100 word limit

scripts/news_digest/main.py

@@ -5,9 +5,13 @@ Recommended: run via ./run.sh, which uses `uv` to handle dependencies
 automatically (no manual venv or pip install needed).
 
 When an `ollama` key is present in config.json, each newly fetched article is
-automatically summarized and the result is stored in the database.  Ollama
-latency provides natural rate limiting between HTTP requests; when Ollama is
-not configured, a 1-second sleep is used instead.
+automatically summarized and the result is stored in the database.  Summaries
+are truncated at the last sentence boundary within 100 words to keep them
+concise.  Ollama latency provides natural rate limiting between HTTP requests;
+when Ollama is not configured, a 1-second sleep is used instead.
+
+The log file (news_digest.log) is cleared at the start of each fetch cycle
+and appended to during the run via run.sh.
 
 Uses a requests.Session with automatic retries and browser-like headers to
 handle transient HTTP errors (429/5xx).  A configurable per-feed article cap
@@ -224,6 +228,23 @@ def get_recent_articles(conn: sqlite3.Connection, hours: int) -> list[dict]:
     return [dict(r) for r in rows]
 
 
+def _truncate_summary(text: str, max_words: int = 100) -> str:
+    """Truncate summary at the last sentence boundary within max_words."""
+    words = text.split()
+    if len(words) <= max_words:
+        return text
+    truncated = " ".join(words[:max_words])
+    # Find the last sentence-ending punctuation
+    last_period = -1
+    for ch in (".", "。", "!", "！", "?", "？"):
+        idx = truncated.rfind(ch)
+        if idx > last_period:
+            last_period = idx
+    if last_period > 0:
+        return truncated[: last_period + 1]
+    return truncated + "..."
+
+
 def generate_summary(title: str, description: str | None, content: str | None, model: str, prompt: str) -> str | None:
     try:
         import ollama as ollama_lib
@@ -242,7 +263,8 @@ def generate_summary(title: str, description: str | None, content: str | None, m
             model=model,
             messages=[{"role": "user", "content": user_message}],
         )
-        return response["message"]["content"]
+        summary = response["message"]["content"]
+        return _truncate_summary(summary)
     except Exception as e:
         logger.warning("Ollama error for '%s': %s", title, e)
         return None