Update utils/summarizer.py

utils/summarizer.py

@@ -1,13 +1,16 @@
 # utils/summarizer.py
 
 from transformers import pipeline
+from typing import List
 
-# Load summarization pipeline
+# ========== Load Summarization Pipeline ==========
 summarizer = pipeline("summarization", model="facebook/bart-large-cnn")
 
-def split_text(text, max_chunk_len=800):
+# ========== Helper Functions ==========
+
+def split_text(text: str, max_chunk_len: int = 800) -> List[str]:
     """
-    Breaks long documents into smaller chunks for summarization.
+    ✂️ Breaks long text into smaller chunks for safe summarization.
     """
     sentences = text.split('. ')
     chunks = []
@@ -25,25 +28,33 @@ def split_text(text, max_chunk_len=800):
 
     return chunks
 
-def summarize_text(text):
+def clean_text(text: str) -> str:
+    """
+    🧹 Remove excessive whitespace and line breaks.
+    """
+    return text.replace("\n", " ").replace("  ", " ").strip()
+
+# ========== Summarization Function ==========
+
+def summarize_text(text: str) -> str:
     """
-    Generate a readable executive summary using bullet points.
+    📄 Generate a readable executive summary using bullet points.
     """
     if not text.strip():
         return "No input provided."
 
-    text = text.replace("\n", " ").replace("  ", " ").strip()
-    chunks = split_text(text)
+    cleaned = clean_text(text)
+    chunks = split_text(cleaned)
 
     bullet_points = []
 
     for chunk in chunks:
         result = summarizer(chunk, max_length=130, min_length=30, do_sample=False)
         summary = result[0]["summary_text"].strip()
-        summary_lines = summary.split('. ')
-        for line in summary_lines:
-            clean = line.strip().rstrip('.')
-            if clean:
-                bullet_points.append(f"• {clean}.")
+        lines = summary.split('. ')
+        for line in lines:
+            cleaned_line = line.strip().rstrip('.')
+            if cleaned_line:
+                bullet_points.append(f"• {cleaned_line}.")
 
     return "📄 Executive Summary:\n" + "\n".join(bullet_points)