Set up reproducible Python environments with conda-lock

.github/workflows/tox-pytest.yml

@@ -35,13 +35,10 @@ jobs:
       - name: Install Conda environment using mamba
         uses: mamba-org/setup-micromamba@v1
         with:
-          environment-file: test/test-environment.yml
+          environment-file: environments/conda-lock.yml
+          environment-name: pudl-dev
           cache-environment: true
-          condarc: |
-            channels:
-            - conda-forge
-            - defaults
-            channel_priority: strict
+          create-args: --category main dev docs test datasette
 
       - name: Log environment details
         run: |
@@ -78,13 +75,10 @@ jobs:
       - name: Install Conda environment using mamba
         uses: mamba-org/setup-micromamba@v1
         with:
-          environment-file: test/test-environment.yml
+          environment-file: environments/conda-lock.yml
+          environment-name: pudl-dev
           cache-environment: true
-          condarc: |
-            channels:
-            - conda-forge
-            - defaults
-            channel_priority: strict
+          create-args: --category main dev docs test datasette
 
       - name: Log environment details
         run: |
@@ -131,13 +125,10 @@ jobs:
       - name: Install Conda environment using mamba
         uses: mamba-org/setup-micromamba@v1
         with:
-          environment-file: test/test-environment.yml
+          environment-file: environments/conda-lock.yml
+          environment-name: pudl-dev
           cache-environment: true
-          condarc: |
-            channels:
-            - conda-forge
-            - defaults
-            channel_priority: strict
+          create-args: --category main dev docs test datasette
 
       - name: Log environment details
         run: |

.github/workflows/update-lockfile.yml

@@ -0,0 +1,77 @@
+---
+name: update-lockfile
+
+on:
+  workflow_dispatch:
+  #schedule:
+  #  - cron: "0 9 * * 1-5" # Weekdays at 9AM UTC
+  #pull_request:
+  #  paths:
+  #   - "pyproject.toml"
+
+jobs:
+  conda-lock:
+    # Don't run scheduled job on forks.
+    if: (github.event_name == 'schedule' && github.repository == 'catalyst-cooperative/pudl') || (github.event_name != 'schedule')
+    defaults:
+      run:
+        shell: bash -l {0}
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+        # If running on a schedule, run on dev.
+        # If running from workflow_dispatch, run on whatever the chosen branch/ref was.
+        # with:
+        #   ref: dev
+      - name: Install Micromamba
+        uses: mamba-org/setup-micromamba@v1
+        with:
+          environment-name: conda-lock
+          create-args: >-
+            python=3.11
+            conda-lock
+
+      - name: Run conda-lock to recreate lockfile from scratch
+        run: |
+          cd environments
+          rm conda-lock.yml
+          conda-lock \
+              --micromamba \
+              --file=../pyproject.toml \
+              --lockfile=conda-lock.yml
+          conda-lock render \
+              --kind explicit \
+              --kind env \
+              --dev-dependencies \
+              --extras docs \
+              --extras datasette \
+              conda-lock.yml
+          conda-lock render \
+              --kind env \
+              --extras docs \
+              --platform linux-64 \
+              --filename-template "readthedocs-{platform}.conda.lock" \
+              conda-lock.yml
+          cd ..
+      - name: Open a pull request
+        uses: peter-evans/create-pull-request@v5
+        with:
+          # # The default GITHUB_TOKEN doesn't allow other workflows to trigger.
+          # # Thus if there are tests to be run, they won't be run. For more info,
+          # # see the note under
+          # # <https://github.com/peter-evans/create-pull-request#action-inputs>.
+          # # One possible workaround is to specify a Personal Access Token (PAT).
+          # # This PAT should have read-write permissions for "Pull Requests"
+          # # and read-write permissions for "Contents".
+          # token: ${{ secrets.GH_PAT_FOR_PR }}
+          commit-message: Update lockfile
+          title: Update Lockfile
+          body: >
+            This pull request relocks the dependencies with conda-lock.
+            It is triggered by [update-lockfile](https://github.com/catalyst-cooperative/pudl/blob/main/.github/workflows/update-lockfile.yml).
+          branch: update-lockfile
+          labels: dependencies, conda-lock
+          reviewers: zaneselvans
+          delete-branch: true
+          # base: dev
+          draft: true

.github/workflows/zenodo-cache-sync.yml

@@ -47,13 +47,10 @@ jobs:
       - name: Install Conda environment using mamba
         uses: mamba-org/setup-micromamba@v1
         with:
-          environment-file: test/test-environment.yml
+          environment-file: environments/conda-lock.yml
+          environment-name: pudl-dev
           cache-environment: true
-          condarc: |
-            channels:
-            - conda-forge
-            - defaults
-            channel_priority: strict
+          create-args: --category main dev docs test datasette
 
       - name: Log environment details
         run: |

.gitignore

@@ -23,8 +23,8 @@ codecov.sh
 .env_pudl/
 *wheel-metadata
 dask-worker-space*
-devtools/user-requirements.txt
-devtools/user-environment.yml
+environments/user-requirements.txt
+environments/user-environment.yml
 .vscode/*
 commit.txt
 devtools/profiles/

.readthedocs.yaml

@@ -14,7 +14,7 @@ build:
 
 # Define the python environment using conda / mamba
 conda:
-  environment: docs/docs-environment.yml
+  environment: environments/readthedocs-linux-64.conda.lock.yml
 
 # Build documentation in the docs/ directory with Sphinx
 sphinx:
@@ -27,5 +27,3 @@ python:
   install:
     - method: pip
       path: .
-      extra_requirements:
-        - doc

docker/Dockerfile

@@ -1,63 +1,54 @@
-FROM condaforge/mambaforge:23.3.1-1
+FROM mambaorg/micromamba:1.5.1
+
+USER root
 
 # Install curl and js
 # awscli requires unzip, less, groff and mandoc
 # hadolint ignore=DL3008
-RUN apt-get update && apt-get install --no-install-recommends -y curl jq unzip less groff mandoc \
-    && apt-get clean \
-    && rm -rf /var/lib/apt/lists/*
+RUN apt-get update && \
+    apt-get install --no-install-recommends -y git curl jq unzip less groff mandoc && \
+    apt-get clean && \
+    rm -rf /var/lib/apt/lists/*
 
 # Configure gsutil authentication
 # hadolint ignore=DL3059
 RUN printf '[GoogleCompute]\nservice_account = default' > /etc/boto.cfg
 
 # Install awscli2
-RUN curl "https://awscli.amazonaws.com/awscli-exe-linux-x86_64.zip" -o "awscliv2.zip" && unzip awscliv2.zip && ./aws/install
-
-# Create a non-root user inside the container
-# hadolint ignore=DL3059
-RUN useradd -Ums /bin/bash catalyst
-
-ENV CONTAINER_HOME=/home/catalyst
+RUN curl "https://awscli.amazonaws.com/awscli-exe-linux-x86_64.zip" -o "awscliv2.zip" && \
+    unzip awscliv2.zip && \
+    ./aws/install
 
-# Switch to being the catalyst user and go into the copied repo
-USER catalyst
+# Switch back to being non-root user and get into the home directory
+USER $MAMBA_USER
+ENV CONTAINER_HOME=/home/$MAMBA_USER
 WORKDIR ${CONTAINER_HOME}
 
 ENV CONDA_PREFIX=${CONTAINER_HOME}/env
 ENV PUDL_REPO=${CONTAINER_HOME}/pudl
-ENV CONDA_RUN="conda run --no-capture-output --prefix ${CONDA_PREFIX}"
-ENV PYTHON_VERSION="3.11"
+ENV CONDA_RUN="micromamba run --prefix ${CONDA_PREFIX}"
 
 ENV CONTAINER_PUDL_WORKSPACE=${CONTAINER_HOME}/pudl_work
-ENV PUDL_INPUT=${CONTAINER_PUDL_WORKSPACE}/data
+ENV PUDL_INPUT=${CONTAINER_PUDL_WORKSPACE}/input
 ENV PUDL_OUTPUT=${CONTAINER_PUDL_WORKSPACE}/output
 ENV DAGSTER_HOME=${CONTAINER_PUDL_WORKSPACE}/dagster_home
 
-# Create data input/output directories
-RUN mkdir -p ${PUDL_INPUT} ${PUDL_OUTPUT} ${DAGSTER_HOME}
-
 # Create a conda environment based on the specification in the repo
-COPY test/test-environment.yml test/test-environment.yml
-RUN mamba create --copy --prefix ${CONDA_PREFIX} --yes python=${PYTHON_VERSION} && \
-    # Then we can use mamba env update, which can parse the environment.yml file:
-    mamba env update --prefix ${CONDA_PREFIX} --file test/test-environment.yml && \
-    conda clean -afy
-
-
+COPY environments/conda-lock.yml environments/conda-lock.yml
+RUN micromamba create --prefix ${CONDA_PREFIX} --yes --category main dev docs test datasette --file environments/conda-lock.yml && \
+    micromamba clean -afy
 # Copy the cloned pudl repository into the user's home directory
-COPY --chown=catalyst:catalyst . ${CONTAINER_HOME}
+COPY --chown=${MAMBA_USER}:${MAMBA_USER} . ${CONTAINER_HOME}
 
 # TODO(rousik): The following is a workaround for sudden breakage where conda
 # can't find libraries contained within the environment. It's unclear why!
 ENV LD_LIBRARY_PATH=${CONDA_PREFIX}/lib
 # We need information from .git to get version with setuptools_scm so we mount that
 # directory without copying it into the image.
 RUN --mount=type=bind,source=.git,target=${PUDL_REPO}/.git \
-    ${CONDA_RUN} pip install --no-cache-dir -e './[dev,doc,test,datasette]' && \
+    ${CONDA_RUN} pip install --no-cache-dir --editable . && \
     # Run the PUDL setup script so we know where to read and write data
     ${CONDA_RUN} pudl_setup
 
-
 # Run the unit tests:
 CMD ["conda", "run", "--no-capture-output", "--prefix", "${CONDA_PREFIX}", "pytest", "test/unit"]