Draft of noaa-oisst recipe

.github/actions/process_recipe/Dockerfile

@@ -0,0 +1,6 @@
+FROM pangeo/pangeo-notebook:latest
+
+COPY action/process_recipe.py /process_recipe.py
+COPY entrypoint.sh /entrypoint.sh
+
+ENTRYPOINT [ "sh", "/entrypoint.sh" ]

.github/actions/process_recipe/README.md

@@ -0,0 +1,34 @@
+# Process Recipe Action
+
+This actions purpose is to take a Recipe contributed via a PR, convert it to a Prefect Flow, register it with the Bakery specified in its metadata file, and run a test invocation of the Flow with a subset of the input data.
+
+## Inputs
+
+### `path_to_recipe_py`
+
+**Required** The path to the `recipe.py` file within the PR. This is relative to the root of the repository.
+
+### `path_to_meta_yaml`
+
+**Required** The path to the `meta.yaml` file within the PR. This is relative to the root of the repository.
+
+
+## Outputs
+
+N/A
+
+## Example usage
+
+```yaml
+# If using this recipe within the pangeo-forge/staged-recipes repository
+uses: ./.github/actions/process_recipe
+with:
+    path_to_recipe_py: "recipes/my_recipe/recipe.py"
+    path_to_meta_yaml: "recipes/my_recipe/meta.yaml"
+
+# If using this recipe in any other repository
+uses: pangeo-forge/staged-recipes/.github/actions/process_recipe@master
+with:
+    path_to_recipe_py: "recipes/my_recipe/recipe.py"
+    path_to_meta_yaml: "recipes/my_recipe/meta.yaml"
+```

.github/actions/process_recipe/action.yaml

@@ -0,0 +1,19 @@
+name: 'Process Recipe'
+description: '''
+For a given recipe.py and its meta.yaml file: convert it to a Prefect Flow,
+register the Flow with the specified Bakery in the meta.yml, and test a pruned run
+of the Flow on the Bakery.
+'''
+inputs:
+  path_to_recipe_py:
+    description: 'Path to the recipe.py file of the recipe to process'
+    required: true
+  path_to_meta_yaml:
+    description: 'Path to the meta.yaml file of the recipe to process'
+    required: true
+runs:
+  using: 'docker'
+  image: 'Dockerfile'
+  args:
+    - ${{ inputs.path_to_recipe_py }}
+    - ${{ inputs.path_to_meta_yaml }}

.github/actions/process_recipe/action/process_recipe.py

@@ -0,0 +1,7 @@
+import sys
+
+def main():
+    print(f"recipe.py: {sys.argv[1]} meta.yaml: {sys.argv[2]}")
+
+if __name__ == "__main__":
+    main()

.github/actions/process_recipe/entrypoint.sh

@@ -0,0 +1,7 @@
+#!/bin/sh -l
+
+# Process dependencies here that need installing - Assuming that the recipe either has:
+# - Conda environment .yaml file attached
+# - Dependencies in a list in the meta.yaml
+
+python3 /process_recipe.py $1 $2

.github/workflows/process-recipe.yaml

@@ -0,0 +1,46 @@
+name: Process PR Recipe
+
+on:
+  pull_request:
+    branches: master
+
+jobs:
+  process-recipe:
+    runs-on: ubuntu-20.04
+    steps:
+      - uses: actions/checkout@v2
+
+      - id: changed-files
+        name: Get the files changed in the Pull Request
+        uses: jitterbit/get-changed-files@v1
+
+      - name: Locate Pull Requests recipe.py
+        id: recipe
+        run: |
+          for changed_file in ${{ steps.changed-files.outputs.all }}; do
+            if [[ ${changed_file} =~ ^.*recipes\/.*recipe\.py$ ]];
+            then
+              echo "::set-output name=RECIPE_PY::${changed_file}";
+              exit 0
+            fi
+          done
+          exit 1 # Fail - recipe.py required
+
+      - name: Locate Pull Requests meta.yaml
+        id: meta
+        run: |
+          for changed_file in ${{ steps.changed-files.outputs.all }}; do
+            if [[ ${changed_file} =~ ^.*recipes\/.*meta\.(yml|yaml)$ ]];
+            then
+              echo "::set-output name=META_YAML::${changed_file}";
+              exit 0
+            fi
+          done
+          exit 1 # Fail - meta.yaml required
+
+      - name: Process Pull Requests recipe
+        if: ${{ steps.recipe.outputs.RECIPE_PY != null && steps.meta.outputs.META_YAML != null }}
+        uses: ./.github/actions/process_recipe
+        with:
+          path_to_recipe_py: ${{ steps.recipe.outputs.RECIPE_PY }}
+          path_to_meta_yaml: ${{ steps.meta.outputs.META_YAML }}

recipes/noaa-oisst/meta.yaml

@@ -0,0 +1,22 @@
+title: "NOAA Optimum Interpolated SST"
+description: "Analysis-ready Zarr datasets derived from NOAA OISST NetCDF"
+pangeo_forge_version: "0.0.1"  # do we need a separate spec version
+recipes:
+  - id: noaa-oisst-avhrr-only
+    module: recipe.py  # the module where to find the recipe
+    name: recipe  # the name of the object to import
+provenance:
+  providers:
+    - name: "NOAA NCEI"
+      description: "National Oceanographic & Atmospheric Administration National Centers for Environmental Information"
+      roles:
+        - producer
+        - licensor
+      url: https://www.ncdc.noaa.gov/oisst
+  license: "CC-BY-4.0"
+maintainers:
+  - name: "Ryan Abernathey"
+    orcid: "0000-0001-5999-4917"
+    github: rabernat
+bakeries:
+  - id: "pangeo-aws-west-1"  # must come from a valid list of bakeries

recipes/noaa-oisst/recipe.py

@@ -0,0 +1,20 @@
+import pandas as pd
+from pangeo_forge.recipe import NetCDFtoZarrSequentialRecipe
+
+input_url_pattern = (
+    "https://www.ncei.noaa.gov/data/sea-surface-temperature-optimum-interpolation"
+    "/v2.1/access/avhrr/{yyyymm}/oisst-avhrr-v02r01.{yyyymmdd}.nc"
+)
+dates = pd.date_range("1981-09-01", "2021-01-05", freq="D")
+input_urls = [
+    input_url_pattern.format(
+        yyyymm=day.strftime("%Y%m"), yyyymmdd=day.strftime("%Y%m%d")
+    )
+    for day in dates
+]
+
+recipe = NetCDFtoZarrSequentialRecipe(
+    input_urls=input_urls,
+    sequence_dim="time",
+    inputs_per_chunk=20
+)