eBay - Scraping & Data Extraction

import urllib.error
 
HEADERS = {
    "User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36",
    "Accept-Language": "en-US,en;q=0.9",
}
 
def is_blocked(html_or_exc):
    """Detect both Akamai edge blocks (HTTPError) and the in-page interstitial."""
    if isinstance(html_or_exc, urllib.error.HTTPError):
        return html_or_exc.code in (403, 429, 503)
    if not isinstance(html_or_exc, str):
        return True
    return 'Pardon Our Interruption' in html_or_exc or len(html_or_exc) < 20_000
 
def safe_get(url, headers):
    """http_get + edge-block detection. Returns html or None."""
    try:
        html = http_get(url, headers=headers)
    except urllib.error.HTTPError as e:
        if is_blocked(e):
            return None
        raise
    return None if is_blocked(html) else html
 
html = safe_get("https://www.ebay.com/sch/i.html?_nkw=laptop&LH_BIN=1", HEADERS)
if html is None:
    raise RuntimeError("eBay block (edge or in-page) — back off, switch IP, or use Chrome fallback")

HEADERS = {
    "User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36",
    "Accept-Language": "en-US,en;q=0.9",
    "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,*/*;q=0.8",
}

# Buy It Now only, sorted by lowest price
"https://www.ebay.com/sch/i.html?_nkw=mechanical+keyboard&LH_BIN=1&_sop=15"
 
# Auctions only
"https://www.ebay.com/sch/i.html?_nkw=vintage+camera&LH_Auction=1"
 
# New condition only, page 2
"https://www.ebay.com/sch/i.html?_nkw=laptop&LH_ItemCondition=1000&_pgn=2"

import re
 
def extract_search_results(html):
    """
    Parse eBay search results HTML into a list of dicts.
    Returns [] if blocked or no results.
    """
    if 'Pardon Our Interruption' in html or len(html) < 20_000:
        return []
 
    cards = re.split(r'(?=<li[^>]+data-listingid=)', html)
    results = []
    seen_ids = set()
 
    for card in cards[1:]:  # skip preamble before first card
        # Listing ID (dedup)
        lid_m = re.search(r'data-listingid=["\']?(\d+)', card)
        if not lid_m:
            continue
        listing_id = lid_m.group(1)
        if listing_id in seen_ids:
            continue
        seen_ids.add(listing_id)
 
        # Item URL (clean, no tracking params)
        url_m = re.search(r'href=["\']?(https://(?:www\.)?ebay\.com/itm/(\d+))', card)
        item_url = url_m.group(1).split('?')[0] if url_m else None
 
        # Title from s-card__title
        title_m = re.search(r's-card__title[^>]*>.*?primary[^>]*>([^<]+)', card, re.DOTALL)
        title = title_m.group(1).strip() if title_m else None
 
        # Skip placeholder "Shop on eBay" stub cards
        if not title or title == 'Shop on eBay':
            continue
 
        # Current price (raw inner text — currency varies by IP geo,
        # e.g. "$159.00", "HKD 297.54", "EUR 49.99"). Do NOT assume "$" prefix.
        price_m = re.search(r's-card__price[^>]*>([^<]+)<', card)
        price = price_m.group(1).strip() if price_m else None
 
        # Original / list price (strikethrough — present when discounted)
        orig_m = re.search(r'strikethrough[^>]*>([^<]+)<', card)
        original_price = orig_m.group(1).strip() if orig_m else None
 
        # Thumbnail image URL
        img_m = re.search(r'class=["\']?s-card__image[^>]*src=["\']?([^\s>"\']+)', card)
        image = img_m.group(1) if img_m else None
 
        results.append({
            'listing_id': listing_id,
            'url': item_url,
            'title': title,
            'price': price,
            'original_price': original_price,  # None if not on sale
            'image': image,
        })
 
    return results

import re
 
HEADERS = {
    "User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36",
    "Accept-Language": "en-US,en;q=0.9",
}
 
html = safe_get("https://www.ebay.com/sch/i.html?_nkw=mechanical+keyboard&LH_BIN=1&_sop=15", headers=HEADERS)
items = extract_search_results(html)
print(f"{len(items)} items")
for item in items[:5]:
    print(f"  {item['listing_id']} | {item['title'][:50]} | {item['price']}")
# Output (confirmed): 60 items
# 168219240588 | One Plus Keyboard 81 Pro Winter Bonfire Mecha... | $159.00
# 167461643107 | Logitech 920-012869 G515 TKL Wired Low Profil... | $49.99
# 167040158614 | Logitech - PRO X TKL LIGHTSPEED Wireless Mech... | $74.99

def has_exact_match(html):
    return 'No exact matches found' not in html
 
# Field-tested 2026-05-15: query "dj posket4" → 0 exact matches, 60 recommended cards.
# 5 of 60 contained "posket" in the title (fuzzy match on the most distinctive token);
# the other 55 were unrelated. Always surface this signal — the user may have typoed
# the query, and silently returning fuzzy results looks like a successful match.

import re, json
 
def extract_item_detail(html):
    """
    Extract structured data from an eBay item page.
    Returns dict or None if blocked.
    """
    if 'Pardon Our Interruption' in html:
        return None
 
    ld_blocks = re.findall(r'application/ld\+json[^>]*>(.*?)</script>', html, re.DOTALL)
    product = None
    breadcrumbs = []
 
    for ld_str in ld_blocks:
        try:
            d = json.loads(ld_str.strip())
        except Exception:
            continue
 
        if d.get('@type') == 'Product':
            product = d
        elif d.get('@type') == 'BreadcrumbList':
            breadcrumbs = [i.get('name') for i in d.get('itemListElement', [])]
 
    if not product:
        return None
 
    offers = product.get('offers', {})
    if isinstance(offers, list):
        offers = offers[0]
 
    # Schema.org condition URL -> human label
    CONDITION_MAP = {
        'NewCondition':          'New',
        'UsedCondition':         'Used',
        'RefurbishedCondition':  'Refurbished',
        'DamagedCondition':      'For Parts / Not Working',
        'LikeNewCondition':      'Like New',
        'VeryGoodCondition':     'Very Good',
        'GoodCondition':         'Good',
        'AcceptableCondition':   'Acceptable',
    }
    cond_url = offers.get('itemCondition', '')
    cond_key = cond_url.split('/')[-1]  # e.g. "RefurbishedCondition"
    condition = CONDITION_MAP.get(cond_key, cond_key)
 
    # List price from priceSpecification (only present when there's a "was" price)
    price_spec = offers.get('priceSpecification', {})
    list_price = price_spec.get('price') if price_spec.get('name') == 'List Price' else None
 
    # Shipping (first destination)
    shipping_details = offers.get('shippingDetails', [])
    if shipping_details:
        shipping_val = shipping_details[0].get('shippingRate', {}).get('value', '')
        shipping = 'Free' if str(shipping_val) in ('0', '0.0') else f"${shipping_val}"
    else:
        shipping = None
 
    # Return policy
    return_policies = offers.get('hasMerchantReturnPolicy', [])
    return_days = return_policies[0].get('merchantReturnDays') if return_policies else None
 
    return {
        'listing_id': offers.get('url', '').split('/itm/')[-1],
        'name': product.get('name'),
        'brand': product.get('brand', {}).get('name') if isinstance(product.get('brand'), dict) else product.get('brand'),
        'price': offers.get('price'),
        'list_price': list_price,     # was-price, None if no discount shown
        'currency': offers.get('priceCurrency'),
        'availability': offers.get('availability', '').split('/')[-1],  # e.g. "InStock"
        'condition': condition,
        'condition_url': cond_url,
        'shipping': shipping,
        'return_days': return_days,
        'images': product.get('image', []),
        'gtin13': product.get('gtin13'),
        'mpn': product.get('mpn'),
        'color': product.get('color'),
        'breadcrumbs': breadcrumbs,
    }

html = safe_get("https://www.ebay.com/itm/167040158614", headers=HEADERS)
if html is None:
    raise RuntimeError("eBay blocked the item request")
detail = extract_item_detail(html)
# {
#   'listing_id':   '167040158614',
#   'name':         'Logitech - PRO X TKL LIGHTSPEED Wireless Mechanical Gaming Keyboard - 920-012118',
#   'brand':        'Logitech',
#   'price':        74.99,
#   'list_price':   '219.99',
#   'currency':     'USD',
#   'availability': 'InStock',
#   'condition':    'Refurbished',
#   'shipping':     'Free',
#   'return_days':  30,
#   'images':       ['https://i.ebayimg.com/images/g/vwsAAeSwEcFpw~hW/s-l1600.jpg', ...],  # 5 images
#   'gtin13':       '097855189066',
#   'mpn':          '920-012118',
#   'color':        'Black',
#   'breadcrumbs':  ['eBay', 'Electronics', 'Computers/Tablets & Networking', ...],
# }

import re
 
def extract_ux_textspans(html):
    """Return list of all ux-textspans text values from an item page."""
    return [m.group(1) for m in re.finditer(r'ux-textspans[^>]*>([^<]+)</span>', html)]
 
# From item 167040158614 (confirmed):
# Index [3]  -> item title
# Index [4]  -> subtitle / seller tagline
# Index [5]  -> seller name ("Logitech")
# Index [6]  -> seller feedback count ("(20742)")
# Index [7]  -> seller feedback % ("99.6% positive")
# Index [10] -> current price ("US $74.99")
# Index [12] -> list price ("US $219.99")
# Index [33] -> condition label ("Excellent - Refurbished")
# Index [36] -> quantity sold ("45 sold")
# Pairs from [105] onward: item specifics as label/value pairs

for page in range(1, 4):
    url = f"https://www.ebay.com/sch/i.html?_nkw=laptop&LH_BIN=1&_sop=15&_pgn={page}"
    html = safe_get(url, headers=HEADERS)
    if html is None:
        break
    items = extract_search_results(html)
    print(f"Page {page}: {len(items)} items")
    # IMPORTANT: add delay between pages to avoid bot detection
    time.sleep(3)

import re, json, time
 
HEADERS = {
    "User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36",
    "Accept-Language": "en-US,en;q=0.9",
}
 
# Step 1: Search
html = safe_get(
    "https://www.ebay.com/sch/i.html?_nkw=mechanical+keyboard&LH_BIN=1&_sop=15&LH_ItemCondition=1000",
    headers=HEADERS
)
if html is None:
    raise RuntimeError("Rate limited — wait 60-120s and retry")
 
items = extract_search_results(html)
print(f"Found {len(items)} items")
 
# Step 2: Fetch details for top results (with delay)
details = []
for item in items[:5]:
    time.sleep(3)
    detail_html = safe_get(item['url'], headers=HEADERS)
    if detail_html is None:
        print(f"Blocked on item {item['listing_id']}, stopping")
        break
    detail = extract_item_detail(detail_html)
    if detail:
        details.append(detail)
        print(f"  {detail['name'][:50]} | {detail['price']} {detail['currency']} | {detail['condition']}")