Latest release

You can install hc by invoking the following commands:

gpg --recv-keys 'C505 B5C9 3B0D B3D3 38A1  B600 5FE9 2C12 EE88 E1F0'
mkdir --parent /tmp/hc && cd /tmp/hc
wget -r -nd -l 1 --accept-regex '^https://(test)?pypi\.python\.org/packages/.*\.whl.*'
current_release="$(find . -type f -name '*.whl' | sort | tail -n 1)"
gpg --verify "${current_release}.asc" "${current_release}" && pip3 install --upgrade "${current_release}"

Refer to Verifying PyPI and Conda Packages for more details. Note that this might pull down dependencies in an unauthenticated way! You might want to install the dependencies yourself beforehand.

Or if you feel lazy and agree that pip/issues/1035 should be fixed you can also install hc like this:

pip3 install hc

Development version

If you want to be more on the bleeding edge of hc development consider cloning the git repository and installing from it:

gpg --recv-keys 'EF96 BC32 AC57 CFC7 2DF0  1D8C 489A 4D5E C353 C98A'
git clone --recursive
cd hc && git verify-commit HEAD
echo 'Check if the HEAD commit has a good signature and only proceed in that case!' && read -r fnord
echo 'Then chose one of the commands below to install hc and its dependencies:'
pip3 install --upgrade .
./ develop --user
./ install --user
./ install

This will also get you the cache which is tracked in git as well to do integration testing over the whole dataset. Please be sure to use the cache by symlinking it to your user cache directory. The following should do the trick:

hc_cache="$(python3 -c 'from appdirs import user_cache_dir; print(user_cache_dir("hc"))')"
ln -sT "$PWD/tests/cache/" "$hc_cache"