mirror of
https://github.com/brues-code/ClassicAPI.git
synced 2026-11-10 20:07:37 +00:00
docs: catalogue the Lua base library the batch scan never saw
BlizzardScriptAPI.md and raw_globals.txt were built by walking the 54 FrameScript_RegisterFunction batches, so the Lua standard libraries — bound by luaL_openlib from luaL_reg tables at state creation — were missing entirely. That gap read as "1.12 has no debugprofilestart", which is wrong: the pair is stock since 1.7.0, in the base-library table at 0x00811E28 (36 entries: setglobal … debugstack). Add the table to both files, with the mechanism (FUN_00703B80 runs the lib open and the alias snippet at .data 0x008722E8 right before FUN_LOAD_SCRIPT_FUNCTIONS), the debugprofile pair's clock (rdtsc delta over a boot-calibrated ticks-per-second, milliseconds as a double), and the rule this taught: check these tables or /dump the name before declaring a global absent.
This commit is contained in:
Brues
@@ -9,7 +9,8 @@ ImageBase `0x00400000`. All VAs in this doc use that base.
|
||||
|
||||
## How Blizzard wires Lua
|
||||
|
||||
There are two registries.
|
||||
There are two Blizzard registries, plus the Lua standard libraries, which are
|
||||
bound a third way (§3) and are NOT in the batch counts below.
|
||||
|
||||
### 1. Global functions — `FrameScript_RegisterFunction` at `0x00704120`
|
||||
|
||||
@@ -57,6 +58,62 @@ C3 ret
|
||||
23 tables found, **481 methods total**. The runtime dispatcher at `0x005368D0`
|
||||
resolves `frame:method(...)` by looking the name up in the matching context.
|
||||
|
||||
### 3. Lua standard libraries — `luaL_openlib` tables
|
||||
|
||||
The Lua 5.0 standard libraries, and Blizzard's replacements inside them, are
|
||||
bound by `luaL_openlib` walking one `luaL_reg { char *name; lua_CFunction
|
||||
func; }[]` table per library when the Lua state is created (`FUN_00703B80`,
|
||||
called immediately BEFORE `FUN_LOAD_SCRIPT_FUNCTIONS`). They never pass
|
||||
through `FrameScript_RegisterFunction`, so none of them appear in the 54
|
||||
batches above or in the loop scan behind `raw_globals.txt` — which is how
|
||||
`debugprofilestart` / `debugprofilestop` (stock since 1.7.0) went
|
||||
uncatalogued here until 2026-09. **Before declaring a global absent, check
|
||||
these tables or `/dump` the name in game.**
|
||||
|
||||
**Base library** — table `[0x00811E28]`, 36 entries, `.rdata`. Blizzard's
|
||||
own base library: the stock 5.0 set minus `dofile`/`loadfile`/`require`/
|
||||
`print` (`print` is a ClassicAPI backport), plus the `debug*` /
|
||||
`*errorhandler` / `date` / `time` additions.
|
||||
|
||||
```
|
||||
0x00702770 setglobal 0x00702DF0 type
|
||||
0x007027B0 getglobal 0x00702E30 tonumber
|
||||
0x007027E0 debugload 0x00702F40 tostring
|
||||
0x007027F0 debuginfo 0x007030A0 unpack
|
||||
0x00702800 debugprint 0x007030F0 rawequal
|
||||
0x00702810 debugdump 0x00703130 rawget
|
||||
0x00702820 debugbreak 0x00703160 rawset
|
||||
0x00702830 debugtimestamp 0x007031A0 assert
|
||||
0x00702840 debugprofilestart 0x00703200 gcinfo
|
||||
0x00702860 debugprofilestop 0x00703250 collectgarbage
|
||||
0x00702900 seterrorhandler 0x00703280 loadstring
|
||||
0x00702950 geterrorhandler 0x007032F0 pcall
|
||||
0x00702970 error 0x00703340 xpcall
|
||||
0x007029F0 getmetatable 0x007033A0 date
|
||||
0x00702A40 setmetatable 0x007035A0 time
|
||||
0x00702AC0 getfenv 0x00703760 debugstack
|
||||
0x00702BE0 setfenv
|
||||
0x00702C90 next
|
||||
0x00702CD0 ipairs
|
||||
0x00702DA0 pairs
|
||||
```
|
||||
|
||||
`debugprofilestart` stores `rdtsc` in `0x00CEEB58` (u64);
|
||||
`debugprofilestop` pushes `|rdtsc − start| / (ticksPerSecond / 1000)` as a
|
||||
double, where `ticksPerSecond` is `FUN_0042C020` = `(int64)(1000.0 /
|
||||
[0x008332C8])` — a TSC scale calibrated at boot (the image holds `1.0`). So:
|
||||
milliseconds with a sub-microsecond fraction, from the CPU counter, unrelated
|
||||
to the engine's millisecond tick and to `GetTime()`.
|
||||
|
||||
The `table`, `string` and `math` libraries have the same shape (the `table`
|
||||
library's `luaL_reg` is at `0x00822D70` — see `Offsets.h`
|
||||
`FUN_LUA_TABLE_INSERT`) and are not enumerated here. The short global
|
||||
aliases (`tinsert`, `getn`, `strlen`, `format`, `floor`, …) are not registered
|
||||
at all: an embedded Lua chunk at `.data 0x008722E8`, run by the same state
|
||||
init, assigns them from the library tables (`tinsert = tab.insert`). A module
|
||||
that registers over a library function must therefore re-bind its alias —
|
||||
`Game::Lua::RegisterGlobalAlias`.
|
||||
|
||||
## Scope
|
||||
|
||||
The 54 global-function batches split into two phases:
|
||||
|
||||
@@ -1259,4 +1259,42 @@
|
||||
2: 0x00706040 CreateFont
|
||||
3: 0x007060B0 CreateFrame
|
||||
|
||||
=== luaL_openlib base library Table 0x00811E28 Count=36 (not a FrameScript_RegisterFunction batch; see BlizzardScriptAPI.md §3) ===
|
||||
0: 0x00702770 setglobal
|
||||
1: 0x007027B0 getglobal
|
||||
2: 0x007027E0 debugload
|
||||
3: 0x007027F0 debuginfo
|
||||
4: 0x00702800 debugprint
|
||||
5: 0x00702810 debugdump
|
||||
6: 0x00702820 debugbreak
|
||||
7: 0x00702830 debugtimestamp
|
||||
8: 0x00702840 debugprofilestart
|
||||
9: 0x00702860 debugprofilestop
|
||||
10: 0x00702900 seterrorhandler
|
||||
11: 0x00702950 geterrorhandler
|
||||
12: 0x00702970 error
|
||||
13: 0x007029F0 getmetatable
|
||||
14: 0x00702A40 setmetatable
|
||||
15: 0x00702AC0 getfenv
|
||||
16: 0x00702BE0 setfenv
|
||||
17: 0x00702C90 next
|
||||
18: 0x00702CD0 ipairs
|
||||
19: 0x00702DA0 pairs
|
||||
20: 0x00702DF0 type
|
||||
21: 0x00702E30 tonumber
|
||||
22: 0x00702F40 tostring
|
||||
23: 0x007030A0 unpack
|
||||
24: 0x007030F0 rawequal
|
||||
25: 0x00703130 rawget
|
||||
26: 0x00703160 rawset
|
||||
27: 0x007031A0 assert
|
||||
28: 0x00703200 gcinfo
|
||||
29: 0x00703250 collectgarbage
|
||||
30: 0x00703280 loadstring
|
||||
31: 0x007032F0 pcall
|
||||
32: 0x00703340 xpcall
|
||||
33: 0x007033A0 date
|
||||
34: 0x007035A0 time
|
||||
35: 0x00703760 debugstack
|
||||
|
||||
|
||||
|
||||
Reference in New Issue
Block a user