https://github.com/nextstrain/conda-base
Conda package build for nextstrain-base
https://github.com/nextstrain/conda-base
conda mamba nextstrain
Last synced: about 2 months ago
JSON representation
Conda package build for nextstrain-base
- Host: GitHub
- URL: https://github.com/nextstrain/conda-base
- Owner: nextstrain
- Created: 2022-10-14T23:55:50.000Z (almost 3 years ago)
- Default Branch: main
- Last Pushed: 2024-11-22T00:36:13.000Z (11 months ago)
- Last Synced: 2024-11-22T01:19:35.546Z (11 months ago)
- Topics: conda, mamba, nextstrain
- Language: Shell
- Homepage: https://anaconda.org/Nextstrain/nextstrain-base
- Size: 89.8 KB
- Stars: 1
- Watchers: 11
- Forks: 1
- Open Issues: 11
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Conda package for nextstrain-base
[](https://anaconda.org/nextstrain/nextstrain-base)
[](https://anaconda.org/nextstrain/nextstrain-base)
This is the source for creating the `nextstrain-base` Conda package.
This meta-package depends on all the other packages needed for a base
Nextstrain runtime installed as a Conda environment. As the nextstrain/base
image is to Nextstrain CLI's Docker runtime, this nextstrain-base package is to
Nextstrain CLI's Conda runtime. The package's dependencies which completely
lock its full transitive dependency tree. This means that if version _X_ of
nextstrain-base worked some way in the past, it'll work the same way in the
future.
Note that this is not a general purpose package for installing Nextstrain.
It's intended for use by Nextstrain CLI's managed Conda runtime and may be
unsuitable for use in a user-managed Conda environment.
## How it works
The meta-package source recipe is in `src/meta.yaml`. This is a [conda-build
recipe spec][]. It defines our runtime's direct dependencies, typically
without version restrictions (pins) or with only loose pinning as necessary.
The recipe declares [`pin_depends: strict`][] in order to fully lock its
dependency tree at package build time.
As the fully locked dependency trees are platform-specific, [CI][] produces
packages for both Linux and macOS (i.e. for Conda's `linux-64` and `osx-64`
subdirs).
[conda-build recipe spec]: https://docs.conda.io/projects/conda-build/en/stable/resources/define-metadata.html
[`pin_depends: strict`]: https://docs.conda.io/projects/conda-build/en/latest/resources/define-metadata.html#pin-runtime-dependencies
## Developing
_You can build this package locally during development, but it's important for
production releases to happen via CI so packages are built for both Linux and
macOS._
First, setup a Conda environment for development in `.dev-env/` so that
[conda-build](https://docs.conda.io/projects/conda-build/) is available.
./devel/setup
You only need to do this once, or whenever you want to refresh your development
environment. Either [Micromamba][], [Mamba][], or [Conda][] must be available
for setup to succeed. You do not need to create a new environment yourself,
the script automatically sets everything up without interfering with your
existing environments.
[Micromamba]: https://mamba.readthedocs.io/page/user_guide/micromamba.html
[Mamba]: https://mamba.readthedocs.io
[Conda]: https://docs.conda.io/projects/conda/
### Building
To build this package locally, run:
./devel/build
The final built package will be written to
`build//nextstrain-base-*.conda`, where `` is a Conda subdir, i.e.
`linux-64`, `osx-64` or `osx-arm64`.
[CI][] builds store the entire `build/` directory as an artifact attached to
each CI run. You can download the artifacts to inspect the built packages or
install them.
[CI]: https://github.com/nextstrain/conda-base/actions/workflows/ci.yaml
### Installing
To install the built package into a new environment, run:
mamba create \
--prefix /path/to/new/env \
--strict-channel-priority \
--override-channels \
--channel conda-forge \
--channel bioconda \
--channel ./build/ \
nextstrain-base
### Uploading
To upload the built package to anaconda.org, run:
./devel/upload
You'll need an appropriate Anaconda API token set in the `ANACONDA_TOKEN`
environment variable. The token must have at least the `api:read`,
`api:write`, and `conda` scopes attached to it. CI uses a token issued for the
[Nextstrain Anaconda organization](https://anaconda.org/Nextstrain/settings/access).
You can adjust the label applied to the uploaded package by setting the `LABEL`
environment variable. By default the Git ref is used to determine the label:
- Uploads from our `main` branch are given the `main` label. These packages
will be found by default for anyone using our Anaconda channel (e.g.
`--channel nextstrain`).
- Uploads from other branches, tags, and PRs will get `branch-`,
`tag-`, and `pull-` labels. These packages can be used by
asking for them explicitly (e.g. `--channel nextstrain/label/pull-123`).
If no label can be worked out, then `dev` is used as a final fallback.
[CI][] uploads the built package if it passes the test phase. You can use the
above labels to install CI-uploaded packages locally without downloading the CI
artifacts.
### Repository layout
`src/` contains the package recipe source. Any files in this directory will be
automatically included in the built package. (`meta.yaml` doesn't live in the
top-level of the repo, and thus `src/` exists, to avoid including the whole
repo in each built package.)
`build/` contains build outputs not tracked in version control.
`devel/` contains programs for development of this package.
`.condarc` is used to specify configuration (e.g. channels) for `conda build`.
## History
- [Initial suggestion of the meta-package][1] on the Nextstrain CLI PR
introducing the Conda runtime.
- [Further discussion about the meta-package][2] on Slack, motivated by the
Conda runtime [breaking between one day and the next][3] due to upstream
changes.
[1]: https://github.com/nextstrain/cli/pull/218#issuecomment-1269082344
[2]: https://bedfordlab.slack.com/archives/C01LCTT7JNN/p1665599068266849
[3]: https://bedfordlab.slack.com/archives/C01LCTT7JNN/p1665594330478279