From b2e74409e31db3ceaf53cb1a61b7534da24f1bf7 Mon Sep 17 00:00:00 2001
From: nrg101 <admin@amazinggoose.com>
Date: Thu, 3 Sep 2026 12:38:23 +0100
Subject: [PATCH] use python cloudscraper with proxies to circumvent Cloudflare
 DDoS protection

---
 scrapers/Affect3DStore.yml               |  20 -----
 scrapers/Affect3DStore/Affect3DStore.py  | 105 +++++++++++++++++++++++
 scrapers/Affect3DStore/Affect3DStore.yml |  11 +++
 3 files changed, 116 insertions(+), 20 deletions(-)
 delete mode 100644 scrapers/Affect3DStore.yml
 create mode 100644 scrapers/Affect3DStore/Affect3DStore.py
 create mode 100644 scrapers/Affect3DStore/Affect3DStore.yml

diff --git a/scrapers/Affect3DStore.yml b/scrapers/Affect3DStore.yml
deleted file mode 100644
index 4acc2ae42..000000000
--- a/scrapers/Affect3DStore.yml
+++ /dev/null
@@ -1,20 +0,0 @@
-# yaml-language-server: $schema=../validator/scraper.schema.json
-name: Affect3DStore
-sceneByURL:
-  - action: scrapeXPath
-    url:
-      - affect3dstore.com
-    scraper: affect3DStore
-xPathScrapers:
-  affect3DStore:
-    scene:
-      Title: //h1/span
-      Details:
-        selector: //div[contains(@class, "info-details-xl")]//div[contains(@class, "description")]/descendant::text()
-        concat: "\n\n"
-      Studio:
-        Name: //div[@class="box-inner1"]//span[@class="artist_name_m"]//a/text()
-      Tags:
-        Name: //div[contains(@class, "info-details-xl")]//table[contains(@class, "additional-attributes")]//td[not(@data-th="Artist/Circle")]/a/text()
-      Image: //img[@alt="main product photo"]/@src
-# Last Updated April 22, 2026
diff --git a/scrapers/Affect3DStore/Affect3DStore.py b/scrapers/Affect3DStore/Affect3DStore.py
new file mode 100644
index 000000000..6e0bf502a
--- /dev/null
+++ b/scrapers/Affect3DStore/Affect3DStore.py
@@ -0,0 +1,105 @@
+"""
+Stash scraper made to circumvent Cloudflare DDOS protection
+"""
+import json
+import sys
+
+from py_common.deps import ensure_requirements
+ensure_requirements("cloudscraper", "fp:free-proxy", "lxml")
+import cloudscraper  # noqa: E402
+from fp.fp import FreeProxy
+from lxml import html
+
+from py_common import log
+from py_common.util import scraper_args
+from py_common.types import ScrapedScene
+
+scraper = cloudscraper.create_scraper()
+
+free_proxies = None
+
+STUDIO_MAPPER = {
+    "Miro Affect3D": "Miro",
+}
+
+def get_proxies() -> dict:
+    proxy = FreeProxy(rand=True).get()
+    log.debug("proxy: %s" % proxy)
+    return { 'http': proxy } if proxy.startswith('http:') else { 'https': proxy }
+
+def scrape_url(_url: str, max_retries=5):
+    "Scrapes a web page and returns a HTML tree"
+    retries = 0
+    while retries < max_retries:
+        try:
+            log.debug('about to execute scraper.get, attempt %d' % (retries + 1))
+            global free_proxies
+            free_proxies = get_proxies()
+            scraped = scraper.get(_url, proxies=free_proxies)
+            if scraped.status_code == 200:
+                log.debug('HTTP Status: 200')
+                return html.document_fromstring(scraped.text)
+            log.error('HTTP Error: %s' % scraped.status_code)
+        except Exception as e:
+            log.error("scraper.get error: %s" % e)
+
+        retries += 1
+        log.debug('Retrying (%d/%d)...' % (retries, max_retries))
+
+    raise Exception(f'Failed to scrape the URL after {max_retries} retries')
+
+def scene_from_url(_url: str) -> ScrapedScene | None:
+    "Scrapes a scene from a URL, running an optional postprocess function on the result"
+    scene: ScrapedScene = {}
+
+    try:
+        tree = scrape_url(_url)
+        # title
+        if title := tree.xpath('//h1/span'):
+            scene['title'] = title[0].text_content()
+        # date
+        if year_released := tree.xpath('//div[contains(@class, "info-details-xl")]//table[contains(@class, "additional-attributes")]//td[@data-th="Year Released"]'):
+            scene['date'] = year_released[0].text_content()
+        # details
+        if description_paragraphs := tree.xpath(
+            '//div[contains(@class, "info-details-xl")]//div[contains(@class, "description")]//p'
+        ):
+            scene['details'] = "\n\n".join([p.text_content() for p in description_paragraphs])
+        # studio
+        if artist_name := tree.xpath(
+            '//div[@class="box-inner1"]//span[@class="artist_name_m"]//a/text()'
+        ):
+            scene['studio'] = {"name": STUDIO_MAPPER.get(artist_name[0]) or artist_name[0]}
+        # tags
+        if tags := tree.xpath(
+            '//div[contains(@class, "info-details-xl")]//table[contains(@class, "additional-attributes")]//td[not(@data-th="Artist/Circle")]/a/text()'
+        ):
+            scene['tags'] = [{"name": t.strip()} for t in tags]
+        # image
+        if main_product_photo := tree.xpath('//img[@alt="main product photo"]/@src'):
+            scene['image'] = main_product_photo[0]
+    except Exception as e:
+        log.error("Error scraping scene from URL: %s" % e)
+        return None
+
+    return scene
+
+
+if __name__ == "__main__":
+    op, args = scraper_args()
+
+    log.debug(f"args: {args}")
+    match op, args:
+        case "scene-by-url", {"url": url} if url:
+            result = scene_from_url(url)
+        # case "scene-by-name", {"name": name} if name:
+        #     result = scene_search(name)
+        # case "scene-by-fragment" | "scene-by-query-fragment", args:
+        #     result = scene_from_fragment(args)
+        case _:
+            log.error(f"Operation: {op}, arguments: {json.dumps(args)}")
+            sys.exit(1)
+
+    log.debug(f"result: {result}")
+
+    print(json.dumps(result))
diff --git a/scrapers/Affect3DStore/Affect3DStore.yml b/scrapers/Affect3DStore/Affect3DStore.yml
new file mode 100644
index 000000000..9e53b71a3
--- /dev/null
+++ b/scrapers/Affect3DStore/Affect3DStore.yml
@@ -0,0 +1,11 @@
+# yaml-language-server: $schema=../validator/scraper.schema.json
+name: Affect3DStore
+sceneByURL:
+  - action: script
+    url:
+      - affect3dstore.com
+    script:
+      - python
+      - Affect3DStore.py
+      - scene-by-url
+# Last Updated September 3, 2026