Generate DB and keep it in repo

.github/workflows/ingest-json-api.yml

@@ -20,3 +20,6 @@ jobs:
           version: '0.33.0'
       - name: Install dependencies
         run: rye sync --no-lock
+      - uses: jorgebg/[email protected]
+      - name: Upsert data from Formula JSON API
+        run: rye run ingest-json-api formula .state/feed.db 'https://formulae.brew.sh/api/formula.json'

.gitignore

@@ -1,3 +1,6 @@
+# Github Action state location
+.state/
+
 # Byte-compiled / optimized / DLL files
 __pycache__/
 *.py[cod]

src/homebrew_new_bot/ingest_json_api.py

@@ -25,7 +25,7 @@ def main() -> None:
     table = typing.cast(sqlite_utils.db.Table, db["feed_items"])
     # Get JSON from API endpoint
     # TODO: use latest added_at from DB and HEAD to check if its even worth getting full thing?
-    r = requests.get("https://formulae.brew.sh/api/formula.json")
+    r = requests.get(args.json_api_url)
     last_modified = email.utils.parsedate_to_datetime(r.headers["last-modified"])
     # TODO: is there a shorter syntax to do this?
     packages = list(