keyword_ranking_crawler/scrapper.py

import os
import pickle
from selenium import webdriver
from selenium.webdriver.chrome.options import Options
from selenium.webdriver.common.by import By
from selenium.webdriver.chrome.service import Service
from webdriver_manager.chrome import ChromeDriverManager

from time import sleep
import json
import time
import re


with open("marketplaces.json", "r", encoding="utf-8") as f:
    data = json.load(f)

with open("keywords.json", "r", encoding="utf-8") as f:
    keywords = json.load(f)

with open("cookies.json", "r", encoding="utf-8") as f:
    cookies_ref = json.load(f)

# Or if it's a Python dict already:
marketplaces = data["marketplaces"]

#BASE_PATH= '/mnt/AmazonReports/Amazon/keyword_ranking'
BASE_PATH= 'data'
MAX_PAGE = 10



def get_driver():
    options = Options()
    #options.add_argument("--headless")
    options.add_argument("--disable-blink-features=AutomationControlled")  # Removes automation flag
    options.add_experimental_option("excludeSwitches", ["enable-automation"])
    options.add_experimental_option('useAutomationExtension', False)
    options.add_argument("user-agent=Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/138.0.0.0 Safari/537.36")
    options.add_argument("--start-maximized")
    options.add_argument("user-data-dir=/home/ec2-user/keyword_ranking_crawler/chrome_path")
    options.add_argument("profile-directory=Default")
    driver = webdriver.Chrome(service=Service(ChromeDriverManager().install()), options=options)
    # Remove navigator.webdriver
    driver.execute_cdp_cmd("Page.addScriptToEvaluateOnNewDocument", {
    "source": """
        Object.defineProperty(navigator, 'webdriver', {
            get: () => undefined
        })
    """
})
    return driver

def save_cookies(driver, path):
    with open(path, "wb") as f:
        pickle.dump(driver.get_cookies(), f)

def parse_number_with_commas(str):
    try:
        return int(str.replace(",", ""))
    except:
        return 0

def parse_rating(ratingHtml,prefix):
    try :
        rating_str = ratingHtml.get_attribute("innerHTML")
        return float(rating_str.split(prefix)[0])
    except:
        return ""


def parse_bought_count(text):
    if not text:
        return 0
    # Extract numeric part using regex
    match = re.search(r'([\d,.]+)([KM]?)\+?', text.upper())
    if not match:
        return 0

    number, unit = match.groups()
    number = float(number.replace(',', ''))

    if unit == 'K':
        return int(number * 1_000)
    elif unit == 'M':
        return int(number * 1_000_000)
    else:
        return int(number)

def save_ranking(rankings, file_path):
    if len(rankings) > 0 :
        with open(file_path, "w", encoding="utf-8") as f:
            json.dump(rankings, f, ensure_ascii=False, indent=4)

def load_cookies(driver, path):
    with open(path, "rb") as f:
        cookies = pickle.load(f)
        for cookie in cookies:
            if 'sameSite' in cookie:
                cookie.pop('sameSite')  # Optional fix if Chrome complains
            driver.add_cookie(cookie)


def check_sponsored(item):
    try:
        # Check if any element inside contains the exact text "Sponsored"
        sponsored_labels = item.find_elements(By.XPATH, './/*[contains(text(), "Sponsored")]')
        for label in sponsored_labels:
            if label.text.strip().lower() == "sponsored":
                return 1
    except:
        return 0

def check_consist_utopia( title ):
    return 1 if "Utopia" in title else 0 



def get_amazon_ranks(url, marketplace, ratingPrefix, keyword, page, count):
    print( '[INFO] Getting Amazon Ranks for: ', marketplace, keyword, page)
    url = f"https://www.{url}/s?k={keyword.replace(' ', '+')}&page={page}"
    driver.get(url)   
    ranks = []
    COOKIE_FILE = f"{cookies_ref[marketplace]['cookies_name']}";
    # Load cookies if available
    if os.path.exists(COOKIE_FILE):
        load_cookies(driver, COOKIE_FILE)
        driver.get(url)
    else:
        print("No cookie file found, visiting fresh")
        driver.get(url)
        sleep(5)  # Give time to solve CAPTCHA manually (if needed)
        save_cookies(driver, COOKIE_FILE)

    sleep(3)  # Wait for JS to load   
    items = driver.find_elements(By.XPATH, '//div[contains(@class,"s-result-item") and @data-asin]')
    for idx, item in enumerate(items, start=1):
        asin = item.get_attribute("data-asin")
        try:
            sponsored = check_sponsored(item)
            title = item.find_element(By.XPATH, './/h2//span').text
            rating = item.find_element(By.XPATH, './/span[@class="a-icon-alt"]')
            reviews_count = item.find_element(By.XPATH, './/span[@class="a-size-base s-underline-text"]').text
            last_month_bought = item.find_element(By.XPATH, './/span[contains(@class, "a-size-base a-color-secondary") and contains(text(), "bought")]').text    
            if title == 'Results':
                continue
            if sponsored == None :
                ranks.append({
                    'rank' : count, 
                    'title' : title, 
                    'marketplace' : marketplace, 
                    'keyword': keyword, 
                    'sponsored' : 0, 
                    'asin' : asin, 
                    'is_utopia' : check_consist_utopia(title),
                    'url' : url,
                    'rating' : parse_rating(rating,ratingPrefix),
                    'reviews_count' : parse_number_with_commas (reviews_count ),
                    'last_month_bought': parse_bought_count (last_month_bought)
                })
                count += 1
        except:
            continue

    file_path = f"{BASE_PATH}/{int(time.time() * 1000)}-{marketplace}-{keyword}.json"
    save_ranking(ranks, file_path )
    if( len(ranks) == 0 ):
        return -1
    return count
   

driver = get_driver()

for keyword in keywords:
    for marketplace, details in marketplaces.items():
        if marketplace == 'AMAZON_USA':
            url = details['url']
            ratingPrefix = details['ratingPrefix']
            count =1
            for i in range(1, MAX_PAGE):
                count = get_amazon_ranks(url, marketplace, ratingPrefix, keyword, i, count)
                if count == -1:
                    break
            sleep(3)
driver.quit()