An automatic web page screenshot service.
Some links are alive, but they have already died;
Some links have died, but they are still alive.
Demo services are not guaranteed to work - they run the latest code from the master branch. Feel free to play with them, but don't take up too much resources.
- Telegram bot demo: @ScreenShooterDemoBot
The pre-built Docker image is very large as it contains noto fonts.
Currently supported:
Screenshot Actuator:
- Chromium (
HeadlessChromeActuator)
Messenging Connector:
- Telegram (
TelegramBotConnector)
- Minimal free memory 512MiB (Chrome eats memory, you know)
- dotnet core 2.1
- Manually install all dependencies of a chromium
On machine with <4GiB memory, please set a smaller ParallelJobs and use a >2GiB swap.
A complete example of deployment can be found at wiki/Deploy-Example
Launch: dotnet ScreenShooter.dll --config path/to/your/config.toml
Config file example:
[GlobalConfig]
# max parallel jobs
ParallelJobs=1
# do not retain saved file on the server
RemoveLocalFile=true
[Actuator]
[[Actuator.HeadlessChromeActuator]]
WindowWidth=1920
WindowHeight=1080
# Wait on initial page load
PageDownloadTimeout=30000
# Wait after page load
DocumentLoadDelay=500
# Wait on every scroll event
PageScrollActionWaitDelay=2000
# Wait after scroll (for lazy loaded images to load)
ExtraDownloadWaitDelay=10000
# How many characters of the page title you want to prepend to the file name
MaxTitlePrependLength=32
[Connector]
[[Connector.TelegramBotConnector]]
ApiKey="your-api-key"
MaxUploadRetries=3
Administrators = [ userid1, userid2, ... ]Note that you can write multiple [[Connector.x]] to get multiple different instances of them.
If this project is helpful to you, please consider buying me a coffee.
