.travis.yml

language: python
python:
    - 2.7

virtualenv:
  system_site_packages: true

branches:
  only:
    - master

before_install:
  # Use miniconda and conda packages to speed up dependency setup (principally 
  # borrowed from https://gist.github.com/dan-blanchard/7045057
  # and https://github.com/Jorge-C/ordination/blob/master/.travis.yml
  - export LD_PRELOAD=$LD_PRELOAD:/usr/lib/x86_64-linux-gnu/libgfortran.so.3
  # The preceding line was added to solve an issue with travis where we
  # got errors like the following:
  # $ nosetests -v --with-coverage pele
  # Failure: ImportError (/home/travis/miniconda/envs/pelenv/bin/../lib/libgfortran.so.3:
  # version `GFORTRAN_1.4' not found (required by
  # /home/travis/miniconda/envs/pelenv/lib/python2.7/site-packages/
  # pele/transition_states/_orthogoptf.so)) ... ERROR
  # It seems that the solution proposed in the link below fixes this:
  # http://code-saturne.org/forum/viewtopic.php?f=3&t=1687&sid=9912a7abac3b92c1b067fc495521ed7a&start=10
  - sudo apt-get update -qq
  - sudo apt-get install -qq libatlas-dev libatlas-base-dev liblapack-dev gfortran python-qt4 lcov
  - gem install coveralls-lcov
  - wget http://repo.continuum.io/miniconda/Miniconda-latest-Linux-x86_64.sh -O miniconda.sh
  - chmod +x miniconda.sh
  - ./miniconda.sh -b -p $HOME/miniconda
  - export PATH="$HOME/miniconda/bin:$PATH"
  # Update conda itself
  - hash -r
  - conda config --set always_yes yes --set changeps1 no
  - conda update conda
  # Useful for debugging any issues with conda
  - conda info -a
  # Additional channel for pyqt4
  - conda config --add channels asmeurer
  # matplotlib needs a DISPLAY
  - "export DISPLAY=:99.0"
  - "sh -e /etc/init.d/xvfb start"


install:
  - conda create --yes -n pelenv python=$TRAVIS_PYTHON_VERSION nose pip cython numpy scipy matplotlib qt pyqt sqlalchemy networkx
  - source activate pelenv
  - pip install munkres hungarian pyro4 brewer2mpl pyopengl coverage coveralls coveralls-merge
  - python setup.py build --fcompiler=gfortran
  - python setup.py install
  - cp .coveragerc $HOME
  - pwd
  - sourcedir=$PWD
  # build the c++ tests
  - cd cpp_tests
  - cpp_test_dir=$PWD
  - ./get_gtest.sh
  - cppbuild=$PWD/build
  - mkdir $cppbuild && cd $cppbuild
  - cmake ../source -DCMAKE_BUILD_TYPE=Coverage
  - make

script:
  # avoid running in the checkout directory so nose finds built modules..
  - rundir=$HOME
  - cd $rundir
  - nosetests -v --with-coverage pele
  # do the c++ tests
  - cd $cppbuild
  - ./test_main

after_success:
  # generate the cpp coverage report
  - cd $cppbuild
  - lcov --capture --directory . -b . --output-file coverage.info
  # Artificially change the directory of the c++ source files to the python
  # site-packages directory.  This is purely so that it has the same root
  # directory as the python files so that the coveralls website can find the
  # source on github. This is an ugly hack, there must be a better way.
  - site_packages_dir=$(dirname `python -c "import pele; print pele.__file__"`)
  - site_packages_dir=$(readlink -f $site_packages_dir/..) # go one dir higher and clean up the dir name
  - sed -i "s=$sourcedir=$site_packages_dir=g" coverage.info
  # coverall-lcov requires the source files to exist, so copy them over
  - cp -r $sourcedir/source $site_packages_dir
  # remove unwanted directories from the coverage report
  - lcov --remove coverage.info 'tests/*' '/usr/*' --output-file coverage.info.cleaned
  - coveralls-lcov -v -n coverage.info.cleaned > coverage.c.json
  - cat coverage.c.json # testing
  # need to be in source directory to call coverage
  - cd $sourcedir # not sure if this is necessary
  - cp $rundir/.coverage .
  - coveralls-merge $cppbuild/coverage.c.json