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: ${{ !github.event.pull_request.head.repo.fork }}
     runs-on: ubuntu-22.04
     steps:
-      -
+      - name: Checkout
-        name: Checkout
+        uses: actions/checkout@v4
-        uses: actions/checkout@v3
+
-      -
+      - name: Downcase repo
-        name: Downcase repo
         run: echo REPOSITORY=$(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 }}
+          images: |
-          tag-sha: false
+            ${{ env.REPOSITORY }},enable=${{ github.event_name != 'pull_request' }}
-      -
+            ghcr.io/${{ env.REPOSITORY }}
-        name: Set up QEMU
+          tags: |
-        uses: docker/setup-qemu-action@v2
+            type=semver,pattern={{version}},prefix=v
-      -
+            type=ref,event=pr
-        name: Set up Docker Buildx
+          flavor: |
-        uses: docker/setup-buildx-action@v2
+            latest=auto
-      -
+
-        name: Login to DockerHub
+      - name: Set up QEMU
-        uses: docker/login-action@v2
+        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
+      - name: Login to GitHub Container Registry
-        uses: docker/login-action@v2
+        uses: docker/login-action@v3
         with:
           registry: ghcr.io
           username: ${{ github.repository_owner }}
           password: ${{ secrets.GH_PAT }}
-      -
+
-        name: Build and push
+      - name: Build and push
-        uses: docker/build-push-action@v3
+        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 +1,107 @@
 # 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
+* Update Cloudflare challenge and checkbox selectors. Thanks @tenettow & @21hsmw
+
+## v3.3.19 (2024/05/23)
+* Fix occasional headless issue on Linux when set to "false". Thanks @21hsmw
+
+## v3.3.18 (2024/05/20)
+
+* Fix LANG ENV for Linux
+* Fix Chrome v124+ not closing on Windows. Thanks @RileyXX
+
+## v3.3.17 (2024/04/09)
+
+* Fix file descriptor leak in service on quit(). Thanks @zkulis
+
+## v3.3.16 (2024/02/28)
+
+* Fix of the subprocess.STARTUPINFO() call. Thanks @ceconelo
+* Add FreeBSD support. Thanks @Asthowen
+* Use headless configuration properly. Thanks @hashworks
+
+## v3.3.15 (2024/02/20)
+
+* Fix looping challenges
+
+## v3.3.14-hotfix2 (2024/02/17)
+
+* Hotfix 2 - bad Chromium build, instances failed to terminate
+
+## v3.3.14-hotfix (2024/02/17)
+
+* Hotfix for Linux build - some Chrome files no longer exist
+
+## v3.3.14 (2024/02/17)
+
+* Update Chrome downloads. Thanks @opemvbs
+
+## v3.3.13 (2024/01/07)
+
+* Fix too many open files error
+
+## v3.3.12 (2023/12/15)
+
+* Fix looping challenges and invalid cookies
+
+## v3.3.11 (2023/12/11)
+
+* Update UC 3.5.4 & Selenium 4.15.2. Thanks @txtsd
+
+## v3.3.10 (2023/11/14)
+
+* Add LANG ENV - resolves issues with YGGtorrent
+
+## v3.3.9 (2023/11/13)
+
+* Fix for Docker build, capture TypeError
+
+## v3.3.8 (2023/11/13)
+
+* Fix headless=true for Chrome 117+. Thanks @NabiKAZ
+* Support running Chrome 119 from source. Thanks @koleg and @Chris7X
+* Fix "OSError: [WinError 6] The handle is invalid" on exit. Thanks @enesgorkemgenc
+
+## v3.3.7 (2023/11/05)
+
+* Bump to rebuild. Thanks @JoachimDorchies
+
+## v3.3.6 (2023/09/15)
+
+* Update checkbox selector, again
+
+## v3.3.5 (2023/09/13)
+
+* Change checkbox selector, support languages other than English
+
+## v3.3.4 (2023/09/02)
+
+* Update checkbox selector
+
+## v3.3.3 (2023/08/31)
+
+* Update undetected_chromedriver to v3.5.3
+
+## v3.3.2 (2023/08/03)
+
+* Fix URL domain in Prometheus exporter
+
+## v3.3.1 (2023/08/03)
+
+* Fix for Cloudflare verify checkbox
+* Fix HEADLESS=false in Windows binary
+* Fix Prometheus exporter for management and health endpoints
+* Remove misleading stack trace when the verify checkbox is not found
+* Revert "Update base Docker image to Debian Bookworm" #849
+* Revert "Install Chromium 115 from Debian testing" #849
+
 ## v3.3.0 (2023/08/02)
 
 * Fix for new Cloudflare detection. Thanks @cedric-bour for #845

Dockerfile

@@ -1,4 +1,4 @@
-FROM python:3.11-slim-bookworm as builder
+FROM python:3.11-slim-bullseye as builder
 
 # Build dummy packages to skip installing them and their dependencies
 RUN apt-get update \
@@ -12,7 +12,7 @@ RUN apt-get update \
     && equivs-build adwaita-icon-theme \
     && mv adwaita-icon-theme_*.deb /adwaita-icon-theme.deb
 
-FROM python:3.11-slim-bookworm
+FROM python:3.11-slim-bullseye
 
 # Copy dummy packages
 COPY --from=builder /*.deb /
@@ -27,16 +27,12 @@ WORKDIR /app
     # Install dummy packages
 RUN dpkg -i /libgl1-mesa-dri.deb \
     && dpkg -i /adwaita-icon-theme.deb \
-    # Use Testing packages. The latest version of Chromium is not available for ARM
-    && sed -i 's/bookworm-updates/bookworm-updates testing/g' /etc/apt/sources.list.d/debian.sources \
     # Install dependencies
     && apt-get update \
-    && apt-get install -y --no-install-recommends -t testing chromium chromium-common chromium-driver xvfb dumb-init \
+    && apt-get install -y --no-install-recommends chromium chromium-common chromium-driver xvfb dumb-init \
-        procps curl vim-tiny xauth \
+        procps curl vim xauth \
     # Remove temporary files and hardware decoding libraries
     && rm -rf /var/lib/apt/lists/* \
-    && rm -f /usr/lib/systemd/systemd* \
-    && rm -f /usr/lib/x86_64-linux-gnu/systemd/* \
     && rm -f /usr/lib/x86_64-linux-gnu/libmfxhw* \
     && rm -f /usr/lib/x86_64-linux-gnu/mfx/* \
     # Create flaresolverr user
@@ -66,17 +62,17 @@ ENTRYPOINT ["/usr/bin/dumb-init", "--"]
 CMD ["/usr/local/bin/python", "-u", "/app/flaresolverr.py"]
 
 # Local build
-# docker build -t ngosang/flaresolverr:3.3.0 .
+# docker build -t ngosang/flaresolverr:3.3.21 .
-# docker run -p 8191:8191 ngosang/flaresolverr:3.3.0
+# 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:3.3.0 --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:3.3.0 --platform linux/arm/v7 --load .
+# 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:3.3.0
+# docker run -p 8191:8191 --platform linux/arm/v7 ngosang/flaresolverr:3.3.21

README.md

@@ -45,7 +45,8 @@ Supported architectures are:
 | ARM32        | linux/arm/v7 |
 | ARM64        | linux/arm64  |
 
-We provide a `docker-compose.yml` configuration file. Clone this repository and execute `docker-compose up -d` to start
+We provide a `docker-compose.yml` configuration file. Clone this repository and execute
+`docker-compose up -d` _(Compose V1)_ or `docker compose up -d` _(Compose V2)_ to start
 the container.
 
 If you prefer the `docker cli` execute the following command.
@@ -78,28 +79,62 @@ This is the recommended way for Windows users.
 
 * Install [Python 3.11](https://www.python.org/downloads/).
 * Install [Chrome](https://www.google.com/intl/en_us/chrome/) (all OS) or [Chromium](https://www.chromium.org/getting-involved/download-chromium/) (just Linux, it doesn't work in Windows) web browser.
-* (Only in Linux / macOS) Install [Xvfb](https://en.wikipedia.org/wiki/Xvfb) package.
+* (Only in Linux) Install [Xvfb](https://en.wikipedia.org/wiki/Xvfb) package.
+* (Only in macOS) Install [XQuartz](https://www.xquartz.org/) package.
 * Clone this repository and open a shell in that path.
 * Run `pip install -r requirements.txt` command to install FlareSolverr dependencies.
 * Run `python src/flaresolverr.py` command to start FlareSolverr.
 
+### From source code (FreeBSD/TrueNAS CORE)
+
+* Run `pkg install chromium python39 py39-pip xorg-vfbserver` command to install the required dependencies.
+* Clone this repository and open a shell in that path.
+* Run `python3.9 -m pip install -r requirements.txt` command to install FlareSolverr dependencies.
+* Run `python3.9 src/flaresolverr.py` command to start FlareSolverr.
+
 ### Systemd service
 
 We provide an example Systemd unit file `flaresolverr.service` as reference. You have to modify the file to suit your needs: paths, user and environment variables.
 
 ## Usage
 
-Example request:
+Example Bash request:
 ```bash
 curl -L -X POST 'http://localhost:8191/v1' \
 -H 'Content-Type: application/json' \
 --data-raw '{
   "cmd": "request.get",
-  "url":"http://www.google.com/",
+  "url": "http://www.google.com/",
   "maxTimeout": 60000
 }'
 ```
 
+Example Python request:
+```py
+import requests
+
+url = "http://localhost:8191/v1"
+headers = {"Content-Type": "application/json"}
+data = {
+    "cmd": "request.get",
+    "url": "http://www.google.com/",
+    "maxTimeout": 60000
+}
+response = requests.post(url, headers=headers, json=data)
+print(response.text)
+```
+
+Example PowerShell request:
+```ps1
+$body = @{
+    cmd = "request.get"
+    url = "http://www.google.com/"
+    maxTimeout = 60000
+} | ConvertTo-Json
+
+irm -UseBasicParsing 'http://localhost:8191/v1' -Headers @{"Content-Type"="application/json"} -Method Post -Body $body
+```
+
 ### Commands
 
 #### + `sessions.create`
@@ -113,7 +148,7 @@ This also speeds up the requests since it won't have to launch a new browser ins
 | Parameter | Notes                                                                                                                                                                                                                                                                                                            |
 |-----------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
 | session   | Optional. The session ID that you want to be assigned to the instance. If isn't set a random UUID will be assigned.                                                                                                                                                                                              |
-| proxy     | Optional, default disabled. Eg: `"proxy": {"url": "http://127.0.0.1:8888"}`. You must include the proxy schema in the URL: `http://`, `socks4://` or `socks5://`. Authorization (username/password) is supported. Eg: `"proxy": {"url": "http://127.0.0.1:8888", username": "testuser", "password": "testpass"}` |
+| proxy     | Optional, default disabled. Eg: `"proxy": {"url": "http://127.0.0.1:8888"}`. You must include the proxy schema in the URL: `http://`, `socks4://` or `socks5://`. Authorization (username/password) is supported. Eg: `"proxy": {"url": "http://127.0.0.1:8888", "username": "testuser", "password": "testpass"}` |
 
 #### + `sessions.list`
 
@@ -232,6 +267,7 @@ This is the same as `request.get` but it takes one more param:
 | LOG_HTML           | false                  | Only for debugging. If `true` all HTML that passes through the proxy will be logged to the console in `debug` level.                                          |
 | CAPTCHA_SOLVER     | none                   | Captcha solving method. It is used when a captcha is encountered. See the Captcha Solvers section.                                                            |
 | TZ                 | UTC                    | Timezone used in the logs and the web browser. Example: `TZ=Europe/London`.                                                                                   |
+| LANG               | none                   | Language used in the web browser. Example: `LANG=en_GB`.                                                                                   |
 | HEADLESS           | true                   | Only for debugging. To run the web browser in headless mode or visible.                                                                                       |
 | BROWSER_TIMEOUT    | 40000                  | If you are experiencing errors/timeouts because your system is slow, you can try to increase this value. Remember to increase the `maxTimeout` parameter too. |
 | TEST_URL           | https://www.google.com | FlareSolverr makes a request on start to make sure the web browser is working. You can change that URL if it is blocked in your country.                      |
@@ -242,8 +278,8 @@ This is the same as `request.get` but it takes one more param:
 
 Environment variables are set differently depending on the operating system. Some examples:
 * Docker: Take a look at the Docker section in this document. Environment variables can be set in the `docker-compose.yml` file or in the Docker CLI command.
-* Linux: Run `export LOG_LEVEL=debug` and then start FlareSolverr in the same shell.
+* Linux: Run `export LOG_LEVEL=debug` and then run `flaresolverr` in the same shell.
-* Windows: Open `cmd.exe`, run `set LOG_LEVEL=debug` and then start FlareSolverr in the same shell.
+* Windows: Open `cmd.exe`, run `set LOG_LEVEL=debug` and then run `flaresolverr.exe` in the same shell.
 
 ## Prometheus exporter
 

package.json

@@ -1,6 +1,6 @@
 {
   "name": "flaresolverr",
-  "version": "3.3.0",
+  "version": "3.3.21",
   "description": "Proxy server to bypass Cloudflare protection",
   "author": "Diego Heras (ngosang / ngosang@hotmail.es)",
   "license": "MIT"

requirements.txt

@@ -1,13 +1,13 @@
 bottle==0.12.25
 waitress==2.1.2
-selenium==4.11.2
+selenium==4.15.2
 func-timeout==4.3.5
 prometheus-client==0.17.1
 # required by undetected_chromedriver
-requests==2.31.0
+requests==2.32.3
-certifi==2023.7.22
+certifi==2024.07.04
 websockets==11.0.3
-# only required for linux
+# only required for linux and macos
-xvfbwrapper==0.2.9
+xvfbwrapper==0.2.9; platform_system != "Windows"
 # only required for windows
-pefile==2023.2.7
+pefile==2023.2.7; platform_system == "Windows"

src/bottle_plugins/prometheus_plugin.py

@@ -2,7 +2,8 @@ import logging
 import os
 import urllib.parse
 
-from dtos import V1ResponseBase
+from bottle import request
+from dtos import V1RequestBase, V1ResponseBase
 from metrics import start_metrics_http_server, REQUEST_COUNTER, REQUEST_DURATION
 
 PROMETHEUS_ENABLED = os.environ.get('PROMETHEUS_ENABLED', 'false').lower() == 'true'
@@ -33,10 +34,18 @@ def prometheus_plugin(callback):
     def export_metrics(actual_response):
         res = V1ResponseBase(actual_response)
 
+        if res.startTimestamp is None or res.endTimestamp is None:
+            # skip management and healthcheck endpoints
+            return
+
         domain = "unknown"
         if res.solution and res.solution.url:
-            parsed_url = urllib.parse.urlparse(res.solution.url)
+            domain = parse_domain_url(res.solution.url)
-            domain = parsed_url.hostname
+        else:
+            # timeout error
+            req = V1RequestBase(request.json)
+            if req.url:
+                domain = parse_domain_url(req.url)
 
         run_time = (res.endTimestamp - res.startTimestamp) / 1000
         REQUEST_DURATION.labels(domain=domain).observe(run_time)
@@ -50,4 +59,8 @@ def prometheus_plugin(callback):
             result = "error"
         REQUEST_COUNTER.labels(domain=domain, result=result).inc()
 
+    def parse_domain_url(url):
+        parsed_url = urllib.parse.urlparse(url)
+        return parsed_url.hostname
+
     return wrapper

src/build_package.py

@@ -25,7 +25,7 @@ def clean_files():
 
 def download_chromium():
     # https://commondatastorage.googleapis.com/chromium-browser-snapshots/index.html?prefix=Linux_x64/
-    revision = "1140001" if os.name == 'nt' else '1140000'
+    revision = "1260008" if os.name == 'nt' else '1260015'
     arch = 'Win_x64' if os.name == 'nt' else 'Linux_x64'
     dl_file = 'chrome-win' if os.name == 'nt' else 'chrome-linux'
     dl_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), os.pardir, 'dist_chrome')
@@ -59,8 +59,7 @@ def download_chromium():
         # Give executable permissions for *nix
         # file * | grep executable | cut -d: -f1
         print("Giving executable permissions...")
-        execs = ['chrome', 'chrome_crashpad_handler', 'chrome_sandbox', 'chrome-wrapper', 'nacl_helper',
+        execs = ['chrome', 'chrome_crashpad_handler', 'chrome_sandbox', 'chrome-wrapper', 'xdg-mime', 'xdg-settings']
-                 'nacl_helper_bootstrap', 'nacl_irt_x86_64.nexe', 'xdg-mime', 'xdg-settings']
         for exec_file in execs:
             exec_path = os.path.join(chrome_path, exec_file)
             os.chmod(exec_path, 0o755)

src/flaresolverr.py

@@ -62,6 +62,12 @@ if __name__ == "__main__":
     if sys.version_info < (3, 9):
         raise Exception("The Python version is less than 3.9, a version equal to or higher is required.")
 
+    # fix for HEADLESS=false in Windows binary
+    # https://stackoverflow.com/a/27694505
+    if os.name == 'nt':
+        import multiprocessing
+        multiprocessing.freeze_support()
+
     # fix ssl certificates for compiled binaries
     # https://github.com/pyinstaller/pyinstaller/issues/7229
     # https://stackoverflow.com/questions/55736855/how-to-change-the-cafile-argument-in-the-ssl-module-in-python3
@@ -95,6 +101,9 @@ if __name__ == "__main__":
     logging.info(f'FlareSolverr {utils.get_flaresolverr_version()}')
     logging.debug('Debug log enabled')
 
+    # Get current OS for global variable
+    utils.get_current_platform()
+
     # test browser installation
     flaresolverr_service.test_browser_installation()
 

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
@@ -40,13 +41,13 @@ CHALLENGE_TITLES = [
 ]
 CHALLENGE_SELECTORS = [
     # Cloudflare
-    '#cf-challenge-running', '.ray_id', '.attack-box', '#cf-please-wait', '#challenge-spinner', '#trk_jschal_js',
+    '#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
     'div.vc div.text-box h2'
 ]
-SHORT_TIMEOUT = 10
+SHORT_TIMEOUT = 1
 SESSIONS_STORAGE = SessionsStorage()
 
 
@@ -119,7 +120,7 @@ def _controller_v1_handler(req: V1RequestBase) -> V1ResponseBase:
         logging.warning("Request parameter 'userAgent' was removed in FlareSolverr v2.")
 
     # set default values
-    if req.maxTimeout is None or req.maxTimeout < 1:
+    if req.maxTimeout is None or int(req.maxTimeout) < 1:
         req.maxTimeout = 60000
 
     # execute the command
@@ -218,9 +219,23 @@ 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 = req.maxTimeout / 1000
+    timeout = int(req.maxTimeout) / 1000
     driver = None
     try:
         if req.session:
@@ -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.')
@@ -245,27 +261,39 @@ def _resolve_challenge(req: V1RequestBase, method: str) -> ChallengeResolutionT:
         raise Exception('Error solving the challenge. ' + str(e).replace('\n', '\\n'))
     finally:
         if not req.session and driver is not None:
+            if utils.PLATFORM_VERSION == "nt":
+                driver.close()
             driver.quit()
             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[@title='Widget containing a Cloudflare security challenge']")
+        iframe = get_shadowed_iframe(driver, "div:not(:has(div))")
         driver.switch_to.frame(iframe)
-        checkbox = driver.find_element(
+        iframe_body = driver.find_element(By.CSS_SELECTOR, "body")
-            by=By.XPATH,
+        if iframe_body:
-            value='//*[@id="cf-stage"]//label[@class="ctp-checkbox-label"]/input',
+            iframe_body.click()
-        )
-        if checkbox:
             actions = ActionChains(driver)
-            actions.move_to_element_with_offset(checkbox, 5, 7)
+            actions.move_to_element_with_offset(iframe_body, 10, 10)
-            actions.click(checkbox)
+            actions.click(iframe_body)
             actions.perform()
-            logging.debug("Cloudflare verify checkbox found and clicked!")
+            logging.debug("Attempted to click on iframe body")
     except Exception as e:
-        logging.debug("Cloudflare verify checkbox not found on the page. Error: " + str(e))
+        logging.debug("Cloudflare verify checkbox not found on the page. %s", repr(e))
     finally:
         driver.switch_to.default_content()
 
@@ -281,24 +309,41 @@ def click_verify(driver: WebDriver):
             actions.click(button)
             actions.perform()
             logging.debug("The Cloudflare 'Verify you are human' button found and clicked!")
-    except Exception as e:
+    except Exception:
-        logging.debug("The Cloudflare 'Verify you are human' button not found on the page. Error: " + str(e))
+        logging.debug("The Cloudflare 'Verify you are human' button not found on the page.")
 
     time.sleep(2)
 
 
+def get_correct_window(driver: WebDriver) -> WebDriver:
+    if len(driver.window_handles) > 1:
+        for window_handle in driver.window_handles:
+            driver.switch_to.window(window_handle)
+            current_url = driver.current_url
+            if not current_url.startswith("devtools://devtools"):
+                return driver
+    return driver
+
+
+def access_page(driver: WebDriver, url: str) -> None:
+    driver.get(url)
+    driver.start_session()
+    driver.start_session()  # required to bypass Cloudflare
+
+
 def _evil_logic(req: V1RequestBase, driver: WebDriver, method: str) -> ChallengeResolutionT:
     res = ChallengeResolutionT({})
     res.status = STATUS_OK
     res.message = ""
 
+
     # navigate to the page
     logging.debug(f'Navigating to... {req.url}')
     if method == 'POST':
         _post_request(req, driver)
     else:
-        with driver:
+        access_page(driver, req.url)
-            driver.get(req.url)
+    driver = get_correct_window(driver)
 
     # set cookies if required
     if req.cookies is not None and len(req.cookies) > 0:
@@ -310,8 +355,8 @@ def _evil_logic(req: V1RequestBase, driver: WebDriver, method: str) -> Challenge
         if method == 'POST':
             _post_request(req, driver)
         else:
-            with driver:
+            access_page(driver, req.url)
-                driver.get(req.url)
+        driver = get_correct_window(driver)
 
     # wait for the page
     if utils.get_config_log_html():
@@ -420,7 +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>
@@ -430,5 +475,6 @@ def _post_request(req: V1RequestBase, driver: WebDriver):
             <script>document.getElementById('hackForm').submit();</script>
         </body>
         </html>"""
-    with driver:
+    driver.get("data:text/html;charset=utf-8,{html_content}".format(html_content=html_content))
-        driver.get("data:text/html;charset=utf-8," + html_content)
+    driver.start_session()
+    driver.start_session()  # required to bypass Cloudflare

src/sessions.py

@@ -66,6 +66,8 @@ class SessionsStorage:
             return False
 
         session = self.sessions.pop(session_id)
+        if utils.PLATFORM_VERSION == "nt":
+            session.driver.close()
         session.driver.quit()
         return True
 

src/undetected_chromedriver/__init__.py

@@ -17,11 +17,12 @@ by UltrafunkAmsterdam (https://github.com/ultrafunkamsterdam)
 from __future__ import annotations
 
 
-__version__ = "3.5.0"
+__version__ = "3.5.5"
 
 import json
 import logging
 import os
+import pathlib
 import re
 import shutil
 import subprocess
@@ -373,6 +374,18 @@ class Chrome(selenium.webdriver.chrome.webdriver.WebDriver):
                 browser_executable_path or find_chrome_executable()
             )
 
+        if not options.binary_location or not \
+                pathlib.Path(options.binary_location).exists():
+                raise FileNotFoundError(
+                    "\n---------------------\n"
+                    "Could not determine browser executable."
+                    "\n---------------------\n"
+                    "Make sure your browser is installed in the default location (path).\n"
+                    "If you are sure about the browser executable, you can specify it using\n"
+                    "the `browser_executable_path='{}` parameter.\n\n"
+                    .format("/path/to/browser/executable" if IS_POSIX else "c:/path/to/your/browser.exe")
+                )
+
         self._delay = 3
 
         self.user_data_dir = user_data_dir
@@ -383,7 +396,7 @@ class Chrome(selenium.webdriver.chrome.webdriver.WebDriver):
         if no_sandbox:
             options.arguments.extend(["--no-sandbox", "--test-type"])
 
-        if headless or options.headless:
+        if headless or getattr(options, 'headless', None):
             #workaround until a better checking is found
             try:
                 v_main = int(self.patcher.version_main) if self.patcher.version_main else 108
@@ -438,8 +451,10 @@ class Chrome(selenium.webdriver.chrome.webdriver.WebDriver):
                 options.binary_location, *options.arguments
             )
         else:
-            startupinfo = subprocess.STARTUPINFO()
+            startupinfo = None
             if os.name == 'nt' and windows_headless:
+                # STARTUPINFO() is Windows only
+                startupinfo = subprocess.STARTUPINFO()
                 startupinfo.dwFlags |= subprocess.STARTF_USESHOWWINDOW
             browser = subprocess.Popen(
                 [options.binary_location, *options.arguments],
@@ -451,11 +466,9 @@ class Chrome(selenium.webdriver.chrome.webdriver.WebDriver):
             )
             self.browser_pid = browser.pid
 
-        # Fix for Chrome 115
+
-        # https://github.com/seleniumbase/SeleniumBase/pull/1967
         service = selenium.webdriver.chromium.service.ChromiumService(
-            executable_path=self.patcher.executable_path,
+            self.patcher.executable_path
-            service_args=["--disable-build-check"]
         )
 
         super(Chrome, self).__init__(
@@ -480,7 +493,7 @@ class Chrome(selenium.webdriver.chrome.webdriver.WebDriver):
         else:
             self._web_element_cls = WebElement
 
-        if options.headless:
+        if headless or getattr(options, 'headless', None):
             self._configure_headless()
 
     def _configure_headless(self):
@@ -758,7 +771,9 @@ class Chrome(selenium.webdriver.chrome.webdriver.WebDriver):
 
     def quit(self):
         try:
+            self.service.stop()
             self.service.process.kill()
+            self.command_executor.close()
             self.service.process.wait(5)
             logger.debug("webdriver process ended")
         except (AttributeError, RuntimeError, OSError):
@@ -773,15 +788,6 @@ class Chrome(selenium.webdriver.chrome.webdriver.WebDriver):
             logger.debug("gracefully closed browser")
         except Exception as e:  # noqa
             pass
-        # Force kill Chrome process in Windows
-        # https://github.com/FlareSolverr/FlareSolverr/issues/772
-        if os.name == 'nt':
-            try:
-                subprocess.call(['taskkill', '/f', '/pid', str(self.browser_pid)],
-                                stdout=subprocess.DEVNULL,
-                                stderr=subprocess.DEVNULL)
-            except Exception:
-                pass
         if (
             hasattr(self, "keep_user_data_dir")
             and hasattr(self, "user_data_dir")
@@ -800,7 +806,11 @@ class Chrome(selenium.webdriver.chrome.webdriver.WebDriver):
                 else:
                     logger.debug("successfully removed %s" % self.user_data_dir)
                     break
-                time.sleep(0.1)
+
+                try:
+                    time.sleep(0.1)
+                except OSError:
+                    pass
 
         # dereference patcher, so patcher can start cleaning up as well.
         # this must come last, otherwise it will throw 'in use' errors
@@ -895,8 +905,6 @@ def find_chrome_executable():
             if item is not None:
                 for subitem in (
                     "Google/Chrome/Application",
-                    "Google/Chrome Beta/Application",
-                    "Google/Chrome Canary/Application",
                 ):
                     candidates.add(os.sep.join((item, subitem, "chrome.exe")))
     for candidate in candidates:

src/undetected_chromedriver/devtool.py

@@ -2,6 +2,7 @@ import asyncio
 from collections.abc import Mapping
 from collections.abc import Sequence
 from functools import wraps
+import os
 import logging
 import threading
 import time
@@ -187,4 +188,6 @@ def test():
 
     time.sleep(10)
 
+    if os.name == "nt":
+        driver.close()
     driver.quit()

src/undetected_chromedriver/dprocess.py

@@ -41,6 +41,7 @@ def start_detached(executable, *args):
     # close pipes
     writer.close()
     reader.close()
+    process.close()
 
     return pid
 

src/undetected_chromedriver/patcher.py

@@ -3,9 +3,11 @@
 
 from distutils.version import LooseVersion
 import io
+import json
 import logging
 import os
 import pathlib
+import platform
 import random
 import re
 import shutil
@@ -19,26 +21,14 @@ from multiprocessing import Lock
 
 logger = logging.getLogger(__name__)
 
-IS_POSIX = sys.platform.startswith(("darwin", "cygwin", "linux", "linux2"))
+IS_POSIX = sys.platform.startswith(("darwin", "cygwin", "linux", "linux2", "freebsd"))
 
 
 class Patcher(object):
     lock = Lock()
-    url_repo = "https://chromedriver.storage.googleapis.com"
-    zip_name = "chromedriver_%s.zip"
     exe_name = "chromedriver%s"
 
     platform = sys.platform
-    if platform.endswith("win32"):
-        zip_name %= "win32"
-        exe_name %= ".exe"
-    if platform.endswith(("linux", "linux2")):
-        zip_name %= "linux64"
-        exe_name %= ""
-    if platform.endswith("darwin"):
-        zip_name %= "mac64"
-        exe_name %= ""
-
     if platform.endswith("win32"):
         d = "~/appdata/roaming/undetected_chromedriver"
     elif "LAMBDA_TASK_ROOT" in os.environ:
@@ -72,13 +62,34 @@ class Patcher(object):
         prefix = "undetected"
         self.user_multi_procs = user_multi_procs
 
+        try:
+            # Try to convert version_main into an integer
+            version_main_int = int(version_main)
+            # check if version_main_int is less than or equal to e.g 114
+            self.is_old_chromedriver = version_main and version_main_int <= 114
+        except (ValueError,TypeError):
+            # Check not running inside Docker
+            if not os.path.exists("/app/chromedriver"):
+                # If the conversion fails, log an error message
+                logging.info("version_main cannot be converted to an integer")
+            # Set self.is_old_chromedriver to False if the conversion fails
+            self.is_old_chromedriver = False
+
+        # Needs to be called before self.exe_name is accessed
+        self._set_platform_name()
+
         if not os.path.exists(self.data_path):
             os.makedirs(self.data_path, exist_ok=True)
 
         if not executable_path:
-            self.executable_path = os.path.join(
+            if sys.platform.startswith("freebsd"):
-                self.data_path, "_".join([prefix, self.exe_name])
+                self.executable_path = os.path.join(
-            )
+                    self.data_path, self.exe_name
+                )
+            else:
+                self.executable_path = os.path.join(
+                    self.data_path, "_".join([prefix, self.exe_name])
+                )
 
         if not IS_POSIX:
             if executable_path:
@@ -97,9 +108,36 @@ class Patcher(object):
             self._custom_exe_path = True
             self.executable_path = executable_path
 
+        # Set the correct repository to download the Chromedriver from
+        if self.is_old_chromedriver:
+            self.url_repo = "https://chromedriver.storage.googleapis.com"
+        else:
+            self.url_repo = "https://googlechromelabs.github.io/chrome-for-testing"
+
         self.version_main = version_main
         self.version_full = None
 
+    def _set_platform_name(self):
+        """
+        Set the platform and exe name based on the platform undetected_chromedriver is running on
+        in order to download the correct chromedriver.
+        """
+        if self.platform.endswith("win32"):
+            self.platform_name = "win32"
+            self.exe_name %= ".exe"
+        if self.platform.endswith(("linux", "linux2")):
+            self.platform_name = "linux64"
+            self.exe_name %= ""
+        if self.platform.endswith("darwin"):
+            if self.is_old_chromedriver:
+                self.platform_name = "mac64"
+            else:
+                self.platform_name = "mac-x64"
+            self.exe_name %= ""
+        if self.platform.startswith("freebsd"):
+            self.platform_name = "freebsd"
+            self.exe_name %= ""
+
     def auto(self, executable_path=None, force=False, version_main=None, _=None):
         """
 
@@ -111,16 +149,15 @@ class Patcher(object):
         Returns:
 
         """
-        # if self.user_multi_procs and \
-        #         self.user_multi_procs != -1:
-        #     # -1 being a skip value used later in this block
-        #
         p = pathlib.Path(self.data_path)
-        with Lock():
+        if self.user_multi_procs:
-            files = list(p.rglob("*chromedriver*?"))
+            with Lock():
-            for file in files:
+                files = list(p.rglob("*chromedriver*"))
-                if self.is_binary_patched(file):
+                most_recent = max(files, key=lambda f: f.stat().st_mtime)
-                    self.executable_path = str(file)
+                files.remove(most_recent)
+                list(map(lambda f: f.unlink(), files))
+                if self.is_binary_patched(most_recent):
+                    self.executable_path = str(most_recent)
                     return True
 
         if executable_path:
@@ -139,26 +176,56 @@ class Patcher(object):
         if force is True:
             self.force = force
 
-        try:
-            os.unlink(self.executable_path)
-        except PermissionError:
-            if self.force:
-                self.force_kill_instances(self.executable_path)
-                return self.auto(force=not self.force)
-            try:
-                if self.is_binary_patched():
-                    # assumes already running AND patched
-                    return True
-            except PermissionError:
-                pass
-            # return False
-        except FileNotFoundError:
-            pass
 
-        release = self.fetch_release_number()
+        if self.platform_name == "freebsd":
-        self.version_main = release.version[0]
+            chromedriver_path = shutil.which("chromedriver")
-        self.version_full = release
+
-        self.unzip_package(self.fetch_package())
+            if not os.path.isfile(chromedriver_path) or not os.access(chromedriver_path, os.X_OK):
+                logging.error("Chromedriver not installed!")
+                return
+
+            version_path = os.path.join(os.path.dirname(self.executable_path), "version.txt")
+
+            process = os.popen(f'"{chromedriver_path}" --version')
+            chromedriver_version = process.read().split(' ')[1].split(' ')[0]
+            process.close()
+
+            current_version = None
+            if os.path.isfile(version_path) or os.access(version_path, os.X_OK):
+                with open(version_path, 'r') as f:
+                    current_version = f.read()
+
+            if current_version != chromedriver_version:
+                logging.info("Copying chromedriver executable...")
+                shutil.copy(chromedriver_path, self.executable_path)
+                os.chmod(self.executable_path, 0o755)
+
+                with open(version_path, 'w') as f:
+                    f.write(chromedriver_version)
+
+                logging.info("Chromedriver executable copied!")
+        else:
+            try:
+                os.unlink(self.executable_path)
+            except PermissionError:
+                if self.force:
+                    self.force_kill_instances(self.executable_path)
+                    return self.auto(force=not self.force)
+                try:
+                    if self.is_binary_patched():
+                        # assumes already running AND patched
+                        return True
+                except PermissionError:
+                    pass
+                # return False
+            except FileNotFoundError:
+                pass
+
+            release = self.fetch_release_number()
+            self.version_main = release.version[0]
+            self.version_full = release
+            self.unzip_package(self.fetch_package())
+
         return self.patch()
 
     def driver_binary_in_use(self, path: str = None) -> bool:
@@ -202,7 +269,11 @@ class Patcher(object):
     def cleanup_unused_files(self):
         p = pathlib.Path(self.data_path)
         items = list(p.glob("*undetected*"))
-        print(items)
+        for item in items:
+            try:
+                item.unlink()
+            except:
+                pass
 
     def patch(self):
         self.patch_exe()
@@ -214,12 +285,32 @@ class Patcher(object):
         :return: version string
         :rtype: LooseVersion
         """
-        path = "/latest_release"
+        # Endpoint for old versions of Chromedriver (114 and below)
-        if self.version_main:
+        if self.is_old_chromedriver:
-            path += f"_{self.version_main}"
+            path = f"/latest_release_{self.version_main}"
-        path = path.upper()
+            path = path.upper()
+            logger.debug("getting release number from %s" % path)
+            return LooseVersion(urlopen(self.url_repo + path).read().decode())
+
+        # Endpoint for new versions of Chromedriver (115+)
+        if not self.version_main:
+            # Fetch the latest version
+            path = "/last-known-good-versions-with-downloads.json"
+            logger.debug("getting release number from %s" % path)
+            with urlopen(self.url_repo + path) as conn:
+                response = conn.read().decode()
+
+            last_versions = json.loads(response)
+            return LooseVersion(last_versions["channels"]["Stable"]["version"])
+
+        # Fetch the latest minor version of the major version provided
+        path = "/latest-versions-per-milestone-with-downloads.json"
         logger.debug("getting release number from %s" % path)
-        return LooseVersion(urlopen(self.url_repo + path).read().decode())
+        with urlopen(self.url_repo + path) as conn:
+            response = conn.read().decode()
+
+        major_versions = json.loads(response)
+        return LooseVersion(major_versions["milestones"][str(self.version_main)]["version"])
 
     def parse_exe_version(self):
         with io.open(self.executable_path, "rb") as f:
@@ -234,10 +325,16 @@ class Patcher(object):
 
         :return: path to downloaded file
         """
-        u = "%s/%s/%s" % (self.url_repo, self.version_full.vstring, self.zip_name)
+        zip_name = f"chromedriver_{self.platform_name}.zip"
-        logger.debug("downloading from %s" % u)
+        if self.is_old_chromedriver:
-        # return urlretrieve(u, filename=self.data_path)[0]
+            download_url = "%s/%s/%s" % (self.url_repo, self.version_full.vstring, zip_name)
-        return urlretrieve(u)[0]
+        else:
+            zip_name = zip_name.replace("_", "-", 1)
+            download_url = "https://storage.googleapis.com/chrome-for-testing-public/%s/%s/%s"
+            download_url %= (self.version_full.vstring, self.platform_name, zip_name)
+
+        logger.debug("downloading from %s" % download_url)
+        return urlretrieve(download_url)[0]
 
     def unzip_package(self, fp):
         """
@@ -245,6 +342,12 @@ class Patcher(object):
 
         :return: path to unpacked executable
         """
+        exe_path = self.exe_name
+        if not self.is_old_chromedriver:
+            # The new chromedriver unzips into its own folder
+            zip_name = f"chromedriver-{self.platform_name}"
+            exe_path = os.path.join(zip_name, self.exe_name)
+
         logger.debug("unzipping %s" % fp)
         try:
             os.unlink(self.zip_path)
@@ -253,10 +356,10 @@ class Patcher(object):
 
         os.makedirs(self.zip_path, mode=0o755, exist_ok=True)
         with zipfile.ZipFile(fp, mode="r") as zf:
-            zf.extract(self.exe_name, self.zip_path)
+            zf.extractall(self.zip_path)
-        os.rename(os.path.join(self.zip_path, self.exe_name), self.executable_path)
+        os.rename(os.path.join(self.zip_path, exe_path), self.executable_path)
         os.remove(fp)
-        os.rmdir(self.zip_path)
+        shutil.rmtree
         os.chmod(self.executable_path, 0o755)
         return self.executable_path
 

src/utils.py

@@ -5,11 +5,13 @@ import re
 import shutil
 import urllib.parse
 import tempfile
+import sys
 
 from selenium.webdriver.chrome.webdriver import WebDriver
 import undetected_chromedriver as uc
 
 FLARESOLVERR_VERSION = None
+PLATFORM_VERSION = None
 CHROME_EXE_PATH = None
 CHROME_MAJOR_VERSION = None
 USER_AGENT = None
@@ -37,6 +39,13 @@ def get_flaresolverr_version() -> str:
         FLARESOLVERR_VERSION = json.loads(f.read())['version']
         return FLARESOLVERR_VERSION
 
+def get_current_platform() -> str:
+    global PLATFORM_VERSION
+    if PLATFORM_VERSION is not None:
+        return PLATFORM_VERSION
+    PLATFORM_VERSION = os.name
+    return PLATFORM_VERSION
+
 
 def create_proxy_extension(proxy: dict) -> str:
     parsed_url = urllib.parse.urlparse(proxy['url'])
@@ -113,7 +122,7 @@ def create_proxy_extension(proxy: dict) -> str:
 
 
 def get_webdriver(proxy: dict = None) -> WebDriver:
-    global PATCHED_DRIVER_PATH
+    global PATCHED_DRIVER_PATH, USER_AGENT
     logging.debug('Launching web browser...')
 
     # undetected_chromedriver
@@ -136,6 +145,14 @@ def get_webdriver(proxy: dict = None) -> WebDriver:
     # https://peter.sh/experiments/chromium-command-line-switches/#use-gl
     options.add_argument('--use-gl=swiftshader')
 
+    language = os.environ.get('LANG', None)
+    if language is not None:
+        options.add_argument('--accept-lang=%s' % language)
+
+    # Fix for Chrome 117 | https://github.com/FlareSolverr/FlareSolverr/issues/910
+    if USER_AGENT is not None:
+        options.add_argument('--user-agent=%s' % USER_AGENT)
+
     proxy_extension_dir = None
     if proxy and all(key in proxy for key in ['url', 'username', 'password']):
         proxy_extension_dir = create_proxy_extension(proxy)
@@ -145,7 +162,7 @@ def get_webdriver(proxy: dict = None) -> WebDriver:
         logging.debug("Using webdriver proxy: %s", proxy_url)
         options.add_argument('--proxy-server=%s' % proxy_url)
 
-    # note: headless mode is detected (options.headless = True)
+    # note: headless mode is detected (headless = True)
     # we launch the browser in head-full mode with the window hidden
     windows_headless = False
     if get_config_headless():
@@ -153,6 +170,10 @@ def get_webdriver(proxy: dict = None) -> WebDriver:
             windows_headless = True
         else:
             start_xvfb_display()
+    # For normal headless mode:
+    # options.add_argument('--headless')
+
+    options.add_argument("--auto-open-devtools-for-tabs")
 
     # if we are inside the Docker container, we avoid downloading the driver
     driver_exe_path = None
@@ -162,10 +183,6 @@ def get_webdriver(proxy: dict = None) -> WebDriver:
         driver_exe_path = "/app/chromedriver"
     else:
         version_main = get_chrome_major_version()
-        # Fix for Chrome 115
-        # https://github.com/seleniumbase/SeleniumBase/pull/1967
-        if int(version_main) > 114:
-            version_main = 114
         if PATCHED_DRIVER_PATH is not None:
             driver_exe_path = PATCHED_DRIVER_PATH
 
@@ -174,9 +191,12 @@ def get_webdriver(proxy: dict = None) -> WebDriver:
 
     # downloads and patches the chromedriver
     # if we don't set driver_executable_path it downloads, patches, and deletes the driver each time
-    driver = uc.Chrome(options=options, browser_executable_path=browser_executable_path,
+    try:
-                       driver_executable_path=driver_exe_path, version_main=version_main,
+        driver = uc.Chrome(options=options, browser_executable_path=browser_executable_path,
-                       windows_headless=windows_headless, headless=windows_headless)
+                           driver_executable_path=driver_exe_path, version_main=version_main,
+                           windows_headless=windows_headless, headless=get_config_headless())
+    except Exception as e:
+        logging.error("Error starting Chrome: %s" % e)
 
     # save the patched driver to avoid re-downloads
     if driver_exe_path is None:
@@ -295,11 +315,15 @@ def get_user_agent(driver=None) -> str:
         if driver is None:
             driver = get_webdriver()
         USER_AGENT = driver.execute_script("return navigator.userAgent")
+        # Fix for Chrome 117 | https://github.com/FlareSolverr/FlareSolverr/issues/910
+        USER_AGENT = re.sub('HEADLESS', '', USER_AGENT, flags=re.IGNORECASE)
         return USER_AGENT
     except Exception as e:
         raise Exception("Error getting browser User-Agent. " + str(e))
     finally:
         if driver is not None:
+            if PLATFORM_VERSION == "nt":
+                driver.close()
             driver.quit()