fix: harden feed timestamp parsing

Preserve item dates from updated timestamps and ignore malformed parsed tuples so feed ingestion keeps working across mixed RSS and Atom metadata.

Author: ArietidsZ

tests/infrastructure/test_feed_source.py

@@ -1,8 +1,10 @@
 from datetime import datetime, timezone
+from types import SimpleNamespace
 
 import pytest
 
 from video_rss_aggregator.application.ports import FetchedFeed, FetchedFeedEntry
+from video_rss_aggregator.infrastructure import feed_source as feed_source_module
 from video_rss_aggregator.infrastructure.feed_source import HttpFeedSource
 
 
@@ -71,3 +73,75 @@ async def test_http_feed_source_fetches_and_maps_entries() -> None:
             ),
         ),
     )
+
+
+@pytest.mark.anyio
+async def test_http_feed_source_falls_back_to_updated_timestamp() -> None:
+    client = FakeAsyncClient(
+        FakeResponse(
+            """
+            <feed xmlns="http://www.w3.org/2005/Atom">
+              <title>Example atom feed</title>
+              <link href="https://example.com" />
+              <entry>
+                <title>First</title>
+                <id>first-guid</id>
+                <updated>2024-01-02T03:04:05Z</updated>
+                <link href="https://example.com/watch?v=1" />
+              </entry>
+            </feed>
+            """
+        )
+    )
+    adapter = HttpFeedSource(client)
+
+    feed = await adapter.fetch("https://example.com/feed.xml")
+
+    assert feed == FetchedFeed(
+        title="Example atom feed",
+        site_url="https://example.com",
+        entries=(
+            FetchedFeedEntry(
+                source_url="https://example.com/watch?v=1",
+                title="First",
+                guid="first-guid",
+                published_at=datetime(2024, 1, 2, 3, 4, 5, tzinfo=timezone.utc),
+            ),
+        ),
+    )
+
+
+@pytest.mark.anyio
+async def test_http_feed_source_ignores_malformed_parsed_dates(monkeypatch) -> None:
+    client = FakeAsyncClient(FakeResponse("ignored"))
+    adapter = HttpFeedSource(client)
+
+    def fake_parse(_text: str) -> SimpleNamespace:
+        return SimpleNamespace(
+            feed={"title": "Example feed", "link": "https://example.com"},
+            entries=[
+                {
+                    "title": "First",
+                    "id": "first-guid",
+                    "link": "https://example.com/watch?v=1",
+                    "published_parsed": (2024, 1),
+                }
+            ],
+        )
+
+    monkeypatch.setattr(feed_source_module.feedparser, "parse", fake_parse)
+
+    feed = await adapter.fetch("https://example.com/feed.xml")
+
+    assert feed == FetchedFeed(
+        title="Example feed",
+        site_url="https://example.com",
+        entries=(
+            FetchedFeedEntry(
+                source_url="https://example.com/watch?v=1",
+                title="First",
+                guid="first-guid",
+                published_at=None,
+            ),
+        ),
+    )

video_rss_aggregator/infrastructure/feed_source.py

@@ -31,22 +31,48 @@ def _map_entry(entry: Any) -> FetchedFeedEntry:
 
 
 def _pick_published_at(entry: Any) -> datetime | None:
-    published = entry.get("published")
-    if published:
-        try:
-            parsed = parsedate_to_datetime(published)
-        except (TypeError, ValueError, IndexError, OverflowError):
-            parsed = None
-        if parsed is not None:
-            if parsed.tzinfo is None:
-                return parsed.replace(tzinfo=timezone.utc)
-            return parsed.astimezone(timezone.utc)
-
-    published_parsed = entry.get("published_parsed")
-    if published_parsed is None:
+    for field in ("published", "updated"):
+        value = entry.get(field)
+        if value:
+            parsed = _parse_datetime_value(value)
+            if parsed is not None:
+                return parsed
+
+    for field in ("published_parsed", "updated_parsed"):
+        value = entry.get(field)
+        if value is not None:
+            parsed = _parse_datetime_tuple(value)
+            if parsed is not None:
+                return parsed
+
+    return None
+
+
+def _parse_datetime_value(value: Any) -> datetime | None:
+    if not isinstance(value, str):
         return None
 
-    return datetime(*published_parsed[:6], tzinfo=timezone.utc)
+    try:
+        parsed = parsedate_to_datetime(value)
+    except (TypeError, ValueError, IndexError, OverflowError):
+        parsed = None
+
+    if parsed is None:
+        try:
+            parsed = datetime.fromisoformat(value.replace("Z", "+00:00"))
+        except ValueError:
+            return None
+
+    if parsed.tzinfo is None:
+        return parsed.replace(tzinfo=timezone.utc)
+    return parsed.astimezone(timezone.utc)
+
+
+def _parse_datetime_tuple(value: Any) -> datetime | None:
+    try:
+        return datetime(*value[:6], tzinfo=timezone.utc)
+    except (TypeError, ValueError, IndexError, OverflowError):
+        return None
 
 
 @dataclass(frozen=True)