Update css selector to get shadowed iframe

*.0 was yanked

.github/workflows/release-docker.yml

@@ -4,50 +4,64 @@ on:
   push:
     tags:
       - 'v*.*.*'
+  pull_request:
+    branches:
+      - master
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
+  cancel-in-progress: true
 
 jobs:
   build-docker-images:
+    if: ${{ !github.event.pull_request.head.repo.fork }}
     runs-on: ubuntu-22.04
     steps:
-      -
-        name: Checkout
-        uses: actions/checkout@v3
-      -
-        name: Downcase repo
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Downcase repo
         run: echo REPOSITORY=$(echo ${{ github.repository }} | tr '[:upper:]' '[:lower:]') >> $GITHUB_ENV
-      -
-        name: Docker meta
+
+      - name: Docker meta
         id: docker_meta
-        uses: crazy-max/ghaction-docker-meta@v3
+        uses: docker/metadata-action@v5
         with:
-          images: ${{ env.REPOSITORY }},ghcr.io/${{ env.REPOSITORY }}
-          tag-sha: false
-      -
-        name: Set up QEMU
-        uses: docker/setup-qemu-action@v2
-      -
-        name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v2
-      -
-        name: Login to DockerHub
-        uses: docker/login-action@v2
+          images: |
+            ${{ env.REPOSITORY }},enable=${{ github.event_name != 'pull_request' }}
+            ghcr.io/${{ env.REPOSITORY }}
+          tags: |
+            type=semver,pattern={{version}},prefix=v
+            type=ref,event=pr
+          flavor: |
+            latest=auto
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v3
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Login to DockerHub
+        uses: docker/login-action@v3
+        if: github.event_name != 'pull_request'
         with:
           username: ${{ secrets.DOCKERHUB_USERNAME }}
           password: ${{ secrets.DOCKERHUB_TOKEN }}
-      -
-        name: Login to GitHub Container Registry
-        uses: docker/login-action@v2
+
+      - name: Login to GitHub Container Registry
+        uses: docker/login-action@v3
         with:
           registry: ghcr.io
           username: ${{ github.repository_owner }}
           password: ${{ secrets.GH_PAT }}
-      -
-        name: Build and push
-        uses: docker/build-push-action@v3
+
+      - name: Build and push
+        uses: docker/build-push-action@v6
         with:
           context: .
           file: ./Dockerfile
           platforms: linux/386,linux/amd64,linux/arm/v7,linux/arm64/v8
-          push: ${{ github.event_name != 'pull_request' }}
+          push: true
           tags: ${{ steps.docker_meta.outputs.tags }}
           labels: ${{ steps.docker_meta.outputs.labels }}

CHANGELOG.md

@@ -1,5 +1,9 @@
 # Changelog
 
+## v3.3.21 (2024/06/26)
+* Add challenge selector to catch reloading page on non-English systems
+* Escape values for generated form used in request.post. Thanks @mynameisbogdan
+
 ## v3.3.20 (2024/06/21)
 * maxTimeout should always be int
 * Check not running in Docker before logging version_main error

Dockerfile

@@ -62,17 +62,17 @@ ENTRYPOINT ["/usr/bin/dumb-init", "--"]
 CMD ["/usr/local/bin/python", "-u", "/app/flaresolverr.py"]
 
 # Local build
-# docker build -t ngosang/flaresolverr:3.3.20 .
-# docker run -p 8191:8191 ngosang/flaresolverr:3.3.20
+# docker build -t ngosang/flaresolverr:3.3.21 .
+# docker run -p 8191:8191 ngosang/flaresolverr:3.3.21
 
 # Multi-arch build
 # docker run --rm --privileged multiarch/qemu-user-static --reset -p yes
 # docker buildx create --use
-# docker buildx build -t ngosang/flaresolverr:3.3.20 --platform linux/386,linux/amd64,linux/arm/v7,linux/arm64/v8 .
+# docker buildx build -t ngosang/flaresolverr:3.3.21 --platform linux/386,linux/amd64,linux/arm/v7,linux/arm64/v8 .
 #   add --push to publish in DockerHub
 
 # Test multi-arch build
 # docker run --rm --privileged multiarch/qemu-user-static --reset -p yes
 # docker buildx create --use
-# docker buildx build -t ngosang/flaresolverr:3.3.20 --platform linux/arm/v7 --load .
-# docker run -p 8191:8191 --platform linux/arm/v7 ngosang/flaresolverr:3.3.20
+# docker buildx build -t ngosang/flaresolverr:3.3.21 --platform linux/arm/v7 --load .
+# docker run -p 8191:8191 --platform linux/arm/v7 ngosang/flaresolverr:3.3.21

package.json

@@ -1,6 +1,6 @@
 {
   "name": "flaresolverr",
-  "version": "3.3.20",
+  "version": "3.3.21",
   "description": "Proxy server to bypass Cloudflare protection",
   "author": "Diego Heras (ngosang / ngosang@hotmail.es)",
   "license": "MIT"

requirements.txt

@@ -4,8 +4,8 @@ selenium==4.15.2
 func-timeout==4.3.5
 prometheus-client==0.17.1
 # required by undetected_chromedriver
-requests==2.31.0
-certifi==2023.7.22
+requests==2.32.3
+certifi==2024.07.04
 websockets==11.0.3
 # only required for linux and macos
 xvfbwrapper==0.2.9; platform_system != "Windows"

src/flaresolverr_service.py

@@ -3,7 +3,8 @@ import platform
 import sys
 import time
 from datetime import timedelta
-from urllib.parse import unquote
+from html import escape
+from urllib.parse import unquote, quote
 
 from func_timeout import FunctionTimedOut, func_timeout
 from selenium.common import TimeoutException
@@ -40,7 +41,7 @@ CHALLENGE_TITLES = [
 ]
 CHALLENGE_SELECTORS = [
     # Cloudflare
-    '#cf-challenge-running', '.ray_id', '.attack-box', '#cf-please-wait', '#challenge-spinner', '#trk_jschal_js', '#turnstile-wrapper',
+    '#cf-challenge-running', '.ray_id', '.attack-box', '#cf-please-wait', '#challenge-spinner', '#trk_jschal_js', '#turnstile-wrapper', '.lds-ring',
     # Custom CloudFlare for EbookParadijs, Film-Paleis, MuziekFabriek and Puur-Hollands
     'td.info #js_info',
     # Fairlane / pararius.com
@@ -218,6 +219,20 @@ def _cmd_sessions_destroy(req: V1RequestBase) -> V1ResponseBase:
         "message": "The session has been removed."
     })
 
+def _init_driver(driver):
+    try:
+        driver.execute_cdp_cmd('Page.enable', {})
+        driver.execute_cdp_cmd('Page.addScriptToEvaluateOnNewDocument', {
+            'source': """
+                Element.prototype._as = Element.prototype.attachShadow;
+                Element.prototype.attachShadow = function (params) {
+                    return this._as({mode: "open"})
+                };
+            """
+        })
+    except Exception as e:
+        logging.debug("Driver init exception: %s", repr(e))
+
 
 def _resolve_challenge(req: V1RequestBase, method: str) -> ChallengeResolutionT:
     timeout = int(req.maxTimeout) / 1000
@@ -238,6 +253,7 @@ def _resolve_challenge(req: V1RequestBase, method: str) -> ChallengeResolutionT:
         else:
             driver = utils.get_webdriver(req.proxy)
             logging.debug('New instance of webdriver has been created to perform the request')
+        _init_driver(driver)
         return func_timeout(timeout, _evil_logic, (req, driver, method))
     except FunctionTimedOut:
         raise Exception(f'Error solving the challenge. Timeout after {timeout} seconds.')
@@ -251,23 +267,33 @@ def _resolve_challenge(req: V1RequestBase, method: str) -> ChallengeResolutionT:
             logging.debug('A used instance of webdriver has been destroyed')
 
 
+def get_shadowed_iframe(driver: WebDriver, css_selector: str):
+    logging.debug("Getting ShadowRoot by selector: %s", css_selector)
+    shadow_element = driver.execute_script("""
+        return document.querySelector(arguments[0]).shadowRoot.firstChild;
+    """, css_selector)
+    if shadow_element:
+        logging.debug("iframe found")
+    else:
+        logging.debug("iframe not found")
+    return shadow_element
+
+
 def click_verify(driver: WebDriver):
     try:
         logging.debug("Try to find the Cloudflare verify checkbox...")
-        iframe = driver.find_element(By.XPATH, "//iframe[starts-with(@id, 'cf-chl-widget-')]")
+        iframe = get_shadowed_iframe(driver, "div:not(:has(div))")
         driver.switch_to.frame(iframe)
-        checkbox = driver.find_element(
-            by=By.XPATH,
-            value='//*[@id="content"]/div/div/label/input',
-        )
-        if checkbox:
+        iframe_body = driver.find_element(By.CSS_SELECTOR, "body")
+        if iframe_body:
+            iframe_body.click()
             actions = ActionChains(driver)
-            actions.move_to_element_with_offset(checkbox, 5, 7)
-            actions.click(checkbox)
+            actions.move_to_element_with_offset(iframe_body, 10, 10)
+            actions.click(iframe_body)
             actions.perform()
-            logging.debug("Cloudflare verify checkbox found and clicked!")
-    except Exception:
-        logging.debug("Cloudflare verify checkbox not found on the page.")
+            logging.debug("Attempted to click on iframe body")
+    except Exception as e:
+        logging.debug("Cloudflare verify checkbox not found on the page. %s", repr(e))
     finally:
         driver.switch_to.default_content()
 
@@ -439,7 +465,7 @@ def _post_request(req: V1RequestBase, driver: WebDriver):
             value = unquote(parts[1])
         except Exception:
             value = parts[1]
-        post_form += f'<input type="text" name="{name}" value="{value}"><br>'
+        post_form += f'<input type="text" name="{escape(quote(name))}" value="{escape(quote(value))}"><br>'
     post_form += '</form>'
     html_content = f"""
         <!DOCTYPE html>
@@ -449,6 +475,6 @@ def _post_request(req: V1RequestBase, driver: WebDriver):
             <script>document.getElementById('hackForm').submit();</script>
         </body>
         </html>"""
-    driver.get("data:text/html;charset=utf-8," + html_content)
+    driver.get("data:text/html;charset=utf-8,{html_content}".format(html_content=html_content))
     driver.start_session()
     driver.start_session()  # required to bypass Cloudflare