hw_import + merge_hw_layout: Hardwareplanung einlesen und mit dem Layout zusammenfuehren

Stufen 2 und 3 aus doc/Konzept_Layout_Hardware_SCL.md:

- lib/hw_import.py: TIA-Import (*_TIA.xlsx, ohne openpyxl gelesen) +
  Positionsdatei (*_positionsconv.json) -> <projekt>_hw.json: je BMK
  Adresse, Index-Konstante, VERW, Station, Lage, Foerderer (K0102, #106 ->
  K0106); Antriebe MA/FC/GF; Gegenpruefung beider Quellen.
- lib/merge_hw_layout.py: <csv>_tro.json + <projekt>_hw.json ->
  <csv>_anlage.json (model.AnlageDocument): bindings je TRO-ID + Slot +
  Rolle mit Konstante und Adresse, Foerderer per Mehrheit (Mubea: K0102 =
  Kreisel3, K0104 = Kreisel2, K0105 = Kreisel1), Handkorrekturen
  cfg/hw_overrides.ini. Zuordnungslogik aus e_planning.py hierher
  verschoben; e_planning re-exportiert sie, Ergebnis unveraendert.
- TRO-JSON traegt optional `parts` (Lage/Huellbox, tro_flow --json),
  PinStore bekommt Slots line<n>.
- Gespeicherter Lagebezug mit SHA-256 der Positionsdatei; Unterordner in
  Eingabenamen relativ zu SKEL_DATA.
- Doku: Python_Scripts.md 5d/5e, Konzept, E-Planung_Bruecke (Abgleich
  mubea_layout.dxf / CSV / E-Planung).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

CLAUDE.md

@@ -36,7 +36,9 @@ Template" below). What exists today is:
   `tro_annotate.py --emit-lisp`)
 - `cfg/dxf_registration.json` — persisted, verified CSV↔DXF coordinate transform;
   `cfg/tro_overrides.ini` — hand-maintained corrections to the derived TROs (see `tro_overrides.py`);
-  `cfg/e_planning.ini` — project-specific rules for assigning E-planning symbols to layout parts;
+  `cfg/e_planning.ini` — project-specific rules for assigning E-planning symbols to layout parts
+  (read by `merge_hw_layout.py`, also for `e_planning.py`);
+  `cfg/hw_overrides.ini` — hand corrections of `merge_hw_layout.py` (bindings per TRO slot role, conveyors);
   `cfg/parameter.ini` — tolerances, thresholds, scales and drawing sizes of the `lib/` modules
   (read by `lib/parameter.py`; no numeric settings in the code, see `doc/Python_Scripts.md` §7a);
   `cfg/<csv>_tro_ids.ini` — per-plant TRO-ID file (versioned, IDs renamed by the SPS
@@ -150,7 +152,7 @@ this order for onboarding:
    `doc/Klassenhierarchie.md`, chapter "Daten".
 10. **`doc/Konzept_Layout_Hardware_SCL.md`** — the target workflow (planned, not built):
     layout → `<csv>_tro.json` (with slots, `jamAreas`), hardware planning →
-    `hw_import.py` → `<projekt>_hw.json`, `hw_assign.py` → `<csv>_anlage.json`,
+    `hw_import.py` → `<projekt>_hw.json`, `merge_hw_layout.py` → `<csv>_anlage.json`,
     then `scl_skeleton.py`; what happens to `e_planning.py` (split),
     `tro_annotate.py` (viewer reading the JSON) and `tro_extract.py` (dropped).
 
@@ -212,7 +214,21 @@ byte — rewrite them (`SKEL_UPDATE_GOLDEN=1`) only for a reviewed, intended cha
    `cfg/e_planning.ini` (distance, target distance, numbering rule `MB = BG + 1`).
    `scl_skeleton.py --eplan` fills separator sensor and stopper valve from it; jam sensors
    and outfeed/switch valves are assigned but not yet placed in the SCL. See
-   `doc/Python_Scripts.md` §5c and `doc/E-Planung_Bruecke.md`.
+   `doc/Python_Scripts.md` §5c and `doc/E-Planung_Bruecke.md`. Since 2026-10-15 its model
+   and assignment logic live in `hw_import.py`/`merge_hw_layout.py` (re-exported here).
+6. **`lib/hw_import.py`** (implemented, stage 2 of `doc/Konzept_Layout_Hardware_SCL.md`) —
+   TIA import `*_TIA.xlsx` (PLC Tags + Constants, read without openpyxl) + positions file
+   `*_positionsconv.json` → `<projekt>_hw.json` (`HardwareDocument`): per BMK address,
+   index constant `cIn…`/`cOut…`, VERW, station, position, conveyor (`K0102`; `#106` →
+   `K0106`); drives MA↔FC (+ encoder GF); cross-check of both sources. EA list and WSCAD
+   are not needed. See `doc/Python_Scripts.md` §5d.
+7. **`lib/merge_hw_layout.py`** (implemented, stage 3) — `<csv>_tro.json` + `<projekt>_hw.json`
+   → `<csv>_anlage.json` (`model.AnlageDocument`: layout unchanged + `bindings[]` keyed by
+   TRO-ID + slot + role, `conveyors[]`). Reads no CSV: the TRO-JSON carries `parts`
+   (positions/bounding boxes, written by `tro_flow --json`). Conveyors are matched by
+   majority vote of the devices naming them; corrections in `cfg/hw_overrides.ini`,
+   numbers in `cfg/parameter.ini [merge_hw_layout]`. `scl_skeleton` does not read the plant JSON
+   yet (stage 4). See `doc/Python_Scripts.md` §5e.
 
 What is **not** built out yet, relative to the original two-tool design: