app.py

import gradio as gr
import pandas as pd
import json
import os
import re
from PyPDF2 import PdfReader
from collections import defaultdict
from typing import Dict, List, Optional, Tuple, Union
import html
from pathlib import Path
import fitz  # PyMuPDF
import pytesseract
from PIL import Image
import io
import secrets
import string
from huggingface_hub import HfApi, HfFolder
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM
import time
import logging
import asyncio
from functools import lru_cache
import hashlib
from concurrent.futures import ThreadPoolExecutor
from pydantic import BaseModel
import plotly.express as px
import pdfplumber
from io import BytesIO
import base64
import matplotlib.pyplot as plt

# ========== CONFIGURATION ==========
PROFILES_DIR = "student_profiles"
ALLOWED_FILE_TYPES = [".pdf", ".png", ".jpg", ".jpeg"]
MAX_FILE_SIZE_MB = 5
MIN_AGE = 5
MAX_AGE = 120
SESSION_TOKEN_LENGTH = 32
HF_TOKEN = os.getenv("HF_TOKEN")
SESSION_TIMEOUT = 3600  # 1 hour session timeout

# Initialize logging
logging.basicConfig(
    level=logging.DEBUG,
    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
    filename='transcript_parser.log'
)

# Model configuration - Using smaller model
MODEL_NAME = "deepseek-ai/deepseek-llm-1.3b"

# Initialize Hugging Face API
if HF_TOKEN:
    try:
        hf_api = HfApi(token=HF_TOKEN)
        HfFolder.save_token(HF_TOKEN)
    except Exception as e:
        logging.error(f"Failed to initialize Hugging Face API: {str(e)}")

# ========== MODEL LOADER ==========
class ModelLoader:
    def __init__(self):
        self.model = None
        self.tokenizer = None
        self.loaded = False
        self.loading = False
        self.error = None
        self.device = "cuda" if torch.cuda.is_available() else "cpu"
    
    def load_model(self, progress: gr.Progress = None) -> Tuple[Optional[AutoModelForCausalLM], Optional[AutoTokenizer]]:
        """Lazy load the model with progress feedback"""
        if self.loaded:
            return self.model, self.tokenizer
            
        if self.loading:
            while self.loading:
                time.sleep(0.1)
            return self.model, self.tokenizer
            
        self.loading = True
        try:
            if progress:
                progress(0.1, desc="Checking GPU availability...")
            
            torch.cuda.empty_cache()
            
            if progress:
                progress(0.2, desc="Loading tokenizer...")
                
            tokenizer = AutoTokenizer.from_pretrained(
                MODEL_NAME,
                trust_remote_code=True
            )
            
            if progress:
                progress(0.5, desc="Loading model (this may take a few minutes)...")
                
            model_kwargs = {
                "trust_remote_code": True,
                "torch_dtype": torch.float16 if self.device == "cuda" else torch.float32,
                "device_map": "auto" if self.device == "cuda" else None,
                "low_cpu_mem_usage": True,
                "offload_folder": "offload"
            }
            
            try:
                model = AutoModelForCausalLM.from_pretrained(
                    MODEL_NAME,
                    **model_kwargs
                )
            except torch.cuda.OutOfMemoryError:
                model_kwargs["device_map"] = None
                model = AutoModelForCausalLM.from_pretrained(
                    MODEL_NAME,
                    **model_kwargs
                ).to('cpu')
                self.device = 'cpu'
                
            test_input = tokenizer("Test", return_tensors="pt").to(self.device)
            _ = model.generate(**test_input, max_new_tokens=1)
            
            self.model = model.eval()
            self.tokenizer = tokenizer
            self.loaded = True
            
            return model, tokenizer
            
        except Exception as e:
            self.error = f"Model loading failed: {str(e)}"
            logging.error(self.error)
            return None, None
        finally:
            self.loading = False

# Initialize model loader
model_loader = ModelLoader()

@lru_cache(maxsize=1)
def get_model_and_tokenizer():
    return model_loader.load_model()

# ========== UTILITY FUNCTIONS ==========
def generate_session_token() -> str:
    alphabet = string.ascii_letters + string.digits
    return ''.join(secrets.choice(alphabet) for _ in range(SESSION_TOKEN_LENGTH))

def sanitize_input(text: str) -> str:
    if not text:
        return ""
    text = html.escape(text.strip())
    text = re.sub(r'<[^>]*>', '', text)
    text = re.sub(r'[^\w\s\-.,!?@#\$%^&*()+=]', '', text)
    return text

def validate_name(name: str) -> str:
    name = name.strip()
    if not name:
        raise ValueError("Name cannot be empty.")
    if len(name) > 100:
        raise ValueError("Name is too long (maximum 100 characters).")
    if any(c.isdigit() for c in name):
        raise ValueError("Name cannot contain numbers.")
    return name

def validate_age(age: Union[int, float, str]) -> int:
    try:
        age_int = int(age)
        if not MIN_AGE <= age_int <= MAX_AGE:
            raise ValueError(f"Age must be between {MIN_AGE} and {MAX_AGE}.")
        return age_int
    except (ValueError, TypeError):
        raise ValueError("Please enter a valid age number.")

def validate_file(file_obj) -> None:
    if not file_obj:
        raise ValueError("Please upload a file first")
    
    file_ext = os.path.splitext(file_obj.name)[1].lower()
    if file_ext not in ALLOWED_FILE_TYPES:
        raise ValueError(f"Invalid file type. Allowed types: {', '.join(ALLOWED_FILE_TYPES)}")
    
    file_size = os.path.getsize(file_obj.name) / (1024 * 1024)
    if file_size > MAX_FILE_SIZE_MB:
        raise ValueError(f"File too large. Maximum size is {MAX_FILE_SIZE_MB}MB.")

# ========== ENHANCED PDF PARSING ==========
def parse_transcript_pdf(file_path: str):
    """Parse the PDF transcript and extract structured data using pdfplumber"""
    student_info = {}
    requirements = []
    courses = []
    
    with pdfplumber.open(file_path) as pdf:
        for page in pdf.pages:
            text = page.extract_text()
            tables = page.extract_tables()
            
            # Parse student information from the first table
            if not student_info and len(tables) > 0:
                header_row = tables[0][0]
                if "Graduation Progress Summary" in header_row[0]:
                    student_info = {
                        'name': tables[0][1][0].split('-')[-1].strip(),
                        'id': tables[0][1][0].split('-')[0].strip(),
                        'school': tables[0][0][0].split('|')[1].strip(),
                        'cohort': tables[0][0][1].replace('Cohort', '').strip(),
                        'grade': tables[0][2][0].replace('Current Grade:', '').strip(),
                        'grad_year': tables[0][2][1].replace('YOG', '').strip(),
                        'gpa_weighted': tables[0][2][2].replace('Weighted GPA', '').strip(),
                        'gpa_unweighted': tables[0][0][2].replace('Un-weighted GPA', '').strip(),
                        'service_hours': tables[0][0][3].replace('Comm Serv Hours', '').strip(),
                        'service_date': tables[0][2][3].replace('Comm Serv Date', '').strip(),
                        'total_credits': tables[0][2][4].replace('Total Credits Earned', '').strip(),
                        'virtual_grade': tables[0][0][4].replace('Virtual Grade', '').strip()
                    }
            
            # Parse requirements table
            if len(tables) > 1 and "Code" in tables[1][0][0]:
                for row in tables[1][1:]:
                    if len(row) >= 6 and row[0] and row[0] != 'Total':
                        requirements.append({
                            'code': row[0],
                            'desc': row[1],
                            'required': float(row[2]) if row[2] else 0,
                            'waived': float(row[3]) if row[3] else 0,
                            'completed': float(row[4]) if row[4] else 0,
                            'status': float(row[5].replace('%', '')) if row[5] and '%' in row[5] else 0
                        })
            
            # Parse course history table
            if len(tables) > 2 and "Requirement" in tables[2][0][0]:
                for row in tables[2][1:]:
                    if len(row) >= 10 and row[0]:
                        courses.append({
                            'requirement': row[0],
                            'year': row[1],
                            'grade': row[2],
                            'course_code': row[3],
                            'course_name': row[4],
                            'term': row[5],
                            'district_num': row[6],
                            'grade_earned': row[7],
                            'included': row[8],
                            'credits': float(row[9]) if row[9] and row[9] not in ['inProgress', ''] else 0,
                            'status': 'Completed' if row[9] and row[9] != 'inProgress' else 'In Progress'
                        })
    
    return student_info, requirements, courses

def analyze_college_readiness(student_info, requirements, courses):
    """Analyze the student's profile for college readiness"""
    analysis = {
        'gpa_rating': '',
        'rigor_rating': '',
        'service_rating': '',
        'recommendations': []
    }
    
    # GPA Analysis
    weighted_gpa = float(student_info.get('gpa_weighted', 0))
    if weighted_gpa >= 4.5:
        analysis['gpa_rating'] = 'Excellent (Highly Competitive)'
    elif weighted_gpa >= 3.8:
        analysis['gpa_rating'] = 'Strong (Competitive)'
    elif weighted_gpa >= 3.0:
        analysis['gpa_rating'] = 'Good'
    else:
        analysis['gpa_rating'] = 'Below Average'
    
    # Course Rigor Analysis
    ap_count = sum(1 for course in courses if 'AP' in course['course_name'])
    de_count = sum(1 for course in courses if 'DE' in course['course_name'])
    honors_count = sum(1 for course in courses if 'Honors' in course['course_name'])
    
    total_rigorous = ap_count + de_count + honors_count
    if total_rigorous >= 10:
        analysis['rigor_rating'] = 'Very High'
    elif total_rigorous >= 6:
        analysis['rigor_rating'] = 'High'
    elif total_rigorous >= 3:
        analysis['rigor_rating'] = 'Moderate'
    else:
        analysis['rigor_rating'] = 'Low'
    
    # Community Service Analysis
    service_hours = int(student_info.get('service_hours', 0))
    if service_hours >= 100:
        analysis['service_rating'] = 'Exceptional'
    elif service_hours >= 50:
        analysis['service_rating'] = 'Strong'
    elif service_hours >= 30:
        analysis['service_rating'] = 'Adequate'
    else:
        analysis['service_rating'] = 'Limited'
    
    # Generate recommendations
    if weighted_gpa < 3.5 and ap_count < 3:
        analysis['recommendations'].append("Consider taking more advanced courses (AP/DE) to strengthen your academic profile")
    if service_hours < 50:
        analysis['recommendations'].append("Additional community service hours could enhance your college applications")
    
    return analysis

def create_requirements_visualization_matplotlib(requirements):
    """Create matplotlib visualization for requirements completion"""
    fig, ax = plt.subplots(figsize=(10, 6))
    req_names = [req['code'] for req in requirements]
    req_completion = [min(req['status'], 100) for req in requirements]
    colors = ['#4CAF50' if x >= 100 else '#FFC107' if x > 0 else '#F44336' for x in req_completion]
    
    bars = ax.barh(req_names, req_completion, color=colors)
    ax.set_xlabel('Completion (%)')
    ax.set_title('Requirement Completion Status')
    ax.set_xlim(0, 100)
    
    # Add value labels
    for bar in bars:
        width = bar.get_width()
        ax.text(width + 1, bar.get_y() + bar.get_height()/2,
                f'{width:.1f}%',
                ha='left', va='center')
    
    plt.tight_layout()
    return fig

def create_credits_distribution_visualization(requirements):
    """Create pie chart for credits distribution"""
    fig, ax = plt.subplots(figsize=(8, 8))
    
    core_credits = sum(req['completed'] for req in requirements if req['code'] in ['A-English', 'B-Math', 'C-Science', 'D-Social'])
    elective_credits = sum(req['completed'] for req in requirements if req['code'] in ['G-Electives'])
    other_credits = sum(req['completed'] for req in requirements if req['code'] in ['E-Arts', 'F-PE'])
    
    credit_values = [core_credits, elective_credits, other_credits]
    credit_labels = ['Core Subjects', 'Electives', 'Arts/PE']
    colors = ['#3498db', '#2ecc71', '#9b59b6']
    
    ax.pie(credit_values, labels=credit_labels, autopct='%1.1f%%', 
           colors=colors, startangle=90)
    ax.set_title('Credit Distribution')
    
    plt.tight_layout()
    return fig

# ========== TEXT EXTRACTION FUNCTIONS ==========
def preprocess_text(text: str) -> str:
    """Normalize text for more reliable parsing"""
    text = re.sub(r'\s+', ' ', text)  # Normalize whitespace
    text = text.upper()                # Standardize case for certain fields
    return text

def extract_text_from_file(file_path: str, file_ext: str) -> str:
    text = ""
    
    try:
        if file_ext == '.pdf':
            try:
                # First try pdfplumber for better table extraction
                student_info, requirements, courses = parse_transcript_pdf(file_path)
                if student_info:
                    # Convert parsed data to text format for compatibility
                    text += f"STUDENT INFORMATION:\n"
                    text += f"Name: {student_info.get('name', '')}\n"
                    text += f"ID: {student_info.get('id', '')}\n"
                    text += f"School: {student_info.get('school', '')}\n"
                    text += f"Grade: {student_info.get('grade', '')}\n"
                    text += f"Graduation Year: {student_info.get('grad_year', '')}\n"
                    text += f"Weighted GPA: {student_info.get('gpa_weighted', '')}\n"
                    text += f"Unweighted GPA: {student_info.get('gpa_unweighted', '')}\n"
                    text += f"Service Hours: {student_info.get('service_hours', '')}\n"
                    text += f"Total Credits: {student_info.get('total_credits', '')}\n\n"
                    
                    text += "GRADUATION REQUIREMENTS:\n"
                    for req in requirements:
                        text += f"{req['code']} | {req['desc']} | Required: {req['required']} | Completed: {req['completed']} | Status: {req['status']}%\n"
                    
                    text += "\nCOURSE HISTORY:\n"
                    for course in courses:
                        text += f"{course['course_code']} | {course['course_name']} | Grade: {course['grade_earned']} | Credits: {course['credits']} | Status: {course['status']}\n"
                    
                    return text
                
                # Fall back to regular text extraction if specialized parsing fails
                import pdfplumber
                with pdfplumber.open(file_path) as pdf:
                    for page in pdf.pages:
                        # Try to extract tables first
                        tables = page.extract_tables({
                            "vertical_strategy": "text", 
                            "horizontal_strategy": "text",
                            "intersection_y_tolerance": 10
                        })
                        
                        if tables:
                            for table in tables:
                                for row in table:
                                    text += " | ".join(str(cell).strip() for cell in row if cell) + "\n"
                        
                        # Fall back to text extraction if tables are empty
                        page_text = page.extract_text()
                        if page_text:
                            text += page_text + "\n"
                            
                if not text.strip():
                    raise ValueError("PDFPlumber returned empty text")
                    
            except Exception as e:
                logging.warning(f"PDFPlumber failed: {str(e)}. Trying PyMuPDF...")
                doc = fitz.open(file_path)
                for page in doc:
                    text += page.get_text("text") + '\n'
                    
        elif file_ext in ['.png', '.jpg', '.jpeg']:
            text = extract_text_with_ocr(file_path)
            
        text = clean_extracted_text(text)
        
        if not text.strip():
            raise ValueError("No text could be extracted.")
            
        return text
    
    except Exception as e:
        logging.error(f"Text extraction error: {str(e)}")
        raise ValueError(f"Failed to extract text: {str(e)}")

def extract_text_with_ocr(file_path: str) -> str:
    try:
        image = Image.open(file_path)
        image = image.convert('L')
        image = image.point(lambda x: 0 if x < 128 else 255, '1')
        custom_config = r'--oem 3 --psm 6'
        text = pytesseract.image_to_string(image, config=custom_config)
        return text
    except Exception as e:
        raise ValueError(f"OCR processing failed: {str(e)}")

def clean_extracted_text(text: str) -> str:
    """Special cleaning for Miami-Dade transcripts"""
    # Normalize whitespace
    text = re.sub(r'\s+', ' ', text).strip()
    
    # Fix common OCR errors
    replacements = {
        'GradeLv1': 'GradeLvl',
        'CrsNu m': 'CrsNum',
        'YOG': 'Year of Graduation',
        'Comm Serv': 'Community Service',
        r'\bA\s*-\s*': 'A-',  # Fix requirement codes
        r'\bB\s*-\s*': 'B-',
        r'\bC\s*-\s*': 'C-',
        r'\bD\s*-\s*': 'D-',
        r'\bE\s*-\s*': 'E-',
        r'\bF\s*-\s*': 'F-',
        r'\bG\s*-\s*': 'G-',
        r'\bZ\s*-\s*': 'Z-'
    }
    
    for pattern, replacement in replacements.items():
        text = re.sub(pattern, replacement, text)
    
    # Fix course codes with spaces
    text = re.sub(r'(\b[A-Z]{2,4})\s(\d{3}[A-Z]?\b)', r'\1\2', text)
    
    # Fix common OCR errors in credits
    text = re.sub(r'in\s*Progress', 'inProgress', text, flags=re.IGNORECASE)
    
    return text

def remove_sensitive_info(text: str) -> str:
    text = re.sub(r'\b\d{3}-\d{2}-\d{4}\b', '[REDACTED]', text)
    text = re.sub(r'\b\d{6,9}\b', '[ID]', text)
    text = re.sub(r'\b[A-Za-z0-9._%+-]+@[A-Za-z9.-]+\.[A-Z|a-z]{2,}\b', '[EMAIL]', text)
    return text

# ========== TRANSCRIPT PARSING ==========
class Course(BaseModel):
    requirement: str
    school_year: str
    grade_level: str
    course_code: str
    description: str
    term: str
    district_number: str
    fg: str
    included: str
    credits: str

class GraduationProgress(BaseModel):
    student_name: str
    student_id: str
    current_grade: str
    year_of_graduation: str
    unweighted_gpa: float
    weighted_gpa: float
    community_service_hours: int
    community_service_date: str
    total_credits_earned: float
    virtual_grade: str
    requirements: Dict[str, Dict[str, float]]
    courses: List[Course]
    assessments: Dict[str, str]

def validate_parsed_data(parsed_data: Dict) -> bool:
    """Ensure all critical fields exist"""
    required_fields = [
        ('student_info', 'name'),
        ('student_info', 'weighted_gpa'),
        ('requirements', 'A-English'),  # Sample requirement
        ('course_history', 0)  # At least one course
    ]
    
    for path in required_fields:
        current = parsed_data
        for key in path:
            if key not in current:
                raise ValueError(f"Missing critical field: {'.'.join(path)}")
            current = current[key]
    return True

class TranscriptParser:
    def __init__(self):
        self.student_data = {}
        self.requirements = {}
        self.current_courses = []
        self.course_history = []
        self.graduation_status = {}
        
    def parse_transcript(self, text: str) -> Dict:
        """Parse transcript text and return structured data"""
        try:
            text = preprocess_text(text)
            
            # First try the specialized Miami-Dade parser
            parsed_data = self._parse_miami_dade_transcript(text)
            if parsed_data:
                validate_parsed_data(parsed_data)
                return parsed_data
                
            # Fall back to simplified parser if detailed parsing fails
            parsed_data = self._parse_simplified_transcript(text)
            if parsed_data:
                return parsed_data
                
            raise ValueError("No data could be parsed from the transcript")
            
        except Exception as e:
            logging.error(f"Error parsing transcript: {str(e)}")
            raise ValueError(f"Couldn't parse transcript content. Error: {str(e)}")

    def _parse_miami_dade_transcript(self, text: str) -> Optional[Dict]:
        """Specialized parser for Miami-Dade County Public Schools transcripts"""
        try:
            parsed_data = {
                'student_info': {},
                'requirements': {},
                'course_history': [],
                'assessments': {}
            }
            
            # Extract student info with more robust pattern
            student_info_match = re.search(
                r"(\d{7})\s*-\s*(.*?)\s*\n.*?Current Grade:\s*(\d+).*?YOG\s*(\d{4})", 
                text, 
                re.DOTALL
            )
            if student_info_match:
                parsed_data['student_info']['id'] = student_info_match.group(1)
                parsed_data['student_info']['name'] = student_info_match.group(2).strip()
                parsed_data['student_info']['grade'] = student_info_match.group(3)
                parsed_data['student_info']['year_of_graduation'] = student_info_match.group(4)
            
            # Extract GPA information
            gpa_matches = re.findall(
                r"(?:Un.?weighted|Weighted)\s*GPA\s*([\d.]+)", 
                text, 
                re.IGNORECASE
            )
            if len(gpa_matches) >= 1:
                parsed_data['student_info']['unweighted_gpa'] = float(gpa_matches[0])
            if len(gpa_matches) >= 2:
                parsed_data['student_info']['weighted_gpa'] = float(gpa_matches[1])
            
            # Extract community service info
            service_hours_match = re.search(r"Comm\s*Serv\s*Hours\s*(\d+)", text, re.IGNORECASE)
            if service_hours_match:
                parsed_data['student_info']['community_service_hours'] = int(service_hours_match.group(1))
                
            service_date_match = re.search(r"Comm\s*Serv\s*Date\s*(\d{2}/\d{2}/\d{4})", text, re.IGNORECASE)
            if service_date_match:
                parsed_data['student_info']['community_service_date'] = service_date_match.group(1)
            
            # Extract credits info
            credits_match = re.search(r"Total\s*Credits\s*Earned\s*([\d.]+)", text, re.IGNORECASE)
            if credits_match:
                parsed_data['student_info']['total_credits'] = float(credits_match.group(1))
            
            # Extract virtual grade
            virtual_grade_match = re.search(r"Virtual\s*Grade\s*([A-Z])", text, re.IGNORECASE)
            if virtual_grade_match:
                parsed_data['student_info']['virtual_grade'] = virtual_grade_match.group(1)
            
            # Extract requirements section - more robust table parsing
            req_section = re.search(
                r"Code\s*Description\s*Required\s*Waived\s*Completed\s*Status(.*?)(?:\n\s*\n|$)", 
                text, 
                re.DOTALL | re.IGNORECASE
            )
            
            if req_section:
                req_lines = [line.strip() for line in req_section.group(1).split('\n') if line.strip()]
                for line in req_lines:
                    if '|' in line:  # Table format
                        parts = [part.strip() for part in line.split('|') if part.strip()]
                        if len(parts) >= 5:  # More lenient check for number of columns
                            try:
                                code = parts[0] if len(parts) > 0 else ""
                                description = parts[1] if len(parts) > 1 else ""
                                required = float(parts[2]) if len(parts) > 2 and parts[2].replace('.','').isdigit() else 0.0
                                waived = float(parts[3]) if len(parts) > 3 and parts[3].replace('.','').isdigit() else 0.0
                                completed = float(parts[4]) if len(parts) > 4 and parts[4].replace('.','').isdigit() else 0.0
                                status = parts[5] if len(parts) > 5 else ""
                                
                                # Extract percentage if available
                                percent = 0.0
                                if status:
                                    percent_match = re.search(r"(\d+)%", status)
                                    if percent_match:
                                        percent = float(percent_match.group(1))
                                
                                parsed_data['requirements'][code] = {
                                    "description": description,
                                    "required": required,
                                    "waived": waived,
                                    "completed": completed,
                                    "percent_complete": percent,
                                    "status": status
                                }
                            except (IndexError, ValueError) as e:
                                logging.warning(f"Skipping malformed requirement line: {line}. Error: {str(e)}")
                                continue
            
            # Extract assessments section
            assess_section = re.search(r"Z-Assessment.*?\n(.*?)(?:\n\s*\n|$)", text, re.DOTALL | re.IGNORECASE)
            if assess_section:
                assess_lines = [line.strip() for line in assess_section.group(1).split('\n') if line.strip()]
                for line in assess_lines:
                    if '|' in line:
                        parts = [part.strip() for part in line.split('|') if part.strip()]
                        if len(parts) >= 5 and parts[0].startswith('Z-'):
                            name = parts[0].replace('Z-', '').strip()
                            status = parts[4] if len(parts) > 4 else ""
                            parsed_data['assessments'][name] = status
            
            # Extract course history with more fault-tolerant parsing
            course_section = re.search(
                r"Requirement.*?School Year.*?GradeLv1.*?CrsNum.*?Description.*?Term.*?DstNumber.*?FG.*?Incl.*?Credits(.*?)(?:Legend|\Z)", 
                text, 
                re.DOTALL | re.IGNORECASE
            )
            
            if course_section:
                course_lines = [
                    line.strip() for line in course_section.group(1).split('\n') 
                    if line.strip() and '|' in line
                ]
                
                for line in course_lines:
                    parts = [part.strip() for part in line.split('|') if part.strip()]
                    
                    # More robust handling of course data
                    try:
                        course = {
                            'requirement': parts[0] if len(parts) > 0 else "",
                            'school_year': parts[1] if len(parts) > 1 else "",
                            'grade_level': parts[2] if len(parts) > 2 else "",
                            'course_code': parts[3] if len(parts) > 3 else "",
                            'description': parts[4] if len(parts) > 4 else "",
                            'term': parts[5] if len(parts) > 5 else "",
                            'district_number': parts[6] if len(parts) > 6 else "",
                            'fg': parts[7] if len(parts) > 7 else "",
                            'included': parts[8] if len(parts) > 8 else "",
                            'credits': parts[9] if len(parts) > 9 else "0"
                        }
                        
                        # Handle "inProgress" and empty credits
                        if "inprogress" in course['credits'].lower() or not course['credits']:
                            course['credits'] = "0"
                        elif not course['credits'].replace('.','').isdigit():
                            course['credits'] = "0"
                            
                        parsed_data['course_history'].append(course)
                    except (IndexError, ValueError) as e:
                        logging.warning(f"Skipping malformed course line: {line}. Error: {str(e)}")
                        continue
            
            return parsed_data
            
        except Exception as e:
            logging.warning(f"Miami-Dade transcript parsing failed: {str(e)}")
            return None

    def _parse_simplified_transcript(self, text: str) -> Dict:
        """Fallback simplified transcript parser with multiple pattern attempts"""
        patterns = [
            (r'(?:COURSE|SUBJECT)\s*CODE.*?GRADE.*?CREDITS(.*?)(?:\n\s*\n|\Z)', 'table'),
            (r'([A-Z]{2,4}\s?\d{3}[A-Z]?)\s+(.*?)\s+([A-F][+-]?)\s+(\d+\.?\d*)', 'line'),
            (r'(.*?)\s+([A-F][+-]?)\s+(\d+\.?\d*)', 'minimal')
        ]
        
        for pattern, pattern_type in patterns:
            try:
                if pattern_type == 'table':
                    # Parse tabular data
                    table_section = re.search(pattern, text, re.DOTALL | re.IGNORECASE)
                    if table_section:
                        courses = re.findall(r'([A-Z]{2,4}\s?\d{3}[A-Z]?)\s+(.*?)\s+([A-F][+-]?)\s+(\d+\.?\d*)', 
                                           table_section.group(1))
                elif pattern_type == 'line':
                    courses = re.findall(pattern, text)
                else:
                    courses = re.findall(pattern, text)
                    
                if courses:
                    parsed_data = {'course_history': []}
                    for course in courses:
                        if len(course) >= 4:
                            parsed_data['course_history'].append({
                                'course_code': course[0].strip(),
                                'description': course[1].strip(),
                                'grade': course[2].strip(),
                                'credits': float(course[3]) if course[3] else 0.0
                            })
                        elif len(course) == 3:
                            parsed_data['course_history'].append({
                                'description': course[0].strip(),
                                'grade': course[1].strip(),
                                'credits': float(course[2]) if course[2] else 0.0
                            })
                    return parsed_data
            except Exception as e:
                logging.warning(f"Pattern {pattern} failed: {str(e)}")
                continue
        
        return None

# ========== ENHANCED ANALYSIS FUNCTIONS ==========
def analyze_gpa(parsed_data: Dict) -> str:
    try:
        gpa = float(parsed_data.get('student_info', {}).get('weighted_gpa', 0))
        if gpa >= 4.5:
            return "🌟 Excellent GPA! You're in the top tier of students."
        elif gpa >= 3.5:
            return "👍 Good GPA! You're performing above average."
        elif gpa >= 2.5:
            return "ℹ️ Average GPA. Consider focusing on improvement in weaker areas."
        else:
            return "⚠️ Below average GPA. Please consult with your academic advisor."
    except (TypeError, ValueError, KeyError, AttributeError):
        return "❌ Could not analyze GPA."

def analyze_graduation_status(parsed_data: Dict) -> str:
    try:
        total_required = sum(
            float(req.get('required', 0))
            for req in parsed_data.get('requirements', {}).values() 
            if req and str(req.get('required', '0')).replace('.', '').isdigit()
        )
        
        total_completed = sum(
            float(req.get('completed', 0))
            for req in parsed_data.get('requirements', {}).values() 
            if req and str(req.get('completed', '0')).replace('.', '').isdigit()
        )
        
        completion_percentage = (total_completed / total_required) * 100 if total_required > 0 else 0
        
        if completion_percentage >= 100:
            return "🎉 You've met all graduation requirements!"
        elif completion_percentage >= 80:
            return f"✅ You've completed {completion_percentage:.1f}% of requirements. Almost there!"
        elif completion_percentage >= 50:
            return f"🔄 You've completed {completion_percentage:.1f}% of requirements. Keep working!"
        else:
            return f"⚠️ You've only completed {completion_percentage:.1f}% of requirements. Please meet with your counselor."
    except (ZeroDivisionError, TypeError, KeyError, AttributeError):
        return "❌ Could not analyze graduation status."

def generate_advice(parsed_data: Dict) -> str:
    advice = []
    
    # GPA advice
    try:
        gpa = float(parsed_data.get('student_info', {}).get('weighted_gpa', 0))
        if gpa < 3.0:
            advice.append("📚 Your GPA could improve. Consider:\n- Seeking tutoring for challenging subjects\n- Meeting with teachers during office hours\n- Developing better study habits")
    except (TypeError, ValueError, KeyError, AttributeError):
        pass
    
    # Community service advice
    try:
        service_hours = int(parsed_data.get('student_info', {}).get('community_service_hours', 0))
        if service_hours < 100:
            advice.append("🤝 Consider more community service:\n- Many colleges value 100+ hours\n- Look for opportunities that align with your interests")
    except (TypeError, ValueError, KeyError, AttributeError):
        pass
    
    # Missing requirements advice
    try:
        missing_reqs = [
            req for code, req in parsed_data.get('requirements', {}).items()
            if req and float(req.get('percent_complete', 0)) < 100 and not code.startswith("Z-Assessment")
        ]
        
        if missing_reqs:
            req_list = "\n- ".join([f"{code}: {req.get('description', '')}" for code, req in missing_reqs])
            advice.append(f"🎓 Focus on completing these requirements:\n- {req_list}")
    except (TypeError, ValueError, KeyError, AttributeError):
        pass
    
    # Course rigor advice
    try:
        ap_count = sum(1 for course in parsed_data.get('course_history', []) 
                      if course and "ADVANCED PLACEMENT" in course.get('description', '').upper())
        if ap_count < 3:
            advice.append("🧠 Consider taking more challenging courses:\n- AP/IB courses can strengthen college applications\n- Shows academic rigor to admissions officers")
    except (TypeError, KeyError, AttributeError):
        pass
    
    return "\n\n".join(advice) if advice else "🎯 You're on track! Keep up the good work."

def generate_college_recommendations(parsed_data: Dict) -> str:
    try:
        gpa = float(parsed_data.get('student_info', {}).get('weighted_gpa', 0))
        ap_count = sum(1 for course in parsed_data.get('course_history', []) 
                      if course and "ADVANCED PLACEMENT" in course.get('description', '').upper())
        service_hours = int(parsed_data.get('student_info', {}).get('community_service_hours', 0))
        
        recommendations = []
        
        if gpa >= 4.0 and ap_count >= 5:
            recommendations.append("🏛️ Reach Schools: Ivy League, Stanford, MIT, etc.")
        if gpa >= 3.7:
            recommendations.append("🎓 Competitive Schools: Top public universities, selective private colleges")
        if gpa >= 3.0:
            recommendations.append("📚 Good Match Schools: State flagship universities, many private colleges")
        if gpa >= 2.0:
            recommendations.append("🏫 Safety Schools: Community colleges, open admission universities")
            
        # Add scholarship opportunities
        if gpa >= 3.5:
            recommendations.append("\n💰 Scholarship Opportunities:\n- Bright Futures (Florida)\n- National Merit Scholarship\n- College-specific merit scholarships")
        elif gpa >= 3.0:
            recommendations.append("\n💰 Scholarship Opportunities:\n- Local community scholarships\n- Special interest scholarships\n- First-generation student programs")
            
        # Add extracurricular advice
        if service_hours < 50:
            recommendations.append("\n🎭 Extracurricular Advice:\n- Colleges value depth over breadth in activities\n- Consider leadership roles in 1-2 organizations")
            
        if not recommendations:
            return "❌ Not enough data to generate college recommendations"
            
        return "Based on your academic profile:\n\n" + "\n\n".join(recommendations)
    except:
        return "❌ Could not generate college recommendations"

def create_gpa_visualization(parsed_data: Dict):
    try:
        gpa_data = {
            "Type": ["Weighted GPA", "Unweighted GPA"],
            "Value": [
                float(parsed_data.get('student_info', {}).get('weighted_gpa', 0)),
                float(parsed_data.get('student_info', {}).get('unweighted_gpa', 0))
            ]
        }
        df = pd.DataFrame(gpa_data)
        fig = px.bar(df, x="Type", y="Value", title="GPA Comparison", 
                     color="Type", text="Value",
                     color_discrete_sequence=["#4C78A8", "#F58518"])
        fig.update_traces(texttemplate='%{text:.2f}', textposition='outside')
        fig.update_layout(yaxis_range=[0,5], uniformtext_minsize=8, uniformtext_mode='hide')
        return fig
    except:
        return None

def create_requirements_visualization(parsed_data: Dict):
    try:
        req_data = []
        for code, req in parsed_data.get('requirements', {}).items():
            if req and req.get('percent_complete'):
                completion = float(req['percent_complete'])
                req_data.append({
                    "Requirement": code,
                    "Completion (%)": completion,
                    "Status": "Complete" if completion >= 100 else "Incomplete"
                })
        
        if not req_data:
            return None
            
        df = pd.DataFrame(req_data)
        fig = px.bar(df, x="Requirement", y="Completion (%)", 
                     title="Graduation Requirements Completion",
                     color="Status",
                     color_discrete_map={"Complete": "#2CA02C", "Incomplete": "#D62728"},
                     hover_data=["Requirement"])
        fig.update_layout(xaxis={'categoryorder':'total descending'})
        return fig
    except:
        return None

def parse_transcript(file_obj, progress=gr.Progress()) -> Tuple[str, Optional[Dict]]:
    """Process transcript file and return analysis results"""
    try:
        if not file_obj:
            raise gr.Error("Please upload a transcript file first (PDF or image)")
            
        validate_file(file_obj)
        file_ext = os.path.splitext(file_obj.name)[1].lower()
        
        # Additional PDF validation
        if file_ext == '.pdf':
            try:
                with open(file_obj.name, 'rb') as f:
                    PdfReader(f)  # Test if PDF is readable
            except Exception as e:
                raise gr.Error(f"Invalid PDF file: {str(e)}. Please upload a non-corrupted PDF.")
        
        if progress:
            progress(0.2, desc="Extracting text from file...")
        
        try:
            text = extract_text_from_file(file_obj.name, file_ext)
        except Exception as e:
            raise ValueError(f"Failed to extract text: {str(e)}. The file may be corrupted or in an unsupported format.")
        
        if not text.strip():
            raise ValueError("The file appears to be empty or contains no readable text.")
        
        if progress:
            progress(0.5, desc="Parsing transcript...")
            
        parser = TranscriptParser()
        try:
            parsed_data = parser.parse_transcript(text)
            if not parsed_data:
                raise ValueError("No data could be parsed from the transcript.")
        except Exception as e:
            raise ValueError(f"Couldn't parse transcript content. Error: {str(e)}")
        
        # Perform enhanced analyses
        gpa_analysis = analyze_gpa(parsed_data)
        grad_status = analyze_graduation_status(parsed_data)
        advice = generate_advice(parsed_data)
        college_recs = generate_college_recommendations(parsed_data)
        gpa_viz = create_gpa_visualization(parsed_data)
        req_viz = create_requirements_visualization(parsed_data)
        
        # Format results for display
        results = [
            f"📊 GPA Analysis: {gpa_analysis}",
            f"🎓 Graduation Status: {grad_status}",
            f"💡 Recommendations:\n{advice}",
            f"🏫 College Recommendations:\n{college_recs}"
        ]
        
        # Store all analysis results in the parsed_data
        parsed_data['analysis'] = {
            'gpa_analysis': gpa_analysis,
            'grad_status': grad_status,
            'advice': advice,
            'college_recs': college_recs,
            'visualizations': {
                'gpa_viz': gpa_viz,
                'req_viz': req_viz
            }
        }
        
        return "\n\n".join(results), parsed_data
        
    except Exception as e:
        error_msg = f"Error processing transcript: {str(e)}"
        logging.error(error_msg)
        raise gr.Error(f"{error_msg}\n\nPossible solutions:\n1. Try a different file format\n2. Ensure text is clear and not handwritten\n3. Check file size (<5MB)")

# ========== LEARNING STYLE QUIZ ==========
class LearningStyleQuiz:
    def __init__(self):
        self.questions = [
            "When you study for a test, you prefer to:",
            "When you need directions to a new place, you prefer:",
            "When you learn a new skill, you prefer to:",
            "When you're trying to concentrate, you:",
            "When you meet new people, you remember them by:",
            "When you're assembling furniture or a gadget, you:",
            "When choosing a restaurant, you rely most on:",
            "When you're in a waiting room, you typically:",
            "When giving someone instructions, you tend to:",
            "When you're trying to recall information, you:",
            "When you're at a museum or exhibit, you:",
            "When you're learning a new language, you prefer:",
            "When you're taking notes in class, you:",
            "When you're explaining something complex, you:",
            "When you're at a party, you enjoy:",
            "When you're trying to remember a phone number, you:",
            "When you're relaxing, you prefer to:",
            "When you're learning to use new software, you:",
            "When you're giving a presentation, you rely on:",
            "When you're solving a difficult problem, you:"
        ]
        
        self.options = [
            ["Read the textbook (Reading/Writing)", "Listen to lectures (Auditory)", "Use diagrams/charts (Visual)", "Practice problems (Kinesthetic)"],
            ["Look at a map (Visual)", "Have someone tell you (Auditory)", "Write down directions (Reading/Writing)", "Try walking/driving there (Kinesthetic)"],
            ["Read instructions (Reading/Writing)", "Have someone show you (Visual)", "Listen to explanations (Auditory)", "Try it yourself (Kinesthetic)"],
            ["Need quiet (Reading/Writing)", "Need background noise (Auditory)", "Need to move around (Kinesthetic)", "Need visual stimulation (Visual)"],
            ["Their face (Visual)", "Their name (Auditory)", "What you talked about (Reading/Writing)", "What you did together (Kinesthetic)"],
            ["Read the instructions carefully (Reading/Writing)", "Look at the diagrams (Visual)", "Ask someone to explain (Auditory)", "Start putting pieces together (Kinesthetic)"],
            ["Online photos of the food (Visual)", "Recommendations from friends (Auditory)", "Reading the menu online (Reading/Writing)", "Remembering how it felt to eat there (Kinesthetic)"],
            ["Read magazines (Reading/Writing)", "Listen to music (Auditory)", "Watch TV (Visual)", "Fidget or move around (Kinesthetic)"],
            ["Write them down (Reading/Writing)", "Explain verbally (Auditory)", "Demonstrate (Visual)", "Guide them physically (Kinesthetic)"],
            ["See written words in your mind (Visual)", "Hear the information in your head (Auditory)", "Write it down to remember (Reading/Writing)", "Associate it with physical actions (Kinesthetic)"],
            ["Read all the descriptions (Reading/Writing)", "Listen to audio guides (Auditory)", "Look at the displays (Visual)", "Touch interactive exhibits (Kinesthetic)"],
            ["Study grammar rules (Reading/Writing)", "Listen to native speakers (Auditory)", "Use flashcards with images (Visual)", "Practice conversations (Kinesthetic)"],
            ["Write detailed paragraphs (Reading/Writing)", "Record the lecture (Auditory)", "Draw diagrams and charts (Visual)", "Doodle while listening (Kinesthetic)"],
            ["Write detailed steps (Reading/Writing)", "Explain verbally with examples (Auditory)", "Draw diagrams (Visual)", "Use physical objects to demonstrate (Kinesthetic)"],
            ["Conversations with people (Auditory)", "Watching others or the environment (Visual)", "Writing notes or texting (Reading/Writing)", "Dancing or physical activities (Kinesthetic)"],
            ["See the numbers in your head (Visual)", "Say them aloud (Auditory)", "Write them down (Reading/Writing)", "Dial them on a keypad (Kinesthetic)"],
            ["Read a book (Reading/Writing)", "Listen to music (Auditory)", "Watch TV/movies (Visual)", "Do something physical (Kinesthetic)"],
            ["Read the manual (Reading/Writing)", "Ask someone to show you (Visual)", "Call tech support (Auditory)", "Experiment with the software (Kinesthetic)"],
            ["Detailed notes (Reading/Writing)", "Verbal explanations (Auditory)", "Visual slides (Visual)", "Physical demonstrations (Kinesthetic)"],
            ["Write out possible solutions (Reading/Writing)", "Talk through it with someone (Auditory)", "Draw diagrams (Visual)", "Build a model or prototype (Kinesthetic)"]
        ]
        
        self.learning_styles = {
            "Visual": {
                "description": "Visual learners prefer using images, diagrams, and spatial understanding.",
                "tips": [
                    "Use color coding in your notes",
                    "Create mind maps and diagrams",
                    "Watch educational videos",
                    "Use flashcards with images",
                    "Highlight important information in different colors"
                ],
                "careers": [
                    "Graphic Designer", "Architect", "Photographer", 
                    "Engineer", "Surgeon", "Pilot"
                ]
            },
            "Auditory": {
                "description": "Auditory learners learn best through listening and speaking.",
                "tips": [
                    "Record lectures and listen to them",
                    "Participate in study groups",
                    "Explain concepts out loud to yourself",
                    "Use rhymes or songs to remember information",
                    "Listen to educational podcasts"
                ],
                "careers": [
                    "Musician", "Journalist", "Lawyer",
                    "Psychologist", "Teacher", "Customer Service"
                ]
            },
            "Reading/Writing": {
                "description": "These learners prefer information displayed as words.",
                "tips": [
                    "Write detailed notes",
                    "Create summaries in your own words",
                    "Read textbooks and articles",
                    "Make lists to organize information",
                    "Rewrite your notes to reinforce learning"
                ],
                "careers": [
                    "Writer", "Researcher", "Editor",
                    "Accountant", "Programmer", "Historian"
                ]
            },
            "Kinesthetic": {
                "description": "Kinesthetic learners learn through movement and hands-on activities.",
                "tips": [
                    "Use hands-on activities",
                    "Take frequent movement breaks",
                    "Create physical models",
                    "Associate information with physical actions",
                    "Study while walking or pacing"
                ],
                "careers": [
                    "Athlete", "Chef", "Mechanic",
                    "Dancer", "Physical Therapist", "Carpenter"
                ]
            }
        }
    
    def evaluate_quiz(self, *answers) -> str:
        """Evaluate quiz answers and return learning style results"""
        answers = list(answers)
        if len(answers) != len(self.questions):
            raise gr.Error("Please answer all questions before submitting")
        
        scores = {style: 0 for style in self.learning_styles}
        
        for i, answer in enumerate(answers):
            if not answer:
                continue
                
            for j, style in enumerate(self.learning_styles):
                if answer == self.options[i][j]:
                    scores[style] += 1
                    break
        
        total_answered = sum(1 for ans in answers if ans)
        if total_answered == 0:
            raise gr.Error("No answers provided")
        
        percentages = {style: (score/total_answered)*100 for style, score in scores.items()}
        sorted_styles = sorted(scores.items(), key=lambda x: x[1], reverse=True)
        
        result = "## Your Learning Style Results\n\n"
        result += "### Scores:\n"
        for style, score in sorted_styles:
            result += f"- **{style}**: {score}/{total_answered} ({percentages[style]:.1f}%)\n"
        
        max_score = max(scores.values())
        primary_styles = [style for style, score in scores.items() if score == max_score]
        
        result += "\n### Analysis:\n"
        if len(primary_styles) == 1:
            primary_style = primary_styles[0]
            style_info = self.learning_styles[primary_style]
            
            result += f"Your primary learning style is **{primary_style}**\n\n"
            result += f"**{primary_style} Characteristics**:\n"
            result += f"{style_info['description']}\n\n"
            
            result += "**Recommended Study Strategies**:\n"
            for tip in style_info['tips']:
                result += f"- {tip}\n"
            
            result += "\n**Potential Career Paths**:\n"
            for career in style_info['careers'][:6]:
                result += f"- {career}\n"
            
            complementary = [s for s in sorted_styles if s[0] != primary_style][0][0]
            result += f"\nYou might also benefit from some **{complementary}** strategies:\n"
            for tip in self.learning_styles[complementary]['tips'][:3]:
                result += f"- {tip}\n"
        else:
            result += "You have multiple strong learning styles:\n"
            for style in primary_styles:
                result += f"- **{style}**\n"
            
            result += "\n**Combined Learning Strategies**:\n"
            result += "You may benefit from combining different learning approaches:\n"
            for style in primary_styles:
                result += f"\n**{style}** techniques:\n"
                for tip in style_info['tips'][:2]:
                    result += f"- {tip}\n"
                
                result += f"\n**{style}** career suggestions:\n"
                for career in style_info['careers'][:3]:
                    result += f"- {career}\n"
        
        return result

learning_style_quiz = LearningStyleQuiz()

# ========== PROFILE MANAGEMENT ==========
class ProfileManager:
    def __init__(self):
        self.profiles_dir = Path(PROFILES_DIR)
        self.profiles_dir.mkdir(exist_ok=True, parents=True)
        self.current_session = None
    
    def set_session(self, session_token: str) -> None:
        self.current_session = session_token
    
    def get_profile_path(self, name: str) -> Path:
        if self.current_session:
            name_hash = hashlib.sha256(name.encode()).hexdigest()[:16]
            return self.profiles_dir / f"{name_hash}_{self.current_session}_profile.json"
        return self.profiles_dir / f"{name.replace(' ', '_')}_profile.json"
    
    def save_profile(self, name: str, age: Union[int, str], interests: str, 
                    transcript: Dict, learning_style: str, 
                    movie: str, movie_reason: str, show: str, show_reason: str, 
                    book: str, book_reason: str, character: str, character_reason: str, 
                    blog: str) -> str:
        try:
            name = validate_name(name)
            age = validate_age(age)
            
            if not interests.strip():
                raise ValueError("Please describe at least one interest or hobby.")
            
            if not transcript:
                raise ValueError("Please complete the transcript analysis first.")
            
            if not learning_style or "Your primary learning style is:" not in learning_style:
                raise ValueError("Please complete the learning style quiz first.")
            
            favorites = {
                "movie": sanitize_input(movie),
                "movie_reason": sanitize_input(movie_reason),
                "show": sanitize_input(show),
                "show_reason": sanitize_input(show_reason),
                "book": sanitize_input(book),
                "book_reason": sanitize_input(book_reason),
                "character": sanitize_input(character),
                "character_reason": sanitize_input(character_reason)
            }
            
            data = {
                "name": name,
                "age": age,
                "interests": sanitize_input(interests),
                "transcript": transcript,
                "learning_style": learning_style,
                "favorites": favorites,
                "blog": sanitize_input(blog) if blog else "",
                "session_token": self.current_session,
                "last_updated": time.time()
            }
            
            filepath = self.get_profile_path(name)
            
            with open(filepath, "w", encoding='utf-8') as f:
                json.dump(data, f, indent=2, ensure_ascii=False)
            
            if HF_TOKEN and 'hf_api' in globals():
                try:
                    hf_api.upload_file(
                        path_or_fileobj=filepath,
                        path_in_repo=f"profiles/{filepath.name}",
                        repo_id="your-username/student-learning-assistant",
                        repo_type="dataset"
                    )
                except Exception as e:
                    logging.error(f"Failed to upload to HF Hub: {str(e)}")
            
            # Return simple confirmation with GPA if available
            confirmation = f"Profile saved successfully for {name}."
            if 'gpa' in data.get('transcript', {}).get('student_info', {}):
                confirmation += f"\nGPA: {data['transcript']['student_info']['gpa']}"
            return confirmation
        
        except Exception as e:
            logging.error(f"Profile validation error: {str(e)}")
            raise gr.Error(f"Couldn't save profile: {str(e)}")

    def load_profile(self, name: str = None, session_token: str = None) -> Dict:
        try:
            if session_token:
                profile_pattern = f"*{session_token}_profile.json"
            else:
                profile_pattern = "*.json"
            
            profiles = list(self.profiles_dir.glob(profile_pattern))
            if not profiles:
                return {}
            
            if name:
                name_hash = hashlib.sha256(name.encode()).hexdigest()[:16]
                if session_token:
                    profile_file = self.profiles_dir / f"{name_hash}_{session_token}_profile.json"
                else:
                    profile_file = self.profiles_dir / f"{name_hash}_profile.json"
                
                if not profile_file.exists():
                    if HF_TOKEN and 'hf_api' in globals():
                        try:
                            hf_api.download_file(
                                path_in_repo=f"profiles/{profile_file.name}",
                                repo_id="your-username/student-learning-assistant",
                                repo_type="dataset",
                                local_dir=self.profiles_dir
                            )
                        except:
                            raise gr.Error(f"No profile found for {name}")
                    else:
                        raise gr.Error(f"No profile found for {name}")
            else:
                profile_file = profiles[0]
            
            with open(profile_file, "r", encoding='utf-8') as f:
                profile_data = json.load(f)
                if time.time() - profile_data.get('last_updated', 0) > SESSION_TIMEOUT:
                    raise gr.Error("Session expired. Please start a new session.")
                return profile_data
        
        except Exception as e:
            logging.error(f"Error loading profile: {str(e)}")
            return {}
    
    def list_profiles(self, session_token: str = None) -> List[str]:
        if session_token:
            profiles = list(self.profiles_dir.glob(f"*{session_token}_profile.json"))
        else:
            profiles = list(self.profiles_dir.glob("*.json"))
        
        profile_names = []
        for p in profiles:
            with open(p, "r", encoding='utf-8') as f:
                try:
                    data = json.load(f)
                    profile_names.append(data.get('name', p.stem))
                except json.JSONDecodeError:
                    continue
        
        return profile_names

profile_manager = ProfileManager()

# ========== AI TEACHING ASSISTANT ==========
class TeachingAssistant:
    def __init__(self):
        self.context_history = []
        self.max_context_length = 5
    
    async def generate_response(self, message: str, history: List[List[Union[str, None]]], session_token: str) -> str:
        try:
            profile = profile_manager.load_profile(session_token=session_token)
            if not profile:
                return "Please complete and save your profile first."
            
            self._update_context(message, history)
            
            # Focus on GPA if mentioned
            if "gpa" in message.lower():
                gpa = profile.get("transcript", {}).get("student_info", {}).get("gpa", "unknown")
                return f"Your GPA is {gpa}. Would you like advice on improving it?"
            
            # Generic response otherwise
            return "I'm your learning assistant. Ask me about your GPA, courses, or study tips."
        
        except Exception as e:
            logging.error(f"Error generating response: {str(e)}")
            return "I encountered an error. Please try again."
    
    def _update_context(self, message: str, history: List[List[Union[str, None]]]) -> None:
        self.context_history.append({"role": "user", "content": message})
        if history:
            for h in history[-self.max_context_length:]:
                if h[0]:
                    self.context_history.append({"role": "user", "content": h[0]})
                if h[1]:
                    self.context_history.append({"role": "assistant", "content": h[1]})
        
        self.context_history = self.context_history[-(self.max_context_length*2):]

teaching_assistant = TeachingAssistant()

# ========== GRADIO INTERFACE ==========
def create_interface():
    with gr.Blocks(theme=gr.themes.Soft(), title="Student Learning Assistant") as app:
        session_token = gr.State(value=generate_session_token())
        profile_manager.set_session(session_token.value)
        
        tab_completed = gr.State({
            0: False,  # Transcript Upload
            1: False,  # Learning Style Quiz
            2: False,  # Personal Questions
            3: False,  # Save & Review
            4: False   # AI Assistant
        })
        
        # Custom CSS
        app.css = """
        .gradio-container { max-width: 1200px !important; margin: 0 auto !important; }
        .tab-content { padding: 20px !important; border: 1px solid #e0e0e0 !important; border-radius: 8px !important; margin-top: 10px !important; }
        .completed-tab { background: #4CAF50 !important; color: white !important; }
        .incomplete-tab { background: #E0E0E0 !important; }
        .nav-message { padding: 10px; margin: 10px 0; border-radius: 4px; background-color: #ffebee; color: #c62828; }
        .file-upload { border: 2px dashed #4CAF50 !important; padding: 20px !important; border-radius: 8px !important; text-align: center; }
        .file-upload:hover { background: #f5f5f5; }
        .progress-bar { height: 5px; background: linear-gradient(to right, #4CAF50, #8BC34A); margin-bottom: 15px; border-radius: 3px; }
        .quiz-question { margin-bottom: 15px; padding: 15px; background: #f5f5f5; border-radius: 5px; }
        .quiz-results { margin-top: 20px; padding: 20px; background: #e8f5e9; border-radius: 8px; }
        .error-message { color: #d32f2f; background-color: #ffebee; padding: 10px; border-radius: 4px; margin: 10px 0; }
        .transcript-results { border-left: 4px solid #4CAF50 !important; padding: 15px !important; background: #f8f8f8 !important; }
        .error-box { border: 1px solid #ff4444 !important; background: #fff8f8 !important; }
        .metric-box { background-color: white; border-radius: 10px; padding: 15px; margin: 10px 0; box-shadow: 0 2px 5px rgba(0,0,0,0.1); }
        .recommendation { background-color: #fff8e1; padding: 10px; border-left: 4px solid #ffc107; margin: 5px 0; }
        
        .dark .tab-content { background-color: #2d2d2d !important; border-color: #444 !important; }
        .dark .quiz-question { background-color: #3d3d3d !important; }
        .dark .quiz-results { background-color: #2e3d2e !important; }
        .dark textarea, .dark input { background-color: #333 !important; color: #eee !important; }
        .dark .output-markdown { color: #eee !important; }
        .dark .chatbot { background-color: #333 !important; }
        .dark .chatbot .user, .dark .chatbot .assistant { color: #eee !important; }
        .dark .metric-box { background-color: #333 !important; }
        """
        
        # Header
        with gr.Row():
            with gr.Column(scale=4):
                gr.Markdown("""
                # Student Learning Assistant
                **Your personalized education companion**  
                Complete each step to get customized learning recommendations.
                """)
            with gr.Column(scale=1):
                dark_mode = gr.Checkbox(label="Dark Mode", value=False)
        
        # Navigation buttons
        with gr.Row():
            with gr.Column(scale=1, min_width=100):
                step1 = gr.Button("1. Transcript", elem_classes="incomplete-tab")
            with gr.Column(scale=1, min_width=100):
                step2 = gr.Button("2. Quiz", elem_classes="incomplete-tab", interactive=False)
            with gr.Column(scale=1, min_width=100):
                step3 = gr.Button("3. Profile", elem_classes="incomplete-tab", interactive=False)
            with gr.Column(scale=1, min_width=100):
                step4 = gr.Button("4. Review", elem_classes="incomplete-tab", interactive=False)
            with gr.Column(scale=1, min_width=100):
                step5 = gr.Button("5. Assistant", elem_classes="incomplete-tab", interactive=False)

        nav_message = gr.HTML(visible=False)

        # Main tabs
        with gr.Tabs(visible=True) as tabs:
            # ===== TAB 1: TRANSCRIPT UPLOAD =====
            with gr.Tab("Transcript", id=0):
                with gr.Row():
                    with gr.Column(scale=1):
                        gr.Markdown("### Step 1: Upload Your Transcript")
                        with gr.Group(elem_classes="file-upload"):
                            file_input = gr.File(
                                label="Drag and drop your transcript here (PDF or Image)",
                                file_types=ALLOWED_FILE_TYPES,
                                type="filepath"
                            )
                            upload_btn = gr.Button("Analyze Transcript", variant="primary")
                            file_error = gr.HTML(visible=False)
                    
                    with gr.Column(scale=2):
                        transcript_output = gr.Textbox(
                            label="Analysis Results", 
                            lines=10,
                            interactive=False,
                            elem_classes="transcript-results"
                        )
                        with gr.Row():
                            gpa_viz = gr.Plot(label="GPA Visualization", visible=False)
                            req_viz = gr.Plot(label="Requirements Visualization", visible=False)
                        transcript_data = gr.State()

                file_input.change(
                    fn=lambda f: (
                        gr.update(visible=False),
                        gr.update(value="File ready for analysis!", visible=True) if f 
                        else gr.update(value="Please upload a file", visible=False)
                    ),
                    inputs=file_input,
                    outputs=[file_error, transcript_output]
                )

                def process_and_visualize(file_obj, tab_status):
                    results, data = parse_transcript(file_obj)
                    
                    # Update visualizations
                    gpa_viz_update = gr.update(visible=data.get('analysis', {}).get('visualizations', {}).get('gpa_viz') is not None)
                    req_viz_update = gr.update(visible=data.get('analysis', {}).get('visualizations', {}).get('req_viz') is not None)
                    
                    # Update tab completion status
                    tab_status[0] = True
                    
                    return results, data, gpa_viz_update, req_viz_update, tab_status

                upload_btn.click(
                    fn=process_and_visualize,
                    inputs=[file_input, tab_completed],
                    outputs=[transcript_output, transcript_data, gpa_viz, req_viz, tab_completed]
                ).then(
                    fn=lambda: gr.update(elem_classes="completed-tab"),
                    outputs=step1
                ).then(
                    fn=lambda: gr.update(interactive=True),
                    outputs=step2
                )

            # ===== TAB 2: LEARNING STYLE QUIZ =====
            with gr.Tab("Learning Style Quiz", id=1):
                with gr.Column():
                    gr.Markdown("### Step 2: Discover Your Learning Style")
                    progress = gr.HTML("<div class='progress-bar' style='width: 0%'></div>")
                    
                    quiz_components = []
                    with gr.Accordion("Quiz Questions", open=True):
                        for i, (question, options) in enumerate(zip(learning_style_quiz.questions, learning_style_quiz.options)):
                            with gr.Group(elem_classes="quiz-question"):
                                q = gr.Radio(
                                    options,
                                    label=f"{i+1}. {question}",
                                    show_label=True
                                )
                                quiz_components.append(q)
                    
                    with gr.Row():
                        quiz_submit = gr.Button("Submit Quiz", variant="primary")
                        quiz_clear = gr.Button("Clear Answers")
                    
                    quiz_alert = gr.HTML(visible=False)
                    learning_output = gr.Markdown(
                        label="Your Learning Style Results",
                        visible=False,
                        elem_classes="quiz-results"
                    )

                for component in quiz_components:
                    component.change(
                        fn=lambda *answers: {
                            progress: gr.HTML(
                                f"<div class='progress-bar' style='width: {sum(1 for a in answers if a)/len(answers)*100}%'></div>"
                            )
                        },
                        inputs=quiz_components,
                        outputs=progress
                    )
                
                quiz_submit.click(
                    fn=lambda *answers: learning_style_quiz.evaluate_quiz(*answers),
                    inputs=quiz_components,
                    outputs=learning_output
                ).then(
                    fn=lambda: gr.update(visible=True),
                    outputs=learning_output
                ).then(
                    fn=lambda: {1: True},
                    inputs=None,
                    outputs=tab_completed
                ).then(
                    fn=lambda: gr.update(elem_classes="completed-tab"),
                    outputs=step2
                ).then(
                    fn=lambda: gr.update(interactive=True),
                    outputs=step3
                )
                
                quiz_clear.click(
                    fn=lambda: [None] * len(quiz_components),
                    outputs=quiz_components
                ).then(
                    fn=lambda: gr.HTML("<div class='progress-bar' style='width: 0%'></div>"),
                    outputs=progress
                )

            # ===== TAB 3: PERSONAL QUESTIONS =====
            with gr.Tab("Personal Profile", id=2):
                with gr.Row():
                    with gr.Column(scale=1):
                        gr.Markdown("### Step 3: Tell Us About Yourself")
                        with gr.Group():
                            name = gr.Textbox(label="Full Name", placeholder="Your name")
                            age = gr.Number(label="Age", minimum=MIN_AGE, maximum=MAX_AGE, precision=0)
                            interests = gr.Textbox(
                                label="Your Interests/Hobbies",
                                placeholder="e.g., Science, Music, Sports, Art..."
                            )
                        
                        save_personal_btn = gr.Button("Save Information", variant="primary")
                        save_confirmation = gr.HTML(visible=False)
                    
                    with gr.Column(scale=1):
                        gr.Markdown("### Favorites")
                        with gr.Group():
                            movie = gr.Textbox(label="Favorite Movie")
                            movie_reason = gr.Textbox(label="Why do you like it?", lines=2)
                            show = gr.Textbox(label="Favorite TV Show")
                            show_reason = gr.Textbox(label="Why do you like it?", lines=2)
                            book = gr.Textbox(label="Favorite Book")
                            book_reason = gr.Textbox(label="Why do you like it?", lines=2)
                            character = gr.Textbox(label="Favorite Character (from any story)")
                            character_reason = gr.Textbox(label="Why do you like them?", lines=2)
                        
                        with gr.Accordion("Personal Blog (Optional)", open=False):
                            blog = gr.Textbox(
                                label="Share your thoughts",
                                placeholder="Write something about yourself...",
                                lines=5
                            )
                
                save_personal_btn.click(
                    fn=lambda n, a, i, ts: (
                        {2: True},
                        gr.update(elem_classes="completed-tab"),
                        gr.update(interactive=True),
                        gr.update(value="<div class='alert-box'>Information saved!</div>", visible=True)
                    ),
                    inputs=[name, age, interests, tab_completed],
                    outputs=[tab_completed, step3, step4, save_confirmation]
                )

            # ===== TAB 4: SAVE & REVIEW =====
            with gr.Tab("Save Profile", id=3):
                with gr.Row():
                    with gr.Column(scale=1):
                        gr.Markdown("### Step 4: Review & Save Your Profile")
                        with gr.Group():
                            load_profile_dropdown = gr.Dropdown(
                                label="Load Existing Profile",
                                choices=profile_manager.list_profiles(session_token.value),
                                visible=False
                            )
                            with gr.Row():
                                load_btn = gr.Button("Load", visible=False)
                                delete_btn = gr.Button("Delete", variant="stop", visible=False)
                        
                        save_btn = gr.Button("Save Profile", variant="primary")
                        clear_btn = gr.Button("Clear Form")
                    
                    with gr.Column(scale=2):
                        output_summary = gr.Markdown(
                            "Your profile summary will appear here after saving.",
                            label="Profile Summary"
                        )
                        with gr.Row():
                            req_viz_matplotlib = gr.Plot(label="Requirements Progress", visible=False)
                            credits_viz = gr.Plot(label="Credits Distribution", visible=False)

                save_btn.click(
                    fn=profile_manager.save_profile,
                    inputs=[
                        name, age, interests, transcript_data, learning_output,
                        movie, movie_reason, show, show_reason,
                        book, book_reason, character, character_reason, blog
                    ],
                    outputs=output_summary
                ).then(
                    fn=lambda td: (
                        gr.update(visible=True),
                        gr.update(visible=True)
                    ) if td and 'requirements' in td else (gr.update(visible=False), gr.update(visible=False)),
                    inputs=transcript_data,
                    outputs=[req_viz_matplotlib, credits_viz]
                ).then(
                    fn=lambda: {3: True},
                    inputs=None,
                    outputs=tab_completed
                ).then(
                    fn=lambda: gr.update(elem_classes="completed-tab"),
                    outputs=step4
                ).then(
                    fn=lambda: gr.update(interactive=True),
                    outputs=step5
                ).then(
                    fn=lambda: profile_manager.list_profiles(session_token.value),
                    outputs=load_profile_dropdown
                ).then(
                    fn=lambda: gr.update(visible=bool(profile_manager.list_profiles(session_token.value))),
                    outputs=load_btn
                ).then(
                    fn=lambda: gr.update(visible=bool(profile_manager.list_profiles(session_token.value))),
                    outputs=delete_btn
                )

                # Create visualizations when profile is loaded
                load_btn.click(
                    fn=lambda name: profile_manager.load_profile(name, session_token.value),
                    inputs=load_profile_dropdown,
                    outputs=None
                ).then(
                    fn=lambda profile: (
                        profile.get('name', ''),
                        profile.get('age', ''),
                        profile.get('interests', ''),
                        profile.get('learning_style', ''),
                        profile.get('favorites', {}).get('movie', ''),
                        profile.get('favorites', {}).get('movie_reason', ''),
                        profile.get('favorites', {}).get('show', ''),
                        profile.get('favorites', {}).get('show_reason', ''),
                        profile.get('favorites', {}).get('book', ''),
                        profile.get('favorites', {}).get('book_reason', ''),
                        profile.get('favorites', {}).get('character', ''),
                        profile.get('favorites', {}).get('character_reason', ''),
                        profile.get('blog', ''),
                        profile.get('transcript', {}),
                        gr.update(value="Profile loaded successfully!"),
                        create_requirements_visualization_matplotlib(profile.get('transcript', {}).get('requirements', [])),
                        create_credits_distribution_visualization(profile.get('transcript', {}).get('requirements', []))
                    ),
                    inputs=None,
                    outputs=[
                        name, age, interests, learning_output,
                        movie, movie_reason, show, show_reason,
                        book, book_reason, character, character_reason,
                        blog, transcript_data, output_summary,
                        req_viz_matplotlib, credits_viz
                    ]
                )

            # ===== TAB 5: AI ASSISTANT =====
            with gr.Tab("AI Assistant", id=4):
                gr.Markdown("## Your Personalized Learning Assistant")
                gr.Markdown("Ask me anything about studying, your courses, grades, or learning strategies.")
                
                async def chat_wrapper(message: str, history: List[List[str]]):
                    response = await teaching_assistant.generate_response(
                        message, 
                        history, 
                        session_token.value
                    )
                    return response
                
                chatbot = gr.ChatInterface(
                    fn=chat_wrapper,
                    examples=[
                        "What's my GPA?",
                        "How should I study for math?",
                        "What courses am I taking?",
                        "Study tips for my learning style"
                    ],
                    title=""
                )

        # Navigation logic
        def navigate_to_tab(tab_index: int, tab_completed_status):
            current_tab = tabs.selected
            
            if tab_index <= current_tab:
                return gr.Tabs(selected=tab_index), gr.update(visible=False)
            
            # Check all previous tabs are completed
            for i in range(tab_index):
                if not tab_completed_status.get(i, False):
                    messages = [
                        "Please complete the transcript analysis first",
                        "Please complete the learning style quiz first",
                        "Please fill out your personal information first",
                        "Please save your profile first"
                    ]
                    return (
                        gr.Tabs(selected=i),
                        gr.update(
                            value=f"<div class='error-message'>⛔ {messages[i]}</div>",
                            visible=True
                        )
                    )
            
            return gr.Tabs(selected=tab_index), gr.update(visible=False)
        
        step1.click(
            lambda idx, status: navigate_to_tab(idx, status),
            inputs=[gr.State(0), tab_completed],
            outputs=[tabs, nav_message]
        )
        step2.click(
            lambda idx, status: navigate_to_tab(idx, status),
            inputs=[gr.State(1), tab_completed],
            outputs=[tabs, nav_message]
        )
        step3.click(
            lambda idx, status: navigate_to_tab(idx, status),
            inputs=[gr.State(2), tab_completed],
            outputs=[tabs, nav_message]
        )
        step4.click(
            lambda idx, status: navigate_to_tab(idx, status),
            inputs=[gr.State(3), tab_completed],
            outputs=[tabs, nav_message]
        )
        step5.click(
            lambda idx, status: navigate_to_tab(idx, status),
            inputs=[gr.State(4), tab_completed],
            outputs=[tabs, nav_message]
        )
        
        # Dark mode toggle
        def toggle_dark_mode(dark):
            return gr.themes.Soft(primary_hue="blue", secondary_hue="gray") if not dark else gr.themes.Soft(primary_hue="blue", secondary_hue="gray", neutral_hue="slate")
        
        dark_mode.change(
            fn=toggle_dark_mode,
            inputs=dark_mode,
            outputs=None
        )
        
        # Load model on startup
        app.load(fn=lambda: model_loader.load_model(), outputs=[])
    
    return app

app = create_interface()

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