feat: adopt timezone-aware timestamps

google-news-drive-sync/.flake8

@@ -0,0 +1,3 @@
+[flake8]
+max-line-length = 100
+extend-ignore = E203

google-news-drive-sync/src/article_repository.py

@@ -5,11 +5,12 @@
 import sqlite3
 from contextlib import contextmanager
 from dataclasses import dataclass
-from datetime import datetime
+from datetime import datetime, timezone
 from pathlib import Path
 from typing import Dict, Iterable, Iterator, List, Optional
 
 from .news_fetcher import NewsArticle
+from .utils import utcnow
 
 
 @dataclass(frozen=True)
@@ -73,13 +74,20 @@ def _article_id(self, article: NewsArticle) -> str:
         return article.url or article.title
 
     def persist(self, articles: Iterable[NewsArticle]) -> None:
-        now = datetime.utcnow().isoformat()
+        now = utcnow().isoformat()
         payload: List[tuple[str, str, str | None, str, str | None, str | None, str]] = []
         for article in articles:
             identifier = self._article_id(article)
             if not identifier:
                 continue
-            published = article.published_at.isoformat() if article.published_at else None
+            published = None
+            if article.published_at is not None:
+                published_dt = article.published_at
+                if published_dt.tzinfo is None:
+                    published_dt = published_dt.replace(tzinfo=timezone.utc)
+                else:
+                    published_dt = published_dt.astimezone(timezone.utc)
+                published = published_dt.isoformat()
             payload.append(
                 (
                     identifier,

google-news-drive-sync/src/cache.py

@@ -5,11 +5,12 @@
 import hashlib
 import sqlite3
 from contextlib import contextmanager
-from datetime import datetime, timedelta
+from datetime import datetime, timedelta, timezone
 from pathlib import Path
 from typing import Iterable, Iterator, List
 
 from .news_fetcher import NewsArticle
+from .utils import utcnow
 
 
 class ArticleCache:
@@ -64,7 +65,7 @@ def filter_new(self, articles: Iterable[NewsArticle]) -> List[NewsArticle]:
         return fresh
 
     def record(self, articles: Iterable[NewsArticle]) -> None:
-        now = datetime.utcnow()
+        now = utcnow()
         cutoff = now - timedelta(days=self.retention_days)
         with self._connect() as conn:
             conn.executemany(
@@ -73,7 +74,7 @@ def record(self, articles: Iterable[NewsArticle]) -> None:
                     (
                         self._article_key(article),
                         article.url or article.title,
-                        (article.published_at or now).isoformat(),
+                        _normalise_datetime(article.published_at, fallback=now).isoformat(),
                     )
                     for article in articles
                 ],
@@ -85,4 +86,12 @@ def record(self, articles: Iterable[NewsArticle]) -> None:
             conn.commit()
 
 
+def _normalise_datetime(value: datetime | None, *, fallback: datetime) -> datetime:
+    if value is None:
+        return fallback
+    if value.tzinfo is None:
+        return value.replace(tzinfo=timezone.utc)
+    return value.astimezone(timezone.utc)
+
+
 __all__ = ["ArticleCache"]

google-news-drive-sync/src/monitor.py

@@ -11,6 +11,11 @@
 from threading import Lock
 from time import perf_counter
 from typing import ContextManager, Dict, Iterator, List, Optional
+codex/implement-google-news-drive-sync-project-pipeline-xuy9hr
+
+from .utils import utcnow
+
+
 
 logger = logging.getLogger(__name__)
 
@@ -60,13 +65,17 @@
 
     def start_run(self) -> None:
         with self._lock:
-            self._last_run_started = datetime.utcnow()
+ codex/implement-google-news-drive-sync-project-pipeline-xuy9hr
+            self._last_run_started = utcnow()
+
+
             self._last_status = "running"
 
     def complete_run(self, *, status: str = "success") -> None:
         with self._lock:
             self._runs += 1
-            self._last_run_completed = datetime.utcnow()
+codex/implement-google-news-drive-sync-project-pipeline-xuy9hr
+            self._last_run_completed = utcnow()
             self._last_status = status
 
     def record_document_upload(self) -> None:

google-news-drive-sync/src/utils.py

@@ -8,6 +8,7 @@
 import logging
 import os
 import time
+from datetime import datetime, timezone
 from pathlib import Path
 from typing import Any, Callable, Iterable, Optional, TypeVar
 
@@ -144,3 +145,9 @@ def decrypt(self, payload: bytes) -> bytes:
         if self._fernet is not None:
             return self._fernet.decrypt(payload)
         return _xor_bytes(payload, self._key)
+
+
+def utcnow() -> datetime:
+    """Return a timezone-aware ``datetime`` in UTC."""
+
+    return datetime.now(timezone.utc)

google-news-drive-sync/tests/test_cache.py

@@ -1,4 +1,4 @@
-from datetime import datetime, timedelta
+from datetime import datetime, timedelta, timezone
 
 from src.cache import ArticleCache
 from src.news_fetcher import NewsArticle
@@ -31,7 +31,7 @@ def test_cache_prunes_expired_entries(tmp_path):
     old = make_article(
         "https://example.com/old",
         "Old",
-        published_at=datetime.utcnow() - timedelta(days=2),
+        published_at=datetime.now(timezone.utc) - timedelta(days=2),
     )
     cache.record([old])
     cache.record([])  # trigger retention cleanup