filter description

backend/kleinanzeigen_scrape.py

@@ -222,6 +222,11 @@ class KleinanzeigenScraper:
             if not zip_code:
                 return None
 
+            desc = None
+            desc_elem = soup.find("p", id="viewad-description-text")
+            if desc_elem:
+                desc = desc_elem.text.strip()
+
             listing = {
                 "title": title,
                 "price": price,
@@ -231,6 +236,7 @@ class KleinanzeigenScraper:
                 "date_added": date_added.isoformat() if date_added else None,
                 "image": first_image,
                 "url": url,
+                "desc": desc,
             }
 
             # Add coordinates
@@ -275,6 +281,7 @@ def main():
     # Search for listings
     print("Step 1: Searching for listing URLs...")
     listing_urls = scraper.search_listings(search_term, max_pages, min_price, max_price)
+    #listing_urls = ["https://www.kleinanzeigen.de/s-anzeige/cube-nuroad-pro-fe-2023-58-cm-l-/3226095826-217-23527"]
     print(f"Found {len(listing_urls)} listings\n")
 
     if len(listing_urls) > 0:
@@ -294,6 +301,7 @@ def main():
             print(f"  Date: {listing['date_added']}")
             print(f"  Coordinates: {listing.get('lat')}, {listing.get('lon')}")
             print(f"  Image: {listing['image']}")
+            print(f"  Description: {listing['desc']}")
         else:
             print("Failed to scrape listing")
 

backend/scrape_proxy.py

@@ -9,6 +9,7 @@ import os
 import uuid
 import threading
 import random
+import re
 
 from kleinanzeigen_scrape import KleinanzeigenScraper
 
@@ -128,6 +129,26 @@ def prefetch_listings_thread(session_id):
     print(f"Prefetch complete for session {session_id}")
 
 
+def filter_listing_search_description(description_term, description):
+    # Normalize the description (lowercase + remove non-alphanumeric chars)
+    text_clean = re.sub(r"[^a-z0-9]+", "", description.lower())
+
+    # Split the search terms by semicolon
+    terms = description_term.split(";")
+
+    # Check each term
+    for term in terms:
+        # Normalize each term
+        term_clean = re.sub(r"[^a-z0-9]+", "", term.lower())
+
+        # If any normalized term is not found in the normalized description → return False
+        if term_clean not in text_clean:
+            return False
+
+    # All terms were found
+    return True
+
+
 @app.route("/api/search", methods=["POST"])
 def api_search():
     """API endpoint for searching listings - returns count and starts prefetch"""
@@ -141,6 +162,7 @@ def api_search():
     num_listings = data.get("num_listings", 25)
     min_price = data.get("min_price", 0)
     max_price = data.get("max_price", 1000000000)
+    power_search_description = data.get("search_term_desc", "")
 
     if not search_term:
         return jsonify({"error": "Search term is required"}), 400
@@ -163,6 +185,7 @@ def api_search():
         # Store session with creation timestamp
         scrape_sessions[session_id] = {
             "urls": listing_urls,
+            "power_search_desc": power_search_description,
             "total": len(listing_urls),
             "scraped": 0,
             "listings": [],
@@ -207,6 +230,11 @@ def api_scrape(session_id):
 
     if len(session["listings"]) > 0:
         listing = session["listings"].pop(0)
+        if session["power_search_desc"]:
+            if not filter_listing_search_description(
+                session["power_search_desc"], listing["desc"]
+            ):
+                listing = None
     else:
         listing = None