Update app.py

app.py

@@ -6,7 +6,6 @@ import json
 import requests
 from PIL import Image
 from openai import OpenAI
-import easyocr
 from io import BytesIO
 
 # ------------------ App Configuration ------------------
@@ -23,7 +22,15 @@ if not OPENAI_API_KEY or not ASSISTANT_ID:
     st.stop()
 
 client = OpenAI(api_key=OPENAI_API_KEY)
-reader = easyocr.Reader(['en'], gpu=False)
+
+# ------------------ Load Structured JSON ------------------
+STRUCTURED_JSON_PATH = "/mnt/data/51940670-Manual-of-Surgical-Pathology-Third-Edition_1_structured_output (1).json"
+try:
+    with open(STRUCTURED_JSON_PATH, "r") as f:
+        structured_data = json.load(f)
+except Exception as e:
+    st.error(f"❌ Failed to load structured summary file: {e}")
+    st.stop()
 
 # ------------------ Session State Initialization ------------------
 if "messages" not in st.session_state:
@@ -46,46 +53,6 @@ if st.sidebar.button("🔄 Clear Chat"):
 
 show_image = st.sidebar.checkbox("📖 Show Document Image", value=True)
 
-# ------------------ OCR + GPT Summary & FAQ Generator ------------------
-def generate_summary_and_faq_from_image_easyocr(image_url):
-    try:
-        response = requests.get(image_url, stream=True)
-        image = Image.open(BytesIO(response.content)).convert("RGB")
-
-        result = reader.readtext(np.array(image), detail=0)
-        extracted_text = "\n".join(result)
-
-        if not extracted_text.strip():
-            return "No readable text found in image.", []
-
-        prompt = f"""
-You are a pathology assistant. Given this OCR-extracted text from a pathology textbook page, do the following:
-1. Provide a concise summary of the main point (1-2 sentences).
-2. Provide two FAQs with brief answers.
-
-Text:
-{extracted_text[:3000]}
-
-Return only JSON:
-{{
-  "summary": "...",
-  "faqs": [
-    {{"question": "...", "answer": "..."}},
-    {{"question": "...", "answer": "..."}}
-  ]
-}}
-"""
-        response = client.chat.completions.create(
-            model="gpt-3.5-turbo",
-            messages=[{"role": "user", "content": prompt}],
-            temperature=0.3
-        )
-        result = json.loads(response.choices[0].message.content)
-        return result.get("summary", "No summary generated."), result.get("faqs", [])
-
-    except Exception as e:
-        return f"Error generating summary: {e}", []
-
 # ------------------ Layout ------------------
 left, center, right = st.columns([1, 2, 1])
 
@@ -162,16 +129,21 @@ with center:
         except Exception as e:
             st.error(f"❌ Error: {str(e)}")
 
-# ------------------ Right Column: OCR-Based Summary + FAQ ------------------
+# ------------------ Right Column: Structured Summary + FAQ ------------------
 with right:
-    st.subheader("📌 Summary & FAQ (via EasyOCR)")
+    st.subheader("📌 Summary & FAQ (from Structured Data)")
+
+    summary_text = "No image selected or page not found."
+    faq_list = []
 
     if st.session_state.image_url:
-        with st.spinner("🔍 Extracting text and generating summary..."):
-            summary_text, faq_list = generate_summary_and_faq_from_image_easyocr(st.session_state.image_url)
-    else:
-        summary_text = "No image selected."
-        faq_list = []
+        match = re.search(r'/(\d{3})\.png', st.session_state.image_url)
+        if match:
+            page_number = match.group(1)
+            page_entry = next((entry for entry in structured_data if entry.get("page_number") == page_number), None)
+            if page_entry:
+                summary_text = page_entry.get("summary", "No summary available.")
+                faq_list = page_entry.get("faqs", [])
 
     st.markdown(summary_text)
 
@@ -180,4 +152,4 @@ with right:
         for faq in faq_list:
             st.markdown(f"**Q:** {faq.get('question', '')}\n\n**A:** {faq.get('answer', '')}")
     else:
-        st.info("No FAQs available or generated from this page.")
+        st.info("No FAQs available for this page.")