revoce extract process to earlier stage

dealfront_enrichment.py

@@ -73,37 +73,45 @@ class DealfrontScraper:
         time.sleep(1)
 
     def extract_current_page_results(self):
-        # kurzer Implicit-Wait für schnelles Fallback
+        # 1) Kurz Implicit-Wait absenken, damit Fehlversuche sofort zurückkehren
         self.driver.implicitly_wait(1)
 
-        # auf ≥1 Zeile warten
+        # 2) Auf erstes Firmen-Element warten (bis zu 15 s), dann kurzen Puffer
-        rows = self.wait.until(EC.presence_of_all_elements_located((
+        first = (By.CSS_SELECTOR, ".sticky-column a.t-highlight-text")
-            By.CSS_SELECTOR, "table#t-result-table tbody tr[id]"
+        self.wait.until(EC.visibility_of_element_located(first))
-        )))
+        time.sleep(1)
 
-        data = []
+        logger.info("Extrahiere Ergebnisse von der aktuellen Seite...")
-        for row in rows:
+        results = []
-            # Name-Element
+
-            ne = row.find_elements(By.CSS_SELECTOR, "a.t-highlight-text.t-highlight-text-snippet")
+        # 3) Auf mindestens eine Tabellenzeile warten
-            if not ne:
+        rows_sel = (By.CSS_SELECTOR, "table#t-result-table tbody tr[id]")
+        self.wait.until(EC.presence_of_all_elements_located(rows_sel))
+        rows = self.driver.find_elements(*rows_sel)
+        logger.info(f"{len(rows)} Firmen-Zeilen gefunden.")
+
+        # 4) Schleife ganz ohne Sleeps oder Implicit-Waits
+        for i, row in enumerate(rows, 1):
+            # Name per bewährtem Selector
+            name_elems = row.find_elements(By.CSS_SELECTOR, ".sticky-column a.t-highlight-text")
+            if not name_elems:
+                logger.warning(f"Zeile {i}: Kein Name-Element gefunden. Überspringe.")
                 continue
-            name = (ne[0].get_attribute("title") or ne[0].text).strip()
+            ne = name_elems[0]
+            company_name = (ne.get_attribute("title") or ne.text).strip()
 
-            # Website-Element
+            # Website per bewährtem Selector
-            we = row.find_elements(By.CSS_SELECTOR, "a.text-gray-400.t-highlight-text")
+            web_elems = row.find_elements(By.CSS_SELECTOR, "a.text-gray-400.t-highlight-text")
-            if we:
+            website = web_elems[0].text.strip() if web_elems else ""
-                site = we[0].get_attribute("href").split("://")[-1].rstrip("/")
-            else:
-                # Fallback: Zellen-Text
-                txt = row.find_elements(By.CSS_SELECTOR, "td:nth-of-type(3)")
-                site = txt[0].text.strip() if txt else ""
 
-            data.append({"name": name, "website": site})
+            results.append({'name': company_name, 'website': website})
 
-        # Implicit-Wait zurücksetzen
+        logger.info(f"Extraktion abgeschlossen: {len(results)} Firmen.")
+        return results
+
+    finally:
+        # 5) Implicit-Wait zurück auf Standard (z. B. 10 s)
         self.driver.implicitly_wait(10)
-        logger.info(f"  Extrahiert: {len(data)} Zeilen")
-        return data
 
     def click_next_page(self):
         # Paginator-Buttons greifen