feat: implement SEC EDGAR 8-K filing collector

2 files changed, 143 insertions, 0 deletions

diff --git a/services/news-collector/src/news_collector/collectors/sec_edgar.py b/services/news-collector/src/news_collector/collectors/sec_edgar.py
new file mode 100644
index 0000000..a00abb5
--- /dev/null
+++ b/services/news-collector/src/news_collector/collectors/sec_edgar.py
@@ -0,0 +1,87 @@
+"""SEC EDGAR filing collector (free, no API key required)."""
+
+import logging
+from datetime import datetime, timezone
+
+import aiohttp
+from nltk.sentiment.vader import SentimentIntensityAnalyzer
+
+from shared.models import NewsCategory, NewsItem
+from news_collector.collectors.base import BaseCollector
+
+logger = logging.getLogger(__name__)
+
+TRACKED_CIKS = {
+    "0000320193": "AAPL", "0000789019": "MSFT", "0001652044": "GOOGL",
+    "0001018724": "AMZN", "0001318605": "TSLA", "0001045810": "NVDA",
+    "0001326801": "META", "0000019617": "JPM", "0000078003": "PFE", "0000021344": "KO",
+}
+
+SEC_USER_AGENT = "TradingPlatform research@example.com"
+
+
+class SecEdgarCollector(BaseCollector):
+    name = "sec_edgar"
+    poll_interval = 1800  # 30 minutes
+
+    def __init__(self) -> None:
+        self._vader = SentimentIntensityAnalyzer()
+
+    async def is_available(self) -> bool:
+        return True
+
+    async def _fetch_recent_filings(self) -> list[dict]:
+        results = []
+        headers = {"User-Agent": SEC_USER_AGENT}
+        async with aiohttp.ClientSession() as session:
+            for cik, ticker in TRACKED_CIKS.items():
+                try:
+                    url = f"https://data.sec.gov/submissions/CIK{cik}.json"
+                    async with session.get(url, headers=headers, timeout=aiohttp.ClientTimeout(total=10)) as resp:
+                        if resp.status == 200:
+                            data = await resp.json()
+                            data["tickers"] = [{"ticker": ticker}]
+                            results.append(data)
+                except Exception as exc:
+                    logger.warning("sec_fetch_failed", cik=cik, error=str(exc))
+        return results
+
+    async def collect(self) -> list[NewsItem]:
+        filings_data = await self._fetch_recent_filings()
+        items = []
+        today = datetime.now(timezone.utc).strftime("%Y-%m-%d")
+
+        for company_data in filings_data:
+            tickers = [t["ticker"] for t in company_data.get("tickers", [])]
+            company_name = company_data.get("name", "Unknown")
+            recent = company_data.get("filings", {}).get("recent", {})
+
+            forms = recent.get("form", [])
+            dates = recent.get("filingDate", [])
+            descriptions = recent.get("primaryDocDescription", [])
+            accessions = recent.get("accessionNumber", [])
+
+            for i, form in enumerate(forms):
+                if form != "8-K":
+                    continue
+                filing_date = dates[i] if i < len(dates) else ""
+                if filing_date != today:
+                    continue
+
+                desc = descriptions[i] if i < len(descriptions) else "8-K Filing"
+                accession = accessions[i] if i < len(accessions) else ""
+                headline = f"{company_name} ({', '.join(tickers)}): {form} - {desc}"
+
+                items.append(NewsItem(
+                    source=self.name,
+                    headline=headline,
+                    summary=desc,
+                    url=f"https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&accession={accession}",
+                    published_at=datetime.strptime(filing_date, "%Y-%m-%d").replace(tzinfo=timezone.utc),
+                    symbols=tickers,
+                    sentiment=self._vader.polarity_scores(headline)["compound"],
+                    category=NewsCategory.FILING,
+                    raw_data={"form": form, "accession": accession},
+                ))
+
+        return items
diff --git a/services/news-collector/tests/test_sec_edgar.py b/services/news-collector/tests/test_sec_edgar.py
new file mode 100644
index 0000000..a10b47a
--- /dev/null
+++ b/services/news-collector/tests/test_sec_edgar.py
@@ -0,0 +1,56 @@
+"""Tests for SEC EDGAR filing collector."""
+
+import pytest
+from datetime import datetime, timezone
+from unittest.mock import AsyncMock, patch, MagicMock
+
+from news_collector.collectors.sec_edgar import SecEdgarCollector
+
+
+@pytest.fixture
+def collector():
+    return SecEdgarCollector()
+
+
+def test_collector_name(collector):
+    assert collector.name == "sec_edgar"
+    assert collector.poll_interval == 1800
+
+
+async def test_is_available(collector):
+    assert await collector.is_available() is True
+
+
+async def test_collect_parses_filings(collector):
+    mock_response = {
+        "filings": {
+            "recent": {
+                "accessionNumber": ["0001234-26-000001"],
+                "filingDate": ["2026-04-02"],
+                "primaryDocument": ["filing.htm"],
+                "form": ["8-K"],
+                "primaryDocDescription": ["Current Report"],
+            }
+        },
+        "tickers": [{"ticker": "AAPL"}],
+        "name": "Apple Inc",
+    }
+
+    mock_datetime = MagicMock(spec=datetime)
+    mock_datetime.now.return_value = datetime(2026, 4, 2, tzinfo=timezone.utc)
+    mock_datetime.strptime = datetime.strptime
+
+    with patch.object(collector, "_fetch_recent_filings", new_callable=AsyncMock, return_value=[mock_response]):
+        with patch("news_collector.collectors.sec_edgar.datetime", mock_datetime):
+            items = await collector.collect()
+
+    assert len(items) == 1
+    assert items[0].source == "sec_edgar"
+    assert items[0].category.value == "filing"
+    assert "AAPL" in items[0].symbols
+
+
+async def test_collect_handles_empty(collector):
+    with patch.object(collector, "_fetch_recent_filings", new_callable=AsyncMock, return_value=[]):
+        items = await collector.collect()
+    assert items == []