fabi/Sylpheed
1
0
Fork 0
Code Issues 9 Pull Requests Actions Packages Projects Releases Wiki Activity

re: the EF_IDX_ prefix - ptc_pack has 727 names, and the map reaches 128 of 137

Browse Source 
Censusing ptc_pack's own naming vocabulary turned up a third variant of the
prefix trap, and this one had been corrupting a number the corpus carried.

268 of ptc_pack's names do not start with eff_ at all. They start with EF_IDX_,
as in EF_IDX_eff_d001_f. A regex anchored at eff_ chops that prefix off and
merges distinct names, which is exactly where the earlier figure of 532 came
from. Enumerating maximal [A-Za-z0-9_] runs gives 727. The two earlier traps were
a STORED name being longer (rot_n001_break) and a BOUND name being a prefix
(eff_f0002 inside eff_f0002_barnhaze); this is the third - a prefix the pattern
cannot see at all, because its anchor sits in the middle of the real name.

Looking each bound name up bare AND under EF_IDX_ resolves 25 of the 34 that were
unlocated. The map is now 128 of 137, and the residual is 9, small enough to
print: eff_e0044, eff_f0002, eff_f0002_barn, eff_h308, eff_j002_e01,
eff_j002_e02, eff_m010_wep_85, eff_m011_wep_85, eff_n0071. All 17 eff_l### are
among the recovered.

This withdraws my own previous correction. I had recorded Base.xpr (53) as
holding more bound effects than ptc_pack (46), and struck out "ptc_pack is the
effect library". With the prefixed keys counted ptc_pack holds 71 - it IS the
larger library, and the 46 was an undercount from the same truncating pattern.
Two shared libraries remains right; which one is bigger does not.

The suffix vocabulary: 106 distinct tokens over the 727 names - IDX 223 (the
prefix above), _f 137, _e 119, _root 87, _col 54, _mdl 45, _break 43, _ring 38,
_ALL 17, _haze 14, _thunder 10. That census counts ALL tokens rather than
trailing ones, which is precisely how the EF_IDX_ PREFIX surfaced inside what I
had first labelled a suffix list - the mislabel found the bug.

Testing the structural candidates the way _hangar was tested, does the suffixed
name have a bare parent:

  _ALL      17 names    17 of 17
  _root     87          64 of 87
  _break    30          15 of 30
  _e        74           0 of 74
  _f        61           0 of 61

_root is strictly terminal - 87 of 87, and it never appears mid-name. The
compound shapes put it outermost: _e_root 19, _f_root 18, _break_root 13, bare
_root 30. So the order is <stem>_[<faction>|<break>]_root and _root reads as a
hierarchy marker rather than a variant - though 64 of 87 having a bare parent
means it is not simply the parent of an existing node, and _break at 15 of 30 is
likewise not a plain destroyed-twin-of-everything.

_e/_f never have a bare parent, 0 of 135. That is independent asset-side
confirmation of the faction law: an effect is authored per faction and there is
no faction-neutral original for either side to derive from.

effect-homes.txt changes 5/30 and every line pairs: five values changed (103->128,
34->9, ptc_pack 46->71 and its sort position, the residual header, 3-digit
80->105 of 110) plus 25 pure deletions, exactly the 25 recovered names. All are
3-digit, so the 4-digit line is unchanged at 23 of 27. The other sixteen
artefacts are byte-identical.
This commit is contained in:
Sylpheed RE agent
2026-08-28 10:01:53 +00:00
parent aa478d9444
commit 23a6cfd979
4 changed files with 121 additions and 35 deletions
Download Patch File Download Diff File Expand all files Collapse all files

32
docs/re/BACKLOG.md
View File

@@ -6794,3 +6794,35 @@ for those kinds; there was nothing to test.
regex requiring `UN_<l>###_<FACTION>_`; the looser form finds **71**. The extra
is **`UN_e910_core_ADAN_GeneratorCore`**, which inserts a sub-kind before the
faction tag. Grammar is `UN_<l>###_[<subkind>_]<FACTION>_<name>`.
---
## ❌✅ 2026-08-28 — `EF_IDX_` prefix: `ptc_pack` has 727 names, and the map reaches 128 of 137
Item (a). Censusing `ptc_pack`'s naming vocabulary found a third prefix trap that
had been corrupting a carried number.
❌ **`ptc_pack` holds 727 distinct names, not 532.** **268 do not start with
`eff_`** — they start with **`EF_IDX_`** (e.g. `EF_IDX_eff_d001_f`). A regex
anchored at `eff_` truncates that and merges distinct names. ⚠️ Third variant of
the trap: the stored name longer (`rot_n001_break`), the bound name a prefix
(`eff_f0002`), and now **a prefix the pattern cannot see because its anchor sits
mid-name**.
✅ **Looking each bound name up bare AND under `EF_IDX_` resolves 25 of the 34.**
The map is now **128 of 137**, residual **9**: `eff_e0044`, `eff_f0002`,
`eff_f0002_barn`, `eff_h308`, `eff_j002_e01`, `eff_j002_e02`, `eff_m010_wep_85`,
`eff_m011_wep_85`, `eff_n0071`.
❌ **Withdraws my own previous correction**: I had recorded `Base.xpr` (53) as
holding more bound effects than `ptc_pack` (46). With prefixed keys `ptc_pack`
holds **71** — it *is* the larger library; 46 was an undercount from the same
pattern.
✅ **Suffix vocabulary** (106 tokens over 727 names): `IDX` 223, `_f` 137, `_e`
119, `_root` 87, `_col` 54, `_mdl` 45, `_break` 43, `_ring` 38, `_ALL` 17.
**`_root` is strictly terminal, 87 of 87, never mid-name**; compounds place it
outermost (`_e_root` 19, `_f_root` 18, `_break_root` 13, bare 30). Bare-parent
test: **`_ALL` 17/17**, `_root` 64/87, `_break` 15/30, and 🔑 **`_e`/`_f` 0 of
135** — asset-side confirmation that effects are authored per faction with no
faction-neutral original.

35
docs/re/data/effect-homes.txt
View File

@@ -1,10 +1,10 @@
effect names bound by a datasheet field : 137
resolved to an .xpr package (exact name): 103
unlocated : 34
resolved to an .xpr package (exact name): 128
unlocated : 9
resolved, by package:
ptc_pack.xpr 71
Base.xpr 53
ptc_pack.xpr 46
Stage_S28.xpr 2
rou_f001_wep_12.xpr 1
rou_f001_wep_27.xpr 1
@@ -13,41 +13,16 @@ resolved, by package:
rou_f001_wep_82.xpr 1
by digit-width:
3-digit: 80 resolved of 110
3-digit: 105 resolved of 110
4-digit: 23 resolved of 27
the 34 unlocated, in full:
the 9 unlocated, in full:
eff_e0044
eff_f0002
eff_f0002_barn
eff_h000
eff_h006
eff_h308
eff_h310
eff_j002_e01
eff_j002_e02
eff_l002_f
eff_l010_e
eff_l011_e
eff_l101_e
eff_l101_f
eff_l102_e
eff_l102_f
eff_l104_e
eff_l104_f
eff_l105_e
eff_l105_f
eff_l106_e
eff_l106_f
eff_l107_e
eff_l108_e
eff_l201_e
eff_l201_f
eff_m010_wep_85
eff_m011_wep_85
eff_n0071
eff_s100
eff_s901_e01_B
eff_s901_e02_D
eff_s901_e03_D
eff_t130

72
docs/re/structures/unit-substructure-records.md
View File

@@ -445,3 +445,75 @@ counted 42 + 26 + 2 = 70 unit objects using a regex that required
extra is **`UN_e910_core_ADAN_GeneratorCore`**, which inserts a sub-kind token
*before* the faction tag. So the ID grammar is `UN_<letter>###_[<subkind>_]<FACTION>_<name>`,
and both counts were right for their own pattern.
### ❌ `ptc_pack` holds 727 names, not 532 — an `EF_IDX_` prefix the pattern truncated
Censusing the package's own naming vocabulary turned up a third variant of the
prefix trap, and this one had been corrupting a number the corpus carried.
**268 of `ptc_pack`'s names do not start with `eff_`** — they start with
**`EF_IDX_`**, e.g. `EF_IDX_eff_d001_f`. A regex anchored at `eff_` chops that
prefix off and merges distinct names, which is where the earlier figure of 532
came from. Enumerating **maximal `[A-Za-z0-9_]` runs** gives **727**.
⚠️ The two earlier traps were a *stored* name being longer (`rot_n001_break`) and
a *bound* name being a prefix (`eff_f0002` ⊂ `eff_f0002_barnhaze`). This is the
third: **a prefix the pattern cannot see at all**, because the anchor sits in the
middle of the real name.
### ✅ The map improves to 128 of 137 — and a claim of mine is withdrawn
Looking each bound name up **bare and under `EF_IDX_`** resolves **25 of the 34**
that were unlocated. The map is now:
| package | bound effects held |
|---|---|
| **`ptc_pack.xpr`** | **71** |
| `Base.xpr` | 53 |
| `Stage_S28.xpr` | 2 |
| five `rou_f001_wep_NN.xpr` | 1 each |
❌ **This withdraws the previous section's correction.** I had recorded that
`Base.xpr` holds *more* bound effects than `ptc_pack` (53 vs 46) and struck out
"`ptc_pack` is the effect library". With the prefixed keys counted, `ptc_pack`
holds **71** — it *is* the larger library, and the 46 was an undercount from the
same truncating pattern. Two shared libraries remains right; which is bigger does
not.
🟡 **Nine names remain unlocated**, small enough to print in full: `eff_e0044`,
`eff_f0002`, `eff_f0002_barn`, `eff_h308`, `eff_j002_e01`, `eff_j002_e02`,
`eff_m010_wep_85`, `eff_m011_wep_85`, `eff_n0071`.
### ✅ The suffix vocabulary, and `_root` as the outermost token
Splitting all 727 names on `_` gives 106 distinct tokens. The frequent ones are
`IDX` 223 (the prefix above), `_f` 137, `_e` 119, `_root` 87, `_col` 54,
`_mdl` 45, `_break` 43, `_ring` 38, `_ALL` 17, `_haze` 14, `_thunder` 10 — a mix
of one structural marker, the faction pair, and descriptive part names.
⚠️ Method note: that census counts **all** tokens, not trailing ones — which is
exactly how the `EF_IDX_` *prefix* surfaced in what I first labelled a suffix
list. The mislabel found the bug.
Testing the two structural candidates the way `_hangar` was tested — does the
suffixed name have a bare parent?
| token | names | bare parent present |
|---|---|---|
| `_ALL` | 17 | **17 of 17** |
| `_root` | 87 | 64 of 87 |
| `_break` | 30 | 15 of 30 |
| `_e` | 74 | **0 of 74** |
| `_f` | 61 | **0 of 61** |
🔑 **`_root` is strictly terminal — 87 of 87, and `_root` never appears
mid-name.** The compound shapes place it outermost: `_e_root` 19, `_f_root` 18,
`_break_root` 13, bare `_root` 30. So the order is
`<stem>_[<faction>|<break>]_root`, and `_root` reads as a hierarchy marker rather
than a variant — though 64 of 87 having a bare parent means it is not simply "the
parent of an existing node", and `_break` at 15 of 30 is likewise not a plain
"destroyed twin of every effect".
🔑 **`_e`/`_f` never have a bare parent — 0 of 135.** That is independent,
asset-side confirmation of the faction law: an effect is authored per faction,
and there is no faction-neutral original for either side to derive from.

17
tools/re-capture/effect_homes.py
View File

@@ -4,7 +4,10 @@
Effect names are matched EXACTLY, not as substrings: Base.xpr contains
eff_f0002_barnhaze, so a substring test for "eff_f0002" reports a false hit for
a resource that is not there. Names are pulled as maximal [A-Za-z0-9_] runs
containing "eff_", which yields bare resource names.
containing "eff_", which yields whole resource names -- 268 of ptc_pack's 727
carry an EF_IDX_ prefix, so a pattern anchored at "eff_" would truncate them and
silently merge distinct names. A bound name is looked up bare AND under that
prefix.
Regenerates docs/re/data/effect-homes.txt.
"""
@@ -44,15 +47,19 @@ def main():
if f and FX.search(f) and isinstance(v, str) and v.startswith('eff_'):
bound.add(v)
res = sorted(n for n in bound if n in home)
un = sorted(n for n in bound if n not in home)
# a name may be stored bare or under the EF_IDX_ index prefix
def where(n):
return home.get(n, set()) | home.get('EF_IDX_' + n, set())
res = sorted(n for n in bound if where(n))
un = sorted(n for n in bound if not where(n))
print('effect names bound by a datasheet field : %d' % len(bound))
print('resolved to an .xpr package (exact name): %d' % len(res))
print('unlocated : %d' % len(un))
print()
per = collections.Counter()
for n in res:
for f in home[n]:
for f in where(n):
per[f] += 1
print('resolved, by package:')
for f, k in sorted(per.items(), key=lambda kv: (-kv[1], kv[0])):
@@ -62,7 +69,7 @@ def main():
for k in (3, 4, 0):
s = [n for n in bound if width(n) == k]
if s:
print(' %d-digit: %d resolved of %d' % (k, len([n for n in s if n in home]), len(s)))
print(' %d-digit: %d resolved of %d' % (k, len([n for n in s if where(n)]), len(s)))
print()
print('the %d unlocated, in full:' % len(un))
for n in un: