"""
Amazon Reviews 基础采集器 — 仅供原理演示
警告：此脚本不适合生产环境，详见下方缺陷分析
依赖: pip install requests beautifulsoup4 lxml
"""

import requests
from bs4 import BeautifulSoup
import time
import random
import json
from dataclasses import dataclass, asdict
from typing import Optional


@dataclass
class Review:
    asin: str
    review_id: str
    title: str
    rating: float
    body: str
    verified: bool
    date: str
    helpful_count: int
    reviewer_name: str
    reviewer_id: Optional[str] = None


class AmazonReviewScraper:
    """
    基础亚马逊评论爬虫
    注意：此实现仅能抓取产品详情页上的"精选评论"（通常 8 条）
    完整评论列表已被亚马逊移入登录墙，此脚本无法访问
    """

    BASE_URL = "https://www.amazon.com/dp/{asin}"

    # 模拟真实浏览器 Headers — 这是能否绕过初步检测的关键
    HEADERS = {
        "User-Agent": (
            "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) "
            "AppleWebKit/537.36 (KHTML, like Gecko) "
            "Chrome/125.0.0.0 Safari/537.36"
        ),
        "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8",
        "Accept-Language": "en-US,en;q=0.5",
        "Accept-Encoding": "gzip, deflate, br",
        "Connection": "keep-alive",
        "Upgrade-Insecure-Requests": "1",
        # 缺少 sec-ch-ua 等现代 Client Hints，TLS 指纹暴露
    }

    def __init__(self, delay_range=(3, 8)):
        self.session = requests.Session()
        self.session.headers.update(self.HEADERS)
        self.delay_range = delay_range

    def _get_page(self, url: str) -> Optional[BeautifulSoup]:
        """获取页面，附加随机延迟"""
        try:
            # 随机延迟模拟人类行为，但频率和节奏仍暴露自动化特征
            time.sleep(random.uniform(*self.delay_range))
            response = self.session.get(url, timeout=15)

            # 检测是否被重定向至登录页或 CAPTCHA
            if "ap/signin" in response.url:
                print(f"[警告] 被重定向至登录页：{response.url}")
                return None
            if response.status_code == 503:
                print(f"[警告] 收到 503，可能触发了 Bot 检测")
                return None
            if response.status_code != 200:
                print(f"[错误] HTTP {response.status_code}: {url}")
                return None

            return BeautifulSoup(response.text, "lxml")

        except requests.exceptions.Timeout:
            print(f"[错误] 请求超时: {url}")
            return None
        except requests.exceptions.ConnectionError as e:
            print(f"[错误] 连接失败: {e}")
            return None

    def scrape_featured_reviews(self, asin: str) -> list:
        """
        仅能抓取产品详情页上的精选评论（8 条）
        完整评论列表（/product-reviews/ 端点）需要登录，此方法无法访问
        """
        url = self.BASE_URL.format(asin=asin)
        soup = self._get_page(url)

        if not soup:
            return []

        reviews = []
        # 亚马逊频繁变更 HTML 结构，以下选择器可能随时失效
        review_divs = soup.select("[data-hook='review']")

        if not review_divs:
            review_divs = soup.select(".review")

        for div in review_divs:
            try:
                review = self._parse_review_div(asin, div)
                if review:
                    reviews.append(review)
            except Exception as e:
                print(f"[解析错误] ASIN {asin}: {e}")
                continue

        return reviews

    def _parse_review_div(self, asin: str, div) -> Optional[Review]:
        """解析单条评论 HTML"""

        review_id_elem = div.get("id", "")
        if not review_id_elem:
            return None

        # 评分：提取星级文本 "4.0 out of 5 stars"
        rating_elem = div.select_one("[data-hook='review-star-rating'] span")
        if not rating_elem:
            rating_elem = div.select_one(".review-rating span")
        rating_text = rating_elem.get_text() if rating_elem else "0"
        try:
            rating = float(rating_text.split()[0])
        except (ValueError, IndexError):
            rating = 0.0

        title_elem = div.select_one("[data-hook='review-title'] span:last-child")
        title = title_elem.get_text(strip=True) if title_elem else ""

        body_elem = div.select_one("[data-hook='review-body'] span")
        body = body_elem.get_text(strip=True) if body_elem else ""

        date_elem = div.select_one("[data-hook='review-date']")
        date = date_elem.get_text(strip=True) if date_elem else ""

        verified_elem = div.select_one("[data-hook='avp-badge']")
        verified = verified_elem is not None

        helpful_elem = div.select_one("[data-hook='helpful-vote-statement']")
        helpful_text = helpful_elem.get_text(strip=True) if helpful_elem else "0"
        try:
            helpful_count = int(helpful_text.split()[0].replace(",", ""))
        except (ValueError, IndexError):
            helpful_count = 0

        reviewer_elem = div.select_one(".a-profile-name")
        reviewer_name = reviewer_elem.get_text(strip=True) if reviewer_elem else "Unknown"

        return Review(
            asin=asin,
            review_id=review_id_elem,
            title=title,
            rating=rating,
            body=body,
            verified=verified,
            date=date,
            helpful_count=helpful_count,
            reviewer_name=reviewer_name,
        )

    def scrape_batch(self, asin_list: list) -> list:
        """批量抓取，返回 JSON 友好的字典列表"""
        all_reviews = []
        for i, asin in enumerate(asin_list):
            print(f"[{i+1}/{len(asin_list)}] 抓取 ASIN: {asin}")
            reviews = self.scrape_featured_reviews(asin)
            all_reviews.extend([asdict(r) for r in reviews])
            print(f"  获取 {len(reviews)} 条精选评论")
        return all_reviews


# ===== 使用示例 =====
if __name__ == "__main__":
    test_asins = ["B08N5WRWNW", "B07XJ8C8F5"]

    scraper = AmazonReviewScraper(delay_range=(4, 10))
    results = scraper.scrape_batch(test_asins)

    print(json.dumps(results[:2], indent=2, ensure_ascii=False))
    print(f"\n总计采集: {len(results)} 条评论")

"""
Pangolinfo Amazon Review API 接入示例
获取完整评论数据，包含 Customer Says、评分分布等字段
文档：https://docs.pangolinfo.com/en-api-reference/
"""

import requests
import json
from typing import Optional


class PangolReviewClient:
    """Pangolinfo Amazon Review API 客户端"""

    API_ENDPOINT = "https://api.pangolinfo.com/v1/amazon/reviews"

    def __init__(self, api_key: str):
        self.api_key = api_key
        self.session = requests.Session()
        self.session.headers.update({
            "Authorization": f"Bearer {api_key}",
            "Content-Type": "application/json"
        })

    def get_reviews(
        self,
        asin: str,
        marketplace: str = "US",
        page: int = 1,
        sort_by: str = "recent",
        star_filter: Optional[int] = None,
        zip_code: Optional[str] = None
    ) -> dict:
        """
        获取指定 ASIN 的评论数据
        返回完整 JSON，含 Customer Says、评分分布、分页信息
        """
        payload = {
            "asin": asin,
            "marketplace": marketplace,
            "page": page,
            "sort_by": sort_by,
        }
        if star_filter:
            payload["star_filter"] = star_filter
        if zip_code:
            payload["zip_code"] = zip_code

        response = self.session.post(self.API_ENDPOINT, json=payload, timeout=30)
        response.raise_for_status()
        return response.json()

    def get_all_reviews(self, asin: str, marketplace: str = "US", max_pages: int = 10) -> list:
        """分页抓取所有评论"""
        all_reviews = []
        for page in range(1, max_pages + 1):
            data = self.get_reviews(asin, marketplace, page=page)
            reviews = data.get("reviews", [])
            if not reviews:
                break
            all_reviews.extend(reviews)
            total_pages = data.get("pagination", {}).get("total_pages", 1)
            print(f"  ASIN {asin} — 第 {page}/{total_pages} 页，已获取 {len(reviews)} 条")
            if page >= total_pages:
                break
        return all_reviews

    def get_customer_says(self, asin: str, marketplace: str = "US") -> Optional[str]:
        """专门抓取 Customer Says（AI 摘要）——大多数 DIY 爬虫无法获取的字段"""
        data = self.get_reviews(asin, marketplace)
        return data.get("customer_says")

    def get_rating_distribution(self, asin: str, marketplace: str = "US") -> dict:
        """获取各星级评分分布"""
        data = self.get_reviews(asin, marketplace)
        return data.get("rating_distribution", {})


# ===== 使用示例 =====
if __name__ == "__main__":
    client = PangolReviewClient(api_key="YOUR_API_KEY")
    asin = "B08N5WRWNW"

    # 1. 获取完整评论（自动分页）
    reviews = client.get_all_reviews(asin, marketplace="US", max_pages=5)
    print(f"共获取 {len(reviews)} 条评论")

    # 2. 查看 Customer Says（AI 摘要）
    customer_says = client.get_customer_says(asin)
    print(f"Customer Says:\n{customer_says}")

    # 3. 查看评分分布
    distribution = client.get_rating_distribution(asin)
    for star, count in sorted(distribution.items(), reverse=True):
        print(f"  {star}星: {count:,} 条")

    # 4. 专门抓取差评（1–2 星），用于产品痛点分析
    negative_reviews = []
    for star in [1, 2]:
        page_data = client.get_reviews(asin, star_filter=star)
        negative_reviews.extend(page_data.get("reviews", []))
    print(f"获取到 {len(negative_reviews)} 条差评")