Yves Gugger
5e0d4c6590
- Remove seed/demo auction endpoint + scripts (no mock data) - Rebuild AuctionScraper: strict validation (no -- bids, requires end_time) - Add robust sources: - ExpiredDomains provider auction pages (GoDaddy/Namecheap/Sedo) - Park.io auctions table - Sav load_domains_ajax table - Simplify hidden API scrapers to Dynadot only - Add unique index on (platform, domain) + safe upsert - Update deployment/docs to reflect real scraping
229 lines
8.5 KiB
Python
229 lines
8.5 KiB
Python
"""
|
|
Hidden JSON API scrapers for auction platforms.
|
|
|
|
Important project rule:
|
|
- We do NOT generate mock/demo/estimated auction values.
|
|
- This module only includes sources that provide verifiable auction fields.
|
|
|
|
Currently enabled:
|
|
- Dynadot hidden JSON API (used by their frontend)
|
|
|
|
Affiliate links:
|
|
- Read from environment variables. If not configured, plain URLs are used.
|
|
- No placeholder affiliate IDs are baked into code.
|
|
"""
|
|
|
|
import logging
|
|
import os
|
|
from datetime import datetime, timedelta
|
|
from typing import Any, Dict, List, Optional
|
|
|
|
import httpx
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
|
|
def build_affiliate_url(platform: str, domain: str, original_url: Optional[str] = None) -> str:
|
|
"""
|
|
Build an affiliate URL for a given platform and domain.
|
|
|
|
If the affiliate program is not configured, returns the plain provider URL.
|
|
If `original_url` is provided, it is preferred (e.g. ExpiredDomains click-through links).
|
|
"""
|
|
if original_url:
|
|
return original_url
|
|
|
|
p = (platform or "").strip()
|
|
d = (domain or "").strip().lower()
|
|
|
|
if not d:
|
|
return ""
|
|
|
|
if p == "Dynadot":
|
|
base = f"https://www.dynadot.com/market/auction/{d}"
|
|
affiliate_id = os.getenv("DYNADOT_AFFILIATE_ID")
|
|
return f"{base}?affiliate_id={affiliate_id}" if affiliate_id else base
|
|
|
|
if p == "GoDaddy":
|
|
base = f"https://auctions.godaddy.com/trpItemListing.aspx?domain={d}"
|
|
isc = os.getenv("GODADDY_ISC")
|
|
return f"{base}&isc={isc}" if isc else base
|
|
|
|
if p == "Namecheap":
|
|
base = f"https://www.namecheap.com/market/domain/{d}"
|
|
aff = os.getenv("NAMECHEAP_AFFILIATE_ID")
|
|
return f"{base}?aff={aff}" if aff else base
|
|
|
|
if p == "Sedo":
|
|
base = f"https://sedo.com/search/details/?domain={d}"
|
|
partner = os.getenv("SEDO_PARTNER_ID")
|
|
return f"{base}&partnerid={partner}" if partner else base
|
|
|
|
if p == "Park.io":
|
|
return f"https://park.io/domain/{d}"
|
|
|
|
if p == "Sav":
|
|
ref = os.getenv("SAV_REF")
|
|
base = f"https://www.sav.com/domain/{d}"
|
|
return f"{base}?ref={ref}" if ref else base
|
|
|
|
return ""
|
|
|
|
|
|
class DynadotApiScraper:
|
|
"""
|
|
Scraper for Dynadot Marketplace using their hidden JSON API.
|
|
|
|
Endpoint:
|
|
- https://www.dynadot.com/dynadot-vue-api/dynadot-service/marketplace-api
|
|
"""
|
|
|
|
BASE_URL = "https://www.dynadot.com"
|
|
MARKETPLACE_API = "/dynadot-vue-api/dynadot-service/marketplace-api"
|
|
|
|
def _parse_end_time(self, item: Dict[str, Any]) -> Optional[datetime]:
|
|
# Dynadot often provides an epoch timestamp in ms
|
|
end_time_stamp = item.get("end_time_stamp")
|
|
if isinstance(end_time_stamp, (int, float)) and end_time_stamp > 0:
|
|
try:
|
|
return datetime.utcfromtimestamp(end_time_stamp / 1000).replace(tzinfo=None)
|
|
except Exception:
|
|
pass
|
|
|
|
# Or a string like "2025/12/12 08:00 PST" (timezone ambiguous)
|
|
end_time_str = item.get("end_time") or item.get("auction_end_time")
|
|
if isinstance(end_time_str, str) and end_time_str.strip():
|
|
raw = end_time_str.strip()
|
|
raw = raw.replace(" PST", "").replace(" PDT", "").replace(" UTC", "")
|
|
for fmt in ("%Y/%m/%d %H:%M", "%Y-%m-%d %H:%M:%S"):
|
|
try:
|
|
return datetime.strptime(raw, fmt).replace(tzinfo=None)
|
|
except Exception:
|
|
continue
|
|
return None
|
|
|
|
async def fetch_auctions(
|
|
self,
|
|
page_size: int = 100,
|
|
page_index: int = 0,
|
|
keyword: Optional[str] = None,
|
|
) -> Dict[str, Any]:
|
|
"""Fetch auctions from Dynadot hidden API."""
|
|
try:
|
|
proxy = os.getenv("SCRAPER_HTTP_PROXY") or os.getenv("SCRAPER_PROXY_URL")
|
|
async with httpx.AsyncClient(timeout=30.0, follow_redirects=True, proxy=proxy) as client:
|
|
params = {
|
|
"command": "get_list",
|
|
"aftermarket_type": "EXPIRED_AUCTION",
|
|
"page_size": page_size,
|
|
"page_index": page_index,
|
|
"lang": "en",
|
|
}
|
|
if keyword:
|
|
params["keyword"] = keyword
|
|
|
|
resp = await client.post(
|
|
f"{self.BASE_URL}{self.MARKETPLACE_API}",
|
|
params=params,
|
|
headers={
|
|
"Accept": "application/json",
|
|
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36",
|
|
"Referer": "https://www.dynadot.com/market",
|
|
},
|
|
)
|
|
|
|
if resp.status_code != 200:
|
|
return {"items": [], "total": 0, "error": f"HTTP {resp.status_code}: {resp.text[:200]}"}
|
|
|
|
data = resp.json()
|
|
listings = data.get("data", {}).get("records", []) or data.get("data", {}).get("list", [])
|
|
|
|
transformed: List[Dict[str, Any]] = []
|
|
for item in listings:
|
|
domain = item.get("domain") or item.get("name") or item.get("utf8_name") or ""
|
|
domain = str(domain).strip().lower()
|
|
if not domain or "." not in domain:
|
|
continue
|
|
|
|
end_time = self._parse_end_time(item)
|
|
if end_time is None:
|
|
# No end time -> skip (no fake)
|
|
continue
|
|
|
|
bid_price = item.get("bid_price") or item.get("current_bid") or item.get("price") or 0
|
|
if isinstance(bid_price, str):
|
|
bid_price = bid_price.replace(",", "").replace("$", "").strip()
|
|
try:
|
|
current_bid = float(bid_price)
|
|
except Exception:
|
|
continue
|
|
if current_bid <= 0:
|
|
continue
|
|
|
|
bids = item.get("bids") or item.get("bid_count") or 0
|
|
try:
|
|
num_bids = int(bids)
|
|
except Exception:
|
|
num_bids = 0
|
|
|
|
tld = domain.rsplit(".", 1)[-1].lower()
|
|
|
|
transformed.append(
|
|
{
|
|
"domain": domain,
|
|
"tld": tld,
|
|
"platform": "Dynadot",
|
|
"current_bid": current_bid,
|
|
"currency": str(item.get("bid_price_currency") or "USD").upper(),
|
|
"num_bids": num_bids,
|
|
"end_time": end_time,
|
|
"auction_url": build_affiliate_url("Dynadot", domain),
|
|
"buy_now_price": float(item.get("accepted_bid_price")) if item.get("accepted_bid_price") else None,
|
|
"age_years": int(item.get("age", 0) or 0) or None,
|
|
"backlinks": int(item.get("links", 0) or 0) or None,
|
|
"scrape_source": "dynadot:hidden_api",
|
|
}
|
|
)
|
|
|
|
return {
|
|
"items": transformed,
|
|
"total": data.get("data", {}).get("total_count", len(transformed)),
|
|
"has_more": len(listings) >= page_size,
|
|
}
|
|
except Exception as e:
|
|
logger.exception(f"Dynadot API scraper error: {e}")
|
|
return {"items": [], "total": 0, "error": str(e)}
|
|
|
|
|
|
class HiddenApiScraperService:
|
|
"""Orchestrates enabled hidden API scrapers."""
|
|
|
|
def __init__(self):
|
|
self.dynadot = DynadotApiScraper()
|
|
|
|
async def scrape_all(self, limit_per_platform: int = 100) -> Dict[str, Any]:
|
|
results: Dict[str, Any] = {"total_found": 0, "platforms": {}, "errors": [], "items": []}
|
|
|
|
try:
|
|
dynadot_data = await self.dynadot.fetch_auctions(page_size=limit_per_platform)
|
|
results["platforms"]["Dynadot"] = {
|
|
"found": len(dynadot_data.get("items", [])),
|
|
"total": dynadot_data.get("total", 0),
|
|
}
|
|
results["items"].extend(dynadot_data.get("items", []))
|
|
results["total_found"] += len(dynadot_data.get("items", []))
|
|
|
|
if dynadot_data.get("error"):
|
|
results["errors"].append(f"Dynadot: {dynadot_data['error']}")
|
|
except Exception as e:
|
|
results["errors"].append(f"Dynadot: {str(e)}")
|
|
|
|
return results
|
|
|
|
|
|
# Export instances
|
|
dynadot_scraper = DynadotApiScraper()
|
|
hidden_api_scraper = HiddenApiScraperService()
|
|
|
|
|