, and visible retained-scene snapshots.
engine/Age.Engine/Vm/VirtualMachine.cs retains VM lifecycle, cross-domain state, and the proven top-level
opcode dispatcher. Its partial-class companion engine/Age.Engine/Vm/VirtualMachine.Audio.cs owns VM audio
state, BGM restart semantics, and the BGM/voice/SFX/mixer opcode handler; Step retains the audio labels and
routes that family into the handler. engine/Age.Engine/Vm/VirtualMachine.Movie.cs owns modal/asynchronous/
positioned movie playback, movie surface metadata/activity queries, and movie-mask transition dispatch; Step
likewise retains and routes the movie labels. engine/Age.Engine/Vm/VirtualMachine.Surface.cs owns surface
allocation/loading, texture binding and sizing, mutable surface fill/copy, render-target control, and transient
surface release; its labels remain at their existing dispatcher positions around the retained numeric-glyph and
object cases. engine/Age.Engine/Vm/VirtualMachine.RetainedObjects.cs owns retained-object registry queries,
default-slot and geometry mutation, direct and range transforms, clone, and erase dispatch; animation, surface,
ADV-binding, and presentation labels remain in their respective dispatcher groups.
engine/Age.Engine/Vm/VirtualMachine.Animation.cs owns retained spritesheet and color channels, timed and cyclic
transforms, per-object animation control, frame-time sampling, and the shared animation-clock opcode handler.
The disposable build/page-map-<SCENE>.jsonl files are produced by editor/development Godot runs and map
runtime ADV page ordinals to their authoritative script offsets for tools/locate_page.py. Packaged exports
have no repository output tree and write their automatic maps below user://diagnostics/page-maps instead.
Conventions
- Three-way separation.
Himegari_Game/= untouched originals;extracted/= game-derived data (regenerable, game-side);age-reimpl/= everything we authored. The first two are consumed, never modified. - Tools never hard-code paths.
tools/paths.pyderivesGAME_DIR,EXTRACTED,DATA1,BUILD, etc. from its own location. To point the tools at a different install, edit that one file. The whole tree can be relocated without touching any other tool. - Path references in docs are
age-reimpl/-relative (e.g.tools/sys4load.py,build/text/dialogue.jsonl) unless they name a game/extracted path explicitly. - Authoritative script copies: where a script exists both as a loose
.BINin the game folder and underextracted/DATA1/, the game-folder copy (patch v1.03) wins.paths.scripts()resolves this automatically (overrides win). build/andextracted/are disposable.build/regenerates viatools/extract_phase2.py(orsys4load.py);extracted/regenerates viabin/BinExtractALF.exeon the.ALFfiles. Safe to delete and rebuild; do not hand-edit.- Opcode knowledge is edited ONLY in
vm-map/opcodes.toml(ABI + semantics + provenance +depends_on). Runtools/opcodes_build.py --buildto regenerate the shim (tools/age_opcodes_himegari.py), machine JSON (build/opcodes.json), reference (docs/opcode-reference.md), and coverage.--lintchecks dangling deps / confidence-ceiling / vocabulary. Kelebek'stools/age_opcodes.pystays pristine. - Global-variable knowledge is edited ONLY in
vm-map/globals.toml(name + category + value_domain + provenance). Runtools/globals_build.py --buildto regeneratebuild/globals.json(sys4load labels) anddocs/global-reference.md;--lintchecks vocabulary / auto-shape≠high / dangling deps. Curated entries override the auto shape map (build/global-var-map.json). Discover story-flag candidates withtools/story_flags.py. - Encoding: all generated text is UTF-8 (source strings are cp932/Shift-JIS, decoded on
extraction). Run Python as
py -3.11 -X utf8. - The game install is a runnable unit — do not relocate
AGE.EXE/*.ALF/DLLs relative to each other, or the game (and any Frida work) breaks.