Simulation (and ArenaSimulation in the balancing tool) now owns the factory's
world data; BuildingSystem holds a reference to it. This is what lets the systems
that operate on the data be handed the same state — phase 3's construction and
deconstruction systems, and later the ecs/system/ classes that today take a
BuildingSystem& only to query it.
reset() clears the state alongside m_admin and m_beltSystem, matching how the
subsystems were already rebuilt from scratch.
Falls short of the tick-argument form I sketched: BuildingSystem still reaches
the data through a member reference rather than a parameter. Making it truly
stateless means the const query surface has to find the data some other way, and
that surface is large — findBuilding alone has 64 call sites, with findSite,
getAllBuildings, getAllSites, isTileOccupied and the rest behind it. Doing that
needs a queries facade behind Simulation::getBuildings() so the callers do not
all move, which is its own decision rather than a side effect of this one.
The constructor gains a parameter, so the four owners and the three test fixtures
that build a BuildingSystem directly are updated; the 33 files that only use one
are untouched.
Verified with a golden-checksum capture before and after — all four sample ticks
identical.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
BuildingSystem is the one system in the codebase that owns the world data it
operates on. The ecs/system/ classes already do the opposite — AiSystem and
SalvagerSystem hold config and their own scratch, and take EntityAdmin and the
other systems as tick arguments — so this is bringing the outlier in line, not
inventing a pattern.
Phase 1 of that: the buildings vector, both work queues and the tile grid move
into a FactoryState struct, still owned by BuildingSystem. Every method reaches
through m_state. The public API is untouched, so none of the 33 files that
reference BuildingSystem needed a change.
DeconstructionEntry moves out of BuildingSystem's private section into
FactoryState.h, since the queue that holds it lives there now.
m_asteroidWidth_tiles stays on the system: it is not checksummed and is a cached
placement bound derived from config and the expansion count, not factory data.
The intent is for Simulation to own FactoryState and pass it into the tick
methods, leaving the systems stateless over it. FactoryState.h notes explicitly
that this is a data/behaviour split and not a step toward putting buildings in
the entity model, which architecture.md rules out.
Verified with a golden-checksum capture before and after — all four sample
ticks identical.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
Eleven methods maintained m_tileOccupancy by hand — place, deconstruct,
removeBuilding, placeImmediate, tickDeconstruction, findRotateInPlaceTarget and
tryDirectCoupleDeposit all indexed a raw std::map<std::pair<int,int>, BuildingId>
directly, so the invariant "occupancy stays in sync with placement" was
re-implemented at every call site. They now ask and tell a small owned index
instead: occupy / release / isOccupied / findOwner.
BuildingGrid is a member of BuildingSystem, not a peer system: it has no
per-tick behaviour and nothing outside BuildingSystem touches it.
The internal keying stays std::pair<int,int> rather than moving to QPoint. The
checksum folds the entries in map iteration order, so the comparator is part of
the determinism contract; changing it is a separate decision, not a side effect
of this move. Verified with a golden-checksum capture before and after — all
four sample ticks identical.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
Four of the five file-local helpers in BuildingSystem.cpp had duplicates
elsewhere: isAutoRecipeBuildingType and isBeltSubsystemType were re-spelled as
isAutoRecipeBuilding and isBeltLike in SelectedBuildingPanel.cpp, and
outputBodyTile was copied verbatim as portBodyTile in GameWorldView.cpp. Same
predicates, different names, so a change to one would silently not reach the
others.
The two BuildingType predicates move to BuildingType.h, which already hosts the
free functions over that enum and is already included by both lib and ui. The
port geometry moves to a new PortGeometry.h; inputBodyTile has no duplicate but
is outputBodyTile's counterpart and belongs beside it — the sim moves items
across the port edge and the renderer draws the virtual belt there, so the two
must agree on which tile a port owns.
inputLaneEntryFree stays file-local: single use, and tied to BeltItemSlot rather
than to building types or port geometry.
Verified the six moved bodies are character-identical to their originals.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
The two selection categories used to arbitrate ownership of the panel by
poking each other's widgets: buildFieldSelection() called clearContent()
and buildEmpty(), buildEmpty() hid the four entity widgets, and
hideAllWidgets() hid the scrap label. Splitting the halves apart without
naming an arbiter would only have spread that across a class boundary.
SelectedBuildingPanel is now the sole arbiter. It still receives all
three selection events, forwards the two field ones to the embedded
FieldSelectionPanel, and drops its own selection and content as soon as
the field panel reports a selection (yieldToFieldSelection), mirroring
what onSelectionChanged() already did in the other direction. The field
panel decides only what to render and whether it is visible at all.
Dropping the field branch of refreshSelectionDisplay() is behaviour
preserving: whenever the field category owns the panel, m_singleBuildingId
is null, so the building refresh returns immediately anyway.
clearContent() and buildEmpty() became identical once the cross-half
hiding was gone, so only buildEmpty() remains.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
SelectedBuildingPanel has grown to 1200 lines by carrying two unrelated
selection categories. Introduce the field half as its own widget first,
so the cut-over is a separate, reviewable step.
The panel owns only its own selection state and widgets: it renders the
single-object stats panel (ship, station, debris) or the multi-object
count summary, subscribes to the tick/commands-applied refresh signals
and the debug-draw toggle, and hides itself while it has no selection.
Which category owns the side panel is not its decision - the parent
feeds it through setSelectedEntities/setSelectedDebris/clearSelection.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
The section documented a 5-method port interface and claimed "no other system
ever asks what is on tile X". The real surface is 15 methods and peekItem does
ask exactly that, so the doc was misleading about a subsystem it exists to
explain. It also showed a tryPutItem signature that no longer matches.
Describes what is there now, grouped by purpose, and separates the two claims
that had been conflated: item transport is still port-only, but tile topology
is genuinely coupled to BuildingSystem because belts are Buildings for cost and
construction. The v2 migration note is qualified accordingly.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
The blanket ban could not be satisfied: a lambda's type is unnameable, and
std::function is the wrong substitute in per-tick hot paths, so around seven
named local lambdas in lib violated the rule with no way to comply. Spelled-out
iterator types were being followed inconsistently in the same file.
Carving out the two cases makes the guideline enforceable rather than silently
broken. No code is changed here; existing spelled-out iterator types stay valid.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
The scripted session only placed buildings, so every unlock container stayed at
its initial value for the whole run and the checksum never saw them change. It
now destroys the enemy stations twice and takes the offered schematic choice,
so awarded groups, per-schematic levels and the derived recipe/item sets are
exercised too.
Adds a test that pins down that unlock state actually reaches the checksum: two
sessions in lockstep, one takes the choice, checksums must diverge. The
scripted-session tests cannot show this themselves — they compare runs against
each other, so they pass whether or not UnlockState is in the fold. Verified by
temporarily removing the fold: the new test fails, the old two do not.
The first choice is asserted rather than assumed, so a config change that stops
offering a group fails loudly instead of silently dropping the coverage.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
Simulation now owns an UnlockState member (constructed before
initializeSubsystems(), since BuildingSystem's spawn-gating lambda
calls into it via isSchematicUnlocked instead of poking the old
m_schematicLevels map directly). The public isXUnlocked accessors
become one-line forwards, applySchematicChoice's group-awarding block
becomes a single awardUnlockGroup() call, and generateSchematicChoices
(which still owns m_rng and must not change call ordering) now reads
group state and builds options through UnlockState.
The checksum fold at computeStateChecksum's schematic/unlock section
had to move together with the containers it reads; UnlockState::
appendChecksum makes the identical seven appendSchematicMap/
appendStringSet calls in the identical order, so the fold is
unaffected. Verified via a temporary golden-checksum test case
(added, checked, then removed) that tick1/100/999/1999 checksums for
seed 12345 are byte-identical to pre-refactor.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
Simulation.h/.cpp had grown a large block of schematic/unlock state
(containers, the implicit-unlock traversal, checksum folding) that has
nothing to do with tick orchestration. Split it into its own class so
Simulation stays legible as "tick orchestration + subsystem handles".
This commit only adds the new UnlockState.h/.cpp (registered in
CMakeLists.txt) with the containers, types, and logic moved in
verbatim; nothing references it yet, so this is a no-op for behavior.
Simulation is wired to use it in the next commit.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
The names are generic (makeError, requireInt, parseFile), and at global scope
with external linkage they would form an overload set with the same-named
anonymous-namespace helpers in VisualsLoader.cpp and BalancingConfig.cpp the
moment either file includes TomlHelpers.h — silently, since the signatures
differ. Namespacing keeps that door shut.
Call sites are qualified explicitly rather than pulled in with a using
directive, matching how utility::getRandomInt and friends are already called.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
Finishes the ConfigLoader.cpp domain split. ConfigLoader.cpp now holds
only the cross-domain validateUnlocks pass and loadFromDirectory
orchestrator, as intended — everything else lives in its own
ConfigLoader<Domain>.cpp. Pure move; no logic change.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
Continues the ConfigLoader.cpp domain split. StatEntry/kKnownStats are
only used by loadModules, so they move along as a domain-local
anonymous-namespace table rather than into TomlHelpers. Pure move; no
logic change.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
Continues the ConfigLoader.cpp domain split. parseRotationString and
parsePlacedModules are only used by loadShips, so they move along as
domain-local anonymous-namespace helpers rather than into TomlHelpers.
Pure move; no logic change.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
Continues the ConfigLoader.cpp domain split. parseRecipeOutputs is only
used by loadRecipes, so it moves along as a domain-local anonymous-
namespace helper rather than into TomlHelpers. Pure move; no logic
change.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
ConfigLoader.cpp had grown to 877 lines by mixing generic TOML-parsing
helpers (used by every per-file loader) with per-domain parsing logic.
Splitting the file per config domain first requires pulling out the
helpers shared by two or more domains, so each domain .cpp can include
them without duplication. Pure move: no logic, message, or ordering
changes.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
TickAdvanced, BuildingBlocksChanged, ExpansionCostChanged, BossWaveUpdated and
ArtifactCountChanged all duplicated state the Simulation already owns. With
every subscriber re-reading from the sim, the fields had no readers left, so
the five events become payload-free refresh signals and the emitters keep the
values only as locals for change detection.
This makes the convention uniform: a state-change event backed by the
simulation carries nothing. Events whose state lives in the view (selection,
game speed, deconstruct and debug-draw modes) keep their payloads, since there
is no sim getter behind them.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
The last three payloads HeaderBar still consumed as truth. All are backed by
Simulation getters (getCurrentTick, getArtifactCount, getBossWaveCounter,
getBossCountdownTicks) and the win count by world.artifacts.artifactWinCount,
so the handlers now re-read rather than trust the event.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
BlueprintPanel was the second panel caching BuildingBlocksChangedEvent's
payload as truth; it already held a Simulation*, so refreshButtonStates()
now re-reads getBuildingBlocksStock() at the point of use, per the "events
are refresh signals" rule.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
ShipLayoutPreview was the one widget the findFooDef sweep missed: it held a
bare const std::vector<ModuleDef>* rather than a config, so the shared
ModulesConfig::findModuleDef was not a drop-in and the file-local copy
survived. Hold the ModulesConfig instead and call the shared finder.
The sole caller already had the ModulesConfig one dereference away.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG
BuildingSystem::findBuildingDef was a byte-equivalent re-implementation of
BuildingsConfig::findBuildingDef. Same cleanup as the GameWorldView copy;
this one sits in the sim layer, so it was missed by both earlier passes.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GH8ZMRY3vhxxXcaUxBqxkk
The Restart button on the Win dialog called Simulation::reset() directly,
while the escape-menu and Game Over restarts enqueue a ResetCommand. That
bypassed the command chokepoint every sim mutation is supposed to flow
through (docs/replay_design.md), so a post-win restart was never recorded
into the replay stream, and it had UI code calling a sim mutator directly.
Mirror the Game Over path instead. The manual resetForNewGame() call goes
away with it: GameWorldView::onFrame already resets the view when it drains
a Reset command.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GH8ZMRY3vhxxXcaUxBqxkk
The two executors were line-for-line duplicates: orbit the behavior target,
then hand it to the owner's in-range modules. Both now call a templated
executeOrbitAndAssign<Behavior, ModuleComponent>; the view types, iteration
order, and sequence of component writes are unchanged.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GH8ZMRY3vhxxXcaUxBqxkk
Four MainWindow sites hand-rolled snapshot speed / setGameSpeed(0) / modal /
restore + resetFrameTimer, with the restore duplicated on early-return paths.
ModalPauseScope does it via RAII, with restore()/release() for the two sites
that must restore early or not at all.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GH8ZMRY3vhxxXcaUxBqxkk
The three restart paths each repeated the same config + visuals reload with
its own try/catch and error dialog. Only that shared part is extracted; each
site keeps its own follow-up (ResetCommand vs. direct Simulation::reset) and
its own error-path cleanup.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GH8ZMRY3vhxxXcaUxBqxkk
Four separate caches rasterized the same item SVGs, one of them rebuilt on
every recipe-dialog open. MainWindow now owns one cache and hands a
non-owning pointer to HeaderBar, BuildButtonGrid, GameWorldView and
RecipeSelectionDialog.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GH8ZMRY3vhxxXcaUxBqxkk
HeaderBar was the only panel caching event payloads as truth. It now holds
a const Simulation* and re-reads getBuildingBlocksStock() /
getCurrentExpansionCost() in the handlers, per the "events are refresh
signals" rule.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GH8ZMRY3vhxxXcaUxBqxkk
The identical 7-line TunnelLookup lambda existed in updateTunnelGhost and
drawSelectedTunnelConnections; it is now GameWorldView::makeTunnelLookup.
The tile key moved from std::pair<int, int> to QPoint with the existing
QPointCompare comparator, dropping the manual packing at every site.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GH8ZMRY3vhxxXcaUxBqxkk
The constructor and reset() held a character-for-character identical
26-line subsystem construction block, including three capturing lambdas.
Both run before the first tick, so the closures can be shared. Order is
unchanged.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GH8ZMRY3vhxxXcaUxBqxkk
Ships/Modules/RecipesConfig now carry lookup helpers mirroring
BuildingsConfig::findBuildingDef. The hand-rolled linear scans in
BuildingSystem, ShipSystem, ShipStatsCalculator, ThreatCostCalculator,
ShipLayoutDialog, SelectedBuildingPanel and SchematicChoiceDialog now
call them instead.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GH8ZMRY3vhxxXcaUxBqxkk
GameWorldView::findBuildingDef was a byte-equivalent re-implementation of
BuildingsConfig::findBuildingDef. All call sites now use the config helper,
matching SelectedBuildingPanel and BlueprintPanel.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GH8ZMRY3vhxxXcaUxBqxkk
rotateInPlace re-implemented the belt-tile re-registration switch inline
instead of calling reregisterBeltTile, and its splitter branch omitted the
setSplitterFilters call the canonical version has. Since an operational
splitter keeps its filters only in BeltSystem, removeTile discarded them and
rotating a configured splitter silently reset it to "accept all".
Replace the duplicated switch with a call to reregisterBeltTile, capturing
the filters beforehand via getSplitterInfo — the same idiom deconstruct
already uses. This removes the second copy of the switch that allowed the
two to drift apart in the first place.
Add a regression test; the existing [rotate-in-place] cases covered belt
tiles only, which is why this went unnoticed.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GH8ZMRY3vhxxXcaUxBqxkk
