Spaces:
Sleeping
Sleeping
File size: 9,367 Bytes
83e8cee | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 | from __future__ import annotations
import json
import os
import re
import urllib.parse
from typing import Any, Dict, Optional
import requests
from smolagents.tools import Tool
SCRAPE_DO_SIGNUP_URL = (
"https://dashboard.scrape.do/sign-up?utm_source=huggingface&utm_medium=amazon-scraper"
)
API_BASE = "https://api.scrape.do/plugin/amazon"
# geocode -> human label, shown in the UI dropdown.
MARKETPLACES: Dict[str, str] = {
"us": "United States (amazon.com)",
"gb": "United Kingdom (amazon.co.uk)",
"de": "Germany (amazon.de)",
"fr": "France (amazon.fr)",
"it": "Italy (amazon.it)",
"es": "Spain (amazon.es)",
"ca": "Canada (amazon.ca)",
"jp": "Japan (amazon.co.jp)",
"in": "India (amazon.in)",
"au": "Australia (amazon.com.au)",
"br": "Brazil (amazon.com.br)",
"mx": "Mexico (amazon.com.mx)",
"nl": "Netherlands (amazon.nl)",
"pl": "Poland (amazon.pl)",
"se": "Sweden (amazon.se)",
"tr": "Turkey (amazon.com.tr)",
"ae": "UAE (amazon.ae)",
"sa": "Saudi Arabia (amazon.sa)",
"sg": "Singapore (amazon.sg)",
"eg": "Egypt (amazon.eg)",
"be": "Belgium (amazon.com.be)",
}
# Marketplaces where Amazon localises delivery by postal code. Everything else
# takes a country-level location instead, via countryName.
ZIP_LEVEL_MARKETS = {"us", "gb", "uk", "de", "fr", "it", "es", "ca", "jp", "in"}
COUNTRY_NAMES: Dict[str, str] = {
"tr": "Turkey",
"br": "Brazil",
"au": "Australia",
"mx": "Mexico",
"nl": "Netherlands",
"pl": "Poland",
"se": "Sweden",
"ae": "United Arab Emirates",
"sa": "Saudi Arabia",
"sg": "Singapore",
"eg": "Egypt",
"be": "Belgium",
}
_ASIN_PATTERN = re.compile(r"(?:/dp/|/gp/product/|/gp/aw/d/|/product/|[?&]asin=)([A-Z0-9]{10})", re.I)
_BARE_ASIN = re.compile(r"^[A-Z0-9]{10}$", re.I)
def extract_asin(value: str) -> Optional[str]:
"""Accept a full Amazon URL or a bare ASIN and return the ASIN."""
if not value:
return None
value = value.strip()
if _BARE_ASIN.match(value):
return value.upper()
match = _ASIN_PATTERN.search(value)
if match:
return match.group(1).upper()
return None
class ScrapeDoAmazonTool(Tool):
name = "amazon_product_scraper"
description = """
Scrape structured product data from any Amazon marketplace: product detail
pages (title, brand, price, rating, images, rankings, technical specs),
all seller offers for an ASIN, or keyword search results with organic and
sponsored positions. Returns clean JSON, no HTML parsing required.
Powered by Scrape.do's Amazon Scraper API, which handles AWS WAF, proxy
rotation, sessions and CAPTCHAs automatically across 21 marketplaces.
Use this when you need Amazon pricing, catalogue or competitive data.
"""
output_type = "string"
def __init__(self, api_token: Optional[str] = None) -> None:
# Token can be passed explicitly (e.g. by a Gradio UI) or read from the
# SCRAPE_DO_API_TOKEN environment variable for programmatic use.
self._api_token = api_token
self.inputs = {
"mode": {
"type": "string",
"description": (
"What to scrape: 'product' for a product detail page, "
"'offers' for every seller offer on an ASIN, or 'search' "
"for keyword/category search results."
),
},
"query": {
"type": "string",
"description": (
"An Amazon product URL or ASIN for 'product' and 'offers'. "
"A search keyword for 'search'."
),
},
"geocode": {
"type": "string",
"description": (
"Marketplace country code: 'us', 'gb', 'de', 'jp', 'tr'… "
"Defaults to 'us'."
),
"nullable": True,
},
"zipcode": {
"type": "string",
"description": (
"Postal code for ZIP-level marketplaces (us, gb, de, fr, it, "
"es, ca, jp, in) to localise price, stock and delivery. "
"Ignored elsewhere, where a country-level location is used."
),
"nullable": True,
},
"page": {
"type": "integer",
"description": "Search results page number. Only used when mode='search'.",
"nullable": True,
},
"language": {
"type": "string",
"description": (
"ISO 639-1 language code in uppercase, e.g. 'EN', 'DE', 'FR'. "
"Must be supported by the target marketplace."
),
"nullable": True,
},
}
super().__init__()
# ------------------------------------------------------------------ token
def _resolve_token(self) -> Optional[str]:
if self._api_token:
return self._api_token.strip()
env_token = os.getenv("SCRAPE_DO_API_TOKEN")
return env_token.strip() if env_token else None
# ---------------------------------------------------------------- request
def request(
self,
mode: str,
query: str,
geocode: str = "us",
zipcode: Optional[str] = None,
page: Optional[int] = None,
language: Optional[str] = None,
) -> Dict[str, Any]:
"""Call the Amazon Scraper API and return a parsed dict."""
token = self._resolve_token()
if not token:
return {
"status": "error",
"errorMessage": "Missing Scrape.do API token.",
"hint": f"Get a free token at {SCRAPE_DO_SIGNUP_URL}",
}
if not query or not query.strip():
return {"status": "error", "errorMessage": "No ASIN, URL or keyword provided."}
geocode = (geocode or "us").strip().lower()
params: Dict[str, Any] = {"token": token, "geocode": geocode}
if mode in ("product", "offers"):
asin = extract_asin(query)
if not asin:
return {
"status": "error",
"errorMessage": (
"Could not read an ASIN from that input. Paste a full "
"Amazon product URL (…/dp/B0BLRJ4R8F) or the 10-character "
"ASIN on its own."
),
}
params["asin"] = asin
endpoint = f"{API_BASE}/pdp" if mode == "product" else f"{API_BASE}/offer-listing"
elif mode == "search":
params["keyword"] = query.strip()
params["page"] = int(page or 1)
endpoint = f"{API_BASE}/search"
else:
return {"status": "error", "errorMessage": f"Unknown mode: {mode}"}
# Location: ZIP for ZIP-level marketplaces, country name for the rest.
if geocode in ZIP_LEVEL_MARKETS:
if zipcode and zipcode.strip():
params["zipcode"] = zipcode.strip()
elif geocode in COUNTRY_NAMES:
# Passing the marketplace's own country is a no-op, so only send it
# when the caller explicitly asked for a different delivery country.
if zipcode and zipcode.strip():
params["countryName"] = zipcode.strip()
if language and language.strip():
params["language"] = language.strip().upper()
url = endpoint + "?" + urllib.parse.urlencode(params, quote_via=urllib.parse.quote)
try:
response = requests.get(url, timeout=90)
except requests.exceptions.RequestException as exc:
return {"status": "error", "errorMessage": f"Request failed: {exc}"}
try:
data = response.json()
except ValueError:
return {
"status": "error",
"errorMessage": f"Unexpected non-JSON response (HTTP {response.status_code}).",
"details": response.text[:500],
}
if response.status_code == 401 or "token" in str(data.get("error", "")).lower():
data.setdefault(
"hint", f"Check your token or create one at {SCRAPE_DO_SIGNUP_URL}"
)
if isinstance(data, dict) and "status" not in data:
data["status"] = "success" if response.ok else "error"
if isinstance(data, dict) and not response.ok and not data.get("errorMessage"):
data["errorMessage"] = data.get("message") or f"HTTP {response.status_code}"
return data
# ---------------------------------------------------------------- forward
def forward(
self,
mode: str,
query: str,
geocode: Optional[str] = None,
zipcode: Optional[str] = None,
page: Optional[int] = None,
language: Optional[str] = None,
) -> str:
result = self.request(
mode=mode,
query=query,
geocode=geocode or "us",
zipcode=zipcode,
page=page,
language=language,
)
return json.dumps(result, ensure_ascii=False, indent=2)
|