EXTRACT band-aware Farbschlag: deep bands (gen>=2) have no colour cell

Julian-confirmed against the real xlsx (Kentucky, Akio Kids): pedigree
blocks have two shapes. EARLY bands (gen 0-1, cols B/E/H) are 5-cell
Name/DOB/Farbschlag/Genotype/Zucht — real Farbschlag present. DEEP bands
(gen>=2, cols K/N/Q...) are 3-cell Name/DOB/Genotype with NO Farbschlag —
the colour is derived from the genotype.

The old extractor grabbed the NEXT block's name or a stray health note
(e.g. "DD-Tumor", "Chronische Ohrenentzündung...") as the deep-band
Farbschlag — source of the D1 'name-in-Farbschlag' conflicts and a big
chunk of farbschlagUnmatched. Fix: gen_of(col)>=2 => never assign a
Farbschlag in that block (Name/DOB/Genotype only); the cell is left for the
next block's name search. Deep-band animals load with empty Farbschlag and
the UI computes 'Errechnet' from the stored genotype (lossless).

Dedup already prefers a real (non-empty) early-band Farbschlag, so an animal
appearing in both an early band (Chesnut -> 'Kohlfuchsschimmel') and deep
bands keeps the real value; Chesnut and Tennessee stay separate (distinct
genotype + name+DOB key).

Result: 248/306 loadable animals now have empty Farbschlag (genotype-
derived), 0 suspicious name/note values remain, 0 Farbschlag-based conflicts
(D1 dissolved). Regression test test_extract.py builds a two-band xlsx and
asserts the deep band yields no Farbschlag + the stray note doesn't bleed.

Extractor-only (no C# change). python test_genotype + test_extract green;
dotnet 118/118. Batches with Kevin's GEN-3f seed for one re-import.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

tools/import/extract.py

@@ -211,6 +211,11 @@ def extract_stammbaum(path):
             farbschlag = ""
             geno = geno0
             breeder = ""
+            # BAND-AWARE (Julian-confirmed): early bands (gen 0-1, cols B/E/H) are 5-cell blocks
+            # WITH a Farbschlag cell; deep bands (gen >= 2, cols K/N/Q...) are 3-cell blocks
+            # (Name/DOB/Genotype) with NO Farbschlag — colour is derived from the genotype. So in
+            # deep bands we must NOT grab the next block's name or a stray health note as Farbschlag.
+            deep_band = gen_of(c) >= 2
             for rr in range(r + 1, r + 4):
                 cell = cells.get((c, rr))
                 if not cell:
@@ -221,7 +226,7 @@ def extract_stammbaum(path):
                 elif re.search(r"\b(Zucht|Privatzucht)\b", cell) or cell.startswith("("):
                     breeder = cell
                     used.add((c, rr))
-                elif not farbschlag and not re.match(r"^\*?\s?\d", cell) \
+                elif not deep_band and not farbschlag and not re.match(r"^\*?\s?\d", cell) \
                         and not looks_like_animal_name(cell):
                     farbschlag = cell
                     used.add((c, rr))

tools/import/test_extract.py

@@ -0,0 +1,83 @@
+"""Zero-dep tests for extract.py band-aware Farbschlag + name-bleed guard.
+
+Run: python test_extract.py   (exit 0 = all pass)
+Covers (PEDIGREE-LINK / Julian-confirmed): deep pedigree bands (gen >= 2, cols K/N/Q...)
+are Name/DOB/Genotype ONLY — no Farbschlag cell — so a stray health note or the next
+block's name must NOT be captured as Farbschlag; early bands (gen 0-1) keep their real
+Farbschlag. Plus the looks_like_animal_name guard (a parent name must not be a Farbschlag).
+"""
+import os
+import sys
+import zipfile
+import tempfile
+import extract as e
+
+failed = 0
+
+
+def check(name, cond):
+    global failed
+    print(("ok:   " if cond else "FAIL: ") + name)
+    if not cond:
+        failed += 1
+
+
+def _cell(ref, text):
+    return f'<c r="{ref}" t="inlineStr"><is><t>{text}</t></is></c>'
+
+
+def _make_xlsx(path, cells):
+    """cells: {(colLetter+row): text}. Build a minimal single-sheet xlsx (no styles)."""
+    rows = {}
+    for ref, text in cells.items():
+        r = int("".join(ch for ch in ref if ch.isdigit()))
+        rows.setdefault(r, []).append(_cell(ref, text))
+    body = "".join(f'<row r="{r}">{"".join(cs)}</row>' for r, cs in sorted(rows.items()))
+    sheet = ('<?xml version="1.0"?><worksheet xmlns="http://x"><sheetData>'
+             + body + "</sheetData></worksheet>")
+    with zipfile.ZipFile(path, "w") as z:
+        z.writestr("xl/worksheets/sheet1.xml", sheet)
+
+
+# --- band-aware Farbschlag ---
+# col E = gen 0 (early, HAS Farbschlag); col K = col 11 = gen 2 (deep, NO Farbschlag).
+tmp = os.path.join(tempfile.gettempdir(), "bandtest.xlsx")
+_make_xlsx(tmp, {
+    # early band (E): Name / *DOB / Farbschlag / Genotype
+    "E10": "Chesnut",
+    "E11": "*13.11.2019",
+    "E12": "Kohlfuchsschimmel",
+    "E13": "aa CC DD ee GG PP spsp rere",
+    # deep band (K): Name / *DOB / Genotype / stray NOTE (must NOT become Farbschlag)
+    "K10": "DeepAnimal",
+    "K11": "*01.01.2020",
+    "K12": "aa CC DD EE GG PP spsp rere",
+    "K13": "DD-Tumor",
+})
+try:
+    animals = e.extract_stammbaum(tmp)
+    by_name = {a["name"]: a for a in animals}
+    check("early band keeps real Farbschlag",
+          by_name.get("Chesnut", {}).get("farbschlag") == "Kohlfuchsschimmel")
+    check("deep band has NO Farbschlag (note not grabbed)",
+          by_name.get("DeepAnimal", {}).get("farbschlag") == "")
+    check("deep-band animal still parsed (Name/DOB/Genotype)",
+          "DeepAnimal" in by_name and by_name["DeepAnimal"]["dob"].startswith("01.01"))
+finally:
+    try: os.remove(tmp)
+    except OSError: pass
+
+gen = e.gen_of
+check("gen_of: early bands < 2 (E,H)", gen(5) < 2 and gen(8) < 2)
+check("gen_of: deep bands >= 2 (K,N,Q)", gen(11) >= 2 and gen(14) >= 2)
+
+# --- name-bleed guard (a parent name is not a Farbschlag) ---
+check("v.d. name rejected", e.looks_like_animal_name("Tennessee von den Kleinen Chaoten"))
+check("gen.+v.d. name rejected", e.looks_like_animal_name("Victoria Welby gen. Welby v.d. Kleinen Chaoten"))
+check("real Farbschlag accepted", not e.looks_like_animal_name("Kohlfuchsschimmel"))
+check("real Farbschlag accepted 2", not e.looks_like_animal_name("Orangeschimmel, hell"))
+
+if failed:
+    print(f"\n{failed} test(s) FAILED")
+    sys.exit(1)
+print("\nALL PASS")