feat: implement Finnhub news collector with VADER sentiment

2 files changed, 156 insertions, 0 deletions

diff --git a/services/news-collector/src/news_collector/collectors/finnhub.py b/services/news-collector/src/news_collector/collectors/finnhub.py
new file mode 100644
index 0000000..13e3602
--- /dev/null
+++ b/services/news-collector/src/news_collector/collectors/finnhub.py
@@ -0,0 +1,88 @@
+"""Finnhub news collector with VADER sentiment analysis."""
+
+import logging
+from datetime import datetime, timezone
+
+import aiohttp
+from nltk.sentiment.vader import SentimentIntensityAnalyzer
+
+from shared.models import NewsCategory, NewsItem
+
+from .base import BaseCollector
+
+logger = logging.getLogger(__name__)
+
+_CATEGORY_KEYWORDS: dict[NewsCategory, list[str]] = {
+    NewsCategory.FED: ["fed", "fomc", "rate", "federal reserve"],
+    NewsCategory.POLICY: ["tariff", "trump", "regulation", "policy", "trade war"],
+    NewsCategory.EARNINGS: ["earnings", "revenue", "profit", "eps", "guidance", "quarter"],
+}
+
+
+def _categorize(text: str) -> NewsCategory:
+    lower = text.lower()
+    for category, keywords in _CATEGORY_KEYWORDS.items():
+        if any(kw in lower for kw in keywords):
+            return category
+    return NewsCategory.MACRO
+
+
+class FinnhubCollector(BaseCollector):
+    name: str = "finnhub"
+    poll_interval: int = 300
+
+    _BASE_URL = "https://finnhub.io/api/v1/news"
+
+    def __init__(self, api_key: str) -> None:
+        self._api_key = api_key
+        self._vader = SentimentIntensityAnalyzer()
+
+    async def is_available(self) -> bool:
+        return bool(self._api_key)
+
+    async def _fetch_news(self) -> list[dict]:
+        url = f"{self._BASE_URL}?category=general&token={self._api_key}"
+        async with aiohttp.ClientSession() as session:
+            async with session.get(url) as resp:
+                resp.raise_for_status()
+                return await resp.json()
+
+    async def collect(self) -> list[NewsItem]:
+        try:
+            raw_items = await self._fetch_news()
+        except Exception as exc:
+            logger.error("Finnhub fetch failed: %s", exc)
+            return []
+
+        items: list[NewsItem] = []
+        for article in raw_items:
+            headline = article.get("headline", "")
+            summary = article.get("summary", "")
+            combined = f"{headline} {summary}"
+
+            sentiment_scores = self._vader.polarity_scores(combined)
+            sentiment = sentiment_scores["compound"]
+
+            ts = article.get("datetime", 0)
+            published_at = datetime.fromtimestamp(ts, tz=timezone.utc)
+
+            related = article.get("related", "")
+            symbols = [t.strip() for t in related.split(",") if t.strip()] if related else []
+
+            category = _categorize(combined)
+
+            items.append(
+                NewsItem(
+                    source=self.name,
+                    headline=headline,
+                    summary=summary or None,
+                    url=article.get("url") or None,
+                    published_at=published_at,
+                    symbols=symbols,
+                    sentiment=sentiment,
+                    category=category,
+                    raw_data=article,
+                )
+            )
+
+        return items
diff --git a/services/news-collector/tests/test_finnhub.py b/services/news-collector/tests/test_finnhub.py
new file mode 100644
index 0000000..74bd5e6
--- /dev/null
+++ b/services/news-collector/tests/test_finnhub.py
@@ -0,0 +1,68 @@
+"""Tests for Finnhub news collector."""
+
+import pytest
+from unittest.mock import AsyncMock, patch
+from datetime import datetime, timezone
+
+from news_collector.collectors.finnhub import FinnhubCollector
+
+
+@pytest.fixture
+def collector():
+    return FinnhubCollector(api_key="test_key")
+
+
+def test_collector_name(collector):
+    assert collector.name == "finnhub"
+    assert collector.poll_interval == 300
+
+
+async def test_is_available_with_key(collector):
+    assert await collector.is_available() is True
+
+
+async def test_is_available_without_key():
+    c = FinnhubCollector(api_key="")
+    assert await c.is_available() is False
+
+
+async def test_collect_parses_response(collector):
+    mock_response = [
+        {
+            "category": "top news",
+            "datetime": 1711929600,
+            "headline": "AAPL beats earnings",
+            "id": 12345,
+            "related": "AAPL",
+            "source": "MarketWatch",
+            "summary": "Apple reported better than expected...",
+            "url": "https://example.com/article",
+        },
+        {
+            "category": "top news",
+            "datetime": 1711929000,
+            "headline": "Fed holds rates steady",
+            "id": 12346,
+            "related": "",
+            "source": "Reuters",
+            "summary": "The Federal Reserve...",
+            "url": "https://example.com/fed",
+        },
+    ]
+
+    with patch.object(collector, "_fetch_news", new_callable=AsyncMock, return_value=mock_response):
+        items = await collector.collect()
+
+    assert len(items) == 2
+    assert items[0].source == "finnhub"
+    assert items[0].headline == "AAPL beats earnings"
+    assert items[0].symbols == ["AAPL"]
+    assert items[0].url == "https://example.com/article"
+    assert isinstance(items[0].sentiment, float)
+    assert items[1].symbols == []
+
+
+async def test_collect_handles_empty_response(collector):
+    with patch.object(collector, "_fetch_news", new_callable=AsyncMock, return_value=[]):
+        items = await collector.collect()
+    assert items == []