[Pass parameters from docker run to python] #docker #dockerfile

command.sh

docker run foo:bar --app_id='abc'

Dockerfile

FROM python:3.6

ENV LC_ALL C.UTF-8
ENV LANG C.UTF-8
ENV GOOGLE_APPLICATION_CREDENTIALS /service/scripts/application_default_credentials.json

COPY . /service/scripts
WORKDIR /service/scripts

RUN pip install pipenv
RUN pipenv install --deploy --system

ENTRYPOINT ["python", "/service/scripts/index_image_analysis.py"]

foo.py

import argparse

parser = argparse.ArgumentParser(description='For index images analysis')
parser.add_argument('--app_id', dest='app_id', type=str, help='app_id', metavar='app_id', default='2760')
parser.add_argument('--sample_n', dest='sample_n', type=int, help='number to sample', default=10000)
parser.add_argument('--top_n', dest='top_n', type=int, help='top n distribution', default=10)

args = parser.parse_args()

@parikshit14
Pretty much exactly as shown in the original. Here is my dockerfile:

FROM python:3.8.16-bullseye
COPY /requirements.txt /requirements.txt
RUN pip install -U pip
RUN pip install -r requirements.txt
WORKDIR historic-employee-count-tool
COPY . .
ENTRYPOINT [ "python", "historic-employee-count-tool/main.py"]

And in the python code here is how to access an arg:

key_arg = sys.argv[1]

Remember with argv, 0 is reserved as the name of the script.

@RheingoldRiver how did you executed it docker run ....what....

@parikshit14
You run it with:

docker run -it container-name-here param1 param2 param3

Here's a link to the full repo including a README in case that's helpful too.