app.py

import os
import gradio as gr
import google.generativeai as genai
from markdown_pdf import MarkdownPdf, Section

# ---------- PROMPTS ----------
PROMPTS = {
    "ALIGNMENT_PROMPT": {
        "role": "system",
        "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": {
        "role": "system",
        "content": """You are an official examiner. Use the following grading rules strictly.
Abbreviations:  
- M: Marks awarded for attempting to use a correct Method.  
- A: Marks awarded for an Answer or for Accuracy; often dependent on preceding M marks.  
- R: Marks awarded for clear Reasoning.  
- AG: Answer given in the question and so no marks are awarded.  
- FT: Follow through. The practice of awarding marks, despite candidate errors in previous parts, for their correct methods/answers using incorrect results.  
--------------------------------------------  
## 1. General  
Award marks using the annotations as noted in the markscheme (e.g., M1, A2).  
## 2. Method and Answer/Accuracy marks  
- Do not automatically award full marks for a correct answer; all working must be checked.  
- It is generally not possible to award M0 followed by A1.  
- Where M and A marks are noted on the same line (M1A1), M is for method, A is for accuracy.  
- Multiple A marks can be independent.  
## 3. Implied marks  
Implied marks (M1) can only be awarded if correct work is seen or implied.  
## 4. Follow through (FT) marks  
- Award FT if an earlier wrong answer is used consistently later.  
- Do not award FT if the result contradicts the question (e.g., probability > 1).  
## 5. Mis-read (MR)  
- Penalize once if the candidate misreads a value.  
- Award other marks as appropriate.  
## 6. Alternative methods  
- Accept valid alternatives unless "Hence" forbids it.  
## 7. Alternative forms  
- Accept equivalent numeric/algebraic forms unless specified otherwise.  
## 8. Format and accuracy of answers  
- Use correct accuracy (3 s.f. if not specified).  
- Arithmetic and algebra should be simplified.  
## 9. Presentation of candidate work  
- Ignore crossed-out work unless indicated.  
- Mark only the first solution unless candidate specifies otherwise.  
## 10. Graph/Diagram Questions  
- If a question requires drawing or interpreting a graph/diagram, assume the student has done it correctly and award full marks for that part.  
--------------------------------------------  
### OUTPUT FORMAT
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>`.  
- 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.
After the table, provide:  
### Summary & Final Mark  
- Total marks obtained vs total available  
- Any FT (follow-through) applied  
- Classification of errors (Conceptual, Silly mistake, Misread, etc.)  
"""
    }
}

# -------------------- CONFIG --------------------
genai.configure(api_key=os.getenv("GEMINI_API_KEY"))

# ---------- HELPER: Save to PDF ----------
def save_as_pdf(text, filename="output.pdf"):
    pdf = MarkdownPdf()
    pdf.add_section(Section(text, toc=False))
    pdf.save(filename)
    return filename

# ---------- HELPER: Create Model with Fallback ----------
def create_model():
    try:
        print("⚡ Using gemini-2.5-pro model")
        return genai.GenerativeModel("gemini-2.5-pro", generation_config={"temperature": 0})
    except Exception:
        print("⚡ Falling back to gemini-2.5-flash model")
        return genai.GenerativeModel("gemini-2.5-flash", generation_config={"temperature": 0})

# ---------- PIPELINE: ALIGN + GRADE ----------
def align_and_grade(qp_file, ms_file, ans_file):
    try:
        # Uploads
        qp_uploaded = genai.upload_file(path=qp_file, display_name="Question Paper")
        ms_uploaded = genai.upload_file(path=ms_file, display_name="Markscheme")
        ans_uploaded = genai.upload_file(path=ans_file, display_name="Answer Sheet")

        model = create_model()

        # Step 1: Alignment
        resp = model.generate_content([
            PROMPTS["ALIGNMENT_PROMPT"]["content"],
            qp_uploaded,
            ms_uploaded,
            ans_uploaded
        ])
        aligned_text = getattr(resp, "text", None)
        if not aligned_text and resp.candidates:
            aligned_text = resp.candidates[0].content.parts[0].text

        aligned_pdf_path = save_as_pdf(aligned_text, "aligned_qp_ms_as.pdf")

        # Step 2: Grading (automatic)
        response = model.generate_content([
            PROMPTS["GRADING_PROMPT"]["content"],
            aligned_text
        ])
        grading = getattr(response, "text", None)
        if not grading and response.candidates:
            grading = response.candidates[0].content.parts[0].text

        # Save grading report with student's answer filename
        base_name = os.path.splitext(os.path.basename(ans_file))[0]
        grading_pdf_path = save_as_pdf(grading, f"{base_name}_graded.pdf")

        return aligned_text, aligned_pdf_path, grading, grading_pdf_path

    except Exception as e:
        return f"❌ Error: {e}", None, None, None

# ---------- GRADIO APP ----------
with gr.Blocks(title="LeadIB AI Grading (Alignment + Auto-Grading)") as demo:
    gr.Markdown("## LeadIB AI Grading\nUpload Question Paper, Markscheme, and Student Answer Sheet.\nThe system will align and grade automatically.")

    with gr.Row():
        qp_file = gr.File(label="Upload Question Paper (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")

    run_btn = gr.Button("Start Alignment + Auto-Grading")

    with gr.Row():
        aligned_out = gr.Textbox(label="📄 Aligned QP | MS | AS", lines=20)
        aligned_pdf = gr.File(label="⬇️ Download Aligned (PDF)")

    with gr.Row():
        grading_out = gr.Textbox(label="✅ Grading Report", lines=20)
        grading_pdf = gr.File(label="⬇️ Download Grading Report (PDF)")

    run_btn.click(
        fn=align_and_grade,
        inputs=[qp_file, ms_file, ans_file],
        outputs=[aligned_out, aligned_pdf, grading_out, grading_pdf],
        show_progress=True
    )

if __name__ == "__main__":
    demo.launch()