#!/usr/bin/env python3 """ Image selector for CleanKitchens articles Matches violations to appropriate stock images """ import json import os import random from collections import defaultdict class ImageSelector: def __init__(self): self.base_url = "/assets/images" self.metadata_path = "/var/www/twin-digital-media/public_html/_sites/cleankitchens/assets/images/generation_metadata.json" # Load metadata if exists self.image_metadata = self.load_metadata() # Track usage for rotation self.usage_counts = defaultdict(int) # Chain restaurant keywords self.chain_mappings = { 'burger_chain': ['mcdonald', 'burger king', 'wendy', 'five guys', 'in-n-out', 'white castle', 'sonic'], 'coffee_chain': ['starbucks', 'dunkin', 'peet', 'caribou', 'tim hortons', 'coffee bean'], 'chicken_chain': ['kfc', 'popeyes', 'chick-fil-a', 'raising cane', 'church'], 'pizza_chain': ['domino', 'pizza hut', 'papa john', 'little caesars', 'marco'], 'mexican_chain': ['chipotle', 'taco bell', 'qdoba', 'moe southwest', 'del taco'], 'sub_chain': ['subway', 'jimmy john', 'jersey mike', 'firehouse', 'quiznos'], 'asian_chain': ['panda express', 'pf chang', 'pick up stix', 'pei wei'], 'casual_chain': ['applebee', 'chili', 'olive garden', 'red lobster', 'outback'] } # Violation keyword mappings self.violation_mappings = { 'rodent': ['rodent', 'mouse', 'mice', 'rat', 'droppings', 'gnaw'], 'roach': ['roach', 'cockroach', 'insect', 'bug', 'pest'], 'temperature': ['temperature', 'cold holding', 'hot holding', 'cooling', 'danger zone'], 'cleanliness': ['clean', 'sanitation', 'dirty', 'soil', 'grease', 'grime'], 'mold': ['mold', 'mildew', 'fungus', 'moisture'], 'cross_contamination': ['cross contaminat', 'raw', 'cooked', 'separate'], 'handwashing': ['hand', 'wash', 'sink', 'soap', 'towel', 'sanitizer'], 'structural': ['ceiling', 'floor', 'wall', 'crack', 'hole', 'damage'], 'sewage': ['sewage', 'drain', 'backup', 'plumbing', 'waste'] # Removed closure category to avoid libel issues } def load_metadata(self): """Load generated image metadata""" if os.path.exists(self.metadata_path): with open(self.metadata_path, 'r') as f: return json.load(f) return None def identify_chain(self, restaurant_name): """Identify if restaurant is a chain and which type""" name_lower = restaurant_name.lower() for chain_type, keywords in self.chain_mappings.items(): for keyword in keywords: if keyword in name_lower: return chain_type return None def identify_violation_type(self, violations_text): """Identify primary violation type from text""" if not violations_text: return 'general' violations_lower = violations_text.lower() scores = {} # Score each violation type based on keyword matches for violation_type, keywords in self.violation_mappings.items(): score = 0 for keyword in keywords: if keyword in violations_lower: score += violations_lower.count(keyword) scores[violation_type] = score # Return highest scoring violation type if scores: best_match = max(scores, key=scores.get) if scores[best_match] > 0: return best_match return 'general' def get_available_images(self, category, subfolder): """Get list of available images for a category""" images = [] # First check metadata if self.image_metadata and 'images' in self.image_metadata: for img in self.image_metadata['images']: if img['category'] == category and img['subfolder'] == subfolder: images.append(img['filename']) # Fallback to directory listing if not images: dir_path = f"/var/www/twin-digital-media/public_html/_sites/cleankitchens/assets/images/{subfolder}" if os.path.exists(dir_path): for file in os.listdir(dir_path): if file.startswith(f"{category}_") and file.endswith('.jpg'): images.append(file) return images def select_least_used_image(self, images, category): """Select the least used image from available options""" if not images: return None # Find image with lowest usage count min_usage = float('inf') best_image = images[0] for image in images: usage_key = f"{category}_{image}" usage = self.usage_counts[usage_key] if usage < min_usage: min_usage = usage best_image = image elif usage == min_usage and random.random() > 0.5: # Random selection among equally used images best_image = image # Increment usage count usage_key = f"{category}_{best_image}" self.usage_counts[usage_key] += 1 return best_image def select_image_for_article(self, inspection_data): """ Select appropriate stock image based on inspection data Priority order: 1. RODENT violations (TOP PRIORITY) 2. Cuisine-specific images (if available) 3. Closures 4. Other specific violations 5. Chain restaurants 6. General violations Args: inspection_data: dict with keys: - facility_name: Restaurant name - violations: Violation text - is_closure: Boolean - article_type: 'violation', 'daily_summary', 'weekly_summary' - cuisine_type: Optional cuisine type Returns: dict with: - image_url: Full URL path to image - image_alt: Alt text for image - category: Category selected """ # Check for summary articles first article_type = inspection_data.get('article_type', 'violation') if article_type == 'daily_summary': images = self.get_available_images('daily_summary', 'summaries') if images: selected = self.select_least_used_image(images, 'daily_summary') return { 'image_url': f"{self.base_url}/summaries/{selected}", 'image_alt': "Daily restaurant inspection summary report - cleankitchens.org", 'category': 'daily_summary' } elif article_type == 'weekly_summary': images = self.get_available_images('weekly_summary', 'summaries') if images: selected = self.select_least_used_image(images, 'weekly_summary') return { 'image_url': f"{self.base_url}/summaries/{selected}", 'image_alt': "Weekly restaurant inspection summary report - cleankitchens.org", 'category': 'weekly_summary' } # For violation articles, check priority order violations_text = inspection_data.get('violations', '') # PRIORITY 1: RODENT violations (TOP PRIORITY) if violations_text: violations_lower = violations_text.lower() if any(word in violations_lower for word in ['rodent', 'mouse', 'mice', 'rat', 'droppings', 'gnaw']): images = self.get_available_images('rodent', 'violations') if images: selected = self.select_least_used_image(images, 'rodent') return { 'image_url': f"{self.base_url}/violations/{selected}", 'image_alt': "Restaurant health violation - rodent activity - cleankitchens.org", 'category': 'rodent' } # PRIORITY 2: Cuisine-specific images (if available) cuisine_type = inspection_data.get('cuisine_type') if cuisine_type: # Check if we have cuisine-specific images images = self.get_available_images(cuisine_type, 'cuisine') if images: selected = self.select_least_used_image(images, cuisine_type) return { 'image_url': f"{self.base_url}/cuisine/{selected}", 'image_alt': f"{cuisine_type.replace('_', ' ').title()} restaurant - cleankitchens.org", 'category': cuisine_type } # PRIORITY 3: Skip closure images - use general instead for safety if inspection_data.get('is_closure'): # Use general images instead of closure to avoid any libel issues images = self.get_available_images('general', 'violations') if images: selected = self.select_least_used_image(images, 'general') return { 'image_url': f"{self.base_url}/violations/{selected}", 'image_alt': "Restaurant health inspection - cleankitchens.org", 'category': 'general' } # PRIORITY 4: Check for other specific violations violation_type = self.identify_violation_type(violations_text) if violation_type != 'general': images = self.get_available_images(violation_type, 'violations') if images: selected = self.select_least_used_image(images, violation_type) alt_text = f"Restaurant health violation - {violation_type.replace('_', ' ')} - cleankitchens.org" return { 'image_url': f"{self.base_url}/violations/{selected}", 'image_alt': alt_text, 'category': violation_type } # PRIORITY 5: Check for chain restaurants restaurant_name = inspection_data.get('facility_name', '') chain_type = self.identify_chain(restaurant_name) if chain_type: images = self.get_available_images(chain_type, 'chains') if images: selected = self.select_least_used_image(images, chain_type) alt_text = f"{chain_type.replace('_', ' ').title()} restaurant inspection - cleankitchens.org" return { 'image_url': f"{self.base_url}/chains/{selected}", 'image_alt': alt_text, 'category': chain_type } # PRIORITY 6: Default to general violation image images = self.get_available_images('general', 'violations') if images: selected = self.select_least_used_image(images, 'general') else: # Ultimate fallback selected = 'general_1.jpg' return { 'image_url': f"{self.base_url}/violations/{selected}", 'image_alt': "Restaurant health inspection violation - cleankitchens.org", 'category': 'general' } def get_usage_stats(self): """Return current usage statistics""" return dict(self.usage_counts) # Example usage if __name__ == "__main__": selector = ImageSelector() # Test with Frankie's data test_data = { 'facility_name': "FRANKIE'S FAST FOOD CORNER", 'violations': "OBSERVED NO HAND WASHING SINK IN REAR DISH WASHING AREA", 'is_closure': False, 'article_type': 'violation' } result = selector.select_image_for_article(test_data) print(f"Selected: {result['image_url']}") print(f"Alt text: {result['image_alt']}") print(f"Category: {result['category']}") # Test with closure test_data2 = { 'facility_name': "McDonald's", 'violations': "Rodent activity found", 'is_closure': True, 'article_type': 'violation' } result2 = selector.select_image_for_article(test_data2) print(f"\nClosure - Selected: {result2['image_url']}") print(f"Alt text: {result2['image_alt']}") print(f"Category: {result2['category']}")