From 31edffb4c5c0f2847f4a4da68f1b570b4b244d45 Mon Sep 17 00:00:00 2001
From: Mustafa Gezen <mustafa@rockylinux.org>
Date: Sun, 7 May 2023 03:44:56 +0200
Subject: [PATCH] Add osv-gcs-exporter and fix before/after for OSV API (#6)

---
 apollo/osv-gcs-exporter/__main__.py      | 77 ++++++++++++++++++++++++
 apollo/osv-gcs-exporter/requirements.txt |  1 +
 apollo/server/routes/api_osv.py          |  6 +-
 3 files changed, 82 insertions(+), 2 deletions(-)
 create mode 100644 apollo/osv-gcs-exporter/__main__.py
 create mode 100644 apollo/osv-gcs-exporter/requirements.txt

diff --git a/apollo/osv-gcs-exporter/__main__.py b/apollo/osv-gcs-exporter/__main__.py
new file mode 100644
index 0000000..45fc2ed
--- /dev/null
+++ b/apollo/osv-gcs-exporter/__main__.py
@@ -0,0 +1,77 @@
+"""
+Export OSV data to GCS.
+
+Individual JSON files are exported to the following path:
+    gs://<BUCKET>/<ADVISORY_ID>.json
+"""
+import argparse
+import requests
+import json
+from datetime import datetime, timedelta
+
+from google.cloud import storage
+
+
+def main(args):
+    # Connect to GCS
+    client = storage.Client()
+    bucket = client.get_bucket(args.bucket)
+
+    time_after = None
+    if args.nightly:
+        time_after = (datetime.now() - timedelta(hours=26)).isoformat("T")
+
+    page = 1
+    while True:
+        r = requests.get(
+            "https://apollo.build.resf.org/api/v3/osv",
+            params={
+                "page": page,
+                "limit": 100,
+                "after": time_after,
+            },
+            timeout=60,
+        )
+        r.raise_for_status()
+        advisories = r.json()["advisories"]
+        if not advisories:
+            break
+
+        for advisory in advisories:
+            advisory_json = json.dumps(advisory)
+
+            file_name = f"{advisory['id']}.json"
+
+            # Skip if file already exists
+            blob = bucket.blob(file_name)
+            if blob.exists():
+                print(f"Skipping {file_name} (already exists)")
+                continue
+
+            # Upload to GCS
+            blob = bucket.blob(file_name)
+            blob.upload_from_string(advisory_json)
+            print(f"Uploaded {file_name}")
+
+        page += 1
+
+    print("Done!")
+
+
+if __name__ == "__main__":
+    # Parse arguments
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--bucket",
+        help="GCS bucket to export to",
+        required=True,
+    )
+    # Flag for nightly job
+    parser.add_argument(
+        "--nightly",
+        help="Run as nightly job",
+        action="store_true",
+    )
+
+    # Send args to main
+    main(parser.parse_args())
diff --git a/apollo/osv-gcs-exporter/requirements.txt b/apollo/osv-gcs-exporter/requirements.txt
new file mode 100644
index 0000000..d265caf
--- /dev/null
+++ b/apollo/osv-gcs-exporter/requirements.txt
@@ -0,0 +1 @@
+google-cloud-storage==2.8.0
\ No newline at end of file
diff --git a/apollo/server/routes/api_osv.py b/apollo/server/routes/api_osv.py
index 950b6ab..54b80e6 100644
--- a/apollo/server/routes/api_osv.py
+++ b/apollo/server/routes/api_osv.py
@@ -1,3 +1,5 @@
+import datetime
+
 from typing import TypeVar, Generic, Optional
 
 from fastapi import APIRouter, Depends
@@ -236,8 +238,8 @@ def to_osv_advisory(ui_url: str, advisory: Advisory) -> OSVAdvisory:
 async def get_advisories_osv(
     params: Params = Depends(),
     product: Optional[str] = None,
-    before: Optional[str] = None,
-    after: Optional[str] = None,
+    before: Optional[datetime.datetime] = None,
+    after: Optional[datetime.datetime] = None,
     cve: Optional[str] = None,
     synopsis: Optional[str] = None,
     keyword: Optional[str] = None,