flowtask 5.8.4__cp312-cp312-manylinux2014_x86_64.manylinux_2_17_x86_64.whl

flowtask/components/Amazon.py

@@ -0,0 +1,523 @@
+"""
+Scrapping a Web Page Using Selenium + ChromeDriver + BeautifulSoup.
+
+
+        Example:
+
+        ```yaml
+        Amazon:
+          type: product_info
+          use_proxies: true
+          paid_proxy: true
+        ```
+
+    """
+from typing import Any
+import asyncio
+from collections.abc import Callable
+import re
+from urllib.parse import urljoin
+from bs4 import BeautifulSoup
+import random
+import httpx
+import pandas as pd
+import backoff
+# Internals
+from ..exceptions import (
+    ComponentError,
+    ConfigError,
+    NotSupported,
+    DataNotFound,
+    DataError
+)
+from ..interfaces.http import ua
+from .reviewscrap import ReviewScrapper, on_backoff, bad_gateway_exception
+
+
+class Amazon(ReviewScrapper):
+    """Amazon.
+
+    Combining API Key and Web Scrapping, this component will be able to extract
+    Amazon Product Information (reviews, etc).
+    """
+    def __init__(
+        self,
+        loop: asyncio.AbstractEventLoop = None,
+        job: Callable = None,
+        stat: Callable = None,
+        **kwargs,
+    ):
+        super(Amazon, self).__init__(
+            loop=loop,
+            job=job,
+            stat=stat,
+            **kwargs
+        )
+        # Always use proxies:
+        self.use_proxy: bool = True
+        self._free_proxy: bool = False
+        self.cookies = {
+            # "aws-session-id": "241-9979986-0092756",
+        }
+        self.headers: dict = {
+            'authority': 'www.amazon.com',
+            'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,*/*;q=0.8',
+            "Accept-Encoding": "gzip, deflate, br, zstd",
+            "Accept-Language": "es-US,es;q=0.9,en-US;q=0.8,en;q=0.7,es-419;q=0.6",
+            "Origin": "https://www.amazon.com",
+            "Referer": "https://www.amazon.com/dp/",
+            "Sec-CH-UA": '"Not A(Brand";v="8", "Chromium";v="132", "Google Chrome";v="132"',
+            "Sec-CH-UA-Mobile": "?0",
+            "Sec-CH-UA-Platform": '"Linux"',
+            'sec-fetch-site': 'none',
+            'sec-fetch-mode': 'navigate',
+            'sec-fetch-dest': 'document',
+            "Sec-Fetch-Site": "same-origin",
+            "User-Agent": random.choice(ua),
+            "Connection": "keep-alive",
+            'dnt': '1',
+            'upgrade-insecure-requests': '1',
+        }
+        self.semaphore = asyncio.Semaphore(10)
+
+    def _extract_reviews_from_page(self, soup: BeautifulSoup) -> list:
+        """
+        Given a BeautifulSoup-parsed Amazon reviews page, extract individual reviews.
+        Returns a list of dictionaries.
+        """
+        reviews = []
+        # Reviews are contained within the element with id 'cm-cr-review_list'
+        reviews_container = soup.find(
+            "ul", id="cm-cr-review_list"
+        ) or soup.find("div", {"data-hook": "reviews-medley-widget"})
+        if reviews_container:
+            # Each review is typically in a <li> element with data-hook "review"
+            for review_el in reviews_container.find_all("li", {"data-hook": "review"}):
+                try:
+                    # Extract review title
+                    title_el = review_el.select_one("[data-hook=review-title] > span")
+                    title = title_el.get_text(strip=True) if title_el else None
+                    # Extract review body text
+                    body_el = review_el.select_one("[data-hook=review-body]")
+                    body = " ".join(body_el.stripped_strings) if body_el else None
+                    # Extract review date and/or location/date information
+                    date_el = review_el.select_one("[data-hook=review-date]")
+                    date_text = date_el.get_text(strip=True) if date_el else None
+                    # Extract rating (look for an element with data-hook containing 'review-star-rating')
+                    if rating_el := review_el.select_one("[data-hook*='review-star-rating'] span.a-icon-alt"):
+                        # Extract numeric rating (first match of digits possibly with a decimal)
+                        import re
+                        rating_match = re.search(r"(\d+\.?\d*) out", rating_el.get_text(strip=True))
+                        rating = rating_match.group(1) if rating_match else None
+                    else:
+                        rating = None
+                    # Extract Verified Purchase badge (if exists)
+                    verified = bool(review_el.select_one("[data-hook=avp-badge]"))
+
+                    review_dict = {
+                        "title": title,
+                        "review": body,
+                        "location_and_date": date_text,
+                        "rating": rating,
+                        "verified": verified
+                    }
+                    reviews.append(review_dict)
+                except Exception as e:
+                    # Log exception for this review, but continue extracting others
+                    self._logger.error(
+                        f"Failed to parse a review: {e}"
+                    )
+        return reviews
+
+    def _extract_next_page_url(self, soup: BeautifulSoup, base_url: str) -> str:
+        """
+        Look for a 'Next' page link in the pagination (typically via the CSS selector
+        '.a-pagination .a-last > a').
+        Returns an absolute URL string if found, otherwise returns None.
+        """
+        pagination_el = soup.select_one(".a-pagination .a-last > a")
+        next_page_relative = pagination_el.get("href") if pagination_el else None
+        return urljoin(base_url, next_page_relative) if next_page_relative else None
+
+    @backoff.on_exception(
+        backoff.expo,
+        (httpx.TimeoutException, httpx.ConnectTimeout, httpx.HTTPStatusError, httpx.HTTPError),
+        max_tries=3,
+        jitter=backoff.full_jitter,
+        on_backoff=on_backoff,
+        giveup=lambda e: not bad_gateway_exception(e) and not isinstance(e, httpx.ConnectTimeout)
+    )
+    async def _fetch_product_page(self, asin: str, cookies: httpx.Cookies, for_reviews: bool = False) -> tuple:
+        product_page_url = f"https://www.amazon.com/dp/{asin}"
+        response = await self._get(url=product_page_url, cookies=cookies, headers=self.headers)
+        if response.status_code != 200:
+            raise DataError(
+                f"Failed to fetch product page, status code: {response.status_code}"
+            )
+        html = response.text
+        soup = BeautifulSoup(html, "html.parser")
+        if for_reviews:
+            if medley := soup.find("div", id="reviewsMedley"):
+                return product_page_url, html, soup
+        elif title_div := soup.find("div", id="title_feature_div"):
+            product_name = title_div.get_text(separator=" ", strip=True)
+            self._logger.info(f"Extracted product name: {product_name} from {product_page_url}")
+            return product_page_url, html, soup
+        else:
+            await asyncio.sleep(1.5)
+            raise httpx.HTTPError(
+                f"Failed to find product name on product page: {product_page_url}"
+            )
+
+    def _extract_reviews_from_product_page(self, url: str, row: Any, soup: BeautifulSoup) -> list:
+        """Extract review snippet(s) from the product page (fallback)."""
+        reviews = []
+        if medley := soup.find("div", id="reviewsMedley"):
+            for li in medley.find_all("li", {"data-hook": "review"}):
+                try:
+                    profile_user = li.find("div", {"class": "a-profile-content"})
+                    profile_name = profile_user.find("span", {"class": "a-profile-name"}).get_text(strip=True)
+                    customer_reviews = ""
+                    title_text = ""
+                    if title := li.find("a", {"data-hook": "review-title"}):
+                        customer_reviews = title["href"]
+                        title_text = title.find_all("span")[-1].text.strip()
+
+                    body = li.select_one("[data-hook=review-body]")
+                    body_text = " ".join(body.stripped_strings) if body else None
+
+                    date_el = li.select_one("[data-hook=review-date]")
+                    date_text = date_el.get_text(strip=True) if date_el else None
+
+                    rating_el = li.select_one("[data-hook*='review-star-rating'] span.a-icon-alt")
+                    rating_match = re.search(r"(\d+\.?\d*) out", rating_el.get_text(strip=True)) if rating_el else None
+                    rating = rating_match.group(1) if rating_match else None
+
+                    verified = bool(li.select_one("[data-hook=avp-badge]"))
+                    _data = row.to_dict()
+                    review_dict = {
+                        "url": url,
+                        "user": profile_name,
+                        "customer_reviews": customer_reviews,
+                        "title": title_text,
+                        "review": body_text,
+                        "location_and_date": date_text,
+                        "rating": rating,
+                        "verified": verified,
+                        **_data
+                    }
+                    reviews.append(review_dict)
+                except Exception as e:
+                    self._logger.error(f"Error parsing a fallback review: {e}")
+        return reviews
+
+    async def _fetch_review_page(self, url: str, cookies: httpx.Cookies) -> str:
+        """
+        Fetches the review page HTML for a given URL.
+        Returns the HTML text.
+        """
+        try:
+            response = await self._get(url=url, cookies=cookies, headers=self.headers)
+            if response.status_code != 200:
+                raise DataError(f"Failed to fetch reviews page (status code: {response.status_code})")
+            return response.text
+        except Exception as e:
+            raise DataError(f"Failed to fetch reviews page: {e}") from e
+
+    async def _product_reviews(self, idx, row, cookies, max_pages: int = 5) -> list:
+        async with self.semaphore:
+            # Prepare payload for the API request
+            asin = row['asin']
+            reviews = []
+            # base_review_url = f"https://www.amazon.com/product-reviews/{asin}/"
+            #
+            # try:
+            #     # Try fetching the reviews page
+            #     html = await self._fetch_review_page(base_review_url, cookies)
+            #     soup = BeautifulSoup(html, "html.parser")
+            #     reviews.extend(self._extract_reviews_from_page(soup))
+            #     self._logger.info(f"Fetched reviews from reviews URL for ASIN {asin}")
+            # except DataError as e:
+            #     # If a redirect (or other error) is detected, log and fall back to the product page.
+            #     self._logger.warning(
+            #         f"Direct reviews page fetch failed ({e}); falling back to product page for ASIN {asin}"
+            #     )
+            try:
+                url, _, soup = await self._fetch_product_page(asin, cookies=cookies, for_reviews=True)
+                reviews.extend(
+                    self._extract_reviews_from_product_page(url, row, soup)
+                )
+            except Exception as ee:
+                self._logger.error(
+                    f"Fallback product page review extraction failed: {ee}"
+                )
+                return []
+            self._logger.info(
+                f"Fetched {len(reviews)} reviews for ASIN {asin}."
+            )
+            await asyncio.sleep(random.randint(3, 5))
+            return reviews
+
+    async def reviews(self):
+        """reviews.
+
+        Target Product Reviews.
+        """
+        httpx_cookies = httpx.Cookies()
+        for key, value in self.cookies.items():
+            httpx_cookies.set(
+                key, value,
+                domain='.amazon.com',
+                path='/'
+            )
+
+        # Iterate over each row in the DataFrame
+        print('starting ...')
+        tasks = [
+            self._product_reviews(
+                idx,
+                row,
+                httpx_cookies,
+                max_pages=2
+            ) for idx, row in self.data.iterrows()
+        ]
+        # Gather results concurrently
+        all_reviews_nested = await self._processing_tasks(tasks)
+        # Flatten the nested list: one item per review, and add the asin as reference.
+        reviews_flat = []
+        for idx, review_list in enumerate(all_reviews_nested):
+            asin = self.data.iloc[idx]['asin']
+            for review in review_list:
+                review['asin'] = asin
+                reviews_flat.append(review)
+
+        reviews_df = pd.DataFrame(reviews_flat)
+        self._logger.notice(f"Extracted total {len(reviews_df)} reviews.")
+
+        # at the end, adding a column for origin of reviews:
+        reviews_df['origin'] = 'amazon'
+        self.data = reviews_df  # or store separately
+        return self.data
+
+    def _extract_product_name(self, soup: BeautifulSoup) -> str:
+        if title_div := soup.find("div", id="title_feature_div"):
+            return title_div.get_text(separator=" ", strip=True)
+        return None
+
+    def _extract_price(self, soup: BeautifulSoup) -> str:
+        price_element = soup.select_one("span.a-offscreen")
+        return price_element.get_text(strip=True) if price_element else None
+
+    def _extract_product_description(self, soup: BeautifulSoup) -> str:
+        if desc_div := soup.find("div", id="productDescription_feature_div"):
+            # Sometimes there is an inner div with id="productDescription"
+            if desc_inner := desc_div.find("div", id="productDescription"):
+                # Join all paragraph texts into one string
+                paragraphs = [p.get_text(separator=" ", strip=True) for p in desc_inner.find_all("p")]
+                product_description = "\n".join([p for p in paragraphs if p])
+            else:
+                product_description = desc_div.get_text(separator=" ", strip=True)
+            return product_description
+        return None
+
+    def _extract_rating(self, soup: BeautifulSoup) -> tuple:
+        """
+        Extract the average rating and review count from an Amazon product page.
+
+        This function parses the BeautifulSoup object to find and extract the average
+        customer rating and the total number of reviews for a product.
+
+        Args:
+            soup (BeautifulSoup): A BeautifulSoup object representing the parsed HTML
+                                    of an Amazon product page.
+
+        Returns:
+            tuple: A tuple containing two elements:
+                - review_rating (str or None): The average rating of the product
+                    (e.g., "4.5 out of 5 stars"), or None if not found.
+                - review_count (str or None): The total number of reviews for the
+                    product (e.g., "1,234"), or None if not found.
+        """
+        review_rating = None
+        review_count = None
+        if acr_div := soup.find("div", id="averageCustomerReviews_feature_div"):
+            # The star rating is contained inside a span within the "acrPopover"
+            if acr_popover := acr_div.find("span", id="acrPopover"):
+                if rating_span := acr_popover.find("span", class_="a-color-base"):
+                    review_rating = rating_span.get_text(strip=True)
+            # The review count is extracted from the anchor "acrCustomerReviewLink"
+            if review_link := acr_div.find("a", id="acrCustomerReviewLink"):
+                if count_span := review_link.find("span", id="acrCustomerReviewText"):
+                    review_count = count_span.get_text(strip=True).replace('ratings', '').strip()
+            return review_rating, review_count
+        return None, None
+
+    def _extract_product_overview(self, soup: BeautifulSoup) -> dict:
+        overview = {}
+        # Check if the overview container is present
+        if overview_container := soup.find("div", id="productOverview_hoc_view_div"):
+            # Iterate over each row in the container. Each row is typically a div with class "a-row"
+            for row in overview_container.find_all("div", class_="a-row"):
+                # Amazon structure: each row contains at least 2 columns.
+                columns = row.find_all("div", class_="a-column")
+                if len(columns) >= 2:
+                    # The first column typically contains the label (e.g., "Screen Size")
+                    label = columns[0].get_text(separator=" ", strip=True)
+                    # The second column typically contains the value (e.g., "86 Inches")
+                    value = columns[1].get_text(separator=" ", strip=True)
+                    if label and value:
+                        overview[label] = value
+        elif overview_div := soup.find("div", id="productOverview_feature_div"):
+            if table := overview_div.find("table", {"class": "a-spacing-micro"}):
+                for row in table.find_all("tr"):
+                    th = row.find("th")
+                    td = row.find("td")
+                    if th and td:
+                        key = th.get_text(separator=" ", strip=True)
+                        value = td.get_text(separator=" ", strip=True)
+                        overview[key] = value
+        return overview
+
+    def _extract_product_details(self, soup: BeautifulSoup) -> tuple:
+        # Extract technical specifications from "productDetails_techSpec_section_1"
+        tech_details = {}
+        if details_table := soup.find("table", id="productDetails_techSpec_section_1"):
+            for tr in details_table.find_all("tr"):
+                th = tr.find("th")
+                td = tr.find("td")
+                if th and td:
+                    key = th.get_text(separator=" ", strip=True)
+                    value = td.get_text(separator=" ", strip=True)
+                    tech_details[key] = value
+
+        # Extract additional product details from "productDetails_detailBullets_sections1"
+        additional_details = {}
+        if additional_table := soup.find("table", id="productDetails_detailBullets_sections1"):
+            for tr in additional_table.find_all("tr"):
+                th = tr.find("th")
+                td = tr.find("td")
+                if th and td:
+                    key = th.get_text(separator=" ", strip=True)
+                    value = td.get_text(separator=" ", strip=True)
+                    additional_details[key] = value
+
+        return tech_details, additional_details
+
+    def _extract_product_info(self, url: str, row: Any, soup: BeautifulSoup) -> dict:
+        """
+        Extract product information from the Amazon product page.
+        Returns a dictionary with:
+            - productName: from 'title_feature_div'
+            - overview: (e.g., screen size, brand, display technology, resolution, refresh rate)
+            - reviewRating: from 'acrPopover' (if available)
+            - reviewCount: from 'acrCustomerReviewText' (if available)
+            - technicalDetails: from table "productDetails_techSpec_section_1"
+            - additionalDetails: from table "productDetails_detailBullets_sections1"
+        """
+        # Extract product information here
+        # Return a dictionary with relevant fields
+        # Extract product name from "title_feature_div"
+        # --- Product Name ---
+        product_name = self._extract_product_name(soup)
+        prince = self._extract_price(soup)
+
+        # Extract review rating and count from "averageCustomerReviews"
+        review_rating, review_count = self._extract_rating(soup)
+
+        # --- Overview (revised for dynamic content) ---
+        overview = self._extract_product_overview(soup)
+
+        # --- Technical Details ---
+        tech_details, additional_details = self._extract_product_details(soup)
+
+        # --- Product Description ---
+        product_description = self._extract_product_description(soup)
+
+        # --- About This Item (feature bullets) ---
+        about_this_item = []
+        if featurebullets_div := soup.find("div", id="featurebullets_feature_div"):
+            if ul := featurebullets_div.find("ul", {"class": "a-unordered-list"}):
+                # Extract each bullet text and add to list
+                for li in ul.find_all("li"):
+                    if text := li.get_text(separator=" ", strip=True):
+                        about_this_item.append(text)
+        _data = row.to_dict()
+        return {
+            "product_name": product_name,
+            "price": prince,
+            "url": url,
+            "about_this_item": about_this_item,
+            "rating": review_rating,
+            "review_count": review_count,
+            "overview": overview,
+            "description": product_description,
+            "tech_details": tech_details,
+            "additional_details": additional_details,
+            **_data
+        }
+
+    async def _product_information(self, idx, row, cookies):
+        async with self.semaphore:
+            # Prepare payload for the API request
+            asin = row['asin']
+            try:
+                # Fetch the product page
+                url, html, soup = await self._fetch_product_page(asin, cookies=cookies, for_reviews=False)
+                if not html:
+                    self._logger.warning(
+                        f"No Product Information found for {asin}."
+                    )
+                    return {}
+            except (httpx.TimeoutException, httpx.HTTPError) as ex:
+                self._logger.warning(f"Request failed: {ex}")
+                return []
+            except Exception as ex:
+                self._logger.error(f"An error occurred: {ex}")
+                return []
+
+            # Extract the product information using BeautifulSoup
+            if product_info := self._extract_product_info(url, row, soup):
+                return product_info
+            raise DataNotFound(
+                f"Failed to extract product information for {asin}"
+            )
+
+    async def product_info(self):
+        """product_info.
+
+        Product Information.
+        """
+        httpx_cookies = httpx.Cookies()
+        for key, value in self.cookies.items():
+            httpx_cookies.set(
+                key, value,
+                domain='.amazon.com',
+                path='/'
+            )
+
+        # Iterate over each row in the DataFrame
+        print('starting ...')
+
+        tasks = [
+            self._product_information(
+                idx,
+                row,
+                httpx_cookies
+            ) for idx, row in self.data.iterrows()
+        ]
+        # Gather results concurrently
+        all_products = await self._processing_tasks(tasks)
+
+        # Convert to DataFrame
+        df = pd.DataFrame(all_products)
+
+        # show the num of rows in final dataframe:
+        self._logger.notice(
+            f"Ending Product Info: {len(df)}"
+        )
+
+        # Override previous dataframe:
+        self.data = df
+
+        # return existing data
+        return self.data