"""
Audio Content Analyzer for GAIA Agent
Provides intelligent content parsing and analysis from audio transcriptions.
Specialized for GAIA evaluation tasks including recipe analysis and educational content.
"""

import logging
import re
from typing import Dict, Any, List, Optional, Tuple
import json

try:
    from .base_tool import SimpleAGNOTool
except ImportError:
    from base_tool import SimpleAGNOTool

logger = logging.getLogger(__name__)


class AudioContentAnalyzer(SimpleAGNOTool):
    """
    Intelligent audio content analyzer for GAIA evaluation tasks.
    
    Specializes in:
    - Recipe ingredient extraction from audio
    - Educational content analysis (homework, page numbers)
    - Structured data extraction from transcriptions
    - Context-aware content understanding
    - High-confidence information extraction
    """
    
    def __init__(self):
        """Initialize the audio content analyzer."""
        super().__init__(
            name="audio_content_analyzer",
            description="Analyze audio transcriptions for structured content extraction and understanding"
        )
        
        # Set availability status
        self.available = True
        
        # Recipe analysis patterns
        self.ingredient_patterns = [
            # Pattern: "2 cups of flour"
            r'(\d+(?:\.\d+)?)\s+(cups?|cup|tablespoons?|tablespoon|tbsp|teaspoons?|teaspoon|tsp|pounds?|pound|lbs?|lb|ounces?|ounce|oz|grams?|gram|g)\s+(?:of\s+)?([a-zA-Z\s]+?)(?=\s*[,.\n]|$)',
            # Pattern: "flour, 2 cups"
            r'([a-zA-Z\s]+?),?\s*(\d+(?:\.\d+)?)\s+(cups?|cup|tablespoons?|tablespoon|tbsp|teaspoons?|teaspoon|tsp|pounds?|pound|lbs?|lb|ounces?|ounce|oz|grams?|gram|g)',
            # Pattern: "add flour"
            r'(?:add|use|mix|combine|include)\s+([a-zA-Z\s]+?)(?=\s*[,.\n]|$)',
        ]
        
        # Common ingredients for validation
        self.common_ingredients = {
            'flour', 'sugar', 'butter', 'eggs', 'egg', 'milk', 'cream', 'vanilla',
            'strawberries', 'strawberry', 'berries', 'berry', 'fruit', 'salt',
            'baking powder', 'baking soda', 'powder', 'soda', 'cinnamon', 'nutmeg',
            'lemon', 'orange', 'chocolate', 'nuts', 'almonds', 'pecans', 'walnuts',
            'honey', 'syrup', 'oil', 'shortening', 'cornstarch', 'gelatin',
            'water', 'juice', 'zest', 'extract', 'spice', 'spices'
        }
        
        # Educational content patterns
        self.education_patterns = {
            'page_numbers': [
                r'page\s+(\d+)',
                r'on\s+page\s+(\d+)',
                r'turn\s+to\s+page\s+(\d+)',
                r'go\s+to\s+page\s+(\d+)',
                r'see\s+page\s+(\d+)',
                r'page\s+number\s+(\d+)'
            ],
            'chapter_numbers': [
                r'chapter\s+(\d+)',
                r'unit\s+(\d+)',
                r'section\s+(\d+)'
            ],
            'exercise_numbers': [
                r'exercise\s+(\d+)',
                r'problem\s+(\d+)',
                r'question\s+(\d+)',
                r'assignment\s+(\d+)'
            ]
        }
    
    def analyze_recipe_content(self, transcription: str) -> Dict[str, Any]:
        """
        Analyze transcription for recipe content and extract ingredients.
        
        Args:
            transcription: Audio transcription text
            
        Returns:
            Dictionary with recipe analysis results
        """
        try:
            logger.info("🍰 Analyzing recipe content from transcription")
            
            analysis = {
                'is_recipe': False,
                'confidence': 0.0,
                'ingredients': [],
                'quantities': [],
                'cooking_methods': [],
                'recipe_type': None,
                'structured_ingredients': []
            }
            
            text_lower = transcription.lower()
            
            # Check if this is likely a recipe
            recipe_indicators = [
                'recipe', 'ingredients', 'cooking', 'baking', 'pie', 'cake',
                'mix', 'stir', 'add', 'combine', 'bake', 'cook', 'prepare'
            ]
            
            recipe_score = sum(1 for indicator in recipe_indicators if indicator in text_lower)
            analysis['is_recipe'] = recipe_score >= 2
            analysis['confidence'] = min(1.0, recipe_score / 5.0)
            
            if not analysis['is_recipe']:
                logger.info("📝 Content does not appear to be a recipe")
                return analysis
            
            # Determine recipe type
            if 'pie' in text_lower:
                analysis['recipe_type'] = 'pie'
            elif 'cake' in text_lower:
                analysis['recipe_type'] = 'cake'
            elif 'cookie' in text_lower:
                analysis['recipe_type'] = 'cookies'
            elif 'bread' in text_lower:
                analysis['recipe_type'] = 'bread'
            
            # Extract ingredients using multiple patterns
            ingredients_found = set()
            structured_ingredients = []
            
            for pattern in self.ingredient_patterns:
                matches = re.findall(pattern, transcription, re.IGNORECASE)
                
                for match in matches:
                    # Handle different match tuple lengths
                    if isinstance(match, tuple):
                        if len(match) == 3:  # quantity, unit, ingredient
                            quantity, unit, ingredient = match
                            ingredient = ingredient.strip().lower()
                            
                            # Validate ingredient
                            if self._is_valid_ingredient(ingredient):
                                ingredients_found.add(ingredient)
                                structured_ingredients.append({
                                    'ingredient': ingredient,
                                    'quantity': quantity,
                                    'unit': unit.lower()
                                })
                        elif len(match) == 1:  # just ingredient
                            ingredient = match[0].strip().lower()
                            if self._is_valid_ingredient(ingredient):
                                ingredients_found.add(ingredient)
                                structured_ingredients.append({
                                    'ingredient': ingredient,
                                    'quantity': None,
                                    'unit': None
                                })
                    else:
                        # Single string match
                        ingredient = str(match).strip().lower()
                        if self._is_valid_ingredient(ingredient):
                            ingredients_found.add(ingredient)
                            structured_ingredients.append({
                                'ingredient': ingredient,
                                'quantity': None,
                                'unit': None
                            })
            
            # Additional ingredient extraction for common items
            for ingredient in self.common_ingredients:
                if ingredient in text_lower and ingredient not in ingredients_found:
                    ingredients_found.add(ingredient)
                    structured_ingredients.append({
                        'ingredient': ingredient,
                        'quantity': None,
                        'unit': None
                    })
            
            analysis['ingredients'] = list(ingredients_found)
            analysis['structured_ingredients'] = structured_ingredients
            
            # Extract cooking methods
            cooking_methods = [
                'bake', 'mix', 'stir', 'whip', 'fold', 'beat', 'combine',
                'add', 'pour', 'melt', 'heat', 'cool', 'chill', 'freeze'
            ]
            
            for method in cooking_methods:
                if method in text_lower:
                    analysis['cooking_methods'].append(method)
            
            # Extract quantities and measurements
            quantity_patterns = [
                r'(\d+(?:\.\d+)?)\s*(cups?|tablespoons?|teaspoons?|pounds?|ounces?)',
                r'(\d+)\s*(degrees?)',
                r'(\d+)\s*(minutes?)',
                r'(\d+)\s*(hours?)'
            ]
            
            for pattern in quantity_patterns:
                matches = re.findall(pattern, text_lower)
                for match in matches:
                    if isinstance(match, tuple) and len(match) == 2:
                        q, u = match
                        analysis['quantities'].append(f"{q} {u}")
                    elif isinstance(match, str):
                        analysis['quantities'].append(match)
            
            logger.info(f"✅ Recipe analysis completed: {len(analysis['ingredients'])} ingredients found")
            
            return analysis
            
        except Exception as e:
            logger.error(f"❌ Recipe analysis failed: {e}")
            return {
                'is_recipe': False,
                'confidence': 0.0,
                'ingredients': [],
                'error': str(e)
            }
    
    def analyze_educational_content(self, transcription: str) -> Dict[str, Any]:
        """
        Analyze transcription for educational content and extract key information.
        
        Args:
            transcription: Audio transcription text
            
        Returns:
            Dictionary with educational analysis results
        """
        try:
            logger.info("📚 Analyzing educational content from transcription")
            
            analysis = {
                'is_educational': False,
                'confidence': 0.0,
                'page_numbers': [],
                'chapter_numbers': [],
                'exercise_numbers': [],
                'subjects': [],
                'assignments': [],
                'key_numbers': []
            }
            
            text_lower = transcription.lower()
            
            # Check if this is educational content
            education_indicators = [
                'homework', 'assignment', 'page', 'chapter', 'exercise',
                'problem', 'study', 'lesson', 'class', 'school', 'teacher',
                'student', 'book', 'textbook', 'worksheet'
            ]
            
            education_score = sum(1 for indicator in education_indicators if indicator in text_lower)
            analysis['is_educational'] = education_score >= 2
            analysis['confidence'] = min(1.0, education_score / 5.0)
            
            if not analysis['is_educational']:
                logger.info("📝 Content does not appear to be educational")
                return analysis
            
            # Extract page numbers with high precision
            for pattern in self.education_patterns['page_numbers']:
                matches = re.findall(pattern, text_lower)
                analysis['page_numbers'].extend(matches)
            
            # Remove duplicates and sort
            analysis['page_numbers'] = sorted(list(set(analysis['page_numbers'])), key=int)
            
            # Extract chapter numbers
            for pattern in self.education_patterns['chapter_numbers']:
                matches = re.findall(pattern, text_lower)
                analysis['chapter_numbers'].extend(matches)
            
            # Extract exercise/problem numbers
            for pattern in self.education_patterns['exercise_numbers']:
                matches = re.findall(pattern, text_lower)
                analysis['exercise_numbers'].extend(matches)
            
            # Identify subjects
            subjects = {
                'math': ['math', 'mathematics', 'algebra', 'geometry', 'calculus', 'arithmetic'],
                'science': ['science', 'physics', 'chemistry', 'biology', 'astronomy'],
                'english': ['english', 'literature', 'reading', 'writing', 'grammar'],
                'history': ['history', 'social studies', 'geography', 'civics'],
                'language': ['spanish', 'french', 'german', 'italian', 'chinese', 'japanese']
            }
            
            for subject, keywords in subjects.items():
                if any(keyword in text_lower for keyword in keywords):
                    analysis['subjects'].append(subject)
            
            # Extract all numbers for potential reference
            all_numbers = re.findall(r'\b\d+\b', transcription)
            analysis['key_numbers'] = list(set(all_numbers))
            
            logger.info(f"✅ Educational analysis completed: {len(analysis['page_numbers'])} page numbers found")
            
            return analysis
            
        except Exception as e:
            logger.error(f"❌ Educational analysis failed: {e}")
            return {
                'is_educational': False,
                'confidence': 0.0,
                'page_numbers': [],
                'error': str(e)
            }
    
    def extract_key_information(self, transcription: str, target_type: str) -> Dict[str, Any]:
        """
        Extract specific key information from transcription based on target type.
        
        Args:
            transcription: Audio transcription text
            target_type: Type of information to extract ('recipe_ingredients', 'page_numbers', 'all')
            
        Returns:
            Dictionary with extracted information
        """
        try:
            logger.info(f"🔍 Extracting key information: {target_type}")
            
            result = {
                'target_type': target_type,
                'success': True,
                'extracted_data': {},
                'confidence': 0.0
            }
            
            if target_type == 'recipe_ingredients' or target_type == 'all':
                recipe_analysis = self.analyze_recipe_content(transcription)
                result['extracted_data']['recipe'] = recipe_analysis
                if recipe_analysis['is_recipe']:
                    result['confidence'] = max(result['confidence'], recipe_analysis['confidence'])
            
            if target_type == 'page_numbers' or target_type == 'all':
                education_analysis = self.analyze_educational_content(transcription)
                result['extracted_data']['education'] = education_analysis
                if education_analysis['is_educational']:
                    result['confidence'] = max(result['confidence'], education_analysis['confidence'])
            
            # Extract the most relevant information based on target type
            if target_type == 'recipe_ingredients':
                if 'recipe' in result['extracted_data'] and result['extracted_data']['recipe']['is_recipe']:
                    result['primary_result'] = result['extracted_data']['recipe']['ingredients']
                else:
                    result['primary_result'] = []
            
            elif target_type == 'page_numbers':
                if 'education' in result['extracted_data'] and result['extracted_data']['education']['is_educational']:
                    result['primary_result'] = result['extracted_data']['education']['page_numbers']
                else:
                    result['primary_result'] = []
            
            else:  # 'all'
                result['primary_result'] = {
                    'recipe_ingredients': result['extracted_data'].get('recipe', {}).get('ingredients', []),
                    'page_numbers': result['extracted_data'].get('education', {}).get('page_numbers', [])
                }
            
            logger.info(f"✅ Key information extraction completed with confidence: {result['confidence']:.2f}")
            
            return result
            
        except Exception as e:
            logger.error(f"❌ Key information extraction failed: {e}")
            return {
                'target_type': target_type,
                'success': False,
                'error': str(e),
                'extracted_data': {},
                'confidence': 0.0
            }
    
    def _is_valid_ingredient(self, ingredient: str) -> bool:
        """Check if a string is likely a valid ingredient."""
        ingredient = ingredient.strip().lower()
        
        # Must be at least 2 characters
        if len(ingredient) < 2:
            return False
        
        # Check against common ingredients
        if ingredient in self.common_ingredients:
            return True
        
        # Check if it contains common ingredient words
        ingredient_words = ingredient.split()
        for word in ingredient_words:
            if word in self.common_ingredients:
                return True
        
        # Check for food-related patterns
        food_patterns = [
            r'.*flour$', r'.*sugar$', r'.*powder$', r'.*extract$',
            r'.*juice$', r'.*zest$', r'.*oil$', r'.*sauce$'
        ]
        
        for pattern in food_patterns:
            if re.match(pattern, ingredient):
                return True
        
        # Exclude common non-ingredients
        non_ingredients = [
            'minutes', 'degrees', 'hours', 'time', 'temperature',
            'oven', 'bowl', 'pan', 'spoon', 'cup', 'tablespoon'
        ]
        
        if ingredient in non_ingredients:
            return False
        
        # If it's a reasonable length and contains letters, consider it valid
        if 2 <= len(ingredient) <= 30 and re.match(r'^[a-zA-Z\s]+$', ingredient):
            return True
        
        return False
    
    def get_tool_functions(self) -> List[Dict[str, Any]]:
        """Get function definitions for AGNO integration."""
        return [
            {
                "name": "analyze_recipe_content",
                "description": "Analyze audio transcription for recipe content and extract ingredients",
                "parameters": {
                    "type": "object",
                    "properties": {
                        "transcription": {
                            "type": "string",
                            "description": "Audio transcription text to analyze for recipe content"
                        }
                    },
                    "required": ["transcription"]
                }
            },
            {
                "name": "analyze_educational_content",
                "description": "Analyze audio transcription for educational content and extract page numbers, assignments",
                "parameters": {
                    "type": "object",
                    "properties": {
                        "transcription": {
                            "type": "string",
                            "description": "Audio transcription text to analyze for educational content"
                        }
                    },
                    "required": ["transcription"]
                }
            },
            {
                "name": "extract_key_information",
                "description": "Extract specific key information from audio transcription",
                "parameters": {
                    "type": "object",
                    "properties": {
                        "transcription": {
                            "type": "string",
                            "description": "Audio transcription text to analyze"
                        },
                        "target_type": {
                            "type": "string",
                            "description": "Type of information to extract",
                            "enum": ["recipe_ingredients", "page_numbers", "all"]
                        }
                    },
                    "required": ["transcription", "target_type"]
                }
            }
        ]


# Create tool instance for AGNO integration
def create_audio_content_analyzer() -> Optional[AudioContentAnalyzer]:
    """Create and return audio content analyzer instance."""
    try:
        tool = AudioContentAnalyzer()
        logger.info("✅ Audio content analyzer created successfully")
        return tool
    except Exception as e:
        logger.error(f"❌ Failed to create audio content analyzer: {e}")
        return None