wow_turbo detours about 380 engine functions to its own reimplementations
from its DllMain. It yields any function whose prologue another mod has
already patched, but ClassicAPI installs from the Load export, which
VanillaFixes calls after every DllMain. So wow_turbo always patches first,
and MinHook copies its jmp into our trampoline: the hook then calls the
reimplementation, not the engine's body. Everything that lives in that body
was lost. The script runners (0x00704D50, 0x00704F10) never reached the
pcall redirects at 0x00704E79 / 0x00705199, so every handler saw self ==
nil. The text builder (0x005CDC20) calls its own emitter, so the inline
texture emitter hook (0x005CCBE0) and the word-wrap redirects at
0x005CDDD3 / 0x005CDE23 never ran.
Before each HookAutoRegister hook is created, a prologue that is a jmp
rel32 into wow_turbo.dll gets its five stock bytes back. The bytes come from
WoW.exe on disk, found through the loaded image's section table. Then the
hook wraps the engine's body. wow_turbo's own rule produces the same state
when it loads after a mod that hooked first. It keeps its original in
place, logs the overwrite and does not re-patch these entries. Release
builds reclaim six: the two runners, the text builder and emitter,
GxSetViewport (0x0058AF60) and the selection-circle update (0x00614A90).
Nothing changes when wow_turbo is not loaded.
Adds the three call stack heights an error handler uses to start its
stack at the code that failed:
debugstack(GetCallstackHeight() - (GetErrorCallstackHeight() - 1))
A height counts levels as lua_getstack (0x006FBAA0) does, and so as
debugstack does: C functions and lost tail calls count, and level 1 is
the caller.
1.12.1 records no error height, but every error report runs one C
function, FrameScriptCallErrorHandler (0x00703B40, __fastcall(L)). It
pushes the seterrorhandler function (registry ref 0x008722CC), moves it
under the message and lua_calls it. Nothing calls it directly: state
open pushes it as a C closure and refs it (0x008722C8). Every script
pcall passes that closure as the errfunc, so Lua calls it at the point
of the error before the stack unwinds; a chunk that fails to compile is
reported by a pcall of it (0x00704AE0). ClassicAPI's securecall,
hooksecurefunc and script runners pass the same ref.
So while any handler runs, that closure's level sits directly above the
code that failed. GetErrorCallstackHeight walks the levels with
lua_getstack and lua_getinfo "f", finds the innermost one whose function
is 0x00703B40 (lua_tocfunction, 0x006F3720) and answers the height below
it, or nil when no error is being reported. There is no hook, and
nothing is recorded, so a handler that fails part way leaves nothing
behind.
SetErrorCallstackHeight sets the height for a handler called with no
error raised. It applies only while no 0x00703B40 level is on the stack,
so a value left set by a failed handler cannot move a later error's
stack. A new Lua state starts with no value set.
Users:
- The Lua error window takes its stack this way, so it starts at the
failing code even when another addon wraps the handler.
- CallErrorHandler and assertsafe set the height around their direct
handler call, so they report from their caller.
- Game::Lua gains GetStack, GetInfo and ToCFunction; SourceLocation uses
them in place of its own copies.
Docs: the xpcall and coroutine notes said error() returns without
raising. Stock error (0x00702970) ends in lua_error (0x006F4440) and
raises. pfUI replaces the global error with one that prints and
returns, which is what those notes described.
Tested in game: with seterrorhandler wrapping the old handler, a nil
index in a local function f showed a stack that starts at f.
Addons can now choose the countdown text and its color for each range of
time left. Cooldown:SetCountdownFormatter(formatter) and
GetCountdownFormatter() take a NumericRuleFormatter from
C_StringUtil.CreateNumericRuleFormatter(). Its breakpoints each give a
threshold, an optional step, rounding (Enum.NumericRuleFormatRounding),
min and max, a format string, and optional div/mod/step components, one
for each format argument. Color codes in the format string pass through.
The formatter is a full Lua userdata, the first in the DLL. Its block
holds a pointer to a reference-counted native object. Every Lua object
and every frame that uses it hold one reference, so the GC finalizer
(__gc) and the frame's deleting destructor can run in any order.
lua_newuserdata (0x6F4560, __fastcall, block at Udata + 0x10, read from
the disassembly) is new in Offsets.h. lua_touserdata reuses
FUN_FRAMESCRIPT_GET_OBJECT, which is that function. One metatable per
Lua state lives in the registry. Its __metatable is false, and the
identity check compares a value's metatable with it. 1.12 has no
newproxy, so the only finalizer an allocation can run is ours. Thus no
addon code can change a formatter while it formats.
The text goes through 1.12's own str_format (0x7FD3D0). Its closure is
made once per state and kept in the registry, because FormatNumber runs
from addon timers and lua_pushcclosure allocates on each push. The
rounding is x - fmod(x, step), plus a step away from zero when the
rounding needs it. Only that arithmetic matches retail to the last bit:
Up of 0.7 by 0.1 gives 0.7000000000000001 and of 0.6 gives 0.6, but
ceil(x / step) * step gives 0.6000000000000001.
The cooldown shares the formatter and gives it the unrounded seconds
left. The formatter replaces the built-in text and both thresholds, and
every gate stays. The cached countdown text is now a heap std::string,
because a formatter's text has no length limit. The frame frees it, and
its reference, in DeletingDestructor.
A probe addon recorded 229 results in retail 12.1.5. The same probe on
1.18.1 matched 197 of them. The other results come from 1.12's own
str_format (%e exponent digits, %q, %d past the int range, its error
wording), from field storage on the object (not ported), and from
addresses and times.
Retail exposes engine constant tables as Constants.<Name>, documented with
Type "Constants" and a Values list (MoneyConstantsDocumentation.lua).
Game::Lua::RegisterConstants builds Constants[sub] with the same integer
entries RegisterIntegerEnum takes, and the documentation recorder has a
constants kind next to the enum kind, so the API browser lists the table as
constants rather than as an enumeration.
Constants.MoneyFormattingConstants.GOLD_REWARD_THRESHOLD_TO_HIDE_COPPER is
10, retail's value. FormattingUtil's GetMoneyString now applies
checkGoldThreshold through retail's FormatDisplayCopper: above 10 gold the
copper is dropped.
Verified in game: the table is present, GetMoneyString(1234567, nil, true)
shows gold and silver only, and GetMoneyString(51234, nil, true) keeps its
copper.
Port of 4.3.4's bag search (CGContainerInfo): C_Container.SetItemSearch,
C_Container.IsContainerFiltered and INVENTORY_SEARCH_UPDATE. An item
matches on its equipment sets' names, its suffixed name, its subclass
short name or its equip slot name, case- and accent-folded per locale.
The keyring is searched; non-ASCII search text is folded whole.
C_Container.GetContainerItemInfo now reports isFiltered and takes an
optional table to fill and return, so bag addons can read every slot
without a new table per call. Item::Location::ResolveBagArgs resolves
the caller's own bag/slot args without wiping the stack.
A handler element with inherit="prepend" now runs the inherited
template's handler first and then its own; "append" runs its own
first. Vanilla's loader let the node's handler replace the template's.
Template nodes load onto the same frame before the node's own content,
so the handler a template set is captured when that template's load
returns, then joined with the node's handler (body, function= or
method=) in a C closure that calls both with the same arguments.
Elements apply in document order. The function= binding skips scripts
that ScriptInherit sets.
Also binds luaL_unref as Game::Lua::Unref and drops an unused local
typedef for it.
Retail's debug name for a region: its global name; else its parent's
debug name, a dot, and the key that holds it in the parent's Lua table,
or its address in hex when no key holds it; "" for an unnamed region
with no parent. preferParentKey puts the key path ahead of a global
name. The key comes from the parent's table, so a plain Lua assignment
resolves once the region is parented, as on retail.
The frame stack tooltip prints names through it, which removes the
debug tools' local _DebugName. Game::Lua::ObjectName exposes the
engine's GetName virtual for unnamed-aware callers.
The four Blizzard string globals still missing from the base library.
strlenutf8 and strconcat are 4.3.4's (a count of non-continuation bytes,
and lua_concat over the whole stack); strcmputf8i and strsplittable are
the newer clients'. Text::Utf8 mirrors 4.3.4 Storm's SStrLenUTF8 and
SStrCmpUTF8I, decoder and case fold included, since 1.12's Storm has
neither. strsplit and strsplittable share one scan; the table form never
grows the Lua stack. string.concat is the by-value twin of strconcat, as
string.split is of strsplit. Adds the lua_concat binding (0x006F44E0).
1.12 has no aura identity: a unit's descriptor holds 48 spell IDs and
nothing that tells one application from the next. Aura::Instance gives
each visible aura an ID when it lands in a slot and keeps it until the
aura leaves; IDs are never reused. One sync step (recorded slot vs live
descriptor) drives everything: the existing Aura::Source add / remove /
stacks observers sync the slot they name, the getters sync before
answering, and a per-frame reconcile catches bulk clears and rebuilt
objects. No new engine hooks.
UNIT_AURA_DELTA(unitTarget, updateInfo) carries retail's UNIT_AURA
payload under its own name, so the engine's UNIT_AURA is untouched. It
fires once per unit per frame, per token, with instance-ID lists only,
in four reused tables, so a fire allocates nothing. With no listener,
no change sets are kept. Timing-only refreshes (cache rewrites, the
player's duration packet) are reported as updated.
Also:
- auraInstanceID on every AuraData table and on C_LossOfControl CC
entries
- C_UnitAuras.GetAuraDataByAuraInstanceID(unit, id [, buf]); the buf
form fills the caller's table
- Game::Lua::Ref, the shared luaL_ref binding, replacing two local
typedefs in the animation code
Twelve files each declared their own SStrCmpI function-pointer type and
cast FUN_SSTR_CMP_I at the call. One wrapper in Game now owns the
__stdcall signature (a __cdecl copy drifts ESP by 12 per call) and the
"whole string" length default. Tooltip::Hyperlink's prefix test moves
from the CRT _strnicmp to the same wrapper, which the engine's SStrCmpI
calls anyway.
Twenty modules each declared their own typedef or wrapper for
FUN_LUA_TO_BOOLEAN_LOOSE (4.3.4's StringToBOOL(L, index, default)) or
FUN_XML_ATTR_TO_BOOL (StringToBOOL(const char *)). Route them all
through one helper each.
frame/Cooldown.cpp had a C++ copy of both that read unknown attribute
text as true; the engine's reader, which 4.3.4's XML loader uses, reads
it as false. SetReverse/SetDrawEdge now also truncate numbers like
every other 1.12 flag method.
1.12's grammar pass is 4.3.4's LanguageProcess with rules 1 and 2 only,
so |4singular:plural; reached the screen as written. The pass is now
4.3.4's loop: |1 and |2 still run 1.12's own functions, and |4 is
4.3.4's LanguageRule4 with its locale plural index. It applies wherever
1.12 runs the pass: SetText, button text, tooltip lines, chat lines,
chat bubbles and name plates.
A code is matched by its single digit, which is all a rule consumes, so
|01 no longer loops forever.
PushLocalizedFormatInt drops its gsub expansion; AddLine now expands |4.
Cross-checked against the 4.3.4 decompilation (wowdecomp15595):
- The handler's first argument is the command name, not unused (the
shared CVar handler looks its CVar up by it); Game::Console::Write
takes 4.3.4's COLOR_T, and Game::Console::Printf is the engine's own
ConsolePrintf (0x0063CD50).
- pitchLimit <deg>: writes the camera pitch clamp, which 1.12 keeps as
two .rdata floats read only by camera code.
- Ctrl+Tab lists every command containing the typed text; Tab after
set/cvar_reset/cvar_default completes CVar names (extension). Our key
handler registers at the console's own priority 7.0, which the
engine's tie rule runs first, so no hook is needed.
- AppendLogToFile [file] [numLines]: implements the contract 4.3.4's
help text documents (the shipped command prints "Access Denied").
- The command-list walk moves into console/Registry.h, and
EnsureParentDirs into Common.
Modified clicks: IsModifiedClick, GetModifiedClick, SetModifiedClick,
GetNumModifiedClickActions and GetModifiedClickAction, with 4.3.4's
<ModifiedClick> Bindings.xml node and its 18 defaults. Each binding set
loads, saves and copies them through co-hooks on the engine's own
binding-set steps, in 4.3.4's "modifiedclick" file lines. [mod:]
asks IsModifiedClick, as 4.3.4's macro conditions do.
Direct actions run as 4.3.4's ExecKey runs them: SPELL / ITEM / MACRO
on key release through 1.12's spellbook, item and macro functions,
cast at you for SELFCAST and at your focus for FOCUSCAST; CLICK sends
the frame a mouse down and up, with 4.3.4's virtual-click button logic
built from the button's own methods. RunBinding no longer runs them.
While a key binding runs, its own modifiers count as released for
every modifier read, the engine's IsKeyDown included.
Overrides follow 4.3.4: two lists per key, checked inside the key
lookup at both of 1.12's lookup steps; GetBindingAction takes
checkOverride; isPriority is loose; "" removes; the owner must be a
frame.
ActionButtonUseKeyDown (default 0) and 4.3.4's action-button key
handlers, ported onto 1.12's bars.
Mirror 4.3.4's CGTooltip::SetFrameStack: walk the engine frame list,
hit-test the layout rect against the cursor, name each frame the way
4.3.4 does, sort by strata and level, and fire the new
OnTooltipSetFrameStack script before Show. Every input already exists
in 1.12; the new offsets are verified against the binary.
GetMouseFoci now replays the engine's mouse-focus pass instead of
calling Lua getters: it walks the per-strata mouse input queues and
calls the engine's hit test, so hit-rect insets and scroll clipping
apply as they do for GetMouseFocus.
Also fix the FUN_REGION_GET_RECT comment: the rect is
{bottom, left, top, right}, not {top, left, bottom, right}.
Compared against 4.3.4's CSimpleFrame attribute code and
CGWorldFrame::OnLayerUpdate:
- OnAttributeChanged gets (self, name, value) as real arguments, like
4.3.4's RunOnAttributeChangedScript, plus the this/arg1/arg2 globals.
It used to fire through the no-argument invoker with globals only, so
a modern function(self, name, value) saw name == nil. The engine's
runners only carry %d/%s values, so Frame::ScriptArgs gains
RunWithValues for arguments already on the Lua stack.
- No recursion guard: a handler that sets attributes fires their
handlers too, as in 4.3.4. The guard silenced every nested
OnAttributeChanged on every frame.
- SetAttribute requires the value argument; the 3-arg GetAttribute form
needs exactly four arguments and treats a nil prefix/suffix as empty;
usage errors name the frame like the engine's (GetName ?: <unnamed>).
- The mouseover unit of a hovered frame comes from FrameXML's
SecureButton_GetModifiedUnit, pcalled under the engine's error
handler, as 4.3.4 does, so unitsuffix and modifier-qualified unit
attributes count. 4.3.4 asks protected frames; with no protection in
1.12 the analog is a frame that has attributes. The mouseover tooltip
stays (4.3.4 shows it only over the world frame).
- SecureStateDriver.lua is 4.3.4's file again: it drives itself through
SecureStateDriverManager attributes and reads OnUpdate's elapsed,
both of which the guard and the old no-argument handlers had ruled
out.
- Game::Lua::ObjectLuaRef / PushObject / ObjectDisplayName replace the
copies in ScriptArgs and Modern.
Verified in-game: handler arguments and globals, a nested handler
firing, both usage errors, the nil-prefix lookup, mouseover following
shift-unit / unit, and a visibility state driver.
Compared against 4.3.4's Script_hooksecurefunc / FrameScript_HookWrapper
(taint bookkeeping aside, which 1.12 has no counterpart for):
- The hook runs under lua_pcall with the engine's frame-script message
handler as errfunc, so a broken hook is reported like any script
error instead of being dropped silently. The original's results still
reach the caller.
- The wrapper is installed with lua_rawset, so a target's __newindex
never runs. The original is still read with lua_gettable, so an
inherited method can be hooked.
- Registered on the glue state too: 4.3.4 registers it from
FrameScript_Initialize, which runs for both states.
- 4.3.4's error texts, and its lua_isstring name check, so a number
name works.
- _G is pushed and inserted at index 1, so both call forms share one
layout. The comment saying pushvalue(GLOBALS_INDEX) crashes was
written in the commit that fixed LUA_PUSH_VALUE (it pointed at
lua_remove); the engine's lua_pushvalue resolves the pseudo-index.
The 11.0 unhookable-name list stays. lua_rawgeti moves into Game::Lua
now that ScriptArgs is not its only caller.
Verified in-game: a hook error is reported and T(3) still returns 3 6;
__newindex does not run and an __index method hooks; a number name
hooks; all three error texts.
The inline-texture backport placed icons with hand-tuned constants: a
0.18 x font lead/trail pad, a 0.6 x font vertical centre, a -1px region
nudge, and written sizes taken as raw screen pixels. Replace all of them
with the rules from the 3.3.5 / 4.3.4 clients (decompiled; the size
formula's helpers confirmed against the real 4.3.4 binary):
- A written size multiplies by fontPx / (14 x ddcWidthScale), so it
follows the text's size, the UI scale and the resolution, and then by
the fontstring's layout scale, as 3.3.5 does. Size defaults follow
ParseEmbeddedTexture: ":0:2" is a width ratio, ":20:0" is font-wide.
- An icon starts at the previous glyph's ink edge (bearing + cell width)
plus the face's outline gap, found by replaying the run's placements
with the emitter's own glyph helpers; the next glyph starts at the
rounded icon edge.
- Vertical alignment follows the node's justify-V (node+0x50).
- Offsets move only the drawn icon; a positive offsetY now moves it up.
GetCoinTextureString now matches 3.3.5: default height 14, a space
between denominations, and retail's 2:0 coin offset in the strings.
The four Coordinate statics at 0x00832A40 get their real names
(VAR_COORD_DDC_WIDTH_SCALE and siblings). Known gap: the fs-level wrap
callers can't see the fontstring, so they budget icons at layout scale 1.
* feat(frame): add SetClampRectInsets / GetClampRectInsets
Later clients let a frame that is clamped to the screen keep a different
rect on screen than its own: SetClampRectInsets(left, right, top, bottom)
grows or shrinks the rect the clamp uses. Zygor Guides Viewer 2.0 sets
them on its frames while building them (a negative top inset lets a
frame's top go off screen), and 3.3.5's own FrameXML uses them too.
The engine clamps a frame to the screen inside its rect calculation
(0x00767A20): it resolves the edges and, for a clamped frame, slides the
rect onto the screen. Hook that calculation: for a frame with insets, take
the unclamped rect, grow it by the insets (converted from the frame's
units to layout units with the inverse of Script GetLeft's scaling), slide
that onto the screen with the engine's own steps and order, and move the
frame's rect by the same amount. Frames without insets go through the
engine's clamp as before. SetClampRectInsets re-lays the frame out, as
SetClampedToScreen does, and (0, 0, 0, 0) clears the insets.
Sign convention from 3.3.5's own calls: WorldMapFrame (0, 0, 0, -60)
"don't overlap the xp/rep bars", chat frames (-35, 35, 26, -50).
Verified in-game (1.12.1 client via VanillaFixes): a clamped 100x100
frame placed 10 above the bottom of the screen stays at 10, moves to 60
with (0, 0, 0, -60) and back to 10 once cleared; a frame placed 20 over
the top is clamped to the edge and, with a top inset of -25, keeps its
20 over. GetClampRectInsets returns the insets set.
* refactor(frame): address review of SetClampRectInsets
Use the existing OFF_REGION_ANCHOR and UI coordinate globals instead of new
duplicate offsets, and move the pixel/anchor-unit conversion into one shared
Game::PixelToInternal / InternalToPixel pair (replacing the copies in
LinePool, Transform and Metrics). Restore only the clamp bit around the
unclamped rect calculation so other layout flag bits survive. Add doc
descriptors for both methods and drop version/implementation notes from
the user docs.
---------
Co-authored-by: Brues <5278969+brues-code@users.noreply.github.com>
The getn heal read a nil at the stored last slot as proof that nobody
maintained the length. For a weak-VALUED table that is false: the
collector clears those slots behind the writer, so `luaL_setn` is still
exactly what the last append left and the holes below it come and go
between two reads.
Compost-2.0 is the casualty. `secondarycache` is `__mode = "v"`, filled
with `table.insert`; `GetTable` walks it with `pairs` and calls
`table.remove(cache, i)` on a live index. Once the GC clears slot 1 and
the stored last slot, the heal bisects to border 0, `luaB_tremove` takes
its `e == 0` empty-table early-out and returns NOTHING, and the next line
indexes with the nil it got back -- "table index is nil" at
Compost-2.0.lua:81, once per recycled table, in every addon that pulls
Compost in through AceLibrary.
The border search is what exposed it (5aa382b's table.insert switch is
not involved -- Compost never appends a nil, so no mark ever applied to
its tables). The walk-down it replaced returned the HIGHEST populated
slot, which is always at or above the live index the pairs loop holds, so
remove still found something; a bisect returns some border and probes its
way down to 0 through the collected slots. That tradeoff was known and
written down -- a weak cache read through table.remove is where it bites.
So gate the heal: `__mode` mentioning "v" returns the stored length
untouched, as rule 3, ahead of the mark. It costs one getmetatable and a
raw `__mode` fetch on the already-rare unhealthy path, so the bisect keeps
its win everywhere else, and the mode is read raw -- the same fetch the
collector's own check uses.
New plumbing: Game::Lua::GetMetatable, from the offset already catalogued
in LuaCAPI.md. LuaEnv::Fenv had rolled its own copy; it now shares this
one.
Verified in game: a three-slot weak table with slots 1 and 3 collected
reports getn 3 and hands back its live slot (0 and nothing before), and
the Compost errors are gone.
docs/API.md is 18.7k lines and its anchors are signature slugs, so editing a
signature breaks the 281 internal and 88 README links into it. Blizzard
solved the same problem by shipping the API as data plus a browser addon,
and warcraft.wiki.gg renders those same tables. One set of descriptors can
feed all three.
Each registration now carries one next to it: typed arguments and returns,
nilability, defaults, and a sentence. Events, enumerations and returned table
shapes are described the same way. No Script_* body changed.
Game.h gains namespace Game::Doc (Field with Req/Opt/Vararg, FieldList,
Function, Method, Structure, Event) and a trailing descriptor argument on all
six Game::Lua::Register* plus Event::Custom::AutoReserve.
api/Documentation.cpp records every registration on the first pass of each
Lua state and seals it, so /reload re-registers but records nothing and the
environment comes from which pass saw the name. Export is lazy: nothing is
materialized until C_APIDocumentation.GetSystem asks for one system, because
eager export would build thousands of tables per reload for a slash command.
_classicapi_UndocumentedAPI lists what still has no descriptor.
AddOns/!!!ClassicAPI/APIDocumentation/ ports Blizzard_APIDocumentation's eight
files. Named /classicapi, not /api: the browser holds only what ClassicAPI
adds, so /api would have implied the whole 1.12 surface. Two deliberate
differences from the retail source, both noted in the file headers: a system
search also matches its namespace, since that is the name people type; and an
undocumented function says so, because an empty signature otherwise reads as
a real one.
LinkUtil gains the SetItemRef to LinkUtil.ProcessLink dispatch. That handler
registry was dead code; a new link type now registers a handler instead of
stacking another SetItemRef override. Also defines assertsafe there, which
was undefined and only mattered once the function went live.
Documented: the C_Spell and C_SpellBook surface, the spell globals, both
spellbook enumerations, the 13 UNIT_SPELLCAST_* events, GameTooltip as the
widget-method example and C_Glue as the glue and mixed-environment example.
66 of 692 function registrations. tools/New-ApiDocSkeletons.ps1 emits
paste-ready skeletons to make the rest of the sweep tractable.
Verified in game: the browser lists, searches, and prints detail; links
render and click through; right-click copies and shift-click dumps.
Table::Length recorded deliberate `table.insert(t, nil)` appends (#36) with
a MinHook on the engine's `luaB_tinsert`. The hook is unnecessary: that
function has exactly one xref — the table-lib luaL_reg entry — so no engine
C code calls it, and every call arrives through one of two Lua values:
`table.insert` (lib open) and the global `tinsert`, which the engine's
embedded compat snippet binds by value (`tinsert = tab.insert`, .data
0x008722E8) from FUN_00703b80 right before FUN_LOAD_SCRIPT_FUNCTIONS
(calls at 0x0048fe97 / 0x0048fe9c; on glue 0x0046a87b / 0x0046a880).
So the marking function is now registered over `table.insert` and calls
`luaB_tinsert` directly, and `tinsert` is re-bound to the same closure by
value through a new `Game::Lua::RegisterGlobalAlias` — the snippet's own
idiom, so `tinsert == table.insert` stays true. Both Lua states get the
pair; the `luaL_getn` heal is unchanged. One MinHook fewer, on a hot Lua
function.
Verified in game: the identity triple (`tinsert == table.insert` and both
`_G.ClassicAPI` mirrors), the #36 reserved slot through both names, the
#39 heal, the positional form, and the error path through the new frame.
Every engine frame method validates its `self` in three steps before
touching it: lua_type == table, resolve to the CFrameScriptObject, then
the IsA(typeId) vmethod at vtable+0x10. We did the first two and skipped
the third, so ResolveObject handed back a pointer for ANY frame and the
method cast it to whatever type it expected.
That was reachable from plain Lua, not just through a mirror or a hook.
The per-type method dispatcher is an __index handler that rawgeti's the
method closure out of the Lua registry, so frame methods are ordinary
first-class values, and
GameTooltip.SetSpellByID(ShapeshiftButton1, 133)
reached the tooltip builder with a Button, writing tooltip fields through
Button memory. The engine refuses that same call with "Wrong object type
for member function".
Game::Lua::ResolveTypedObject runs the engine's prologue and raises the
engine's own three error strings, read from .data rather than reinvented.
FrameScriptTypeId mirrors the lazy id assignment, so a check is valid
before any stock method of the type has run. Seven per-type wrappers stop
a call site pairing a method with the wrong class's id, and
ResolveTypedObjectAny covers a method that more than one type reaches.
45 call sites across 19 files now gate self. model/DisplayInfo.cpp had
already worked this out for Model alone and loses its private copy.
Type ids are per CLASS, one shared by all of that class's methods. Each
was read from the prologue of a method that can only belong to its class:
a base-class method registered into several tables carries the BASE id, so
sampling GetObjectType or GetDrawLayer would have produced a gate that
accepts unrelated types. Frame comes from Script_RegisterEvent and Texture
from Script_SetDesaturated, both cross-checked against the first entry of
their table.
Deliberately left ungated: the binding `owner` in bindings/Api.cpp and the
three co-hook sites in texture/Atlas.cpp. Each only uses the pointer as an
identity key and never casts it, so a gate would tighten the contract for
no safety gain -- and raising inside a co-hook would fire before the
original ran.
Verified in-game: the Button-as-tooltip call now raises the engine's
error; a Button still reaches a Frame-registry method
(button:SetAttribute round-trips), so IsA does accept derived types; and
fontstring:GetRotation() still answers -- that reader is registered on
both the Texture and FontString registries and needs the id pair rather
than the Texture id alone.
Registering a Lua name is not the same as owning it. The engine's
registrars write _G, FrameXML and every addon load after that, and
whoever writes last wins with no error and no way for the loser to
notice. GetServerTime is the case on record: a client whose GameTime.lua
declares its own replaced ours on every login, and no addon could reach
the C function again.
Every name RegisterGlobalFunction, RegisterGlueFunction and
RegisterTableFunction bind is now also bound under _G.ClassicAPI, which
nothing else writes -- ClassicAPI.GetServerTime,
ClassicAPI.C_Item.IsBound, namespaces kept as subtables. All three
registrars are single chokepoints, so this is one helper covering 637
in-game registrations plus 26 glue ones, and anything added later needs
no extra call.
Two properties, both of which depend on doing this in the registrar
rather than from addon Lua:
- It runs at registration, so the mirror captures OUR function. A
Lua-side `ClassicAPI.X = X` in the embedded addon would run after
FrameXML and faithfully copy the clobbered version.
- It mirrors by VALUE, reading the name back rather than making a
second closure over the same C function. So `ClassicAPI.X == X`
holds until something replaces the global, and the pair doubles as a
clobber check:
if GetServerTime ~= ClassicAPI.GetServerTime then ... end
EnsureSubTable uses raw get/set so a namespace carrying a metatable
cannot divert either side, and MirrorRegistration restores the stack via
an absolute gettop rather than counting pops.
Not covered, and documented as such: frame methods, which live in the
per-frame-type registry rather than _G, and RegisterIntegerEnum values,
where a mirror would be a stale copy rather than a reference.
Verified in-game: ClassicAPI.GetServerTime == GetServerTime is false on
a client whose FrameXML takes that global, while
ClassicAPI.GetTimeCached == GetTimeCached is true.
The v1.12.8 gap heuristic (keep the stored length when it overshoots the
border by exactly one) protected `table.insert(t, nil)` slots (#36) but
broke every 5.1-dialect table that is stale by exactly one — a
pairs-cleared one-element table, or a recycled table whose previous life
was one slot longer. Tablet-2.0's del()/copy() pool produces the latter on
every tooltip refresh, so FuBar/Ace2/Tablet menus and tooltips shifted by
one with a nil hole at slot 1. The two states are byte-identical, and both
consumers reach the heal through table.insert — neither the table nor the
caller can split them.
Only the writer knows which dialect the table speaks, and the write is
observable: co-hook the engine's table.insert (FUN_LUA_TABLE_INSERT =
0x7FB6B0, verified against the table-lib luaL_reg pair in .data and a
decompile matching 5.0.2 tinsert line for line) and record
{table -> stored length} in a weak-keyed registry table whenever the
two-arg form appends a literal nil. Rule 3 now heals every stale table to
the border (the 5.1 answer) and keeps the stored length only when the gap
is one AND the mark still matches. Stale marks die on their own: any later
insert/remove moves the stored length off the recorded value, and the weak
key collects with the table.
Verified in-game with a 21-test suite (directed regressions for
#29/#36/#39, byte-identical twin discrimination, metamethod isolation,
coroutine-side marking, and a seeded differential fuzzer against an
executable model of the contract): 15/21 on the broken build, 21/21 with
the fix, #36's nil-append cases among the passes both times.
New plumbing: LUA_SET_METATABLE (0x6F4020, from LuaCAPI.md) and
Game::Lua::SetMetatable for building the weak mark table from C.
Fixes#39
Add `UnitCreatedBySpell(unit)` — the spell that summoned a unit (a totem's
totem-drop spell, a pet/guardian's summon spell), read from the unit's
UNIT_CREATED_BY_SPELL descriptor field (byte +0x230). It is the same field
the engine's unit-title builder consults to pick the Pet/Minion/Guardian/
Creation title, and it is broadcast, so it resolves for any unit in range.
This is the summoning spell, distinct from the (server-only) spell a totem
casts.
Also add a single `Game::ResolveUnitToken(token)` helper that wraps the
engine token->CGUnit* resolver, and route unit/Pet.cpp through it plus the
Game::Read<T> field accessors — removing the hand-rolled ResolveUnitToken_t
typedef and raw reinterpret_casts from this module. The name was already
referenced in Offsets.h and CLAUDE.md but had never been implemented.
Modules that keep reload-fragile file-static state (Lua registry refs,
frame/object-pointer-keyed maps, cached event-slot indices) each expose a
PrepareForReload() that DllMain called from the FrameScript_Initialize hook.
That list was hand-maintained, so a new module could silently forget to wire
itself in -- the failure mode behind the Frame::Attributes stale-ref bug and
Tooltip::SetEvents issue #33.
Add Game::ReloadAutoRegister + Game::RunReloadCleanups(), mirroring the
existing ModuleAutoRegister / HookAutoRegister / WorldTick::AutoSubscribe
patterns. Each module self-registers with one file-scope line after its
PrepareForReload:
static const Game::ReloadAutoRegister _reloadReg{&PrepareForReload};
DllMain's 11-call manual list collapses to a single RunReloadCleanups(); a
new module can no longer be forgotten. Player::NameCache::Flush stays an
explicit call -- it's a save-before-teardown, not a stale-state clear, and
also runs on the glue-return path.
A sweep of all file-static state found one uncovered module: Time::Timer
pins C_Timer.After/NewTimer/NewTicker callbacks in the Lua registry but never
released them on /reload -- a surviving/stale ref could fire a dead closure.
Add Time::Timer::PrepareForReload (release each pending timer's registry
callback, clear the queue) and self-register it.
- Game.{h,cpp}: ReloadAutoRegister + RunReloadCleanups.
- DllMain.cpp: collapse the manual list to RunReloadCleanups(); drop 6
now-dead includes.
- 11 existing reload modules + Time::Timer: add the self-register line.
Behavior-preserving (same cleanups, before the Lua teardown); auto-register
order is unspecified across TUs but every callback only clears its own state.
DllMain enabled ~90 hooks individually, and every MH_EnableHook freezes all
process threads (CreateToolhelp32Snapshot + SuspendThread each) under the
Windows loader lock. On machines whose security stack intercepts thread
suspension, 90 serialized freezes ran past VanillaFixes' 10-second
remote-LoadLibrary deadline, so VF read the still-running injection thread as
STILL_ACTIVE (259) and showed its generic "compatible client" error — a false
diagnosis of a slow load that blocked launch entirely for affected users.
Move the whole install off the loader lock. DllMain now only spawns a worker
thread and returns immediately. Hooks install via a latched EnsureInitialized()
reached from two triggers: VanillaFixes' `Load` export (called on the game's
main thread after injection — VF's sanctioned extension point) or the fallback
worker (for any non-VF injector). Whichever arrives first installs; the other
observes the cached result. All hooks are queued and activated in a single
MH_ApplyQueued, so ~90 thread-freezes collapse to one.
- src/DllMain.cpp: minimal DllMain + InstallHooks/EnsureInitialized/Load export
- src/Game.cpp: RunHookRegistrations queues instead of enabling
- src/ClassicAPI.def + CMake /DEF: export `Load` undecorated
- src/Common.h: drop the now-dead HOOK_FUNCTION/UNHOOK_FUNCTION macros
Replace the raw `*reinterpret_cast<const T *>(base + Offsets::OFF_X)`
idiom (and its double-cast variant when base wasn't already a byte
pointer) with three typed accessors in Game.h:
Game::Read<T>(base, offset) field value read
Game::Ref<T>(base, offset) writable field lvalue
Game::Ptr<T>(base, offset) pointer to a field (arrays / out-params)
Game::Read<T>(address) / Ref<T>(address) fixed-VA globals
Pure mechanical refactor, zero behavior change: ~330 cast sites across
84 files collapse to 651 helper calls. Offsets stay in Offsets.h per the
single-source-of-truth rule — the helpers only remove the cast noise and
fold away the inner `reinterpret_cast<uint8_t*>(obj)` base casts.
Left as raw casts by design: FUN_* function-pointer casts, void*/char*
VA handles passed as arguments (caches, method registries), subobject-
anchor pointers fed to __thiscall helpers, `void *volatile` player-ptr
readiness gates (converting would strip volatile), and VA-as-array-base
captures that are subscripted later.
* frame: backport Frame:SetAttribute/GetAttribute + unit-frame mouseover/click
Adds native SetAttribute / SetAttributeNoHandler / GetAttribute on the base
Frame registry: a per-frame, case-insensitive value store with the 3.3.5
prefix/name/suffix getter precedence, stored on the frame's own Lua table.
A `unit` attribute makes the frame a mouseover source. A WorldTick poll of the
engine's mouse-focus frame (GetMouseFocus's global) calls the real mouseover
setter (FUN_00492890) for its unit -- 1:1 with hovering the 3D model: model
highlight, mouseover tooltip, UPDATE_MOUSEOVER_UNIT, and the native GUID slot.
Stomp-proof (the engine's 3D-hover setter is event-driven) and independent of
the frame's scripts, so an addon setting its own OnEnter can't break it.
A `type1`/`type2`/`type` attribute installs a chained OnClick on that frame only
(PCall-guarded, Button-only) that dispatches target / assist / focus on the
unit at click time, then runs any prior handler.
No global hooks: mouseover reads the mouse-focus global; clicks use the frame's
own OnClick. Deliberately does NOT MinHook the button-click dispatcher
(FUN_00779540) -- SuperWoW's click-casting owns that prologue and a second
inline hook corrupts the trampoline (ERROR #132).
* frame: extend attribute backport (one-verb click dispatcher + OnAttributeChanged)
The `type*` click path now mirrors retail's SecureActionButton_OnClick:
resolve ONE verb per click from the modifier/button-qualified `type`
attribute and perform it, instead of a fixed `type1=target`.
- Modifier/button resolution: `[alt-/ctrl-/shift-]type[1..5]` with
prefix..name..suffix -> name..suffix -> name precedence, read from the
live key state and the click's button.
- Verbs: target (with the engine's default-interaction precedence -
cast a pending spell / drop a cursor item on the unit instead of
switching target), assist, focus, spell, stopcasting, macro, and menu.
- Own-the-click: when a verb resolves we perform it and do NOT chain the
frame's previous OnClick, so a configured `type1` no longer double-
dispatches alongside an addon's own conditional click handler.
- Self-healing wire: addons that re-SetScript OnClick (pfUI on every raid
relayout) clobber our closure; WireOnClick re-wraps a clobbered Lua
handler but skips its own C closure, so re-setting `type*` recovers
without double-chaining.
Dispatch goes straight to C++ where we own the code - no Lua round-trip
through the global table: spell -> Spell::AtUnit::CastByName (a native
unit-targeted cast; the unit's GUID is fed to the engine's dispatcher so
there's no target juggling, and ground-target spells land at the unit's
feet), focus -> Unit::Focus::Set, and the target verb's predicates ->
Spell::AtCursor::IsPlacementActive / Cursor::Info::HasItem. Only genuine
engine entries (TargetUnit/AssistUnit/SpellTargetUnit/DropItemOnUnit/
SpellStopCasting) and addon-provided functions still go through Lua
globals. Spell::AtUnit gains a C++ header sharing its CastCore with the
C_Spell.CastAtUnit Lua entry.
macro verb: takes the content from the `macrotext`/`macro` attribute and
prefers an addon-provided RunMacro (SuperCleveRoidMacros etc. - handles
named macros and extended macro text), falling back to running the text
natively when no RunMacro global is present. The native path runs each
line through the stock FrameXML ChatEdit_ParseText via a throwaway
edit-box (GetText closure + no-op __index), so it needs no addon.
menu/togglemenu verb: pops the standard unit dropdown at the cursor via
the new ClassicAPI_ToggleUnitMenu (Util/SecureUnitMenu.lua), resolving
the unit -> menu type with Blizzard's own generic resolver
(self->SELF, pet->PET, grouped player->PARTY, other player->PLAYER,
NPC->RAID_TARGET_ICON). Keeps the FrameXML dropdown logic in Lua with
the C side as a thin call.
OnAttributeChanged is now a real SetScript/GetScript/HookScript-able
frame script on every frame: SetAttribute fires it, SetAttributeNoHandler
suppresses it. Made settable by co-hooking the base-frame script-name
resolver FUN_0076a0d0 (the one every frame type delegates through) and
handing back an external per-frame {handler, context} cell for that name -
frames are immortal in 1.12, so the cell never goes stale. One hook does
it (vs the tooltip OnTooltipSet* work's five) because we already own the
firing side. 1.12 convention: the handler reads this = frame, arg1 = name,
arg2 = value. Recursion-guarded so a handler may itself SetAttribute.
* frame: add Frame:ClearAttribute (retail 11.2.0)
`cleared = frame:ClearAttribute(name)` removes an attribute and returns
whether it was set. Registered on the base Frame method registry next to
SetAttribute/GetAttribute, so every frame has it.
Routes through DoSet's nil-value path (removes the stored key, drops a
`unit` mouseover binding) but does NOT fire OnAttributeChanged — verified
against a live retail 11.2 client: SetAttribute fires the handler,
ClearAttribute fires nothing, and the return is true for a set attribute /
false when there was nothing to clear.
* frame: store attributes in the Lua registry, hidden from Lua
Attributes were kept as a subtable on the frame's own Lua table under a
control-char key ("\1ClassicAPIAttributes"). That key, while collision-
safe, is still enumerable via pairs(frame) and readable/writable from Lua.
Move the store to registry[frame] in the Lua registry (LUA_REGISTRYINDEX),
which has no _G path — so the attribute map is invisible to pairs(frame)
and can't be read or tampered with from script, matching how retail keeps
attributes C-side. Only PushSub changed (create/fetch now key the registry
by the frame table); GetAttribute/DoSet/CopyAttr sit on top unchanged.
1.12 frames are immortal, so the strong registry key never leaks. The dead
kAttrKey constant is removed.
Verified in-game: SetAttribute/GetAttribute still round-trip, and a
pairs(frame) scan no longer finds any ClassicAPI key.
* frame: add the `item` click verb (SecureActionButton item action)
type*="item" uses the modifier-qualified `item` attribute:
- a "bag slot" string (e.g. "0 1") -> UseContainerItem(bag, slot);
- otherwise an item name / itemID / link -> C_Item.UseItemByName(item, unit),
passing the unit as the use target (self-use items ignore it, targeted
ones like scrolls/bandages apply to it);
- the deprecated `bag` + `slot` attributes are honored when `item` is unset.
Both bag paths go through lua_call (new CallTableFunc2Str / CallGlobalNum2)
rather than a direct C++ call: the item-use bag walk does SetTop(L, 0), so
it needs its own Lua stack frame — a direct call would wipe the dispatcher's
stack. (spell stays direct because Spell::AtUnit is stack-free; the helper
comment records the distinction.) Retail's "bag slot" string is the
preferred form; the deprecated per-attribute form is supported too.
* docs: format the attribute click verbs as a table
The verbs were one long paragraph of nested parentheticals - hard to scan.
Replace it with a Verb | Extra attributes | Action table so each verb's
inputs and behavior are legible at a glance. Content-equivalent; no API change.
* frame: clear target on the `target` verb when unit is "none"
Retail's SecureActionButton target action treats a "none" unit as "clear
target". Match it: the `target` verb calls ClearTarget() when the resolved
`unit` is "none" (case-insensitive), before the normal cursor/spell
precedence. Mouseover already handles "none" implicitly (it resolves to
GUID 0 = no unit). Docs' target-verb row updated.
* frame: pass modern positional args to script handlers
Vanilla always invokes frame-script handlers with zero Lua args; the
handler reads this/arg1/event as globals. Modern addon ports write
SetScript("OnMouseWheel", function(self, delta) ...) and get nothing.
Co-hook both script runners -- FUN_FRAME_INVOKE_SCRIPT (no-arg scripts)
and FUN_FRAME_RUN_SCRIPT_ARGS (scripts with values) -- and reimplement
their pcall tail: keep the engine's exact this/argN global save-set-
restore and message-handler errfunc, but additionally push self +
arg1..argN as real positional arguments and pcall with nargs = 1+N.
Purely additive: a zero-param vanilla handler ignores the extras and
reads globals; a modern (self, delta) handler binds them.
OnEvent gets (self, event, arg1..argN), detected structurally by
matching the invoked handler ref against the frame OnEvent slot
(frame+0x0C) -- exact, since each SetScript makes a distinct ref.
Runtime switch SetModernScriptArgs()/GetModernScriptArgs(), default on;
when off both detours are a straight passthrough to the original runner,
which matters because FUN_FRAME_RUN_SCRIPT_ARGS is the hottest Lua path
in the engine (every OnUpdate frame). Verified in-game across
OnMouseWheel/OnClick/OnValueChanged/OnUpdate/OnEnter/OnLeave/OnEvent.
* net: consolidate spell SMSG hooks onto one dispatch funnel
The spell subsystem's per-opcode SMSG handlers (SpellGo, SpellStart,
SpellDelayed, ChannelStart/Update, SpellFailure, SpellFailedOther,
SpellCooldown) were each co-hooked directly -- and nampower (hadesmem)
plus SuperWoW detour the same prologues. Two inline hookers on one
prologue is the documented MinHook collision risk, and a whole-protocol
bisect fingered exactly this cluster as the crash-suspect surface.
Add Net::PacketDispatch: a receive-side mirror of Net::SendObserver that
co-hooks the single incoming-SMSG dispatch funnel FUN_NET_MESSAGE_DISPATCH
(0x00537AA0) -- a chokepoint nampower does NOT hook (it detours the leaf
handlers) -- and fans each packet out to opcode subscribers with the
cursor at the body and centralized save/restore.
Migrate all eight handlers to subscribers: aura/Source (SpellGo, which
re-decodes the packed guids + spellId the leaf handler used to hand it
pre-decoded), spell/Cast (Start/Delayed/Channel*/Failure*), and
lossofcontrol (Cooldown). The parse bodies are unchanged; only the hook
site moved. Subscribers run before the engine's leaf handler instead of
after a co-hook's original -- verified order-independent for all eight
(self-contained cache/state updates; SpellGo's aura-application handoff
rides a separate SMSG_UPDATE_OBJECT).
Cuts the ClassicAPI-nampower shared spell prologues from 10 to 3 (the
OnAura* method hooks, which have no packet chokepoint to relocate to).
One hook on an uncontended function replaces eight on contended ones.
Verified in-game: cast bars, channels (incl. Mind Control cast-then-
channel), pushback, remote-cast interrupt eviction, and C_LossOfControl
school lockouts all intact.
* ScriptArgs is back to opt-in
* frame: fix ScriptArgs comments to match opt-in default
* frame/net: dedupe packet observers, reuse Lua/token helpers, fix attr edge cases
Post-review cleanup of the attribute + SMSG-consolidation work.
- net: extract the shared packet-observer plumbing (PacketObserver.h) — the
intrusive subscriber list, Subscribe(), and the read-opcode / reset-cursor /
fan-out / restore loop — so Net::PacketDispatch (SMSG, u16) and
Net::SendObserver (CMSG, u32) each reduce to a one-line hook thunk over
FanOutByOpcode<OpcodeT>. Drops the dead per-namespace Callback aliases
rather than keeping them as compat shims.
- Game::Lua: promote PushGlobalFunction / CallGlobal / CallGlobalString into
the shared Lua helpers (a pattern several modules open-code); Frame
attributes use them instead of private copies.
- frame/Attributes: resolve unit tokens through Unit::Identity::GuidForToken
(engine resolver + player fast-path) instead of a duplicate raw-cast wrapper.
- frame/Attributes: wire the chained OnClick for every "type" action attribute
the click resolver reads (IsTypeAttr: modifier prefixes + type1..type5), not
just type/type1/type2 — a middle-click or modifier-only binding now installs.
- frame/Attributes: FireAttributeChanged snapshots and restores the caller's
arg1/arg2 around the fire, so calling SetAttribute from inside a running
script handler no longer clobbers its globals.
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.
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.
The repo was internally inconsistent: the LICENSE file is GPL v3 but every
src/ header carried the LGPL v3 notice and the README said "LGPL v3 or
later". LGPL was also the wrong fit — this is an injected mod plus a
bundled addon, not a library other code links against, so LGPL's linking
exception buys nothing.
It's also required: src/voicechat/Tts.cpp is ported from awesome_wotlk's
VoiceChat.cpp, which is GPL v3, and a derivative of GPL code cannot be
relicensed under more permissive terms. Dependencies stay compatible —
MinHook (BSD-2) and the nampower-mirrored packet shapes (BSD) are
GPL-compatible, and SAPI is covered by the system-library exception.
Swept "GNU Lesser General Public License" -> "GNU General Public License"
across all 257 src/ headers and changed the README license line to match
the GPL v3 LICENSE file. Comment-only; no build impact.
Backport of the 4.3.4 developer-console command: dumps Blizzard's UI
files out of the mounted MPQs to disk (`ExportInterfaceFiles code`
writes .lua/.xml/.toc/.xsd to BlizzardInterfaceCode\; `art` writes
.blp/.tga to BlizzardInterfaceArt\), preserving the Interface\ subtree
and skipping the user's own AddOns\.
- New src/interface/Export.cpp: enumerates each archive's (listfile)
via the MPQ walker FUN_00401470, reads each entry through
FUN_FILE_READ, and writes it out with Win32.
- New generic Game::Console facility (RegisterCommand / Write / Category
enum) wrapping the engine's console-command registrar and output
paths, so future commands register in one line.
- Offsets.h: FUN_MPQ_ENUM_FILES, FUN_CONSOLE_COMMAND_REGISTER,
FUN_CONSOLE_WRITE; corrected FUN_FILE_READ to __stdcall and documented
its archive-handle arg0.
- Docs: README "Console commands" section + docs/API.md "Console" entry.
Re-registers the standard five entries (create / resume / yield /
status / wrap) on the global `coroutine` table. The C-level
machinery (lua_resume, lua_yield, lua_newthread, threads as tag 8)
was always linked into the engine; only the script-side library
was stripped from the registration tables at build time.
Two WoW-specific quirks discovered along the way and worked
around:
- `lua_resume` returns 1 for BOTH yield and caught runtime error.
Disambiguated by checking `co->ci > co->base_ci` — a yielded
coroutine still has its body frame on the call stack; an
errored one has been unwound to base_ci.
- The byte at L+0x31 is `allowhook` (saved/restored around
protected calls), not `L->status` as it initially looked.
Status reading is replaced by the ci comparison above.
Registers `SetEquipmentSet` on the GameTooltip method registry,
mirroring the 4.3.4 implementation at 0x0046E690. Builds a tooltip
with header + counts (equipped / in inventory / ignored) + one line
per missing item.
Storage format now persists itemID alongside GUID per slot
(`slot N guid=0xHEX item=ID`), so missing items can be looked up by
name in the item cache even when the live CGItem is no longer in the
client's object map. Backward compat: pre-itemID files load fine and
fall back to a `%d missing` summary line for those slots.
Adds `Game::Lua::PushLocalizedString` and `PushLocalizedFormatInt`
helpers — generic lookup of `_G[name]` with a C-string fallback for
servers stripped of standard GlobalStrings. The tooltip uses these
for the four count lines via Blizzard's `ITEMS_VARIABLE_QUANTITY`,
`ITEMS_EQUIPPED`, `ITEMS_IN_INVENTORY`, `ITEM_SLOTS_IGNORED` keys,
and for the per-missing-item line via `ITEM_MISSING`.
AddOn side: a `GlobalStrings.lua` companion provides English defaults
for those five keys plus a `ShowEquipmentSet()` test helper in
Constants.lua.
First implementation passed `nargs=4` to `lua_call` where the stack
needed 5 (self + text + r + g + b for `tooltip:SetText`) — Lua then
treated `self` as the function, surfacing as
`<in C code>: in function 'SetEquipmentSet'` with no Lua frame above.
Fixed before this commit.
Module-self-registration pattern for the glue Lua state, mirroring the
existing in-game ModuleAutoRegister. Modules opt in with a file-scope
GlueModuleAutoRegister; the hook on FUN_0046ABB0 (master glue init,
single caller, 35-byte linear function) fires their callbacks after
the engine finishes registering its own 109 glue functions onto the
freshly-created glue state.
Prior attempt hooked the inner FUN_0046DDF0 60-fn batch trampoline and
crashed on Octo, likely from sibling-DLL collision on a small target.
The linear caller is larger and less obvious; no crash, hook fires
cleanly on initial launch and every world->glue return.
Three sites were doing the same 3-call (PushString + PushNumber + SetTable
on GLOBALS_INDEX) to declare integer globals. Promote to a shared helper
that uses RawSet (globals have no __newindex in vanilla 1.12, so the
metatable check SetTable does is dead).
Also lift CLASSIC_API_VERSION out of DllMain.cpp into api/Version.cpp
so it self-registers via ModuleAutoRegister like every other Lua surface,
instead of being a special case wired into FrameScript_Initialize_h.