Michael/Priceflow
1
0 Fork 0
mirror of https://github.com/R0m1k3/Priceflow.git synced 2026-10-12 01:39:25 +02:00
Code Issues Packages Projects Releases Wiki Activity
Files
f20bae52bdcecbda3dce8eefe3d23837cccfca02
Priceflow/app/core/search_config.py
T

231 lines
9.2 KiB
Python
Raw Blame History

"""
Search Configuration
Centralizes all configuration for the search system:
- Site definitions (selectors, URLs)
- Proxy configurations
- User Agents
"""
import os
import random
# === BROWSERLESS CONFIGURATION ===
BROWSERLESS_URL = os.getenv("BROWSERLESS_URL", "ws://browserless:3000")
DEBUG_DUMPS_DIR = "/app/debug_dumps"
# === PROXY CONFIGURATION ===
# Format from user: ip:port:username:password
AMAZON_PROXY_LIST_RAW = [
"142.111.48.253:7030:jasuwwjr:elbsx170nmnl",
"31.59.20.176:6754:jasuwwjr:elbsx170nmnl",
"23.95.150.145:6114:jasuwwjr:elbsx170nmnl",
"198.23.239.134:6540:jasuwwjr:elbsx170nmnl",
"107.172.163.27:6543:jasuwwjr:elbsx170nmnl",
"198.105.121.200:6462:jasuwwjr:elbsx170nmnl",
"64.137.96.74:6641:jasuwwjr:elbsx170nmnl",
"84.247.60.125:6095:jasuwwjr:elbsx170nmnl",
"216.10.27.159:6837:jasuwwjr:elbsx170nmnl",
"142.111.67.146:5611:jasuwwjr:elbsx170nmnl",
]
def get_amazon_proxies() -> list[dict]:
"""Convert raw proxy list to Playwright format"""
proxies = []
for proxy in AMAZON_PROXY_LIST_RAW:
parts = proxy.split(":")
if len(parts) == 4:
proxies.append({
"server": f"http://{parts[0]}:{parts[1]}",
"username": parts[2],
"password": parts[3]
})
return proxies
# === USER AGENTS ===
USER_AGENT_POOL = [
# Chrome Windows
"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36",
"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/130.0.0.0 Safari/537.36",
# Chrome Mac
"Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36",
# Firefox Windows
"Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:133.0) Gecko/20100101 Firefox/133.0",
"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36 Edg/131.0.0.0",
# Recent Chrome
"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/129.0.0.0 Safari/537.36",
"Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/129.0.0.0 Safari/537.36",
]
def get_random_user_agent() -> str:
return random.choice(USER_AGENT_POOL)
# === SITE CONFIGURATIONS ===
SITE_CONFIGS = {
# === MAGASINS DISCOUNT ===
"gifi.fr": {
"name": "Gifi",
"search_url": "https://www.gifi.fr/resultat-recherche?q={query}",
"product_selector": "a.link",
"product_image_selector": "img.tile-image, img[class*='product'], picture img",
"wait_selector": ".product-tile",
"category": "Discount",
"requires_proxy": False,
},
"stokomani.fr": {
"name": "Stokomani",
"search_url": "https://www.stokomani.fr/search?options%5Bprefix%5D=last&q={query}",
"product_selector": "a.reversed-link.block, a[href*='/products/']",
"product_image_selector": "img[loading='lazy'], img[class*='object'], img[sizes]",
"wait_selector": "a.reversed-link.block, .product-card",
"category": "Discount",
"requires_proxy": False,
},
"action.com": {
"name": "Action",
"search_url": "https://www.action.com/fr-fr/search/?q={query}",
"product_selector": "a.group[href^='/fr-fr/p/']",
"product_image_selector": "img[loading='lazy'], img[src*='product'], picture img",
"wait_selector": "a.group[href^='/fr-fr/p/']",
"category": "Discount",
"requires_proxy": False,
},
"cdiscount.com": {
"name": "Cdiscount",
"search_url": "https://www.cdiscount.com/search/10/{query}.html",
"product_selector": "a.prdtBILnk, a[href*='/f-'][href*='.html']",
"product_image_selector": "img.lazy, img[data-src], img[src*='image']",
"wait_selector": ".prdtBILDetails, .prdtBIL",
"category": "E-commerce",
"requires_proxy": False,
},
# === ÉLECTRONIQUE / HIGH-TECH ===
"darty.com": {
"name": "Darty",
"search_url": "https://www.darty.com/nav/recherche?text={query}",
"product_selector": "a[href*='/nav/achat/'][href*='.html']",
"product_image_selector": "img.product_img",
"wait_selector": ".product-card, .product-list",
"category": "Électronique",
"requires_proxy": False,
},
"boulanger.com": {
"name": "Boulanger",
"search_url": "https://www.boulanger.com/resultats?tr={query}",
"product_selector": "a[href*='/ref/'][href*='_']",
"product_image_selector": "img.product-image",
"wait_selector": ".product-list, .product-card",
"category": "Électronique",
"requires_proxy": False,
},
"fnac.com": {
"name": "Fnac",
"search_url": "https://www.fnac.com/SearchResult/ResultList.aspx?Search={query}",
"product_selector": "a.Article-title[href*='/a']",
"product_image_selector": "img.Article-itemVisual, img[data-src], img[src*='product'], picture img",
"wait_selector": ".Article-item, .Article-list",
"category": "Culture & Tech",
"requires_proxy": False,
},
# === NOUVEAUX SITES ===
"bmstores.fr": {
"name": "B&M",
"search_url": "https://bmstores.fr/module/ambjolisearch/jolisearch?s={query}",
"product_selector": "a.thumbnail.product-thumbnail",
"product_image_selector": "img.product-thumbnail",
"wait_selector": ".products, .product-miniature",
"category": "Discount",
"requires_proxy": False,
},
"centrakor.com": {
"name": "Centrakor",
"search_url": "https://www.centrakor.com/search/{query}",
"product_selector": "a.link.link--block, a.product-item__link",
"product_image_selector": "img.product-card__image, img[loading='lazy'], noscript img",
"wait_selector": ".product-list, .collection__products",
"category": "Discount",
"requires_proxy": False,
},
"lincroyable.fr": {
"name": "L'Incroyable",
"search_url": "https://www.lincroyable.fr/recherche?query={query}",
"product_selector": "a.product-link, a[href*='/p/']",
"product_image_selector": "img[loading='lazy'], img.product-image__img, img[src*='product']",
"wait_selector": ".product-card, .products-grid",
"category": "Discount",
"requires_proxy": False,
},
# === GRANDES SURFACES ===
"e-leclerc.com": {
"name": "E.Leclerc",
"search_url": "https://www.e-leclerc.com/recherche?text={query}",
"product_selector": "a[data-testid='product-link'], a[href*='/cat/'][href*='/sku/']",
"product_image_selector": "img[data-testid='product-image'], img[src*='scene7'], img[loading='eager']",
"wait_selector": "[data-testid='product-card'], [data-testid='product-link']",
"category": "Grande Surface",
"requires_proxy": False,
},
"auchan.fr": {
"name": "Auchan",
"search_url": "https://www.auchan.fr/search?text={query}",
"product_selector": "article.product-item a.product-item__name, a[class*='product-item'][href]",
"product_image_selector": "img.product-thumbnail__img, img[class*='product'][class*='img']",
"wait_selector": "article.product-item, .product-list",
"category": "Grande Surface",
"requires_proxy": False,
},
"carrefour.fr": {
"name": "Carrefour",
"search_url": "https://www.carrefour.fr/s?q={query}",
"product_selector": "a[data-testid='product-card-link'], a.product-card__link",
"product_image_selector": "img[data-testid='product-card-image'], img[fetchpriority], img.ds-product-image__img",
"wait_selector": "[data-testid='product-card'], article[data-testid]",
"category": "Grande Surface",
"requires_proxy": False,
},
}
# === COOKIE BANNERS ===
COOKIE_ACCEPT_SELECTORS = [
"#sp-cc-accept", # Amazon
"#onetrust-accept-btn-handler", # Many sites
".onetrust-accept-btn-handler",
"#popin_tc_privacy_button_2", # Auchan/Cdiscount
".popin_tc_privacy_button",
"#didomi-notice-agree-button", # Many French sites
".didomi-continue-without-agreeing",
"[data-testid='accept-cookies-button']", # Carrefour
".bcom-consent-accept-all", # Boulanger
"#cookieBanner-accept",
".action-primary.action-accept", # Magento
"#btn-cookie-allow",
"button:has-text('Tout accepter')",
"button:has-text('Accepter tout')",
"button:has-text('Accepter')",
"button:has-text('J\\'accepte')",
# Promotional Popups / Modals
".modal-close",
".close-modal",
".popup-close",
"button[aria-label='Close']",
"button[aria-label='Fermer']",
".js-modal-close",
"div[class*='popup'] button[class*='close']",
"div[class*='modal'] button[class*='close']",
# Specific Gifi/Discount popups
".popin-close",
"#popin-close",
"a.close-popin",
# Centrakor Store Overlay
"div.mod-shops button.close",
".mod-shops .close",
"#shops-modal .close-button",
".ui-dialog-titlebar-close",
# Action
"#cc-accept",
".cc-btn-accept",
"button[data-testid='uc-accept-all-button']",
"button:has-text('Accepter tout')",
"button:has-text('Accepter')",
"button:has-text('Refuser')",
]
Reference in new issue View Git Blame Copy Permalink