infologistix/docker-selenium-python repository overview

⁠docker-selenium-python

Docker base image to run python based scripts for selenium in one container. Compiled on top of base python image build on alpine. This method uses chromium in headless.

⁠Usage

Build docker-image from scratch. This may take a while, because numpy and pandas must be compiled from source. This is indeed a point, where improvements can begin.

Basic requirements are pandas, numpy and selenium. Simply change your requirements to your needs.

• Create a python script to trigger selenium based events or crawl the web
• Copy your files to the specified locations
• Run docker run -it --rm --shm-size=128m infologistix/docker-selenium-python:latest python main.py

Intended usage is for crawling scheduled sites on AWS ECS and Azure ACR. Simply build the container as base and add your scripts to it. Add --shm.size=128m to it to ensure running correctly. See Docker Run specifications. For some webpages Chrome needs extra memory.

⁠Example

To run the given example execute the following:

docker run -it --rm --shm-size=128m infologistix/docker-selenium-python:latest python ./examples/main.py

This example scrapes our homepage and writes the services of our company human readable to console. This can be extended using a MS Teams Channel do be sent to. Documentation can be found here⁠.

⁠Advanced Usage

Triggering your script directly at the start of the container requires you to fiddle with the Dockerfile. Simply add your requirements to a requirements.txt and copy your source files to the image.

We provide an example script located in examples folder. Use this as a reference guide provided.

FROM infologistix/docker-selenium-python:alpine

WORKDIR /usr/src/app

COPY requirements.txt ./
RUN pip install --no-cache-dir -r requirements.txt

COPY your/app/ ./

CMD "python main.py"

Simply build this file to suit your applicational use.

When using Machine learning tools use one of our other selenium based Containers by changing the tag:

• bullseye, latest, 3.10, 3.10-bullseye
• alpine, 3.10-alpine

For Machine Learning applications we recommend the bullseye tag to install every library there is. Make sure to install only necessary things.

This image was developed for infologistix GmbH⁠