Floke/Brancheneinstufung2
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

duplicate_checker.py aktualisiert

Browse Source
This commit is contained in:
Floke
2025-08-01 11:10:42 +00:00
parent 0960f87409
commit 343f62cbdb
1 changed files with 37 additions and 35 deletions
Download Patch File Download Diff File Expand all files Collapse all files

72
duplicate_checker.py
View File

@@ -1,4 +1,4 @@
# duplicate_checker.py
# duplicate_checker.py (v2.0 - mit Blocking-Strategie)
import logging
import pandas as pd
@@ -10,42 +10,33 @@ from google_sheet_handler import GoogleSheetHandler
# --- Konfiguration ---
CRM_SHEET_NAME = "CRM_Accounts"
MATCHING_SHEET_NAME = "Matching_Accounts"
SCORE_THRESHOLD = 80 # Mindest-Score, um als "wahrscheinlicher Treffer" zu gelten
SCORE_THRESHOLD = 80
# --- Logging einrichten ---
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
def calculate_similarity(record1, record2):
"""Berechnet einen gewichteten Ähnlichkeits-Score zwischen zwei Datensätzen."""
total_score = 0
# 1. Website-Domain (stärkstes Signal)
if record1['normalized_domain'] and record1['normalized_domain'] == record2['normalized_domain']:
total_score += 100
# 2. Firmenname (Fuzzy-Signal)
if record1['normalized_name'] and record2['normalized_name']:
name_similarity = fuzz.token_set_ratio(record1['normalized_name'], record2['normalized_name'])
total_score += name_similarity * 0.7 # Gewichtung: 70%
# 3. Standort (Bestätigungs-Signal)
total_score += name_similarity * 0.7
if record1['CRM Ort'] and record1['CRM Ort'] == record2['CRM Ort']:
if record1['CRM Land'] and record1['CRM Land'] == record2['CRM Land']:
total_score += 20 # Bonus für vollen Standort-Match
total_score += 20
return round(total_score)
def main():
"""Hauptfunktion zum Laden, Vergleichen und Schreiben der Daten."""
logging.info("Starte den Duplikats-Check...")
logging.info("Starte den Duplikats-Check (v2.0 mit Blocking)...")
try:
sheet_handler = GoogleSheetHandler()
except Exception as e:
logging.critical(f"FEHLER bei der Initialisierung des GoogleSheetHandler: {e}")
logging.critical(f"FEHLER bei Initialisierung des GoogleSheetHandler: {e}")
return
# 1. Daten laden
logging.info(f"Lade Master-Daten aus '{CRM_SHEET_NAME}'...")
crm_df = sheet_handler.get_sheet_as_dataframe(CRM_SHEET_NAME)
if crm_df is None or crm_df.empty:
@@ -58,29 +49,41 @@ def main():
logging.critical(f"Konnte keine Daten aus '{MATCHING_SHEET_NAME}' laden. Breche ab.")
return
# 2. Daten normalisieren
logging.info("Normalisiere Daten für den Vergleich...")
crm_df['normalized_name'] = crm_df['CRM Name'].astype(str).apply(normalize_company_name)
crm_df['normalized_domain'] = crm_df['CRM Website'].astype(str).apply(simple_normalize_url)
crm_df['CRM Ort'] = crm_df['CRM Ort'].astype(str).str.lower().str.strip()
crm_df['CRM Land'] = crm_df['CRM Land'].astype(str).str.lower().str.strip()
for df in [crm_df, matching_df]:
df['normalized_name'] = df['CRM Name'].astype(str).apply(normalize_company_name)
df['normalized_domain'] = df['CRM Website'].astype(str).apply(simple_normalize_url)
df['CRM Ort'] = df['CRM Ort'].astype(str).str.lower().str.strip()
df['CRM Land'] = df['CRM Land'].astype(str).str.lower().str.strip()
# Blocking Key: Das erste Wort des normalisierten Namens
df['block_key'] = df['normalized_name'].apply(lambda x: x.split()[0] if x else None)
matching_df['normalized_name'] = matching_df['CRM Name'].astype(str).apply(normalize_company_name)
matching_df['normalized_domain'] = matching_df['CRM Website'].astype(str).apply(simple_normalize_url)
matching_df['CRM Ort'] = matching_df['CRM Ort'].astype(str).str.lower().str.strip()
matching_df['CRM Land'] = matching_df['CRM Land'].astype(str).str.lower().str.strip()
# 3. Matching-Prozess
logging.info("Starte Matching-Prozess... Dies kann einige Zeit dauern.")
# --- NEUE, SCHNELLE BLOCKING-STRATEGIE ---
logging.info("Erstelle Index für CRM-Daten zur Beschleunigung...")
crm_index = {}
for index, row in crm_df.iterrows():
key = row['block_key']
if key:
if key not in crm_index:
crm_index[key] = []
crm_index[key].append(row)
logging.info("Starte Matching-Prozess...")
results = []
total_matches = len(matching_df)
for index, match_row in matching_df.iterrows():
best_score = 0
best_match_name = ""
logging.info(f"Prüfe: {match_row['CRM Name']}...")
logging.info(f"Prüfe {index + 1}/{total_matches}: {match_row['CRM Name']}...")
for _, crm_row in crm_df.iterrows():
# Finde den Block von Kandidaten
block_key = match_row['block_key']
candidates = crm_index.get(block_key, [])
# Führe den teuren Vergleich nur für die Kandidaten in diesem Block durch
for crm_row in candidates:
score = calculate_similarity(match_row, crm_row)
if score > best_score:
best_score = score
@@ -89,17 +92,16 @@ def main():
if best_score >= SCORE_THRESHOLD:
results.append({'Potenzieller Treffer im CRM': best_match_name, 'Ähnlichkeits-Score': best_score})
else:
results.append({'Potenzieller Treffer im CRM': '', 'Ähnlichkeits-Score': best_score})
# Wenn nichts im Block gefunden wurde, trotzdem den besten Treffer (kann 0 sein) anzeigen
results.append({'Potenzieller Treffer im CRM': '' if not best_match_name else best_match_name, 'Ähnlichkeits-Score': best_score})
# 4. Ergebnisse zusammenführen und schreiben
logging.info("Matching abgeschlossen. Schreibe Ergebnisse zurück ins Sheet...")
result_df = pd.DataFrame(results)
output_df = pd.concat([matching_df.reset_index(drop=True), result_df], axis=1)
# Entferne die temporären normalisierten Spalten für eine saubere Ausgabe
output_df = output_df.drop(columns=['normalized_name', 'normalized_domain'])
# Die ursprünglichen Spalten aus matching_df für die Ausgabe nehmen
output_df = matching_df[['CRM Name', 'CRM Website', 'CRM Ort', 'CRM Land']].copy()
output_df = pd.concat([output_df.reset_index(drop=True), result_df], axis=1)
# Konvertiere DataFrame in Liste von Listen für den Upload
data_to_write = [output_df.columns.values.tolist()] + output_df.values.tolist()
success = sheet_handler.clear_and_write_data(MATCHING_SHEET_NAME, data_to_write)