Build angular project in Dockerfile-CodePudding

i have Angular/Node Project in Gitlab.. I Have Dockerfile

FROM node:12 as builder    
WORKDIR /app
COPY package.json package-lock.json ./
ENV CI=1
RUN npm ci    
COPY . .
RUN npm run build-web --output-path=/dist

this Dockerfile is for Master branch , i want if i push code in develop branch, want to define in Dockerfile how to build , for example if i push in develop , in dockerfile want be

RUN npm run build-dev

How i can do this ? can i use if else statement in dockerfile ? or yaml

CodePudding user response：

Technically you should use single artefact and have it take the configuration at runtime. We do that using angular & environments with a envsubs script on the launch.

You could also check this SO question which makes a good example of it.

We are using:

Inside index.html you add script source to /assets/env.js , which gets pulled on page load from static files inside container (assets) and in container runtime script (ENTRYPOINT) just execute:

#!/bin/sh

# window env substitutions
envsubst < /usr/share/nginx/html/de-DE/assets/data/env.template.js > /usr/share/nginx/html/de-DE/assets/data/env.js
envsubst < /usr/share/nginx/html/en-US/assets/data/env.template.js > /usr/share/nginx/html/en-US/assets/data/env.js

exec "$@"

which then fills env.js with env.template.js (which has $PLACEHOLDER signs, which are overriden by env variables of container).

and then load it into window object.

And to even make it type safe we created:

export interface RuntimeEnvironmentInterface {
  baseUrl: string;
}

which then gets global override on the Window object:

declare global {
  interface Window {
    envVariables: RuntimeEnvironmentInterface;
  }
}

And inside application's environment settings (environment.prod.ts for angular):

export const environment: EnvironmentInterface = {
  production: true,
  api: window.envVariables.baseUrl   '/api/v1',
};

We use it in production for almost two years now and it works as a charm (knocking on the wood)

CodePudding user response：

workflow:
  rules:
    - if: $CI_COMMIT_BRANCH == "develop"
      variables:
        SSH_IP: "[email protected]"
        ENVIRONMENT: "develop"
    - if: $CI_COMMIT_BRANCH == "master"
      variables:
        SSH_IP: "[email protected]"
        ENVIRONMENT: "production"

I achieve this by defining a variable in my CI file that checks for which environment i'm trying to build, based on the branch i'm committing from (You could change this to fit your situation).

A bit unrelated if you specifically require the image locally, but might be the solution you're looking for if you're trying to then deploy the image somewhere.