Update app.py

app.py

@@ -3,39 +3,46 @@ import fitz  # PyMuPDF
 from transformers import pipeline
 import re
 
-# Use a faster and lighter summarization model
+# Use faster summarization model
 summarizer = pipeline("summarization", model="sshleifer/distilbart-cnn-12-6")
 
 def extract_text_from_pdf(pdf_file):
-    doc = fitz.open(stream=pdf_file.read(), filetype="pdf")
-    text = "".join(page.get_text() + "\n" for page in doc)
-    return text
+    try:
+        doc = fitz.open(pdf_file.name)  # FIXED: Use .name instead of .read()
+        text = ""
+        for page in doc:
+            text += page.get_text()
+        return text
+    except Exception as e:
+        return None
 
 def chunk_text(text, max_words=500):
     words = text.split()
     return [" ".join(words[i:i + max_words]) for i in range(0, len(words), max_words)]
 
 def simplify_summary(summary):
-    # Remove repetitive hospital info
-    summary = re.sub(r"\b(?:Mayo Hospital|Lahore Hospital|submitted by Dr\.).+\n?", "", summary, flags=re.IGNORECASE)
+    summary = re.sub(r"(Mayo Hospital|Lahore Hospital|Dr\.\s+\w+)+.*", "", summary, flags=re.IGNORECASE)
     return "🩺 In simple terms:\n" + summary.strip()
 
 def process_report(pdf_file):
-    text = extract_text_from_pdf(pdf_file)
-    if not text.strip():
-        return "❌ Couldn't extract text from the PDF.", ""
+    raw_text = extract_text_from_pdf(pdf_file)
+    if not raw_text:
+        return "❌ Could not read PDF file.", ""
     
-    # Remove irrelevant boilerplate
-    header, *rest = text.split("\n\n", 1)
-    core_text = rest[0] if rest else text
+    chunks = chunk_text(raw_text, max_words=600)
+    summaries = []
+    for chunk in chunks:
+        try:
+            result = summarizer(chunk, max_length=150, min_length=30, do_sample=False)
+            summaries.append(result[0]['summary_text'])
+        except:
+            continue
 
-    chunks = chunk_text(core_text, max_words=600)
-    summaries = [summarizer(chunk, max_length=150, min_length=30, do_sample=False)[0]['summary_text']
-                 for chunk in chunks]
-    
-    final_summary = " ".join(summaries)
-    simple = simplify_summary(final_summary)
-    return final_summary, simple
+    if not summaries:
+        return "❌ Summarization failed. Try a smaller or clearer PDF.", ""
+
+    full_summary = " ".join(summaries)
+    return full_summary, simplify_summary(full_summary)
 
 demo = gr.Interface(
     fn=process_report,
@@ -45,7 +52,7 @@ demo = gr.Interface(
         gr.Textbox(label="Simplified Explanation", lines=8)
     ],
     title="🏥 Medical Report Summarizer",
-    description="Speeds up summarization by chunking text & uses a lighter distil-BART model, focusing on core medical findings."
+    description="Upload a medical report PDF to get an AI summary and non-medical explanation."
 )
 
 if __name__ == "__main__":