Merge remote-tracking branch 'refs/remotes/origin/master' into fork-master

# Conflicts:
#	.github/workflows/release-docker.yml
This commit is contained in:
Kennedy Oliveira 2024-08-04 18:57:58 -03:00
commit 95cf12a23b
No known key found for this signature in database
14 changed files with 254 additions and 100 deletions

View File

@ -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.GITHUB_TOKEN }}
-
name: Build and push
uses: docker/build-push-action@v3
password: ${{ secrets.GH_PAT }}
- 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 }}

View File

@ -1,5 +1,48 @@
# 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

View File

@ -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:3.3.13 .
# docker run -p 8191:8191 ngosang/flaresolverr:3.3.13
# 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:3.3.13 --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:3.3.13 --platform linux/arm/v7 --load .
# docker run -p 8191:8191 --platform linux/arm/v7 ngosang/flaresolverr:3.3.13
# 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

View File

@ -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.
@ -78,11 +79,19 @@ 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.
@ -115,6 +124,17 @@ 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`
@ -259,8 +279,8 @@ 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.
* Windows: Open `cmd.exe`, run `set 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 run `flaresolverr.exe` in the same shell.
## Prometheus exporter

View File

@ -1,6 +1,6 @@
{
"name": "flaresolverr",
"version": "3.3.13",
"version": "3.3.21",
"description": "Proxy server to bypass Cloudflare protection",
"author": "Diego Heras (ngosang / ngosang@hotmail.es)",
"license": "MIT"

View File

@ -4,10 +4,10 @@ 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
xvfbwrapper==0.2.9
# only required for linux and macos
xvfbwrapper==0.2.9; platform_system != "Windows"
# only required for windows
pefile==2023.2.7
pefile==2023.2.7; platform_system == "Windows"

View File

@ -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',
'nacl_helper_bootstrap', 'nacl_irt_x86_64.nexe', 'xdg-mime', 'xdg-settings']
execs = ['chrome', 'chrome_crashpad_handler', 'chrome_sandbox', 'chrome-wrapper', 'xdg-mime', 'xdg-settings']
for exec_file in execs:
exec_path = os.path.join(chrome_path, exec_file)
os.chmod(exec_path, 0o755)

View File

@ -101,6 +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()

View File

@ -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',
'#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
@ -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
@ -220,7 +221,7 @@ def _cmd_sessions_destroy(req: V1RequestBase) -> V1ResponseBase:
def _resolve_challenge(req: V1RequestBase, method: str) -> ChallengeResolutionT:
timeout = req.maxTimeout / 1000
timeout = int(req.maxTimeout) / 1000
driver = None
try:
if req.session:
@ -245,6 +246,8 @@ 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')
@ -256,7 +259,7 @@ def click_verify(driver: WebDriver):
driver.switch_to.frame(iframe)
checkbox = driver.find_element(
by=By.XPATH,
value='//*[@id="challenge-stage"]/div/label/input',
value='//*[@id="content"]/div/div/label/input',
)
if checkbox:
actions = ActionChains(driver)
@ -287,20 +290,35 @@ def click_verify(driver: WebDriver):
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}')
driver.get(req.url)
driver.start_session() # required to bypass Cloudflare
if method == 'POST':
_post_request(req, driver)
else:
driver.get(req.url)
driver.start_session() # required to bypass Cloudflare
access_page(driver, req.url)
driver = get_correct_window(driver)
# set cookies if required
if req.cookies is not None and len(req.cookies) > 0:
@ -312,8 +330,8 @@ def _evil_logic(req: V1RequestBase, driver: WebDriver, method: str) -> Challenge
if method == 'POST':
_post_request(req, driver)
else:
driver.get(req.url)
driver.start_session() # required to bypass Cloudflare
access_page(driver, req.url)
driver = get_correct_window(driver)
# wait for the page
if utils.get_config_log_html():
@ -427,7 +445,7 @@ 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>
@ -437,5 +455,6 @@ 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

View File

@ -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

View File

@ -17,7 +17,7 @@ by UltrafunkAmsterdam (https://github.com/ultrafunkamsterdam)
from __future__ import annotations
__version__ = "3.5.4"
__version__ = "3.5.5"
import json
import logging
@ -451,8 +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],
@ -769,7 +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):
@ -784,15 +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")

View File

@ -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()

View File

@ -21,7 +21,7 @@ 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):
@ -68,8 +68,10 @@ class Patcher(object):
# 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):
# If the conversion fails, print an error message
print("version_main cannot be converted to an integer")
# 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
@ -80,6 +82,11 @@ class Patcher(object):
os.makedirs(self.data_path, exist_ok=True)
if not executable_path:
if sys.platform.startswith("freebsd"):
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])
)
@ -127,6 +134,9 @@ class Patcher(object):
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):
"""
@ -166,6 +176,35 @@ class Patcher(object):
if force is True:
self.force = force
if self.platform_name == "freebsd":
chromedriver_path = shutil.which("chromedriver")
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:
@ -186,6 +225,7 @@ class Patcher(object):
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:
@ -290,7 +330,7 @@ class Patcher(object):
download_url = "%s/%s/%s" % (self.url_repo, self.version_full.vstring, zip_name)
else:
zip_name = zip_name.replace("_", "-", 1)
download_url = "https://edgedl.me.gvt1.com/edgedl/chrome/chrome-for-testing/%s/%s/%s"
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)

View File

@ -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'])
@ -138,7 +147,7 @@ def get_webdriver(proxy: dict = None) -> WebDriver:
language = os.environ.get('LANG', None)
if language is not None:
options.add_argument('--lang=%s' % language)
options.add_argument('--accept-lang=%s' % language)
# Fix for Chrome 117 | https://github.com/FlareSolverr/FlareSolverr/issues/910
if USER_AGENT is not None:
@ -164,6 +173,8 @@ def get_webdriver(proxy: dict = None) -> WebDriver:
# 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
version_main = None
@ -180,9 +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
try:
driver = uc.Chrome(options=options, browser_executable_path=browser_executable_path,
driver_executable_path=driver_exe_path, version_main=version_main,
windows_headless=windows_headless, headless=windows_headless)
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:
@ -308,6 +322,8 @@ def get_user_agent(driver=None) -> str:
raise Exception("Error getting browser User-Agent. " + str(e))
finally:
if driver is not None:
if PLATFORM_VERSION == "nt":
driver.close()
driver.quit()