.gitlab-ci.yml

# This file is a template, and might need editing before it works on your project.
# This template uses the java:8 docker image because there isn't any
# official Gradle image at this moment
#
# This is the Gradle build system for JVM applications
# https://gradle.org/
# https://github.com/gradle/gradle
image: java:8


variables:
  VERSION: "0.1.7"

# Make the gradle wrapper executable. This essentially downloads a copy of
# Gradle to build the project with.
# https://docs.gradle.org/current/userguide/gradle_wrapper.html
# It is expected that any modern gradle project has a wrapper
before_script:
  - chmod +x gradlew

# We redirect the gradle user home using -g so that it caches the
# wrapper and dependencies.
# https://docs.gradle.org/current/userguide/gradle_command_line.html
#
# Unfortunately it also caches the build output so
# cleaning removes reminants of any cached builds.
# The assemble task actually builds the project.
# If it fails here, the tests can't run.
build:
  stage: build
  script:
    - ./gradlew -g /cache/.gradle clean assemble
  allow_failure: false
  artifacts:
    paths:
    - gradle/build/artifacts/mbeddr-arduino/
    expire_in: 1 week

build lib with GCC 4.9:
  stage: build
  image: coolya/gcc-avr:4.9
  artifacts:
    paths:
    - lib/build/
    expire_in: 1 week
  script:
    - cd lib
    - make

build lib with GCC 5.4:
  stage: build
  image: coolya/gcc-avr:5.4
  artifacts:
    paths:
    - lib/build/
    expire_in: 1 week
  script:
    - cd lib
    - make

# Use the generated build output to run the tests.
test:
  stage: test
  script:
    - ./gradlew -g /cache./gradle check

deploy github release:
  stage: deploy
  before_script:
    # Install ssh-agent if not already installed, it is required by Docker.
    # (change apt-get to yum if you use a CentOS-based image)
    - 'which ssh-agent || ( apt-get update -y && apt-get install openssh-client zip -y )'
    # Run ssh-agent (inside the build environment)
    - eval $(ssh-agent -s)
    # For Docker builds disable host key checking. Be aware that by adding that
    # you are suspectible to man-in-the-middle attacks.
    # WARNING: Use this only with the Docker executor, if you use it with shell
    # you will overwrite your user's SSH config.
    - mkdir -p ~/.ssh
    - '[[ -f /.dockerenv ]] && echo -e "Host *\n\tStrictHostKeyChecking no\n\n" > ~/.ssh/config'
    - git remote add github https://$GITHUB_TOKEN@github.com/coolya/mbeddr.arduino.git

  after_script:
    - git remote remove github

  script:
  - mkdir tmp
  - cd tmp
  - wget https://github.com/aktau/github-release/releases/download/v0.6.2/linux-amd64-github-release.tar.bz2
  - tar -xjvf linux-amd64-github-release.tar.bz2
  - mv bin/linux/amd64/github-release /usr/bin/
  - cd ..
  - rm -rf tmp/
  - git tag $VERSION HEAD
  - git push github HEAD:release
  - git push github $VERSION
  - github-release release -s $GITHUB_TOKEN -u coolya -r mbeddr.arduino -t $VERSION
  - github-release upload -s $GITHUB_TOKEN  -u coolya -r mbeddr.arduino -t $VERSION -f gradle/build/artifacts/mbeddr-arduino/mbeddr-arduino.zip -n mbeddr-arduino.zip
  only:
    - release

deploy github:
  stage: deploy
  before_script:
    # Install ssh-agent if not already installed, it is required by Docker.
    # (change apt-get to yum if you use a CentOS-based image)
    - 'which ssh-agent || ( apt-get update -y && apt-get install openssh-client -y )'
    # Run ssh-agent (inside the build environment)
    - eval $(ssh-agent -s)
    # For Docker builds disable host key checking. Be aware that by adding that
    # you are suspectible to man-in-the-middle attacks.
    # WARNING: Use this only with the Docker executor, if you use it with shell
    # you will overwrite your user's SSH config.
    - mkdir -p ~/.ssh
    - '[[ -f /.dockerenv ]] && echo -e "Host *\n\tStrictHostKeyChecking no\n\n" > ~/.ssh/config'
    - git remote add github https://$GITHUB_TOKEN@github.com/coolya/mbeddr.arduino.git

  after_script:
    - git remote remove github
  only:
    - master
  script:
      - git push github HEAD:master