Merge pull request #20 from pangeo-forge/integration-test

Drop custom base image, add integration tests, install runner from release

.github/workflows/test-action.yaml

@@ -0,0 +1,55 @@
+name: Test action
+
+on:
+  push:
+    branches: ["main"]
+  pull_request:
+    branches: [ "*" ]
+
+jobs:
+  test:
+    name: (recipes@${{ matrix.recipes-version }})
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        recipes-version: ["0.10.3"]  # , "main"]
+    steps:
+      - uses: actions/checkout@v4
+      - name: "Clone test feedstock"
+        # Fetches test feedstock (containing example test recipes) from pangeo-forge-recipes
+        run: |
+          git clone --depth 1 --branch ${{ matrix.recipes-version }} https://github.com/pangeo-forge/pangeo-forge-recipes.git
+      - name: "Add requirements.txt"
+        # The clone step above gives us recipe modules and meta.yaml, but does not contain a requirements file,
+        # so we add that here, providing the action with the correct version of pangeo-forge-recipes to install
+        # in the container at action runtime.
+        run: |
+          echo "pangeo-forge-recipes==${{ matrix.recipes-version }}" > pangeo-forge-recipes/examples/feedstock/requirements.txt
+      - name: "Overwrite meta.yaml"
+        # The example feedstock contains multiple recipes, but we just want to test gpcp
+        run: |
+          grep -E 'recipes|gpcp' pangeo-forge-recipes/examples/feedstock/meta.yaml > temp.yaml \
+          && mv temp.yaml pangeo-forge-recipes/examples/feedstock/meta.yaml
+      - name: "Deploy recipes"
+        uses: ./
+        with:
+          # select_recipe_by_label: true
+          pangeo_forge_runner_config: >
+            {
+            "BaseCommand": {
+              "feedstock_subdir": "pangeo-forge-recipes/examples/feedstock"
+            },
+            "Bake": {
+              "prune": true,
+              "bakery_class": "pangeo_forge_runner.bakery.local.LocalDirectBakery"
+            },
+            "TargetStorage": {
+              "fsspec_class": "fsspec.implementations.local.LocalFileSystem",
+              "root_path": "./target"
+            },
+            "InputCacheStorage": {
+              "fsspec_class": "fsspec.implementations.local.LocalFileSystem",
+              "root_path": "./cache"
+            }
+            }

Dockerfile

@@ -1,10 +1,12 @@
-FROM quay.io/pangeo/forge:5e51a29
+FROM ubuntu:22.04
 
-RUN conda run -n notebook pip install git+https://github.com/pangeo-forge/pangeo-forge-runner@main
+RUN apt-get update && apt-get install -y python3 python3-pip
+
+# install pangeo-forge-runner. to make this faster, let's drop the pangeo-forge-recipes
+# dependency from -runner (because recipes is always installed at runtime anyway)
+RUN python3 -m pip install pangeo-forge-runner
 
 COPY action/deploy_recipe.py /deploy_recipe.py
 COPY entrypoint.sh /entrypoint.sh
 
-ENV CONDA_ENV=notebook
-
 ENTRYPOINT [ "sh", "/entrypoint.sh" ]

action/deploy_recipe.py

@@ -27,15 +27,19 @@ def deploy_recipe_cmd(cmd: list[str]):
     """Wrapper for `call_subprocess_run` with extra stdout parsing when deploying recipes."""
     stdout = call_subprocess_run(cmd)
     lastline = json.loads(stdout.splitlines()[-1])
-    job_id = lastline["job_id"]
-    job_name = lastline["job_name"]
-    print(f"Job submitted with {job_id = } and {job_name = }")
+    if all([k in lastline for k in ("job_id", "job_name")]):
+        job_id = lastline["job_id"]
+        job_name = lastline["job_name"]
+        print(f"Job submitted with {job_id = } and {job_name = }")
+    else:
+        print(
+            "Keys 'job_id' and/or 'job_name' missing from deploy recipe process stdout, "
+            "but deploy command did not fail. Perhaps the configured bakery type does not "
+            "provide this logging information?"
+        )
 
 
 def main():
-    # set in Dockerfile
-    conda_env = os.environ["CONDA_ENV"]
-
     # injected by github actions
     repository = os.environ["GITHUB_REPOSITORY"]  # will this fail for external prs?
     api_url = os.environ["GITHUB_API_URL"]
@@ -50,7 +54,6 @@ def main():
     select_recipe_by_label = os.environ["INPUT_SELECT_RECIPE_BY_LABEL"]
 
     # log variables to stdout
-    print(f"{conda_env = }")
     print(f"{head_ref = }")
     print(f"{sha = }")
     print(f"{config = }")
@@ -79,18 +82,18 @@ def main():
     # if calling `pangeo-forge-runner` directly, `--feedstock-subdir` can be passed as a CLI arg.
     # in the action context, users do not compose their own `pangeo-forge-runner` CLI calls, so if
     # they want to use a non-default value for feedstock-subdir, it must be passed via the long-form
-    # name in the config JSON (i.e, `{"BaseCommand": "feedstock-subdir": ...}}`).
+    # name in the config JSON (i.e, `{"BaseCommand": "feedstock_subdir": ...}}`).
     feedstock_subdir = (
-        config["BaseCommand"]["feedstock-subdir"]
-        if "BaseCommand" in config and "feedstock-subdir" in config["BaseCommand"]
+        config["BaseCommand"]["feedstock_subdir"]
+        if "BaseCommand" in config and "feedstock_subdir" in config["BaseCommand"]
         else "feedstock"
     )
     # because we've run the actions/checkout step before reaching this point, our current
     # working directory is the root of the feedstock repo, so we can list feedstock repo
     # contents directly on the filesystem here, without requesting it from github.
     if "requirements.txt" in os.listdir(feedstock_subdir):
         call_subprocess_run(
-            f"mamba run -n {conda_env} pip install -Ur {feedstock_subdir}/requirements.txt".split()
+            f"python3 -m pip install -Ur {feedstock_subdir}/requirements.txt".split()
         )
 
     with tempfile.NamedTemporaryFile("w", suffix=".json") as f:

entrypoint.sh

@@ -1,3 +1,3 @@
 #!/bin/sh -l
 
-conda run --no-capture-output -n notebook python3 /deploy_recipe.py
+python3 /deploy_recipe.py

tests/test_main.py

@@ -27,7 +27,6 @@ def select_recipe_by_label(request):
 @pytest.fixture
 def env(select_recipe_by_label, head_ref):
     return {
-        "CONDA_ENV": "notebook",
         "GITHUB_REPOSITORY": "my/repo",
         "GITHUB_API_URL": "https://api.github.com",
         # fixturing of `head_ref` reflects that on `push`
@@ -56,8 +55,8 @@ class MockCompletedProcess:
     returncode: int
 
 
-@pytest.fixture
-def subprocess_run_side_effect():
+@pytest.fixture(params=[True, False], ids=["has_job_id", "no_job_id"])
+def subprocess_run_side_effect(request):
     def _get_mock_completed_proc(cmd: list[str], *args, **kwargs):
         # `subprocess.run` is called a few ways, so use a side effect function
         # to vary the output depending on what arguments it was called with.
@@ -69,8 +68,11 @@ def _get_mock_completed_proc(cmd: list[str], *args, **kwargs):
                 returncode=returncode,
             )
         elif "bake" in " ".join(cmd):
+            # not all bakery types have a job_id, represent that here
+            has_job_id = request.param
+            stdout = b'{"job_id": "foo", "job_name": "bar"}' if has_job_id else b'{}'
             return MockCompletedProcess(
-                stdout=b'{"job_id": "foo", "job_name": "bar"}',
+                stdout=stdout,
                 stderr=b"",
                 returncode=0,
             )
@@ -134,7 +136,7 @@ def test_main(
 
     if pip_install_raises:
         config: dict = json.loads(env["INPUT_PANGEO_FORGE_RUNNER_CONFIG"])
-        config.update({"BaseCommand": {"feedstock-subdir": "broken-requirements-feedstock"}})
+        config.update({"BaseCommand": {"feedstock_subdir": "broken-requirements-feedstock"}})
         env["INPUT_PANGEO_FORGE_RUNNER_CONFIG"] = json.dumps(config)
 
     with patch.dict(os.environ, env):
@@ -156,8 +158,7 @@ def test_main(
                     else "feedstock"
                 )
                 expected_cmd = (
-                    f"mamba run -n {env['CONDA_ENV']} "
-                    f"pip install -Ur {feedstock_subdir}/requirements.txt"
+                    f"python3 -m pip install -Ur {feedstock_subdir}/requirements.txt"
                 ).split()
                 subprocess_run.assert_any_call(expected_cmd, capture_output=True)
             else: