Initial commit — Prism financial analysis dashboard

Streamlit app with market bar, price chart, financial statements,
DCF valuation engine, comparable companies, and news feed.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

1 files changed, 93 insertions, 0 deletions

diff --git a/components/news.py b/components/news.py
new file mode 100644
index 0000000..cea678e
--- /dev/null
+++ b/components/news.py
@@ -0,0 +1,93 @@
+"""News feed with sentiment badges."""
+import streamlit as st
+from datetime import datetime
+from services.news_service import get_company_news, get_news_sentiment
+from services.fmp_service import get_company_news as get_fmp_news
+
+
+def _sentiment_badge(sentiment: str) -> str:
+    badges = {
+        "bullish": "🟢 Bullish",
+        "bearish": "🔴 Bearish",
+        "neutral": "⚪ Neutral",
+    }
+    return badges.get(sentiment.lower(), "⚪ Neutral")
+
+
+def _classify_sentiment(article: dict) -> str:
+    """Classify sentiment from Finnhub article data."""
+    # Finnhub doesn't return per-article sentiment, use headline heuristics
+    positive = ["beats", "surges", "rises", "gains", "profit", "record", "upgrade",
+                "buy", "outperform", "growth", "strong", "higher", "rally"]
+    negative = ["misses", "falls", "drops", "loss", "cut", "downgrade", "sell",
+               "underperform", "weak", "lower", "decline", "warning", "layoff"]
+    headline = (article.get("headline") or article.get("title") or "").lower()
+    summary = (article.get("summary") or "").lower()
+    text = headline + " " + summary
+    pos = sum(1 for w in positive if w in text)
+    neg = sum(1 for w in negative if w in text)
+    if pos > neg:
+        return "bullish"
+    if neg > pos:
+        return "bearish"
+    return "neutral"
+
+
+def _fmt_time(timestamp) -> str:
+    try:
+        if isinstance(timestamp, (int, float)):
+            return datetime.utcfromtimestamp(timestamp).strftime("%b %d, %Y")
+        return str(timestamp)[:10]
+    except Exception:
+        return ""
+
+
+def render_news(ticker: str):
+    # Overall sentiment summary from Finnhub
+    sentiment_data = get_news_sentiment(ticker)
+    if sentiment_data:
+        buzz = sentiment_data.get("buzz", {})
+        score = sentiment_data.get("sentiment", {})
+        col1, col2, col3 = st.columns(3)
+        col1.metric("Articles (7d)", buzz.get("articlesInLastWeek", "—"))
+        bull_pct = score.get("bullishPercent")
+        bear_pct = score.get("bearishPercent")
+        col2.metric("Bullish %", f"{bull_pct * 100:.1f}%" if bull_pct else "—")
+        col3.metric("Bearish %", f"{bear_pct * 100:.1f}%" if bear_pct else "—")
+        st.divider()
+
+    # Fetch articles — Finnhub first, FMP as fallback
+    articles = get_company_news(ticker)
+    if not articles:
+        articles = get_fmp_news(ticker)
+
+    if not articles:
+        st.info("No recent news found.")
+        return
+
+    for article in articles:
+        headline = article.get("headline") or article.get("title", "No title")
+        source = article.get("source") or article.get("site", "")
+        url = article.get("url") or article.get("newsURL") or article.get("url", "")
+        timestamp = article.get("datetime") or article.get("publishedDate", "")
+        summary = article.get("summary") or article.get("text") or ""
+
+        sentiment = _classify_sentiment(article)
+        badge = _sentiment_badge(sentiment)
+        time_str = _fmt_time(timestamp)
+
+        with st.container():
+            col1, col2 = st.columns([5, 1])
+            with col1:
+                if url:
+                    st.markdown(f"**[{headline}]({url})**")
+                else:
+                    st.markdown(f"**{headline}**")
+                meta = " · ".join(filter(None, [source, time_str]))
+                if meta:
+                    st.caption(meta)
+                if summary:
+                    st.caption(summary[:200] + ("…" if len(summary) > 200 else ""))
+            with col2:
+                st.markdown(badge)
+        st.divider()