Update app.py

app.py

@@ -3,35 +3,43 @@ import gradio as gr
 import google.generativeai as genai
 from markdown_pdf import MarkdownPdf, Section
 
-# -------------------- CONFIG --------------------
-genai.configure(api_key=os.getenv("GEMINI_API_KEY"))
-
-# ---------- PROMPTS (JSON style) ----------
-
+# ---------- PROMPTS ----------
 PROMPTS = {
-    "TRANSCRIPTION_PROMPT": {
-        "role": "system",
-        "content": """Your Role: You are an expert technical transcriber specializing in mathematical and scientific documents. 
-Your mission is to convert handwritten solutions from a provided image or PDF into a clean, accurate, and logically structured Markdown format.
-Instructions:
-- Use ## for questions, ### for subquestions.
-- Transcribe only the corrected, final version of the solution (ignore scribbles, cancellations, mistakes).
-- Keep mathematical expressions in fenced code blocks.
-- If something is illegible, use [illegible].
-- Do not recreate graphs, only describe them.
-"""
-    },
-    "MARKSCHEME_TRANSCRIPTION_PROMPT": {
+    "ALIGNMENT_PROMPT": {
         "role": "system",
-        "content": """Your Role: You are an expert transcriber. 
-Convert the official marking scheme from the provided PDF into clean, structured Markdown.
-Instructions:
-- Preserve all structure (questions, subquestions).
-- Keep M, A, R annotations exactly as written.
-- Represent alternative methods clearly (METHOD 1, METHOD 2, etc.).
-- Preserve any accuracy requirements.
-- Format in Markdown using ## and ### for hierarchy.
-- Use code blocks for equations.
+        "content": """Your Role: You are an expert examiner and transcription specialist.  
+Your task is to **align three sources**:
+- Question Paper (QP)
+- Markscheme (MS)
+- Student Answer Sheet (AS)
+
+### Instructions
+1. Parse all documents carefully and align them **per question and sub-question**.
+2. For each question/sub-question, produce a structured block:
+
+---
+## Question X (and sub-question if applicable)  
+**QP:** [Insert the exact question text]  
+**MS:** [Insert the relevant part of the markscheme]  
+**AS:** [Insert the student's final cleaned answer transcription]  
+---
+
+3. Formatting Rules:
+- Use `##` for main questions and `###` for sub-questions.  
+- Write **QP | MS | AS** exactly in that order.  
+- Preserve all mathematical expressions inside fenced code blocks.  
+- Do not re-create diagrams/graphs. Write `[Graph omitted]`.  
+- If part of the student's answer is unreadable, write `[illegible]`.  
+- If a student skipped a question, write `[No response]`.  
+- Keep MS annotations (M1, A1, R1, etc.) exactly as in the original.  
+
+4. Output must be **clean, deterministic, and consistent** — so that another model can grade directly using this aligned representation.
+
+### Example
+## Question 1  
+**QP:** Expand `(1+x)^3`  
+**MS:** M1 for binomial expansion, A1 for coefficients, A1 for final form  
+**AS:**  
 """
     },
     "GRADING_PROMPT": {
@@ -60,7 +68,7 @@ Implied marks (M1) can only be awarded if correct work is seen or implied.
 - Penalize once if the candidate misreads a value.  
 - Award other marks as appropriate.  
 ## 6. Alternative methods  
-- Accept valid alternatives unless \"Hence\" forbids it.  
+- Accept valid alternatives unless "Hence" forbids it.  
 ## 7. Alternative forms  
 - Accept equivalent numeric/algebraic forms unless specified otherwise.  
 ## 8. Format and accuracy of answers  
@@ -75,7 +83,7 @@ Produce a GitHub-flavored Markdown table with 3 columns:
 | Student wrote | Marks Awarded | Reason |  
 |---------------|---------------|--------|  
 Special Formatting Rule:  
-- Whenever a mark is lost (M0, A0, R0 etc.), wrap it in red using: `<span style=\"color:red\">M0</span>`.  
+- Whenever a mark is lost (M0, A0, R0 etc.), wrap it in red using: `<span style="color:red">M0</span>`.  
 - Also wrap the corresponding Reason in red color.
 - Keep awarded marks (M1, A1, etc.) in plain text.  
 - If mixed (e.g., M1A0A1), only highlight the lost marks (`A0`) and its reason.
@@ -88,6 +96,9 @@ After the table, provide:
     }
 }
 
+# -------------------- CONFIG --------------------
+genai.configure(api_key=os.getenv("GEMINI_API_KEY"))
+
 # ---------- HELPER: Save to PDF ----------
 def save_as_pdf(text, filename="output.pdf"):
     pdf = MarkdownPdf()
@@ -95,103 +106,90 @@ def save_as_pdf(text, filename="output.pdf"):
     pdf.save(filename)
     return filename
 
-# ---------- STEP 1: TRANSCRIBE STUDENT ----------
-def transcribe_student(ans_file):
+# ---------- HELPER: Create Model with Fallback ----------
+def create_model():
     try:
-        ans_uploaded = genai.upload_file(path=ans_file, display_name="Answer Sheet")
-        model = genai.GenerativeModel("gemini-2.5-flash", generation_config={"temperature": 0})
-
-        resp = model.generate_content([PROMPTS["TRANSCRIPTION_PROMPT"]["content"], ans_uploaded])
-        transcription = getattr(resp, "text", None)
-        if not transcription and resp.candidates:
-            transcription = resp.candidates[0].content.parts[0].text
-
-        pdf_path = save_as_pdf(transcription, "student_transcription.pdf")
-        return transcription, pdf_path
-    except Exception as e:
-        return f"❌ Error during transcription: {e}", None
+        return genai.GenerativeModel("gemini-2.5-pro", generation_config={"temperature": 0})
+    except Exception:
+        return genai.GenerativeModel("gemini-2.5-flash", generation_config={"temperature": 0})
 
-# ---------- STEP 2: TRANSCRIBE MARKSCHEME ----------
-def transcribe_ms(ms_file):
+# ---------- STEP 1: ALIGNMENT ----------
+def align_documents(qp_file, ms_file, ans_file):
     try:
+        qp_uploaded = genai.upload_file(path=qp_file, display_name="Question Paper")
         ms_uploaded = genai.upload_file(path=ms_file, display_name="Markscheme")
-        model = genai.GenerativeModel("gemini-2.5-flash", generation_config={"temperature": 0})
+        ans_uploaded = genai.upload_file(path=ans_file, display_name="Answer Sheet")
 
-        resp = model.generate_content([PROMPTS["MARKSCHEME_TRANSCRIPTION_PROMPT"]["content"], ms_uploaded])
-        ms_transcription = getattr(resp, "text", None)
-        if not ms_transcription and resp.candidates:
-            ms_transcription = resp.candidates[0].content.parts[0].text
+        model = create_model()
+        resp = model.generate_content([
+            PROMPTS["ALIGNMENT_PROMPT"]["content"],
+            qp_uploaded,
+            ms_uploaded,
+            ans_uploaded
+        ])
 
-        pdf_path = save_as_pdf(ms_transcription, "ms_transcription.pdf")
-        return ms_transcription, pdf_path
+        aligned_text = getattr(resp, "text", None)
+        if not aligned_text and resp.candidates:
+            aligned_text = resp.candidates[0].content.parts[0].text
+
+        pdf_path = save_as_pdf(aligned_text, "aligned_qp_ms_as.pdf")
+        return aligned_text, pdf_path
     except Exception as e:
-        return f"❌ Error during MS transcription: {e}", None
+        return f"❌ Error during alignment: {e}", None
 
-# ---------- STEP 3: GRADING ----------
-def grade(qp_file, ms_transcription, student_transcription):
+# ---------- STEP 2: GRADING ----------
+def grade(aligned_text, ans_file):
     try:
-        qp_uploaded = genai.upload_file(path=qp_file, display_name="Question Paper")
-        model = genai.GenerativeModel("gemini-2.5-flash", generation_config={"temperature": 0})
-
+        model = create_model()
         response = model.generate_content([
             PROMPTS["GRADING_PROMPT"]["content"],
-            qp_uploaded,
-            "### Markscheme Transcription:\n" + ms_transcription,
-            "### Student Transcription:\n" + student_transcription
+            aligned_text
         ])
 
         grading = getattr(response, "text", None)
         if not grading and response.candidates:
             grading = response.candidates[0].content.parts[0].text
 
-        pdf_path = save_as_pdf(grading, "grading.pdf")
+        # Use answer sheet filename for graded PDF
+        base_name = os.path.splitext(os.path.basename(ans_file))[0]
+        pdf_name = f"{base_name}_graded.pdf"
+
+        pdf_path = save_as_pdf(grading, pdf_name)
         return grading, pdf_path
     except Exception as e:
         return f"❌ Error during grading: {e}", None
 
 # ---------- GRADIO APP ----------
-with gr.Blocks(title="LeadIB AI Grading (3-Step)") as demo:
-    gr.Markdown("## LeadIB AI Grading (3-Step)\nUpload exam documents to transcribe and grade step by step.")
+with gr.Blocks(title="LeadIB AI Grading (Alignment + Grading)") as demo:
+    gr.Markdown("## LeadIB AI Grading\nUpload Question Paper, Markscheme, and Student Answer Sheet to align and grade.")
 
     with gr.Row():
         qp_file = gr.File(label="Upload Question Paper (PDF)", type="filepath")
-        ms_file = gr.File(label="Upload Mark Scheme (PDF)", type="filepath")
+        ms_file = gr.File(label="Upload Markscheme (PDF)", type="filepath")
         ans_file = gr.File(label="Upload Student Answer Sheet (PDF)", type="filepath")
 
-    # Step 1: Transcribe Student
-    transcribe_student_btn = gr.Button("Step 1: Transcribe Student Answer Sheet")
-    with gr.Row():
-        student_out = gr.Textbox(label="📄 Student Transcription", lines=20)
-        student_pdf = gr.File(label="⬇️ Download Student Transcription (PDF)")
-
-    # Step 2: Transcribe Markscheme
-    transcribe_ms_btn = gr.Button("Step 2: Transcribe Markscheme")
+    # Step 1: Alignment
+    align_btn = gr.Button("Step 1: Align QP | MS | AS")
     with gr.Row():
-        ms_out = gr.Textbox(label="📄 Markscheme Transcription", lines=20)
-        ms_pdf = gr.File(label="⬇️ Download Markscheme Transcription (PDF)")
+        aligned_out = gr.Textbox(label="📄 Aligned QP | MS | AS", lines=20)
+        aligned_pdf = gr.File(label="⬇️ Download Aligned (PDF)")
 
-    # Step 3: Grading
-    grade_btn = gr.Button("Step 3: Grade the Student")
+    # Step 2: Grading
+    grade_btn = gr.Button("Step 2: Grade the Student")
     with gr.Row():
-        grading_out = gr.Textbox(label="✅ Grading Report (Step-by-Step)", lines=20)
+        grading_out = gr.Textbox(label="✅ Grading Report", lines=20)
         grading_pdf = gr.File(label="⬇️ Download Grading Report (PDF)")
 
     # Button Logic
-    transcribe_student_btn.click(
-        fn=transcribe_student,
-        inputs=[ans_file],
-        outputs=[student_out, student_pdf],
-        show_progress=True
-    )
-    transcribe_ms_btn.click(
-        fn=transcribe_ms,
-        inputs=[ms_file],
-        outputs=[ms_out, ms_pdf],
+    align_btn.click(
+        fn=align_documents,
+        inputs=[qp_file, ms_file, ans_file],
+        outputs=[aligned_out, aligned_pdf],
         show_progress=True
     )
     grade_btn.click(
         fn=grade,
-        inputs=[qp_file, ms_out, student_out],
+        inputs=[aligned_out, ans_file],
         outputs=[grading_out, grading_pdf],
         show_progress=True
     )