improve speed by baking the catalogs into the docker image

Author: N-Clerkx

.github/workflows/deploy_function.yml

@@ -47,8 +47,10 @@ name: Build and Deploy Report function to Cloud Run
 on:
   push:
     branches:
-      - "main"
-      - report-pdf-cloud-run
+      - main
+  pull_request:
+    branches:
+      - main
 
 env:
   PROJECT_ID: dgds-i1000482-002
@@ -97,6 +99,7 @@ jobs:
       # END - Docker auth and build
 
       - name: Deploy to Cloud Run
+        if: github.ref == 'refs/heads/main' && github.event_name == 'push'
         id: deploy
         uses: google-github-actions/deploy-cloudrun@v2
         with:
@@ -111,4 +114,5 @@ jobs:
 
       # If required, use the Cloud Run url output in later steps
       - name: Show Output
+        if: github.ref == 'refs/heads/main' && github.event_name == 'push'
         run: echo ${{ steps.deploy.outputs.url }}

App/functions/report-python-cloud-run/.dockerignore

@@ -8,3 +8,4 @@
 *.pyj
 *.pyx
 *.pyd
+data/catalogs

App/functions/report-python-cloud-run/Dockerfile

@@ -48,4 +48,21 @@ RUN apt-get update && \
 # Place executables in the environment at the front of the path
 ENV PATH="/app/.venv/bin:$PATH"
 
+# Clone the coclicodata and global-coastal-atlas repositories STAC catalogs
+RUN mkdir -p /app/data/catalogs
+RUN cd /app/data/catalogs && \
+    git clone -n --depth=1 --filter=tree:0 https://github.com/openearth/coclicodata.git && \
+    cd coclicodata && \
+    git sparse-checkout set --no-cone /current && \
+    git checkout && \
+    cd /app/data/catalogs && \
+    git clone -b subsidence_etienne -n --depth=1 --filter=tree:0 https://github.com/openearth/global-coastal-atlas.git && \
+    cd global-coastal-atlas && \
+    git sparse-checkout set --no-cone /STAC/data/current && \
+    git checkout && \
+    cd /app
+
+ENV STAC_ROOT_DEFAULT="./data/catalogs/global-coastal-atlas/STAC/data/current/catalog.json"
+ENV STAC_COCLICO="./data/catalogs/coclicodata/current/catalog.json"
+
 CMD uv run --with gunicorn gunicorn --bind :8080 --workers 1 --threads 8 --timeout 0 main:app

App/functions/report-python-cloud-run/data/.gitignore

@@ -0,0 +1 @@
+catalogs/

App/functions/report-python-cloud-run/report/report.py

@@ -1,6 +1,7 @@
 # %%
 from dataclasses import dataclass
 from io import BytesIO
+import os
 from pathlib import Path
 
 # import fitz  # type: ignore
@@ -19,9 +20,15 @@
 from datetime import datetime
 
 POLYGON_DEFAULT = """{"coordinates":[[[2.3915028831735015,51.7360381463356],[5.071438932343227,50.89406012060684],[6.955992986278972,51.49577449585874],[7.316959036046541,53.18700330195111],[6.636226617140238,53.961350092621075],[3.8631377106468676,54.14643052276938],[2.1218958391276317,53.490771261555096],[2.3915028831735015,51.7360381463356]]],"type":"Polygon"}"""
-STAC_ROOT_DEFAULT = "https://raw.githubusercontent.com/openearth/global-coastal-atlas/subsidence_etienne/STAC/data/current/catalog.json"
-STAC_COCLICO = (
-    "https://raw.githubusercontent.com/openearth/coclicodata/main/current/catalog.json"
+# Get stac catalog location from environment variable if available otherwise use default
+# this is used to use local catalog instead of remote versions in deployed versions
+STAC_ROOT_DEFAULT = os.getenv(
+    "STAC_ROOT_DEFAULT",
+    "https://raw.githubusercontent.com/openearth/global-coastal-atlas/subsidence_etienne/STAC/data/current/catalog.json",
+)
+STAC_COCLICO = os.getenv(
+    "STAC_COCLICO",
+    "https://raw.githubusercontent.com/openearth/coclicodata/main/current/catalog.json",
 )