Backport the WoW global strsplit(sep, str [, pieces]) to 1.12, ported from
3.3.5's strsplit (FUN_00816a60): split str on any character in sep, returning
the pieces as multiple values; pieces>0 caps the count with the unsplit
remainder as the final piece (0/omitted = unlimited, 1 = whole string).
Consecutive/trailing delimiters yield empty pieces, matching retail.
Two deliberate changes from the 3.3.5 source:
- Skip its lua_settop(L, 0). Popping the args un-roots the source string, so a
GC step during a pushlstring could free the buffer we're still scanning.
Leaving the args on the stack keeps the source GC-rooted; Lua returns the
top N (the pieces) regardless of the args below.
- Bind lua_checkstack (0x006F2F30, __fastcall(L, size) -> int, found via its
cmp ...,0x800 LUA_MAXCSTACK test) as Game::Lua::CheckStack and guard every
push with it, erroring "strsplit(): Stack overflow" like 3.3.5 on genuine
exhaustion (strsplit pushes an unbounded number of results). Reusable for
any future multi-return C function.
Registered on both Lua states. Verified in-game across basic split, pieces
cap, multi-char separator set, empty middle piece, and empty-separator cases.
Post numStacks auctions of `quantity` each from one source stack in a
single call (retail multi-sell style). Vanilla's CMSG_AUCTION_SELL_ITEM
(opcode 0x256, verified from Script_StartAuction 0x004CE770) has no count
field like WotLK's, so the server posts the whole item object a GUID points
at; posting a partial quantity requires the client to split it into a real
bag item first. Both the split and the post are server-authoritative and
async, so this runs a WorldTick state machine:
Idle -> (source > quantity) split `quantity` into a reused free general
bag slot via Item::Swap::MoveCount (CMSG_SPLIT_ITEM)
-> SplitWait until the stack materializes
-> point the sell-slot globals at it and call the engine's own
Script_StartAuction (reusing all its per-stack validation)
-> PostWait until the server pulls it from the slot -> next stack.
The final exact stack posts the source object directly, no split.
At most one reused free slot is needed (each post frees it for the next
split); a whole-stack post needs none. Per-step ~10s timeout aborts
gracefully. Progress via retail event names AUCTION_MULTISELL_START /
_UPDATE(done,total) / _FAILURE. Immediate rejections return (false, reason).
New offsets: FUN_SCRIPT_START_AUCTION, VAR_AUCTION_AUCTIONEER_GUID_LO/HI,
VAR_AUCTION_DURATION_TABLE ({120,480,1440} minutes = 2h/8h/24h).
v1 limits: single source stack (no cross-bag aggregation), general-purpose
bags only for the temp split, one job at a time.
Backport math.fmod to 1.12's Lua 5.0. Pure 5.0->5.1 rename: 5.0 ships the
identical C function as math.mod (mathlib entry, FUN 0x007FAFD0 —
checknumber(1/2) -> fmod -> pushnumber); 5.1 renamed the Lua-facing entry
to math.fmod. Registered as a direct alias of that function pointer on both
the in-game and glue states (BaseLib::MathLib), no wrapper. math.mod stays
available, so code written for either name works.
Backport the Lua 5.1 pattern helpers to 1.12's Lua 5.0 (BaseLib::StringLib),
both reusing pattern machinery 5.0 already ships:
- string.match(s, pattern [, init]) calls the engine's str_find
(0x007FC3B0) and reshapes the result — captures, or the whole match when
the pattern has none, or nil. match and find are the same operation in
lstrlib.c; find returns start/end, match returns the captures.
- string.gmatch is 5.0's string.gfind verbatim (renamed in 5.1), registered
as a direct alias of the engine gfind (0x007FCFA0) — no wrapper.
Registered on both the in-game and glue Lua states. Only the
string.foo(s, ...) form works, not ("x"):foo(...): WoW's VM strips
type-metatables for non-table/userdata values (both lua_setmetatable and
luaT_gettmbyobj hard-return no-metamethod for strings), so string method
sugar can't be enabled client-side without hooking a hot VM core function.
Documented as a known limitation.
Drops the Util/String.lua polyfill (was just string.gmatch = string.gfind),
now redundant, and its .toc entry.
Backport the modern structured-table accessor. Vanilla's global
GetContainerItemInfo (Script_GetContainerItemInfo, 0x004F9670) returns
only a flat texture/itemCount/locked/quality/readable tuple; assemble the
full ContainerItemInfo table from the same descriptor + item-cache sources
plus the fields 1.12 never surfaced (hyperlink, hasNoValue, itemID,
isBound, itemName).
Live per-instance fields (stackCount, isLocked, isBound, itemID, hyperlink)
are always present; cache-derived fields (iconFileID, quality, hasLoot,
hasNoValue, itemName) are omitted/nil until the item's static data lands.
Reuses Item::StatAccum::FetchRecord and Item::Link for the cache peek and
link/name build (no new duplication); gathers all data into locals before
building the table so the engine link/name builders never interleave with
the half-built table on the Lua stack.
readable mirrors the engine branch at 0x004F97BD (static page text OR the
per-instance readable-text id at descriptor +0x9C, added as
OFF_DESCRIPTOR_READABLE_TEXT_ID). hasLoot is a best-effort static LOOTABLE
approximation (vanilla can't track post-loot emptiness client-side).
GetSpellBonusDamage(school) reads the player's per-school spell damage
directly from the CGPlayer sub-struct at [player+0xE68]+0xFD4 (POS) / +0xFF0
(NEG), net = POS-NEG. That field isn't in the broadcast descriptor (+0x110
reads zero — the vanilla server never sends it there), but the client keeps a
fully-computed copy at +0xE68 with gear/enchants/buffs/talents/sets baked in.
Same source nampower's GetSpellPower reads; verified 613 generic / 647 Fire on
a geared mage. 1-based school, matching 3.3.5's Script_GetSpellBonusDamage.
GetSpellBonusHealing() has no such field (vanilla lacks a healing-done field
entirely), so it's derived to mirror the server's SpellBaseHealingBonusDone:
- flat healing: SPELL_AURA_MOD_HEALING_DONE (135) off gear equip-effects,
item enchants, random suffixes, and active buffs
- Spirit conversion: SPELL_AURA_MOD_SPELL_HEALING_OF_STAT_PERCENT (175) x
total Spirit / 100 (Priest Spiritual Guidance)
- Armor conversion: SPELL_AURA_MOD_SPELL_HEALING_OF_ARMOR_PERCENT (199) x
armor / 100 (Turtle Ironclad)
Percents are read straight from Spell.dbc, so it's generic across any
aura-175/199 talent/buff including Turtle customs, and gated on SPELL_ATTR_
PASSIVE for the known-spell-bitmap walk. Spirit/Armor come from UNIT_FIELD_
STAT4 (+0x250) / UNIT_FIELD_RESISTANCES[0] (+0x254).
Item flat healing routes through the shared Item::StatAccum engine rather than
re-implementing the item spell/suffix/enchant walk: extract a reusable
ApplyEnchant() (from ApplyRandomSuffix) plus a Value() reader, and read
ITEM_MOD_SPELL_HEALING_DONE_SHORT.
The healing sum is expensive (StatAccum over 19 items + buff table + the 27k
known-spell bitmap), so it's cached and recomputed only when a player stat
input changes. New Player::StatSignal (a pull-based epoch) is bumped from the
aura hooks (player-guarded), the equipment observer, and the spell learn/
unlearn writers (0x4B25B0 / 0x4B2C50) -- so buffs, gear, and talent respecs
all invalidate instantly. A coarse world-tick recompute remains only as a
backstop against a missed signal.
The PlayerLocation-taking C_PlayerInfo accessors. Resolve the location to a
unit token - unit locations use their token, GUID locations via
Unit::Identity::TokenFromGUID (while the GUID has a live token) - then:
- GetName / GetSex / IsConnected tail-call the engine's Script_UnitName /
Script_UnitSex / Script_UnitIsConnected (the GetFactionInfoByID
forward-the-return pattern), reusing all vanilla unit logic. GetSex must go
through the engine because the raw UNIT_FIELD_BYTES_0 gender byte is 0/1
while the API reports 2/3. IsConnected defaults to the local player when
called with no location.
- GetRace / GetClass read UNIT_FIELD_BYTES_0 for the numeric ids vanilla's
UnitRace / UnitClass don't return (login-time mirror globals for "player",
else the descriptor bytes); GetClass resolves className / classFile from
ChrClasses.dbc like C_CreatureInfo.GetClassInfo. classID/raceID round-trip
with C_CreatureInfo.
Only the unit and guid location kinds resolve on 1.12; the chat/who/BG/
community/voice kinds return nil. New file - player/Info.cpp is the separate
NameCache/GetPlayerInfoByGUID module. Adds the three Script_Unit* offsets
from raw_globals.txt.
Backport the modern C_NewItems system - IsNewItem / RemoveNewItem /
ClearAll - flagging items freshly added to bags as "new", cleared once
acknowledged. Vanilla has none of the machinery (no DBC, no server data),
so it is pure client-side bookkeeping, done entirely in C++ like retail.
Newness is keyed on each item's instance GUID (read off the CGItem), so a
flag survives the item being moved between slots - impossible for a Lua
addon, which can only track by (bag, slot). Bags 0..4 only; equipment and
bank are out of scope, matching retail.
The feed is event-driven, not polled: it subscribes to Bag::UpdateDelayed
(which co-hooks the engine's three BAG_UPDATE fire sites), so a bag rescan
runs only on frames where a bag actually changed. A one-time baseline taken
~1.5s after login (re-armed on character switch) keeps items owned at login
from being flagged. BAG_NEW_ITEMS_UPDATED fires when the set changes.
Consolidations along the way:
- New Bag::UpdateDelayed::AutoSubscribe exposes the authoritative
"a bag changed this frame" signal as a C++ subscription.
- New Item::Location::ContainerInventory factors the CGContainer -> its
inventory (vtable +0x10) dispatch hand-rolled in item/Count,
item/AverageLevel, and equipmentset/Locations; all now share it.
- New Item::Location::EquippedBagInventory (equipped bag -> inventory) plus
the FUN_GET_EQUIPPED_BAG_GUID / OFF_CONTAINER_GET_INVENTORY offsets.
- NewItems resolves the player via the shared Unit::Identity::PlayerObject.
Raw CreatureType id (CreatureType.dbc row: 1=Beast … 11=Totem), the
counterpart to the family additions:
UnitCreatureTypeID(unit) -> id, nil if none
C_CreatureInfo.GetCreatureTypeInfo(id) -> { id, name }
C_CreatureInfo.GetCreatureTypeIDs() -> ascending id array (1..11)
UnitCreatureTypeID calls the engine's own type resolver at 0x00605570
(Script_UnitCreatureType's inner helper) directly — it handles all three
engine paths (display-override, creature-cache +0x18, player-race
ChrRaces+0x24), so it matches UnitCreatureType for every unit kind
(players resolve to Humanoid). The C_ lookups read CreatureType.dbc
(records 0x00C0DE2C, count 0x00C0DE30, name +0x04).
Also documented that 0x00605570 is really the creature-type resolver
(new canonical alias FUN_UNIT_CREATURE_TYPE); the existing
FUN_UNIT_PET_MINION_CLASS name is a historical label — Pet.cpp's
"!= 1" test works because it is effectively "is this a Beast (type 1)":
hunter pets are Beasts, warlock minions are Demons (3). Verified
in-game: target -> 7 Humanoid, IDs[8] -> 8 Critter, round-trips.
The by-raceID form of vanilla's UnitFactionGroup, returning
{ name (localized), groupTag (english "Alliance"/"Horde") } for a race
with no unit required. Mirrors the player branch of Script_UnitFactionGroup
(0x00516630) exactly: ChrRaces[race]+0x08 -> FactionTemplate id ->
FactionTemplate[id]+0x0C group mask -> scan FactionGroup.dbc for the row
whose bit the mask sets and whose localized name is non-empty (the
non-empty test is how the engine skips the always-set "Player" bit and
the nameless "Monster" row to land on Alliance/Horde). nil for
non-numeric / non-positive / unknown race or a race with no named group.
Every field offset ground-truthed against ChrRaces/FactionTemplate/
FactionGroup .dbc files. FactionGroup is stored contiguously (base +
i*0x30, direct field reads), so it's iterated by hand rather than via
DBC::Record. Its records-base/count globals (0x00C0DD5C / 0x00C0DD60)
are taken from the disassembly, which is authoritative over DBCs.md's
standard-shape tabulation for this table (0x00C0DD64 / 0x00C0DD68) --
verified in-game: race 1 -> Alliance, race 2 -> Horde.
Retail backports over CreatureFamily.dbc (records 0x00C0DE7C, count
0x00C0DE80 — the globals Script_UnitCreatureFamily already indexes), on
the existing C_CreatureInfo table:
GetCreatureFamilyInfo(id) -> { id, name, iconFile }
GetCreatureFamilyIDs() -> ascending array of populated family ids
Localized name at record +0x20 (via LocalizedField), icon at +0x44 (via
StringField). GetCreatureFamilyIDs skips the sparse unused rows
(10/13/14/22/...); each element round-trips with GetCreatureFamilyInfo.
Faithful deviation: retail's iconFile is a numeric fileID, but vanilla's
DBC stores the icon as a texture-path string, so iconFile is a string
(usable directly with SetTexture) and "" for families with no icon
(warlock pets). Verified in-game: 27 -> Wind Serpent + icon path, 15 ->
Felhunter + "", IDs array sparse, unused id 10 -> nil.
Extension exposing the numeric CreatureFamily id (CreatureFamily.dbc
row) for a unit, the value addons want for stable checks — stock
UnitCreatureFamily only returns the locale-dependent display name.
It's the same id the engine's Script_UnitCreatureFamily (0x0051A310)
resolves internally: its family helper FUN_006055e0 returns
[unit+0xB30] ? [row+0x1C] : 0, then maps it through CreatureFamily.dbc
for the name. We stop at the id. Resolves the token via the same
FUN_RESOLVE_UNIT_TOKEN, returns nil for family 0 (players, non-beast
NPCs) and unsynced/unresolvable units — matching UnitCreatureFamily.
Shares UnitSubName's +0xB30 creature-cache read path. Verified in-game:
target (cat) -> 2, targetpet -> 2.
GetTaxiNodesForMap(mapID) reads TaxiNodes.dbc: real flight masters on a
continent, faction-tagged, discovered or not -- the static DB pfQuest's
pfDB.meta.flight wants. Non-flight rows are filtered out: a node must
have a flight mount (horde @+0x38 / alliance @+0x3C -- drops transports/
boats/zeppelins) AND be a TaxiPath.dbc endpoint (drops Northshire Abbey,
spurious duplicate hub rows, standalone markers with no connecting
path), plus a guard for Blizzard's lone "Generic, World target"
placeholder. Returns retail nodeID/name/faction/position (continent
0-1) + ext mapID/x/y/z and derived areaID/mapX/mapY (zone-relative).
GetAllTaxiNodes([uiMapID]) wraps the open flight master's live node
list (VAR_TAXI_NODE_ARRAY stride 0x20: slot+0 = TaxiNodes record ptr,
+0x04/+0x08 = flight-map x/y; count VAR_TAXI_NODE_COUNT; type via
FUN_004dc4e0): slotIndex (TakeTaxiNode), name, nodeID, state
(Enum.FlightPathState). Empty when no taxi map is open.
Registers Enum.FlightPathFaction / Enum.FlightPathState (retail values).
Extracts the zone-percent projection into shared Map::Area::ZonePercent
+ adds ContinentPercent; refactors AreaTriggers onto ZonePercent.
Verified: EK 48 real masters (Northshire/EPL/transports gone), Stormwind
Alliance / Undercity Horde, continent position matches projection; live
24 nodes with Stormwind state=Current.
GetExploredMapTextures(areaID) backports the retail (9.0-era) call:
only overlays the local player has discovered. GetUnexploredMapTextures
is the ClassicAPI-extension complement (the reveal-over-fog pieces); its
inverse, C_Map.GetMapOverlays, still returns all overlays.
Refactor the overlay walk into a shared Map::Overlays::PushZoneOverlays
with a Filter {All, Explored, Unexplored}; GetMapOverlays is now the
All wrapper. "Explored" is computed per-areaID from the player's
explored-areas bitfield ([player+0xE68]+0xE6C, indexed by AreaTable
AreaBit +0x0C, gated by +0x28) -- mirrors the engine's discovered-
overlay rebuild FUN_004a67a0 -- so the areaID arg works for any zone,
not just the current view. Also add a fileDataIDs field (ordered tile
paths, retail's field name) to every overlay entry.
New offsets: OFF_AREATABLE_AREA_BIT, OFF_AREATABLE_EXPLORE_GATE,
OFF_PLAYER_EXPLORED_BITS. Verified: Elwynn explored(10) + unexplored(2)
== all(12).
Each overlay now carries hitRectTop/Left/Bottom/Right (WorldMapOverlay.dbc
+0x34..+0x40) -- the tight clickable landmass bounds in world-map canvas
px (~1002x668), the subzone-center source map addons scrape into their
zone tables. Verified: areaID 12 overlay 8 (NorthshireValley, areaID 9)
-> hitRect 164/400/388/617.
Also stop dropping texture-less rows: a row with an areaID + hit rect but
no texture now appears with textureName == "", zero size, and an empty
tiles table (ipairs-safe), instead of being skipped. No-op on stock/Octo
data (every overlay has a texture) but matches the schema and the "every
overlay of a zone" contract.
Each overlay table now carries areaID (primary AreaTable id revealed,
first nonzero) and areaIDs (the full nonzero list -- WorldMapOverlay.dbc
areaID[4]; 52 of 707 overlays span 2-3 subzones). Lets consumers map an
overlay to its subzone(s) / exploration state. Verified: overlay 1 of
areaID 85 -> areaID 157.
Expose AreaTrigger.dbc (517 static trigger volumes: subzone-entry /
exploration / teleport), which the client loads but never surfaces to
Lua -- the reason map addons ship hand-scraped trigger tables. Same
"read the hidden DBC" pattern as C_Map.GetMapOverlays.
Each trigger returns both the authoritative raw geometry (mapID, world
x/y/z, radius, isBox + box length/width/height/yaw) and a derived
consumable form (areaID + mapX/mapY as a 0..100 zone-relative percent)
via the WorldMapArea rect, picking the smallest zone whose rect
contains the point. GetAreaTriggerInfo(id) returns one trigger;
GetAreaTriggers([mapID]) enumerates all, optionally filtered to a map.
Adds the AreaTrigger.dbc record offsets and the WorldMapArea mapID /
loc-rect offsets to Offsets.h.
C_Item.PickupItem(itemInfo) picks the matching item up onto the cursor.
itemInfo is an itemID / "item:NNN" / link / name (same input family as
UseItemByName / EquipItemByName): resolved via Item::Arg::Resolve and
found by Item::Location::FindByArg across equipment (1..19, checked first)
and bags (0..4) using the shared MatchesArg predicate (itemID, or the
decorated name for a name arg). An equipped match is lifted off the
paperdoll; a bag match off the container. Verified in-game (bag + equipped,
by id and by name).
Drives the engine's cursor-pickup primitives DIRECTLY rather than the Lua
Script_Pickup* state-machine wrappers (no faking the Lua stack), via a new
shared Item::Cursor:
- PickupEquipmentSlot -> FUN_004C7300 (self-contained; own guards)
- PickupBagItem -> FUN_00494B60, replicating the two guards the
Script wrapper applies (empty cursor + unlocked item) and the
post-pickup source lock. The empty-cursor check reuses a new shared
Cursor::Info::HasItem() (the same cursor-GUID globals GetCursorInfo
reads), rather than duplicating it.
Supporting consolidation in Item::Location:
- ResolveBagSlotDetail exposes the container object + linear slot the
cursor primitive needs; ResolveBagSlot now delegates to it.
- FindByArg (equipment-then-bags, MatchesArg) reuses FindByArgInBags for
the bag half; FindByItemID is now its itemID-only specialization.
C_EquipmentSet.UseEquipmentSet was considered but already uses the atomic
FUN_INVENTORY_SWAP (no cursor). EquipItemByName's auto-slot path still uses
the Script wrapper; migrating it to Item::Cursor::PickupBagItem is a
follow-up.
The location form of C_Item.GetItemName points at a real item instance,
so its name should carry random-suffix decoration ("... of the Owl") the
way modern WoW does -- but we were reading m_name[0] off the itemID-keyed
ItemStats_C cache, which only holds the base name.
Resolve the location to its CGItem and build the name via the engine's
own per-instance name builder (FUN_005D8BC0): it reads the suffix ID off
the descriptor (+0x98) and formats base + ItemRandomProperties localized
suffix through the engine's ITEM_SUFFIX_TEMPLATE. This is the same
builder FUN_GAMETOOLTIP_BUILD_ITEM_LINK wraps in brackets, so the output
now matches GetItemLink's bracketed name and modern C_Item.GetItemName
exactly (verified in-game: "Iridium Chain of the Owl"). Falls back to the
base ItemStats name (and warms the cache) when the instance build yields
nothing.
GetItemNameByID is unchanged -- an itemID has no instance, so base name
only is correct.
ClassicAPI extension (retail has no such global): true if the player
has clear LoS to unit, false if world geometry blocks it, nil when the
check can't apply (absent/unresolvable token, or non-unit object).
Traces a ray between the two units through the client's world-collision
core CWorld::Intersect (0x672170, verified in-binary: __fastcall(start,
end, unused, hitOut, tOut, flags), tail-calls the real intersect
FUN_0069BFF0). Flags 0x100111 = terrain + WMO (0x100171 also tests M2
doodads but crashes in some dungeons; matches vanilla server LoS which
ignores doodads). Endpoints are raised by each unit's collision-box
height (CMovement+0xB4) so foot-to-foot rays don't false-block on
terrain, with a two-pass level/look-up trace for height mismatches and
a 150yd guard against a long-segment crash.
Mirrors UnitXP_SP3's proven implementation (same VA, flags, guard, and
two-pass logic) but native, so it works with or without that addon --
we call the intersect rather than hooking it, and our flag isn't one
UnitXP_SP3's detour acts on. Gated on object type (unit/player) so it
never reads CMovement off a GameObject.
Adds FUN_CWORLD_INTERSECT, OFF_CMOVEMENT_COLLISION_BOX_HEIGHT,
OFF_CGOBJECT_TYPE_ID, OBJECT_TYPE_UNIT/PLAYER (reuses the existing
OFF_UNIT_MOVEMENT_INFO_PTR = 0x118), plus API.md and README entries.
Returns true/false for in/out of range, nil when the check does not
apply (rangeless spell, or unresolvable spell/unit). spellIdentifier
accepts a spellID, link, or spellbook name; targetUnit is a unit token.
Backed by the engine's own geometric range core FUN_006E47B0 -- the
same test the action-button range coloring uses (Script_IsActionInRange
to FUN_004E56F0 to FUN_006E4440 to FUN_006E47B0). It derives the
spell min/max range via FUN_006E3480 (which folds in the target's
bounding radius AND applies the range SpellMod op 5, so range-extending
talents/items like Hawk Eye and Nature's Reach count) then compares
center-to-center distance, so the boundary matches the client exactly,
melee and min-range spells included.
We call the pure core rather than FUN_006E4440: that wrapper also gates
on target type and folds range-not-applicable into the same truthy
return as in-range, so it cannot yield a clean true/false/nil. The core
is range-only, matching retail (ignores line of sight; does not reject
wrong-faction targets, documented). Positions are guarded via
Unit::Position::Read before the engine call, which dereferences the
GetPosition result without a null check.
Adds Offsets::FUN_SPELL_RANGE_CHECK (0x006E47B0) with derivation notes,
plus API.md and README entries.
table.wipe nil'd every key via lua_next + rawset(nil) but left the
Lua 5.0 length counter untouched. In 5.0 a table's table.insert/getn
length lives out of band from its keys (a t.n field, else a weak
sizes[t] entry in the registry's LUA_SIZES table), so it survived the
clear. After a wipe, getn stayed at its old value and subsequent
table.insert appended past the wiped slots -- t[1] stayed nil. The
3.3.5 (Lua 5.1) twipe this was ported from has no getn legacy, which
is why the ported logic looked complete.
Fix: call luaL_setn(L, 1, 0) after the clear loop. Bound the engine's
luaL_setn (FUN_006f4ea0, verified against 5.0.2; located via
table.insert FUN_007fb6b0) as Game::Lua::SetN.
Repro before: table.insert(t,1); table.insert(t,2); table.wipe(t);
table.insert(t,3) -> getn 3, t[1] nil, t[3]=3.
After: getn 2, t[1]=3, t[2]=4. Verified in-game.
Player-minion classification for a unit, all owner-based:
- UnitIsMinion(unit): owned by a player (pet/guardian/totem). Primary
check is the owner GUID, since summoned guardians ("X's Minion") carry
an owner but not UNIT_FLAG_PLAYER_CONTROLLED; the flag is a fallback.
- UnitIsPet(unit): a player's controllable pet specifically -- owned by a
player AND the engine's creature-family pet/minion classifier
FUN_00605570 == 1, the same signal driving the UNITNAME_TITLE_PET vs
_MINION tooltip label. The GUID can't be used (pets and guardians share
the 0xF14 prefix) nor can PLAYER_CONTROLLED (a Minion can have it).
ClassicAPI extension.
- UnitIsOtherPlayersPet(unit): owner is a player other than you; matches
5.4.8's owner-based Script_UnitIsOtherPlayersPet.
Owner GUID read from CHARMEDBY (+0x10, verified via Script_UnitIsCharmed)
and CREATEDBY (+0x20, verified via the unit title builder FUN_0052FD30).
Verified in-game: Salvatory's Pet is a pet; Jimmyf's/Methl's Minion is a
minion but not a pet.
The proficiency/level/class/skill/spell/rep gate that answers "can this
player equip or use this item" -- the tooltip's red-requirement logic,
distinct from IsEquippableItem (fits a slot) and IsUsableItem (on-use
castable). Mirrors the requirement checks FUN_0052B650 inlines:
1. proficiency - VAR_PROFICIENCY_TABLE[m_class] & (1<<m_subClass)
2. RequiredLevel
3. AllowableClass / AllowableRace
4. RequiredSkill - effective rank via FUN_SKILL_LINE_TO_SLOT
5. RequiredSpell - spell-knowledge bitmap (crafting specializations
like Armorsmith, proficiency spells)
6. RequiredHonorRank / RequiredCityRank - PvP honor rank + medal mask
7. RequiredReputation - current standing vs the reaction-min band
Verified in-game: mail boots false (proficiency), Riding mount false
without skill 150, Plans: Lionheart Helm false until Armorsmith learned.
Residual gap: a ranked RequiredSpell known only at a different rank.
Adds OFF_CGPLAYER_HONOR_RANK / OFF_CGPLAYER_PVP_MEDALS.
Item analog of IsUsableSpell, returning (usable, noMana) for an item's
on-use ability -- not equippability. Gates on the item having an ON_USE
spell, its use requirements (RequiredLevel, AllowableClass/Race), then
the on-use spell's castability via the engine helper FUN_SPELL_IS_USABLE
(0x006E3D60), which sets noMana on power shortfall.
Shares the on-use spell-slot walk with C_Item.GetItemSpell: promoted
Item::Spell::FindOnUseSpellIDInRecord to the header so both use it
instead of re-deriving the trigger scan.
Backport the optional-deps getter vanilla never exposed (it ships
GetAddOnDependencies for the required array only). Resolves the arg
the same way stock Script_GetAddOnDependencies does -- numeric via
FUN_ADDON_GET_BY_INDEX, string via the by-name hash lookup 0x0051E350
(returns record+0x48; subtract to recover the base) -- then walks the
OptionalDeps descriptor at record+0x38 (count +0x3C, data +0x40),
verified against the TOC parser FUN_0051C9B0. Arg/error behavior
matches stock: out-of-range index or a non-string/number arg errors;
an unknown name returns nothing.
Backport of the MoP C_LootHistory namespace. Vanilla 1.12 has no
loot-history store, but it receives the live group-loot roll traffic,
which the stock UI only turns into chat. This reconstructs the history
client-side (like Aura::Source): four co-hooks on the loot-roll packet
handlers accumulate a ring (128) of rolled items with per-player
Need/Greed/Pass results and the winner, exposed through the namespace.
Handlers, from the opcode dispatch FUN_005E6010:
- START_ROLL 0x2A1 FUN_0061B310 (__fastcall(packet), RET 0)
- ALL_PASSED 0x29E FUN_0061B640 (msg=EDX, RET 0)
- ROLL_WON 0x29F FUN_0061B9E0 (ctx=ECX, msg=EDX, +stack, RET 4)
- ROLL 0x2A2 FUN_0061C0B0 (msg=ECX, ctx=EDX, +stack, RET 4)
The per-handler ABI (message register, stack arg, RET) is load-bearing:
matching it exactly is what keeps the co-hooks from corrupting ESP.
Names and class come from the engine NameCache (FUN_0055F080, null-
callback read of a loaded entry), captured at roll time — so they
resolve even for a roller who has left the group by the time their roll
arrives. The item's random suffix is only in SMSG_LOOT_START_ROLL (the
result packets send 0), so it's read off the store-head node right after
that handler runs, giving a correct item:id:0:suffix:unique link.
Surface: C_LootHistory.GetNumItems / GetItem / GetPlayerInfo (name,
class token, rollType, roll, isWinner, isMe) + LOOT_HISTORY_ROLL_CHANGED
/ _ROLL_COMPLETE events (_FULL_UPDATE reserved). The MoP master-loot
functions (SetExpiration / GiveMasterLoot / CanMasterLoot) are omitted —
they need server support 1.12 lacks. Documented in docs/API.md + README.
Vanilla has RegisterEvent / UnregisterEvent as frame methods but not the
query IsEventRegistered (added in a later expansion). There's no engine
Script_* to delegate to, so we replicate the membership check already
embedded in Frame::RegisterEvent (FUN_00702140): find the event by name
in the live event table (SStrCmpI), then walk that entry's subscriber
chain (head@+0xC, next@+4, frame@+8, low-bit/null sentinel) for this
frame. self is resolved via Game::Lua::ResolveObject, yielding the same
CFrameScriptObject* the chain stores.
Registered on the Frame method registry (VAR_FRAME_METHOD_REGISTRY,
alongside RegisterEvent) so it resolves on any frame. Pure read, no new
hooks. Returns a boolean, matching the 3.3.5-era contract; verified
in-game (true for a registered event, false otherwise). Documented in
docs/API.md and README.md.
The old IsStealthed read descriptor +0x17C & 0x02, which is actually
UNIT_FIELD_AURALEVELS (per-slot caster-level bytes) - it only worked by
coincidence on the original test char and read false at level 60
(0x3C & 0x02 == 0), so Prowl/Stealth reported not-stealthed.
_classicapi_DiffAll on a Prowl toggle from a Cat-Form baseline showed
the real signal: UNIT_FIELD_BYTES_1 (+0x210) byte 3 flips 0x00->0x02
(the sneak/anim-tier flag the client render path reads to draw a
stealthed unit semi-transparent); everything else that changed was
aura-array slot metadata. IsStealthed now reads that broadcast bit
directly - verified in-game for Rogue Stealth, Druid Prowl, and
Shadowmeld, with no mount false-positive.
Also refactored the shared buff-slot scan into FindPlayerBuffWithEffect
(Dismount still uses it for the mount aura) and dropped the bogus
OFF_PLAYER_FIELD_VIS_BYTES / PLAYER_VIS_BIT_STEALTH constants.
Synthesize the modern C_LossOfControl surface for the local player, which
vanilla lacks entirely (a Mists-era aggregation layer):
- SCHOOL_INTERRUPT: co-hook the SMSG_SPELL_COOLDOWN handler (FUN_006e9460,
same __stdcall(opCode*, CDataStore*) shape/registrar as Spell::Cast's
failure/channel hooks). Turtle's Player::ProhibitSpellSchool sends the
lockout as ONE cooldown packet listing every spell of the interrupted
school at one uniform duration -- a shape nothing else emits (verified
against the server source) -- so a player-GUID batch of >=2 same-school
spells with equal duration authoritatively yields the locked school
(1<<Spell.dbc School) + duration. The interrupter isn't in the packet, so
spellID=0 / displayText="Interrupted".
- CC types (STUN/FEAR/ROOT/CONFUSE/CHARM/POSSESS/SILENCE/PACIFY/
PACIFYSILENCE/DISARM): scan the player's debuffs, classify each by its
EffectApplyAuraName (values authoritative from the server's SPELL_AURA
table). Timing is best-effort from the Aura::Source cache (nil when the
applying cast wasn't observed -- the fields are nullable).
- LOSS_OF_CONTROL_ADDED / LOSS_OF_CONTROL_UPDATE: a WorldTick poll-and-diff
of the active set fires ADDED (with the 1-based eventIndex) per new effect
and one coalesced UPDATE on any change. Chosen over event-driven aura hooks
because effect EXPIRY has no engine packet (a tick is needed regardless),
and the WorldTick tail keeps Lua handlers out of the mid-packet contexts
the state updates in. Scan is bounded (16 debuff slots) and idle-cheap.
Placeholders with no vanilla equivalent: priority=0, displayType=2,
auraInstanceID=nil; iconTexture is the icon path (no FileDataIDs in vanilla).
Verified in-game: Polymorph -> CONFUSE, correct name/icon/duration, nil once
worn off; a counterspelled heal -> SCHOOL_INTERRUPT (lockoutSchool + 10s) +
SILENCE (4s), with two ADDED at cast and UPDATE as each expires.
Both resolve by name through the engine's own template lookup
(FUN_006ee6f0) -- the same path inherits= uses -- so they're
case-insensitive and cover Blizzard FrameXML plus addon templates.
GetTemplateInfo returns { type, width, height, keyValues, inherits } or
nil if the template doesn't exist. type is the element tag; width/height
are the RAW values from a <Size> element declared directly on the
template (0 if none / inherited) -- read straight from the attributes
rather than via the engine's Size parser (FUN_006f1eb0), which applies
UI-scale conversion we don't want here. inherits is the inherits=
attribute or nil. keyValues is always empty (vanilla's XML schema has no
<KeyValues> element) and sourceLocation is omitted (a 10.2 feature;
vanilla records no per-node file/line) -- both parity-safe.
DoesTemplateExist returns a bool.
Verified in-game: DoesTemplateExist true/false as expected; GetTemplate-
Info nil for a missing name; StatFrameTemplate -> 104x13; FriendsFrame-
TabTemplate -> inherits "CharacterFrameTabButtonTemplate".
Backport the modern C_XMLUtil.GetTemplates() -- enumerate every
registered virtual XML template as { name, type }.
The engine keeps virtual templates (the <Frame virtual="true"> etc. that
inherits= targets) in a Storm hash table. The XML parser (FUN_006ede10)
registers each via FUN_006ee500 and the inherits resolver reads the same
table via FUN_006ee6f0, so it's the authoritative template store. The
new module walks it: each node carries the template name (+0x14) and its
parsed definition node (+0x18), whose element tag string (+0x8) is the
frame type ("Frame" / "Button" / ...).
Covers both Blizzard FrameXML and addon templates; rebuilt on /reload.
Fonts are a separate registry and correctly excluded (matching retail).
Returns an empty table before the first template registers (mask ==
0xFFFFFFFF). Pure read, no hooks. Verified in-game: 212 templates,
entry 1 = { name="AtlasLootDefaultFrameButtonTemplate", type="Button" }.
Every overlay of a zone regardless of exploration, replacing the
hand-measured tables map-reveal addons ship (pfUI pfMapOverlayData).
The engine's GetNumMapOverlays/GetMapOverlayInfo only expose
DISCOVERED overlays -- the exploration gate is a server-fed ID list at
0x00B6E630 -- so unexplored areas were previously undrawable without
per-server data. Reads WorldMapOverlay.dbc directly instead, scoped by
the zone's WorldMapArea row (never by folder contents).
No argument = the world map's currently viewed zone (same
continent/zone resolution GetMapInfo performs, globals 0x0084506C /
0x00845070 / 0x00B6E668); numeric argument = AreaTable areaID, the
same identity C_Map.GetBestMapForUnit returns.
Tile grids are resolved in the DLL rather than left to consumers,
because Octo's shipped data breaks the canonical ceil(w/256) layout
three ways (mirroring the InkLab map generator's empirically-derived
disambiguation):
- sliver columns the DBC rect rounds away (Icepoint's Kaneq'nuun has a
real 8px third column; DBC w=512 implies 2 -- naive layout shears
the overlay into misplaced strips, the one visibly broken overlay);
- foreign tiles appended to an overlay's number sequence (Kaneqnuun
ships 12 tiles, the true grid is 3x2 = 6);
- same-rect higher-resolution re-exports (Stonetalon's WindshearCrag:
4 real tiles for a 1-tile rect, kept and scaled).
The true grid comes from the actual BLP header dimensions (width @
+0x0C, height @ +0x10, BLP1 and BLP2 alike) read through the engine's
own VFS reader, since only a row's last tile may be narrower than 256.
A sweep of all 707 overlays found 108 art/DBC mismatches; Kaneq'nuun
is the only shear-class one, matching the single visible artifact.
Resolution is cached per overlay per session.
Each overlay returns the DBC fields plus a ready-to-draw `tiles`
array (file / width / height / texCoordX / texCoordY / offsetX /
offsetY) so consumers need no grid math at all. WorldMapArea.dbc and
WorldMapOverlay.dbc schemas verified against fresh extractions and
the engine's own Script_GetMapOverlayInfo (0x004A8A00).
ClassicAPI advertises WOW_PROJECT_ID = WOW_PROJECT_CLASSIC, which
steers modern addon ports onto their Classic Era code paths -- where
they also assume Era's C-level frame behaviors. Supply the missing
ones (surfaced by AI_VoiceOver's Era path, but generic to every port):
- region:SetPoint("RIGHT") -- the one-arg form. Vanilla's parser
accepts every other modern shape but raises the usage error when
arg 3 is fully omitted (LUA_TNONE fails its type validation); a
co-hook on Script_SetPoint (0x007A2540, the ONE implementation in
the Region base table) normalizes it to (point, 0, 0) -- vanilla's
parent-relative form, exactly the modern semantic.
- region:SetSize / region:GetSize -- registered on the REGION base
method registry (ctx 0x00CF54B4), so they resolve on frames,
buttons, textures, fontstrings -- including engine-created textures
(GetNormalTexture():SetSize(...)).
- frame:SetShown -- Show/Hide are per-branch script functions with
their own object typechecks, so it's registered separately on
Frame / Texture / FontString with matching delegates.
- frame:SetResizeBounds -- modern rename of SetMinResize/SetMaxResize.
- frame:HookScript -- absent in vanilla entirely. Chains the old
handler (invoked vanilla-style, reading the this/event/argN
globals), then invokes the hook handler with modern positional args
((frame, event, arg1..arg9) for OnEvent, (frame, arg1..arg9)
otherwise) via a C closure over (handler, old, isOnEvent).
All methods delegate to the engine's own Script_* implementations by
reshaping the Lua stack -- no re-derived setters, so UI-scale
conversion, object resolution, and type checking stay engine code.
Method-registry contexts beyond GameTooltip's (Region / Frame /
Texture / FontString) are now in Offsets.h per the
docs/BlizzardScriptAPI.md catalogue.
Verified in-game: GameTooltip:GetSize() returns real dimensions,
CreateFrame + SetSize(50,50) reads back 50, and AI_VoiceOver's Era
path runs through its entire frame construction (SetPoint one-arg,
SetSize, SetShown, SetResizeBounds, 14 HookScript sites) where it
previously errored on the first call.
Rupture at 4 combo points reported duration=6 (the SpellDuration.dbc
base) instead of the real 14s: finisher durations are computed
server-side at cast time and 1.12 never transmits them for debuffs the
player puts on other units. Verified with a whole-protocol sniff --
SMSG_UPDATE_AURA_DURATION (0x137) is self-scoped (slot+duration for the
player's own aura array), 0x2DD has no client handler and never
arrives, and nothing else carries the value. (pfUI issue #17.)
Mirror the server's computation instead (new Aura::ComboDuration):
- Snapshot combo points ([player+0xE68]+0x1029, the field behind
Script_GetComboPoints) the instant CMSG_CAST_SPELL leaves, via a
co-hook on the NetClient send funnel (FUN_NET_SEND 0x005379A0). The
server consumes CP before SMSG_SPELL_GO returns, so reading at
SpellGo time is too late; the send funnel also covers
nampower-queued casts, which bypass any Spell_C_CastSpell co-hook
via trampoline re-entry.
- Compute base + (max - base) * CP / 5 from the spell's own
SpellDuration.dbc row ((v)mangos CalculateSpellDuration), then apply
the player's duration SpellMods, matching server order. DBC-driven
and spell-agnostic: finishers are exactly the rows with base != max
(Rupture 6/16s, Kidney Shot 1/6s, r1 0/5s, Slice and Dice 6/21s;
Expose Armor 30/30 self-disables). Consumed by Aura::Source's
SpellGo hook; 0 falls through to the existing base path.
- Turtle's reworked Rip (8s + 2s per CP) has NO client data: all six
ranks' DurationIndex points at SpellDuration row 87, which Turtle
added server-side only -- the client table jumps 86 -> 105. Built-in
values cover it, gated on the dangling-row condition itself (the
only dangling cross-reference in the entire client Spell.dbc, per a
full DurationIndex/CastingTimeIndex/RangeIndex sweep) rather than on
realm detection: stock clients resolve Rip's index normally, and if
Turtle ships the missing row the DBC takes over automatically.
- C_UnitAuras.RegisterComboDuration(spellID, baseSec, maxSec) as an
escape hatch for other servers with the same class of client-data
gap; a registered override beats DBC and built-ins.
Verified in-game: Rupture at 4 CP now reads duration=14.
Turtle's core doesn't broadcast SMSG_SPELL_FAILURE or
SMSG_SPELL_FAILED_OTHER on interrupt (verified by logging both
handlers through live tests), so the SPELL_FAILED_OTHER co-hook never
fired and interrupted enemy casts (Kick / Earth Shock / Counterspell)
lingered as ghost bars. What actually stops the cast visual on Turtle
is SuperWoW calling CGUnit_C::ClearCastingSpell (0x0060D040) directly.
Co-hook that choke point instead: every unit-stopped-casting path
funnels through it (failure packets, SPELL_GO completion, movement,
death, animation events, local cancel), gated on the unit's live
current-cast field at +0xC8C, so eviction is independent of which
signal the server sends. Both failure-packet handlers stay co-hooked
as backstops for servers that do broadcast them. Verified in-game:
all three interrupts evict; non-interrupting damage doesn't
false-evict.
Also confirmed via a whole-protocol sniff on the NetClient message
dispatch (FUN_NET_MESSAGE_DISPATCH = 0x00537AA0, documented for
future use): SMSG_SPELL_DELAYED goes only to the affected caster --
observers receive no packet of any kind for pushback, so remote bars
can't stretch. Documented as an inherent 1.12 protocol limitation.
Fires (no payload) when any equipped item's durability changes -
combat damage, death, repairs - via a co-hook on the engine's
inventory-alerts recompute (FUN_004C7EE0, the routine behind vanilla's
UPDATE_INVENTORY_ALERTS). The engine calls it whenever an owned item's
fields change (the item post-update handler FUN_005D9A90 routes every
item values-update through it) plus equip/enter-world paths, so it is
the engine's own durability-may-have-changed signal. Post-original we
diff a per-slot {itemGuid, durability} snapshot and fire once per
recompute with a real change - a death dinging all armor produces one
event, not nineteen.
The snapshot seeds silently on the first recompute of the process and
is deliberately NOT re-seeded per enter-world: death/release/graveyard
teleports re-run the setup, and a per-setup reset swallowed the
death's own durability loss (verified in-game; repairs fired, deaths
didn't until this was fixed).
Dead end, kept documented at OFF_DESC_ITEM_DURABILITY: bank-1 (ITEM
field) descriptor observers register without error but are never
invoked - item values-updates demonstrably arrive and player-bank
observers fire in the same session; registration arithmetic fully
verified (bank base FUN_00465690(1)=0x18, bank walk item 0->1->2).
Prime suspect is the registrar's internal wrapper lookup
(FUN_00464890, the observer registry, distinct from the object
manager) silently no-op'ing for item objects.
Also in Player::Equipment: resolve the player via the non-throwing
GUID resolver (FUN_00468460, what the engine's own bag observer uses)
instead of FUN_RESOLVE_UNIT_TOKEN, which raises a Lua error when the
token doesn't resolve - unsafe inside SMSG_UPDATE_OBJECT processing
where the player can be mid-create.
The earlier investigation tried to hook a firer of
UNIT_INVENTORY_CHANGED; none of the three known 0xBA fire sites
trigger on equip because equipment descriptor writes are silent — the
engine registers descriptor-field observers (the __AUCMirrorHandler__
system: registrar 0x00467E70, dispatcher 0x00465570 with per-node
mirror + memcmp change detection) for the bag/backpack/bank/keyring
GUID ranges but never for equipment 0x4A8..0x538.
So register our own: co-hook the engine's inventory observer setup
(FUN_004F8CC0, once per enter-world) and, post-original, register one
observer per equipment GUID field through the engine's registrar —
same bank/size/GUID shape as the engine's bag observers, engine-
managed lifetime, no polling, no hot-function hook. The dispatcher
invokes us only on a real change.
Callback ABI verified from the dispatcher call site at 0x004655DC and
the engine callback FUN_004F8DB0's RET 0x10:
__fastcall(fieldOffset /*ecx*/, size /*edx*/, guidLo, guidHi,
oldValue*, userArg) -> ret 0x10
equipmentSlot is the 1-based Lua inventory slot (1 head .. 19 tabard);
hasCurrent (1/nil) is read from the player invMgr GUID array — the
same post-change truth the engine's bag observer reads.
Vanilla has two disjoint questgiver paths. Gossip NPCs send
SMSG_GOSSIP_MESSAGE into the gossip arrays this namespace reads; but
quest-only NPCs send SMSG_QUESTGIVER_QUEST_LIST (the QUEST_GREETING
event / QuestFrameGreetingPanel), whose data lands in entirely
separate storage C_GossipInfo never read — every getter returned
empty for those NPCs. GetText was also ungated: the gossip text
buffer persists after close, so it served the previous gossip NPC's
greeting.
- Gate all gossip reads on the live gossip NPC GUID (0x00BC3F58,
cleared by the close worker at 0x004E1FA0 before GOSSIP_CLOSED
fires). Stale text/options/quests are no longer visible, and
selectors can't send a CMSG from stale slots.
- Serve the quest-greeting session when one is live instead: the
questgiver GUID (0x00BE0810) must match the greeting-fill GUID
(0x00C4D740) so leftovers can't leak when a quest session starts
from a gossip-menu click. GetText returns the greeting text;
Get{Available,Active}Quests / GetNum* read the greeting arrays
(32 x 0x4C entries {questID, level, title}) with isComplete
recovered from the packet handler's scratch status array
(status 4 = turn-in-ready); Select{Available,Active}Quest route to
the engine's greeting selectors (0x005012A0 / 0x00501320).
- Neither session live: empty tables, 0 counts, empty text.
GetOptions/GetNumOptions stay gossip-only — greetings have no
options. Retail keeps the two APIs disjoint; merging is deliberate so
modern-addon ports work on greeting-path NPCs (common on 1.12
servers).
Introduce QUEST_REMOVED (fires with questID) and document it in API.md and README. Rename Quest::Accepted → Quest::LogEvents and update the quest-log rebuild hook to compute pre/post deltas for both QUEST_ACCEPTED (questLogIndex, questID) and QUEST_REMOVED (questID). Add suppression rules to avoid firing on login/resync bulk changes. Update Offsets.h commentary to reflect the new LogEvents module and its responsibilities.
Mind Control is a 3 s cast on a spell flagged CHANNELED with a 60 s
duration. HandleSpellStart branched on the channel attribute alone, so
at cast start it stamped g_channel with the 60 s duration and killed
the client hook's correct 3 s cast stamp — the bar showed a premature
60 s countdown (pfUI #9). Same story remotely: an observed MC stored a
60 s channel entry instead of the 3 s cast.
- Treat a channeled spell with castTime > 0 as its CAST phase in
HandleSpellStart (player and remote paths); instant channels keep
the immediate stamp.
- Co-hook the MSG_CHANNEL_START handler (0x006E7550, layout from
nampower's SpellChannelStartHandlerHook: spellId u32, duration u32).
The server sends it to the caster the moment the channel begins —
after MC's cast completes — with the server-authoritative duration.
Stamp g_channel there, before the original fires
SPELLCAST_CHANNEL_START, so pollers see fresh data. For instant
channels it re-stamps what the SpellStart path just wrote.
Verified: MC shows 3 s cast then 60 s channel; Mind Flay-style instant
channels unaffected.
Remote casts cached from SMSG_SPELL_START were expired only by their
computed end time — an interrupted remote cast lingered as a ghost bar
for every UnitCastingInfo consumer (pfUI #11). The interrupt signal
does exist at the packet layer: the server broadcasts
SMSG_SPELL_FAILED_OTHER to observers whenever a started cast aborts
(interrupt, death, movement, fizzle).
Co-hook its handler (0x006E8E40, same __stdcall shape as
FUN_SPELL_DELAYED; address + layout mirrored from nampower's
SpellFailedOtherHandlerHook — body is plain casterGuid(u64),
spellId(u32)) and:
- evict the matching g_remoteCasts entry (guid + spellID), so remote
cast/channel bars drop the same tick the cast dies;
- clear the player's own g_cast on a guid match — covers the
server-stamped chained recast that the WorldTick VAR==0 poll can't
clear (self-receipt is server-dependent; the endMs self-expiry
backstop remains if it never fires).
Update the stale 'no per-unit interrupt signal in 1.12' notes in
Cast.cpp and docs/API.md.
GameTooltip:GetItem() returned the previous item after SetUnitAura /
SetEquipmentSet (and any tooltip that shows no item). The engine's per-tooltip
clear FUN_00530050 zeroes the unit / gameobject / itemID / spell identity fields
but leaves the item GUID (+0x380/+0x384) intact -- the field CGItem tooltips
(SetBagItem/SetInventoryItem) set and that Item::TooltipItem::CurrentID resolves
when the itemID field is 0. So a stale GUID lingered across a switch to non-item
content and GetItem/HasItem/IsEquippedItem resolved the old item.
Co-hook FUN_00530050 and zero +0x380/+0x384 after the original. All 14
Set*/builder paths route through this clear (verified via xrefs; the aura
builder FUN_0052f880 and the item builder FUN_0052b650 both call it), and the
item builder calls it before writing the GUID -- so every fresh tooltip resets
the item GUID and only a real item build re-sets it. The hook runs no Lua, just
zeroes 8 bytes; the CurrentID unit/GO/spell gating is now belt-and-suspenders.
Verified in-game: bag/equipped items still resolve; SetUnitAura / SetEquipmentSet
no longer leak the previous item.
The initial fire used FUN_00702690, whose exec-context mutation (DAT_00ceeac0)
is only valid from the engine's top-level frame-script dispatch; firing it from
inside the item builder (mid-Script_Set*, and in the wild under a stack of addon
Set* wrappers) raised through the outer wrappers. Two fixes make the fire safe
for normal handlers:
- Call the inner invoker FUN_00704d50(handler, tooltip) directly. It binds the
global `this` = tooltip and runs the handler under its own protected pcall
(0 args), self-contained and context-independent.
- Snapshot lua_gettop before the fire and lua_settop back after, so the still-
running outer C code sees exactly the stack it left. Without this the invoker
left the stack shifted and the outer code / other DLLs' hooks indexed a stale
slot -> ACCESS_VIOLATION.
Verified in-game: SetScript / HookScript / GetScript all work, `this` is the
tooltip (a1 is nil -- vanilla convention, not modern `self`), and no-op /
print / AddLine handlers fire cleanly on bag / equipped / merchant / auction.
Docs: correct the handler example to use `this` (not a `self` arg), and add the
caveat that re-entrant tooltip rebuilds from the handler (ShowCompareItem, Set*
on another tooltip) can crash in a hook-heavy client and must be deferred
(C_Timer.After(0, ...)) -- the event fires mid-build.
Backports the modern OnTooltipSetItem tooltip script to 1.12 as a real
frame script, settable via the standard SetScript / GetScript / HookScript
and firing whenever a tooltip's item is set (with the tooltip as self). Lets
addons annotate item tooltips by hooking one script instead of wrapping every
Set* method -- the destructive wrap-and-rebuild pattern (AtlasLoot's
AddSourceLine) that clobbers the tooltip's item linkage.
Fully native, two Game::HookAutoRegister co-hooks in Tooltip::SetItemEvent:
- Resolver (FUN_005295d0, the CGGameTooltip vtable script-name -> slot
vmethod): for "OnTooltipSetItem" (a name the engine doesn't know) return a
pointer to a C-side per-tooltip 8-byte {handler, context} cell. The tooltip
object (alloc size 0x460) has no free slot, and the resolver's only contract
is "return where the handler lives", so external storage behaves exactly
like the engine's own slots -- proven by OnTooltipCleared (+0x44c).
- Item builder (FUN_0052B650, the shared funnel for SetBagItem /
SetInventoryItem / SetHyperlink / SetMerchantItem / SetAuctionItem / ...):
after the build, fire the handler via FUN_00702690 (the no-arg runner the
clear uses for OnTooltipCleared). Recursion-guarded, and suppressed via
Tooltip::SetItemEvent::Suppressor during Tooltip::Compare's internal
equipped-item build so a handler doesn't run mid line-shift.
Item only for now; OnTooltipSetSpell / OnTooltipSetUnit follow the same
pattern with their own builders.
Introduce Item::TooltipItem::CurrentID -- the single "what item is this
tooltip showing" resolver. It trusts the link-path itemID (+0x398), and
otherwise resolves the CGItem GUID (+0x380) but only when no unit /
gameobject / spell id is set, since the clear does not zero the item GUID
and it goes stale after the tooltip moves on.
- GameTooltip:GetItem / HasItem now route through CurrentID: fixes GetItem
returning nil for CGItem tooltips (bag / equipped / merchant / loot) and
HasItem false-positiving on a unit tooltip after an item hover.
- Add GameTooltip:IsEquippedItem() (3.3.5 backport, registry index 63),
resolving the displayed item via CurrentID + the equipment-slot walk.
- SetHyperlinkCompareItem: the comparisonTooltip (4th arg) fallback uses
CurrentID; stat deltas now show by default and hide only for an explicit
false / 0 (was: show only for a truthy shift).
- GetItem round-trips the random-suffix id in the link for auction/compare
items (no CGItem instance) via the compare descriptor.
- FrameXML GameTooltip_ShowCompareItem falls back to self.compareLink, so a
caller can hand it an explicit link when the tooltip's item state was
rebuilt out from under it (e.g. an AH addon appending source lines).
- Offsets: compare-descriptor + FontString offsets; drop the duplicate
tooltip item-id/guid offsets in favor of the canonical OFF_TOOLTIP_*.
Backports the 3.3.5+ shopping-tooltip method, fully native (no lua_pcall):
fills a tooltip with the item currently equipped in the given item's slot
— a grey "Currently Equipped" header, the equipped item's tooltip, then a
green/red per-stat delta breakdown gated on shiftButton.
- Extract the stat engine (base record + on-equip-spell auras + random
suffix) out of item/Stats.cpp into the shared Item::StatAccum, reused by
C_Item.GetItemStats/GetItemStatDelta and the compare deltas.
- Build via the engine's own item-tooltip builder; prepend the grey header
by shifting the line FontStrings, replicating the show/hide protocol,
keying on the visible flag (no stale text) and resetting cell color.
- Delta colors sourced from ColorData.h via a new constexpr ByTag helper.
- offset selects the slot for two-slot items (rings/trinkets/1H weapons);
shiftButton gates the deltas (true/1 vs false/0/nil); a comparisonTooltip
4th arg lets it run link-free, 1:1 with the retail call.
- FrameXML GameTooltip_ShowCompareItem consumer with a dynamic,
backdrop-derived anchor gap.
Vanilla GameTooltip:AddLine silently drops any line past the 30
TextLeftN/TextRightN FontStrings the FrameXML template declares, so
stat-heavy tooltips (pfUI eqcompare, AtlasLoot) lose their tail.
Tooltip::LinePool co-hooks the tooltip's XML-OnLoad region scan
(FUN_00529650, CGameTooltip vtable slot 9) via HookAutoRegister. After
the engine sets up the template's 30 lines it creates TextLeft/Right
31..60 in C++ (alloc from the CSimpleFontString pool -> ctor(parent) ->
name -> font copied from the previous line -> anchor) and appends each
pair to the tooltip's three line arrays, bumping numLinesAllocated. Runs
once per GameTooltip-family frame at creation and per /reload, so it
auto-covers GameTooltip, ShoppingTooltip1/2, ItemRefTooltip, AtlasLoot,
WorldMapTooltip, etc. No addon Lua and no per-line AddLine hook.
SetPoint stores offsets in internal coordinates, not pixels, so the
anchor offsets are scaled the way Script_SetPoint does
(pixel * [0x00832A44] / ([0x00832A4C] * 1024)) — without that a -2px gap
rendered ~3950px off.
Adds the tooltip line-pool, FontString-creation, and UI-scale offsets to
Offsets.h; documents the behavior in the README and marks TODO #97 done.