conflict-decisions: correctDob remaps a wrong-birthdate duplicate before dedup
god added a `correctDob` (DD.MM.YYYY) decisions field: the matched (name+dob) record is a DUPLICATE with a wrong birthdate → remap its DOB to correctDob so dedup MERGES it into the canonical same-named animal. apply_dob_remaps runs BEFORE dedup (it changes the dedup identity); tolerates a missing file; logged as "DOB-Remaps: N". First use: Chelsea *15.10.2021 → *02.04.2021 (merges into the canonical record). test_extract covers the remap + that both records then share one name+dob identity. Extractor-only. python + dotnet 121/121 green. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
@@ -883,6 +883,31 @@ def write_report(merged, conflicts, orphans, raw_count, litters, photo_count,
|
||||
|
||||
|
||||
# ------------------------------------------------------------------------ main
|
||||
def apply_dob_remaps(raw_animals, path):
|
||||
"""PRE-dedup: a conflict-decision carrying `correctDob` marks a record as a DUPLICATE with a
|
||||
wrong birthdate — remap that raw record's DOB to correctDob so dedup MERGES it into the
|
||||
canonical same-named animal (e.g. Chelsea *15.10.2021 -> *02.04.2021). Match =
|
||||
norm_name(name)+norm_dob(dob). Tolerates a missing/garbled file. Returns the remap count.
|
||||
Must run BEFORE dedup (it changes the dedup identity). (god/HUMANQUESTION D — Dubletten.)"""
|
||||
remaps = {}
|
||||
try:
|
||||
with open(path, encoding="utf-8") as fh:
|
||||
for r in (json.load(fh).get("resolutions") or []):
|
||||
if r.get("correctDob"):
|
||||
remaps[(norm_name(r.get("name", "")), norm_dob(r.get("dob", "")))] = r["correctDob"]
|
||||
except (OSError, ValueError):
|
||||
return 0
|
||||
if not remaps:
|
||||
return 0
|
||||
n = 0
|
||||
for a in raw_animals:
|
||||
new = remaps.get((norm_name(a.get("name", "")), norm_dob(a.get("dob", ""))))
|
||||
if new and a.get("dob") != new:
|
||||
a["dob"] = new
|
||||
n += 1
|
||||
return n
|
||||
|
||||
|
||||
def apply_conflict_decisions(merged, conflicts, path):
|
||||
"""Consume human conflict resolutions (tools/import/conflict-decisions.json) so the wife's
|
||||
answers UN-QUARANTINE animals. Schema: {"resolutions":[{name, dob, decision, genotype?,
|
||||
@@ -953,8 +978,9 @@ def main():
|
||||
litters = extract_wurfchronik(args.wurfchronik)
|
||||
print(f"Wurfchronik: {len(litters)} Würfe")
|
||||
|
||||
merged, conflicts, orphans, zucht_splits = dedup(raw_animals)
|
||||
decisions_path = os.path.join(HERE, "conflict-decisions.json")
|
||||
dob_remaps = apply_dob_remaps(raw_animals, decisions_path) # before dedup (changes identity)
|
||||
merged, conflicts, orphans, zucht_splits = dedup(raw_animals)
|
||||
resolved_by_decision = apply_conflict_decisions(merged, conflicts, decisions_path)
|
||||
match_stats = match_litters(merged, litters)
|
||||
photo_count = sum(len(a["photos"]) for a in merged)
|
||||
@@ -973,7 +999,7 @@ def main():
|
||||
|
||||
print(f"\nRoh: {len(raw_animals)} → eindeutig: {len(merged)} "
|
||||
f"| Konflikte: {len(conflicts)} | per Entscheidung gelöst: {resolved_by_decision} "
|
||||
f"| Zucht-Splits: {len(zucht_splits)} "
|
||||
f"| DOB-Remaps: {dob_remaps} | Zucht-Splits: {len(zucht_splits)} "
|
||||
f"| Orphans: {len(orphans)} | Fotos: {photo_count}")
|
||||
print(f"Wurf-Verknüpfung: {match_stats['parents']} (Datum+Eltern), "
|
||||
f"{match_stats['dateOnly']} (nur Datum), {match_stats['ambiguous']} mehrdeutig "
|
||||
|
||||
@@ -5,10 +5,10 @@ _Automatisch erzeugt von `tools/import/extract.py` — **noch nichts in die Date
|
||||
## Überblick
|
||||
|
||||
- Rohe Tier-Einträge aus den Stammbäumen: **950**
|
||||
- Nach Zusammenführung (eindeutige Tiere): **622**
|
||||
- davon mit Geburtsdatum: 327
|
||||
- Nach Zusammenführung (eindeutige Tiere): **621**
|
||||
- davon mit Geburtsdatum: 326
|
||||
- in mehreren Dateien gefunden (Dubletten zusammengeführt): 158
|
||||
- Konflikte zur Klärung: **15**
|
||||
- Konflikte zur Klärung: **9**
|
||||
- Mehrdeutige / unvollständige Einträge (ohne Name+Datum): **310**
|
||||
- Fotos zugeordnet: **137**
|
||||
- Würfe aus der Wurfchronik: **752**
|
||||
@@ -26,14 +26,8 @@ Gleiches Tier (Name+Datum), aber widersprüchliche Angaben in verschiedenen Date
|
||||
| Tier | Geburtsdatum | abweichende Genotypen | abweichende Farbschläge | Sterbedaten | Dateien |
|
||||
|---|---|---|---|---|---|
|
||||
| Ella | 10.06.2019 | Aa C D- ee[f] GG P- spsp // Aa Cc[chm] D- ee[f] UwUw P- spsp | Algierfuchsschimmel, hell | 03.02.2023 | Stammbaum von Akio Kids, Stammbaum von CP-Fuchs, CP-Sa Sp von Unity, Stammbaum von Valentino Firehearts Kids |
|
||||
| Zuleika von den Kleinen Chaoten | 24.10.2015 | aa c[chm]c[h] D- E G P- spsp // aa c[chm]c[h] D- Ee Gg P- spsp // aa c[chm]c[h] DD Ee Gg P- spsp | — | 24.02.2019 | Stammbaum von Akio Kids, Stammbaum von CP-Fuchs, CP-Sa Sp von Unity, Stammbaum von Valentino Firehearts Kids |
|
||||
| Vestra von den Schlossmäusen | 08.02.2019 | Aa Cc[chm] D- EE GG PP Spsp [WP] // Aa Cc[chm] DD EE GG PP Spsp [WP] | — | 26.05.2023 | Stammbaum von Akio Kids, Stammbaum von CP-Fuchs, CP-Sa Sp von Unity, Stammbaum von Fire Kids, Stammbaum von Valentino Firehearts Kids |
|
||||
| Flint von den Kleinen Chaoten | 23.12.2017 | aa Cc[chm] D- ee Gg P- spsp | — | 10.05.2021 // 10.05.2022 | Stammbaum von Akio Kids, Stammbaum von CP-Fuchs, CP-Sa Sp von Unity |
|
||||
| Kazu von den Kleinen Chaoten | 23.04.2013 | Aa Cc[chm] DD e[f]e[f] Gg P Spsp // Aa Cc[chm] DD ee[f] UwUw PP Spsp | — | 03.09.2017 | Stammbaum von Akio Kids, Stammbaum von Vance |
|
||||
| Milka of LennyLengo | 09.12.2018 | aa C- dd E- Gg P- Spsp // aa Cc[h] dd EE Gg P- Spsp | — | 22.12.2021 | Stammbaum von Alberto Kids, Stammbaum von Stella Kids |
|
||||
| Silvain von den Kleinen Chaoten | 27.03.2022 | aa c[chm]c[chm] Dd Ee[-] Gg P- Spsp // aa c[chm]c[chm] Dd ee[-] Gg Pp Spsp | — | 31.12.2024 | Stammbaum von Alberto Kids, Stammbaum von CP-Fuchs, CP-Sa Sp von Unity |
|
||||
| Enya von den Kleinen Chaoten | 01.11.2017 | Aa c[chm]c[chm] D- ee[-] G- P- spsp // Aa c[chm]c[chm] D- ee[-] Uwuw[d] P- spsp | — | — | Stammbaum von Alberto Kids, Stammbaum von Fire Kids, Stammbaum von Kohlief, Goldfuchsef Sp von Chrissi, Stammbaum von Stella Kids |
|
||||
| Molly of Black Forest | 13.09.2021 | /+, Aa Cc[chm] D- Ee gg P- spsp // Aa Cc[chm] Dd Ee gg Pp spsp | — | 03.05.2021 | Stammbaum von Alberto Kids, Stammbaum von CP-Fuchs, CP-Sa Sp von Unity |
|
||||
| Little Runner's Big Ben | 03.02.2020 | Aa Cc[chm] DD Ee Gg PP Spsp // Aa Cc[chm] DD Ee Gg Pp Spsp | — | 14.10.2023 | Stammbaum von CP-Fuchs, CP-Sa Sp von Unity, Stammbaum von Fire Kids, Stammbaum von Goldfuchs Sp (Pikachu) Kids, Stammbaum von Kohlief, Goldfuchsef Sp von Chrissi, Stammbaum von Valentino Firehearts Kids, Stammbaum von Watarus Kids |
|
||||
| Vance Jr. von den Kleinen Chaoten | 10.04.2022 | aa Cc[hm] Dd Ee gg P- Spsp // aa Cc[hm] Dd Ee gg P- spsp | Kohlfuchs, hell | — | Stammbaum von Fire Kids, Stammbaum von Stella Kids |
|
||||
| Victoria Welby gen. Welby v.d. Kleinen Chaoten | 16.01.2023 | Aa CC D- Ee[f] Gg pp Spsp [DP] // Aa CC D- ee[f] Gg pp Spsp [DP] | Goldfuchsschimmel Punktschecke DP | 17.02.2026 | Stammbaum von Goldfuchs Sp (Pikachu) Kids, Stammbaum von Kohlief, Goldfuchsef Sp von Chrissi, Stammbaum von Watarus Kids |
|
||||
@@ -97,7 +91,7 @@ Gleiches Tier (Name+Datum), aber widersprüchliche Angaben in verschiedenen Date
|
||||
- „Hagrid Rubeus of Black Forest“ → Hagrid Rubeus of Black Forest (*18.07.2019)
|
||||
- „Charly of Golden Lights“ → Charly of Golden Lights (*05.04.2016)
|
||||
- „Ziwa of Golden Lights“ → Ziwa of Golden Lights (*29.04.2016)
|
||||
- „Chelsea von den Kleinen Chaoten“ → Chelsea von den Kleinen Chaoten (*02.04.2021); Chelsea von den Kleinen Chaoten (*15.10.2021)
|
||||
- „Chelsea von den Kleinen Chaoten“ → Chelsea von den Kleinen Chaoten (*02.04.2021)
|
||||
- „Pinto of Fiomi“ → Pinto of Fiomi (*28.08.2016)
|
||||
- „Living Force's Idefix“ → Living Force's Idefix (*05.04.2016)
|
||||
- „Scarlett of Samsimar“ → Scarlett of Samsimar (*05.09.2018)
|
||||
@@ -139,12 +133,12 @@ Diese Tokens stehen weiter in `rawGenotype`/`unmappedTokens` — Entscheidung (M
|
||||
| `/+` | 7 | ? |
|
||||
| `-g` | 2 | ? |
|
||||
| `C(C)` | 2 | Schreibweise (C trägt c) |
|
||||
| `chmchm` | 2 | Schreibweise (c[chm]c[chm]) |
|
||||
| `Cc[]` | 1 | ? |
|
||||
| `-psp` | 1 | ? |
|
||||
| `G(G)` | 1 | ? |
|
||||
| `/` | 1 | ? |
|
||||
| `+2018` | 1 | ? |
|
||||
| `chmchm` | 1 | Schreibweise (c[chm]c[chm]) |
|
||||
| `c[chm]chm]` | 1 | ? |
|
||||
| `Dea/dea]` | 1 | ? |
|
||||
| `DD-Tumor` | 1 | ? |
|
||||
|
||||
@@ -101,6 +101,30 @@ check("decision removes both entries from conflicts list", conflicts == [])
|
||||
check("apply_conflict_decisions returns resolved count", n == 2)
|
||||
check("missing decisions file tolerated (returns 0)",
|
||||
e.apply_conflict_decisions([], [], os.path.join(tempfile.gettempdir(), "does-not-exist.json")) == 0)
|
||||
|
||||
# --- correctDob: a wrong-birthdate duplicate is remapped BEFORE dedup so it merges ---
|
||||
dec2 = os.path.join(tempfile.gettempdir(), "decisions-dob.json")
|
||||
_json.dump({"resolutions": [
|
||||
{"name": "Chelsea von den Kleinen Chaoten", "dob": "15.10.2021",
|
||||
"decision": "duplicate wrong birthdate", "correctDob": "02.04.2021", "source": "test"},
|
||||
]}, open(dec2, "w", encoding="utf-8"))
|
||||
raw = [
|
||||
{"name": "Chelsea von den Kleinen Chaoten", "dob": "15.10.2021"}, # the wrong-dob duplicate
|
||||
{"name": "Chelsea von den Kleinen Chaoten", "dob": "02.04.2021"}, # canonical
|
||||
{"name": "Other Animal", "dob": "01.01.2020"},
|
||||
]
|
||||
rn = e.apply_dob_remaps(raw, dec2)
|
||||
check("correctDob remaps the wrong-dob record", raw[0]["dob"] == "02.04.2021")
|
||||
check("correctDob leaves the canonical record alone", raw[1]["dob"] == "02.04.2021")
|
||||
check("correctDob leaves unrelated records alone", raw[2]["dob"] == "01.01.2020")
|
||||
check("apply_dob_remaps returns remap count", rn == 1)
|
||||
check("after remap both Chelsea share one dedup identity (name+dob)",
|
||||
e.norm_dob(raw[0]["dob"]) == e.norm_dob(raw[1]["dob"]))
|
||||
check("missing decisions file tolerated for dob remaps (returns 0)",
|
||||
e.apply_dob_remaps([], os.path.join(tempfile.gettempdir(), "nope.json")) == 0)
|
||||
try: os.remove(dec2)
|
||||
except OSError: pass
|
||||
|
||||
try: os.remove(dec_path)
|
||||
except OSError: pass
|
||||
|
||||
|
||||
Reference in New Issue
Block a user