Old Headless mode has been removed from the Chrome binary #174
Comments
|
I resolved this issue by following a post from the Chrome for Developers Blog: Download old Headless Chrome as chrome-headless-shell. First, I installed npx @puppeteer/browsers install chrome-headless-shell@stableThen, in my code, I specified the hti = Html2Image(
browser_executable="chrome-headless-shell"
)The next step is to find a way to obtain chrome-headless-shell without needing to use npx. |
|
I'll add my two cents here: for those who don't want to download it using The usage is simple: python download-chrome-headless-shell.py
"""This module handles downloading and extracting Chrome Headless builds for testing purposes.
It includes functionality to identify the current operating system, resolve download URLs based on
platform and release channel, and manage the extraction of downloaded archives.
Features:
- Automatic identification of the operating system.
- Downloading Chrome Headless builds from Google Chrome's public testing storage.
- Extraction of ZIP archives to a specified directory, with proper permission settings.
Dependencies:
- httpx: For HTTP requests.
- zipfile: For handling ZIP archives.
- platform: For system identification.
Usage:
Run the module directly to download and extract the latest stable Chrome Headless build for the detected platform.
Example:
python download-chrome-headless-shell.py
"""
import os
import httpx
from enum import Enum
import platform
import zipfile
import stat
import tempfile
import argparse
CHROME_LAST_KNOWN_GOOD_VERSION = "https://googlechromelabs.github.io/chrome-for-testing/last-known-good-versions.json"
CHROME_FOR_TESTING_PUBLIC_BASE_URL = "https://storage.googleapis.com/chrome-for-testing-public"
def ensure_directory_exists(directory: str):
"""Ensures that a directory exists, creating it if necessary.
Args:
directory (str): Path to the directory.
"""
if not os.path.exists(directory):
os.makedirs(directory)
class BrowserPlatform(Enum):
"""Enum representing supported browser platforms."""
LINUX = "linux"
MAC_ARM = "mac-arm"
MAC = "mac"
WIN32 = "win32"
WIN64 = "win64"
@staticmethod
def identify_current_os() -> "BrowserPlatform":
"""Identifies the current operating system and returns the corresponding BrowserPlatform.
Returns:
BrowserPlatform: Enum value representing the detected OS.
Raises:
ValueError: If the operating system is unsupported.
"""
system = platform.system()
machine = platform.machine().lower()
if system == "Linux":
return BrowserPlatform.LINUX
elif system == "Darwin":
if machine == "arm64":
return BrowserPlatform.MAC_ARM
else:
return BrowserPlatform.MAC
elif system == "Windows":
if "64" in machine:
return BrowserPlatform.WIN64
else:
return BrowserPlatform.WIN32
else:
raise ValueError("Unsupported operating system")
def get_folder(self) -> str:
"""Returns the folder name corresponding to the platform.
Returns:
str: Folder name for the platform.
Raises:
ValueError: If the platform is unsupported.
"""
try:
return {
BrowserPlatform.LINUX: "linux64",
BrowserPlatform.MAC_ARM: "mac-arm64",
BrowserPlatform.MAC: "mac-x64",
BrowserPlatform.WIN32: "win32",
BrowserPlatform.WIN64: "win64",
}.get(self)
except KeyError:
raise ValueError(f"Unsupported platform: {self}")
class ChromeReleaseChannel(Enum):
"""Enum representing Chrome release channels."""
STABLE = "Stable"
BETA = "Beta"
DEV = "Dev"
CANARY = "Canary"
def resolve_download_url(platform: BrowserPlatform, build_id: str) -> str:
"""Constructs the download URL for Chrome Headless based on platform and build ID.
Args:
platform (BrowserPlatform): The platform enum value.
build_id (str): The build ID to download.
Returns:
str: Full URL for the Chrome Headless download.
"""
return f"{CHROME_FOR_TESTING_PUBLIC_BASE_URL}/{'/'.join(resolve_download_path(platform, build_id))}"
def resolve_download_path(platform: BrowserPlatform, build_id: str) -> list[str]:
"""Resolves the download path components for Chrome Headless.
Args:
platform (BrowserPlatform): The platform enum value.
build_id (str): The build ID to download.
Returns:
list[str]: List of path components for the download.
"""
return [
build_id,
platform.get_folder(),
f"chrome-headless-shell-{platform.get_folder()}.zip",
]
def download_file(url: str, download_path: str):
"""Downloads a file from a URL and saves it to the specified path.
Args:
url (str): The URL to download from.
download_path (str): Path to save the downloaded file.
"""
with httpx.stream("GET", url) as response:
response.raise_for_status()
with open(download_path, "wb") as download_file:
for chunk in response.iter_bytes():
download_file.write(chunk)
def extract_zip(zip_path: str, extract_to: str):
"""Extracts the contents of a ZIP file to a specified directory and adjusts permissions for executables.
Args:
zip_path (str): Path to the ZIP file.
extract_to (str): Directory to extract contents into.
"""
with zipfile.ZipFile(zip_path, "r") as zip_ref:
zip_ref.extractall(extract_to)
for item in os.listdir(extract_to):
item_path = os.path.join(extract_to, item)
if os.path.isdir(item_path) and "chrome-headless-shell" in item:
new_folder_name = "chrome-headless-shell"
new_folder_path = os.path.join(extract_to, new_folder_name)
os.rename(item_path, new_folder_path)
break
if platform.system() != "Windows":
for root, _, files in os.walk(extract_to):
for file in files:
file_path = os.path.join(root, file)
os.chmod(file_path, os.stat(file_path).st_mode | stat.S_IEXEC)
def get_last_known_good_release_for_channel(channel: ChromeReleaseChannel) -> dict:
"""Retrieves the last known good release for a specific Chrome channel.
Args:
channel (ChromeReleaseChannel): The release channel.
Returns:
dict: Version and revision of the last known good release.
"""
response = httpx.get(CHROME_LAST_KNOWN_GOOD_VERSION)
response.raise_for_status()
data = response.json()
channels = data.get("channels", {})
channel_data = channels.get(channel.value, {})
return {
"version": channel_data.get("version"),
"revision": channel_data.get("revision"),
}
def resolve_build_id(channel: ChromeReleaseChannel) -> str:
"""Resolves the build ID for a specific Chrome release channel.
Args:
channel (ChromeReleaseChannel): The release channel.
Returns:
str: Build ID.
"""
release_info = get_last_known_good_release_for_channel(channel)
return release_info["version"]
def relative_executable_path(platform: BrowserPlatform) -> str:
"""Returns the relative path to the Chrome Headless executable.
Args:
platform (BrowserPlatform): The platform enum value.
build_id (str): The build ID.
Returns:
str: Relative path to the executable.
Raises:
ValueError: If the platform is unsupported.
"""
executable_file = "chrome-headless-shell"
if platform in [BrowserPlatform.WIN32, BrowserPlatform.WIN64]:
executable_file += ".exe"
return os.path.join("chrome-headless-shell", executable_file)
def find_executable_path(extract_to: str, platform: BrowserPlatform) -> str:
"""Finds the absolute path to the Chrome Headless executable.
Args:
extract_to (str): Directory where the Chrome Headless build is extracted.
platform (BrowserPlatform): The platform enum value.
Returns:
str: Absolute path to the executable.
"""
relative_path = relative_executable_path(platform)
absolute_path = os.path.abspath(os.path.join(extract_to, relative_path))
if not os.path.exists(absolute_path):
raise FileNotFoundError(f"Executable not found at {absolute_path}")
return absolute_path
def main(browser_platform: BrowserPlatform, channel: ChromeReleaseChannel, output_path: str):
"""Main function to handle Chrome Headless download and extraction.
Args:
browser_platform (BrowserPlatform): The platform enum value.
channel (ChromeReleaseChannel): The release channel.
output_path (str): Directory to extract the Chrome Headless build into.
Raises:
FileNotFoundError: If the executable is not found.
Returns:
str: Absolute path to the Chrome Headless executable.
"""
with tempfile.TemporaryDirectory() as download_directory:
ensure_directory_exists(output_path)
build_id = resolve_build_id(channel)
url = resolve_download_url(browser_platform, build_id)
download = os.path.join(download_directory, os.path.basename(url))
download_file(url, download)
extract_zip(download, output_path)
executable_path = find_executable_path(output_path, browser_platform)
print(executable_path) # noqa: T201
if __name__ == "__main__":
parser = argparse.ArgumentParser(
description=(
"This script downloads and extracts Chrome Headless builds for testing purposes. "
"It identifies the current operating system, determines the appropriate Chrome Headless build "
"based on the specified release channel, and extracts the binary to a specified directory."
),
epilog="Example usage: python download-chrome-headless-shell.py --output ./chrome-bin",
)
parser.add_argument(
"-b",
"--browser_platform",
type=BrowserPlatform,
required=False,
default=BrowserPlatform.identify_current_os(),
help=(
"Specify the target browser platform for the Chrome Headless build. "
"Defaults to the platform detected on the current system."
),
)
parser.add_argument(
"-c",
"--channel",
type=ChromeReleaseChannel,
required=False,
default=ChromeReleaseChannel.STABLE,
help=(
"Specify the Chrome release channel to download from. "
"Options: Stable, Beta, Dev, Canary. Defaults to Stable."
),
)
parser.add_argument(
"-o",
"--output",
type=str,
required=False,
default=".bin",
help=("Specify the output directory where the Chrome Headless build will be extracted. Defaults to '.bin'."),
)
args = parser.parse_args()
main(browser_platform=args.browser_platform, channel=args.channel, output_path=args.output) |
|
Hi, I am using below docker file FROM python:3.10 WORKDIR /app Update and install dependenciesRUN apt-get update && Install chrome-headless-shell using npxRUN npx @puppeteer/browsers install chrome-headless-shell@stable Copy necessary filesCOPY ./img-obj_100px_300px.html ./ Install Python dependenciesRUN pip install html2image==2.0.5 fastapi==0.115.6 uvicorn==0.28.0 httpx==0.27.0 Expose a port if your application listens on oneEXPOSE 8080 Define entrypoint commandCMD ["python3", "./htmlimage.py"] =========================================================== and for generating screenshot, i am using hti = Html2Image(custom_flags=['--virtual-time-budget=10000', '--hide-scrollbars', '--no-sandbox', '--disable-font-subpixel-positioning', '--disable-gpu', '--disable-web-security', '--enable-font-antialiasing', '--disable-dev-shm-usage'], size=(width, height), output_path=destination_folder_files, temp_path= destination_folder_files, keep_temp_files=False, browser_executable = "chrome-headless-shell") hti.screenshot(html_str=modified_html, save_as=screenshot_name) =========================================================== Can you please help me with a full working code, this is super urgent as my production enviroment is suddenly down. Please let me know if any other information is required. |
|
Hey @suman-saurav! I believe the problem is in the path you are setting for the executable. This code npx @puppeteer/browsers install chrome-headless-shell@stableOn MacOS, running today for example, it saves in Depending on the operating system it will save in a different place. In my script above, you can specify the folder where the binary will be saved. python download-chrome-headless-shell.py -o /path/to/folder/After you can use the path hti = Html2Image(
browser_executable="/path/to/folder/chrome-headless-shell/chrome-headless-shell"
) |
|
Another option is to use an older version of Chromium. See this StackOverflow answer. |
|
As a temporary workaround, you may set the from html2image import Html2Image
hti = Html2Image()
hti.browser.use_new_headless = None # defaults to `False`, setting it to `True` also works
hti.screenshot(url='http://www.python.org', save_as='pyorg.png')Details on what's going on behind the scene : hti.browser.print_command = True
hti.browser.use_new_headless = False # default value
hti.screenshot(url='http://www.python.org', save_as='pyorg.png')
# Runs chrome.exe --headless=old
hti.browser.use_new_headless = True
hti.screenshot(url='http://www.python.org', save_as='pyorg.png')
# Runs chrome.exe --headless=new
hti.browser.use_new_headless = None
hti.screenshot(url='http://www.python.org', save_as='pyorg.png')
# Runs chrome.exe --headlessI'll look into future proofing this |
|
@vgalin this actually worked until some previous versions of Chrome. But now Chrome has removed the old headless. There is no way (not even by passing the flag) to use the old headless. |
|
@vgalin did you update Chrome? Try updating it and running the code (I think it’ll break). |
|
@satheler chrome://settings/help declares Chrome up to date (Version 132.0.6834.160 (Official Build) (64-bit)).
Do you happen to have different results? |
Old Headless mode has been removed from the Chrome binary. Please use the new Headless mode (https://developer.chrome.com/docs/chromium/new-headless) or the chrome-headless-shell which is a standalone implementation of the old Headless mode (https://developer.chrome.com/blog/chrome-headless-shell).
The text was updated successfully, but these errors were encountered: