File size: 22,518 Bytes
1227ff0
f746366
 
 
 
1227ff0
 
 
 
 
 
 
 
 
 
f746366
1227ff0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f746366
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b82d15d
1227ff0
f746366
1227ff0
 
f746366
 
1227ff0
 
f746366
1227ff0
 
 
 
 
 
 
 
 
b82d15d
 
1227ff0
f746366
b82d15d
f746366
 
b82d15d
f746366
b82d15d
f746366
 
 
1227ff0
f746366
 
 
 
1227ff0
f746366
 
1227ff0
f746366
 
 
 
 
1227ff0
 
 
 
 
 
 
b82d15d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e53d7f7
 
 
f746366
 
e53d7f7
f746366
77acbe4
0d84ddc
 
b82d15d
 
 
f746366
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e53d7f7
 
 
 
f746366
 
 
 
e53d7f7
 
 
 
f746366
e53d7f7
 
 
 
 
 
f746366
 
 
 
 
e53d7f7
f746366
 
 
e53d7f7
 
f746366
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b82d15d
 
 
 
f746366
 
 
 
e53d7f7
 
f746366
 
 
 
 
 
 
 
 
e53d7f7
 
 
b82d15d
 
f746366
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1227ff0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e53d7f7
 
 
1227ff0
 
 
 
 
f746366
1227ff0
 
 
f746366
1227ff0
 
 
 
 
f746366
1227ff0
 
 
f746366
 
 
 
1227ff0
f746366
b82d15d
f746366
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b82d15d
 
f746366
b82d15d
f746366
b82d15d
f746366
b82d15d
1227ff0
 
f746366
 
1227ff0
f746366
 
 
b82d15d
f746366
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b82d15d
f746366
1227ff0
f746366
b82d15d
f746366
 
 
 
 
1227ff0
f746366
1227ff0
 
 
 
 
f746366
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1227ff0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f746366
1227ff0
 
f746366
1227ff0
 
 
 
 
f746366
1227ff0
f746366
 
1227ff0
f746366
1227ff0
 
f746366
 
 
1227ff0
 
f746366
 
 
 
 
1227ff0
f746366
1227ff0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f746366
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1227ff0
 
 
 
 
 
 
 
f746366
1227ff0
 
 
f746366
 
 
1227ff0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
"""
Utility functions for Phramer AI
By Pariente AI, for MIA TV Series

Enhanced with professional cinematography knowledge and multi-engine optimization
"""

import re
import logging
import gc
from typing import Optional, Tuple, Dict, Any, List
from PIL import Image
import torch
import numpy as np

from config import PROCESSING_CONFIG, FLUX_RULES, PROFESSIONAL_PHOTOGRAPHY_CONFIG

# Configure logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)


def setup_logging(level: str = "INFO") -> None:
    """Setup logging configuration"""
    logging.basicConfig(
        level=getattr(logging, level.upper()),
        format='%(asctime)s - %(name)s - %(levelname)s - %(message)s'
    )


def optimize_image(image: Any) -> Optional[Image.Image]:
    """
    Optimize image for processing
    
    Args:
        image: Input image (PIL, numpy array, or file path)
        
    Returns:
        Optimized PIL Image or None if failed
    """
    if image is None:
        return None
        
    try:
        # Convert to PIL Image if necessary
        if isinstance(image, np.ndarray):
            image = Image.fromarray(image)
        elif isinstance(image, str):
            image = Image.open(image)
        elif not isinstance(image, Image.Image):
            logger.error(f"Unsupported image type: {type(image)}")
            return None
        
        # Convert to RGB if necessary
        if image.mode != 'RGB':
            image = image.convert('RGB')
        
        # Resize if too large
        max_size = PROCESSING_CONFIG["max_image_size"]
        if image.size[0] > max_size or image.size[1] > max_size:
            image.thumbnail((max_size, max_size), Image.Resampling.LANCZOS)
            logger.info(f"Image resized to {image.size}")
        
        return image
        
    except Exception as e:
        logger.error(f"Image optimization failed: {e}")
        return None


def validate_image(image: Any) -> bool:
    """
    Validate if image is processable
    
    Args:
        image: Input image to validate
        
    Returns:
        True if valid, False otherwise
    """
    if image is None:
        return False
        
    try:
        optimized = optimize_image(image)
        return optimized is not None
    except Exception:
        return False


def clean_memory() -> None:
    """Clean up memory and GPU cache"""
    try:
        gc.collect()
        if torch.cuda.is_available():
            torch.cuda.empty_cache()
            torch.cuda.synchronize()
        logger.debug("Memory cleaned")
    except Exception as e:
        logger.warning(f"Memory cleanup failed: {e}")


def detect_scene_type_from_analysis(analysis_metadata: Dict[str, Any]) -> str:
    """Detect scene type from BAGEL analysis metadata"""
    try:
        # Check if BAGEL provided scene detection
        if "scene_type" in analysis_metadata:
            return analysis_metadata["scene_type"]
        
        # Check camera setup for scene hints
        camera_setup = analysis_metadata.get("camera_setup", "").lower()
        
        if any(term in camera_setup for term in ["portrait", "85mm", "135mm"]):
            return "portrait"
        elif any(term in camera_setup for term in ["landscape", "wide", "24mm", "phase one"]):
            return "landscape"
        elif any(term in camera_setup for term in ["street", "35mm", "documentary", "leica"]):
            return "street"
        elif any(term in camera_setup for term in ["cinema", "arri", "red", "anamorphic"]):
            return "cinematic"
        elif any(term in camera_setup for term in ["architecture", "building", "tilt"]):
            return "architectural"
        
        return "default"
        
    except Exception as e:
        logger.warning(f"Scene type detection failed: {e}")
        return "default"


def apply_flux_rules(prompt: str, analysis_metadata: Optional[Dict[str, Any]] = None) -> str:
    """
    Apply enhanced prompt optimization rules for multi-engine compatibility
    
    Args:
        prompt: Raw prompt text from BAGEL analysis
        analysis_metadata: Enhanced metadata with cinematography suggestions
        
    Returns:
        Optimized prompt for Flux, Midjourney, and other generative engines
    """
    if not prompt or not isinstance(prompt, str):
        return ""
    
    # Clean the prompt from unwanted elements
    cleaned_prompt = prompt
    for pattern in FLUX_RULES["remove_patterns"]:
        cleaned_prompt = re.sub(pattern, '', cleaned_prompt, flags=re.IGNORECASE)
    
    # Extract description part only (remove CAMERA_SETUP section if present)
    description_part = _extract_description_only(cleaned_prompt)
    
    # Check if BAGEL provided intelligent camera setup with cinematography context
    camera_config = ""
    scene_type = "default"
    
    if analysis_metadata and analysis_metadata.get("has_camera_suggestion") and analysis_metadata.get("camera_setup"):
        # Use BAGEL's intelligent camera suggestion - enhanced with cinematography knowledge
        bagel_camera = analysis_metadata["camera_setup"]
        scene_type = detect_scene_type_from_analysis(analysis_metadata)
        camera_config = _format_professional_camera_suggestion(bagel_camera, scene_type)
        logger.info(f"Using BAGEL cinematography suggestion: {camera_config}")
    else:
        # Enhanced fallback with professional cinematography knowledge
        scene_type = _detect_scene_from_description(description_part.lower())
        camera_config = _get_enhanced_camera_config(scene_type, description_part.lower())
        logger.info(f"Using enhanced cinematography configuration for {scene_type}")
    
    # Add enhanced lighting with cinematography principles
    lighting_enhancement = _get_cinematography_lighting_enhancement(description_part.lower(), camera_config, scene_type)
    
    # Add style enhancement for multi-engine compatibility
    style_enhancement = _get_style_enhancement(scene_type, description_part.lower())
    
    # Build final prompt: Description + Camera + Lighting + Style
    final_prompt = description_part + camera_config + lighting_enhancement + style_enhancement
    
    # Clean up formatting
    final_prompt = _clean_prompt_formatting(final_prompt)
    
    return final_prompt


def _extract_description_only(prompt: str) -> str:
    """Extract only the description part, removing camera setup sections"""
    # Remove CAMERA_SETUP section if present
    if "CAMERA_SETUP:" in prompt:
        parts = prompt.split("CAMERA_SETUP:")
        description = parts[0].strip()
    elif "2. CAMERA_SETUP" in prompt:
        parts = prompt.split("2. CAMERA_SETUP")
        description = parts[0].strip()
    else:
        description = prompt
    
    # Remove "DESCRIPTION:" label if present
    if description.startswith("DESCRIPTION:"):
        description = description.replace("DESCRIPTION:", "").strip()
    elif description.startswith("1. DESCRIPTION:"):
        description = description.replace("1. DESCRIPTION:", "").strip()
    
    # Clean up any remaining camera recommendations from the description
    description = re.sub(r'For this type of scene.*?shooting style would be.*?\.', '', description, flags=re.DOTALL)
    description = re.sub(r'I would recommend.*?aperture.*?\.', '', description, flags=re.DOTALL)
    description = re.sub(r'Professional Context:.*?\.', '', description, flags=re.DOTALL)
    description = re.sub(r'Cinematography context:.*?\.', '', description, flags=re.DOTALL)
    
    # Remove numbered section residues
    description = re.sub(r'\s*\d+\.\s*,?\s*$', '', description)
    description = re.sub(r'\s*\d+\.\s*,?\s*', ' ', description)
    
    return description.strip()


def _detect_scene_from_description(description_lower: str) -> str:
    """Enhanced scene detection from description with cinematography knowledge"""
    scene_keywords = PROFESSIONAL_PHOTOGRAPHY_CONFIG.get("scene_detection_keywords", {})
    
    # Score each scene type
    scene_scores = {}
    for scene_type, keywords in scene_keywords.items():
        score = sum(1 for keyword in keywords if keyword in description_lower)
        if score > 0:
            scene_scores[scene_type] = score
    
    # Additional cinematography-specific detection
    if any(term in description_lower for term in ["film", "movie", "cinematic", "dramatic lighting", "anamorphic"]):
        scene_scores["cinematic"] = scene_scores.get("cinematic", 0) + 2
    
    if any(term in description_lower for term in ["studio", "controlled lighting", "professional portrait"]):
        scene_scores["portrait"] = scene_scores.get("portrait", 0) + 2
    
    # Return highest scoring scene type
    if scene_scores:
        return max(scene_scores.items(), key=lambda x: x[1])[0]
    else:
        return "default"


def _format_professional_camera_suggestion(bagel_camera: str, scene_type: str) -> str:
    """Format BAGEL's camera suggestion with enhanced cinematography knowledge"""
    try:
        camera_text = bagel_camera.strip()
        camera_text = re.sub(r'^CAMERA_SETUP:\s*', '', camera_text)
        
        # Enhanced extraction patterns for cinema equipment
        cinema_patterns = {
            'camera': r'(ARRI [^,]+|RED [^,]+|Canon EOS [^,]+|Sony A[^,]+|Leica [^,]+|Hasselblad [^,]+|Phase One [^,]+)',
            'lens': r'(\d+mm[^,]*(?:anamorphic)?[^,]*|[^,]*(?:anamorphic|telephoto|wide-angle)[^,]*)',
            'aperture': r'(f/[\d.]+[^,]*)'
        }
        
        extracted_parts = []
        for key, pattern in cinema_patterns.items():
            match = re.search(pattern, camera_text, re.IGNORECASE)
            if match:
                extracted_parts.append(match.group(1).strip())
        
        if extracted_parts:
            camera_info = ', '.join(extracted_parts)
            # Add scene-specific enhancement
            if scene_type == "cinematic":
                return f", Shot on {camera_info}, cinematic photography"
            elif scene_type == "portrait":
                return f", Shot on {camera_info}, professional portrait photography"
            else:
                return f", Shot on {camera_info}, professional photography"
        else:
            return _get_enhanced_camera_config(scene_type, camera_text.lower())
                
    except Exception as e:
        logger.warning(f"Failed to format professional camera suggestion: {e}")
        return _get_enhanced_camera_config(scene_type, "")


def _get_enhanced_camera_config(scene_type: str, description_lower: str) -> str:
    """Get enhanced camera configuration with cinematography knowledge"""
    # Enhanced camera configurations with cinema equipment
    enhanced_configs = {
        "cinematic": ", Shot on ARRI Alexa LF, 35mm anamorphic lens, cinematic photography",
        "portrait": ", Shot on Canon EOS R5, 85mm f/1.4 lens at f/2.8, professional portrait photography",
        "landscape": ", Shot on Phase One XT, 24-70mm f/4 lens at f/8, epic landscape photography",
        "street": ", Shot on Leica M11, 35mm f/1.4 lens at f/2.8, documentary street photography",
        "architectural": ", Shot on Canon EOS R5, 24-70mm f/2.8 lens at f/8, architectural photography",
        "commercial": ", Shot on Hasselblad X2D 100C, 90mm f/2.5 lens, commercial photography"
    }
    
    # Use enhanced config if available, otherwise fall back to FLUX_RULES
    if scene_type in enhanced_configs:
        return enhanced_configs[scene_type]
    elif scene_type in FLUX_RULES["camera_configs"]:
        return FLUX_RULES["camera_configs"][scene_type]
    else:
        return FLUX_RULES["camera_configs"]["default"]


def _get_cinematography_lighting_enhancement(description_lower: str, camera_config: str, scene_type: str) -> str:
    """Enhanced lighting with cinematography principles"""
    # Don't add lighting if already mentioned
    if any(term in description_lower for term in ["lighting", "lit", "illuminated"]) or 'lighting' in camera_config.lower():
        return ""
    
    # Enhanced lighting based on scene type and cinematography knowledge
    if scene_type == "cinematic":
        if any(term in description_lower for term in ["dramatic", "moody", "dark"]):
            return ", dramatic cinematic lighting with practical lights"
        else:
            return ", professional cinematic lighting"
    elif scene_type == "portrait":
        return ", professional studio lighting with subtle rim light"
    elif "dramatic" in description_lower or "chaos" in description_lower:
        return FLUX_RULES["lighting_enhancements"]["dramatic"]
    else:
        return FLUX_RULES["lighting_enhancements"]["default"]


def _get_style_enhancement(scene_type: str, description_lower: str) -> str:
    """Get style enhancement for multi-engine compatibility"""
    style_enhancements = FLUX_RULES.get("style_enhancements", {})
    
    if scene_type == "cinematic":
        if "film grain" not in description_lower:
            return ", " + style_enhancements.get("cinematic", "cinematic composition, film grain")
    elif scene_type in ["portrait", "commercial"]:
        return ", " + style_enhancements.get("photorealistic", "photorealistic, ultra-detailed")
    elif "editorial" in description_lower:
        return ", " + style_enhancements.get("editorial", "editorial photography style")
    
    return ""


def _clean_prompt_formatting(prompt: str) -> str:
    """Clean up prompt formatting"""
    if not prompt:
        return ""
    
    # Ensure it starts with capital letter
    prompt = prompt.strip()
    if prompt:
        prompt = prompt[0].upper() + prompt[1:] if len(prompt) > 1 else prompt.upper()
    
    # Clean up spaces and commas
    prompt = re.sub(r'\s+', ' ', prompt)
    prompt = re.sub(r',\s*,+', ',', prompt)
    prompt = re.sub(r'^\s*,\s*', '', prompt)  # Remove leading commas
    prompt = re.sub(r'\s*,\s*$', '', prompt)  # Remove trailing commas
    
    # Remove redundant periods
    prompt = re.sub(r'\.+', '.', prompt)
    
    return prompt.strip()


def calculate_prompt_score(prompt: str, analysis_data: Optional[Dict[str, Any]] = None) -> Tuple[int, Dict[str, int]]:
    """
    Calculate enhanced quality score with professional cinematography criteria
    
    Args:
        prompt: The prompt to score
        analysis_data: Enhanced analysis data with cinematography context
        
    Returns:
        Tuple of (total_score, breakdown_dict)
    """
    if not prompt:
        return 0, {"prompt_quality": 0, "technical_details": 0, "professional_cinematography": 0, "multi_engine_optimization": 0}
    
    breakdown = {}
    
    # Enhanced Prompt Quality (0-25 points)
    length_score = min(15, len(prompt) // 10)  # Reward appropriate length
    detail_score = min(10, len(prompt.split(',')) * 1.5)  # Reward structured detail
    breakdown["prompt_quality"] = int(length_score + detail_score)
    
    # Technical Details with Cinematography Focus (0-25 points)
    tech_score = 0
    
    # Cinema equipment (higher scores for professional gear)
    cinema_equipment = ['ARRI', 'RED', 'Canon EOS R', 'Sony A1', 'Leica', 'Hasselblad', 'Phase One']
    for equipment in cinema_equipment:
        if equipment.lower() in prompt.lower():
            tech_score += 6
            break
    
    # Lens specifications
    if re.search(r'\d+mm.*f/[\d.]+', prompt):
        tech_score += 5
    
    # Anamorphic and specialized lenses
    if 'anamorphic' in prompt.lower():
        tech_score += 4
    
    # Professional terminology
    tech_keywords = ['shot on', 'lens', 'photography', 'lighting', 'cinematic']
    for keyword in tech_keywords:
        if keyword in prompt.lower():
            tech_score += 2
    
    # Bonus for BAGEL cinematography suggestions
    if analysis_data and analysis_data.get("has_camera_suggestion"):
        tech_score += 8
        
    breakdown["technical_details"] = min(25, tech_score)
    
    # Professional Cinematography (0-25 points) - NEW CATEGORY
    cinema_score = 0
    
    # Professional lighting techniques
    lighting_terms = ['cinematic lighting', 'dramatic lighting', 'studio lighting', 'rim light', 'practical lights']
    cinema_score += sum(3 for term in lighting_terms if term in prompt.lower())
    
    # Composition techniques
    composition_terms = ['composition', 'framing', 'depth of field', 'bokeh', 'rule of thirds']
    cinema_score += sum(2 for term in composition_terms if term in prompt.lower())
    
    # Cinematography style elements
    style_terms = ['film grain', 'anamorphic', 'telephoto compression', 'wide-angle']
    cinema_score += sum(3 for term in style_terms if term in prompt.lower())
    
    # Professional context bonus
    if analysis_data and analysis_data.get("cinematography_context_applied"):
        cinema_score += 5
    
    breakdown["professional_cinematography"] = min(25, cinema_score)
    
    # Multi-Engine Optimization (0-25 points)
    optimization_score = 0
    
    # Check for multi-engine compatible elements
    multi_engine_terms = ['photorealistic', 'ultra-detailed', 'professional photography', 'cinematic']
    optimization_score += sum(3 for term in multi_engine_terms if term in prompt.lower())
    
    # Technical specifications for generation
    if any(camera in prompt for camera in FLUX_RULES["camera_configs"].values()):
        optimization_score += 5
        
    # Lighting configuration
    if any(lighting in prompt for lighting in FLUX_RULES["lighting_enhancements"].values()):
        optimization_score += 4
        
    # Style enhancements
    if any(style in prompt for style in FLUX_RULES.get("style_enhancements", {}).values()):
        optimization_score += 3
    
    breakdown["multi_engine_optimization"] = min(25, optimization_score)
    
    # Calculate total with enhanced weighting
    total_score = sum(breakdown.values())
    
    return total_score, breakdown


def calculate_professional_enhanced_score(prompt: str, analysis_data: Optional[Dict[str, Any]] = None) -> Tuple[int, Dict[str, int]]:
    """
    Enhanced scoring with professional cinematography criteria
    
    Args:
        prompt: The prompt to score
        analysis_data: Analysis data with cinematography context
        
    Returns:
        Tuple of (total_score, breakdown_dict)
    """
    # Use the enhanced scoring system
    return calculate_prompt_score(prompt, analysis_data)


def get_score_grade(score: int) -> Dict[str, str]:
    """
    Get grade information for a score
    
    Args:
        score: Numeric score
        
    Returns:
        Dictionary with grade and color information
    """
    from config import SCORING_CONFIG
    
    for threshold, grade_info in sorted(SCORING_CONFIG["grade_thresholds"].items(), reverse=True):
        if score >= threshold:
            return grade_info
    
    # Default to lowest grade
    return SCORING_CONFIG["grade_thresholds"][0]


def format_analysis_report(analysis_data: Dict[str, Any], processing_time: float) -> str:
    """
    Format analysis data into a readable report with cinematography insights
    
    Args:
        analysis_data: Analysis results with cinematography context
        processing_time: Time taken for processing
        
    Returns:
        Formatted markdown report
    """
    model_used = analysis_data.get("model", "Unknown")
    prompt_length = len(analysis_data.get("prompt", ""))
    has_cinema_context = analysis_data.get("cinematography_context_applied", False)
    scene_type = analysis_data.get("scene_type", "general")
    
    report = f"""**🎬 PHRAMER AI ANALYSIS COMPLETE**
**Model:** {model_used} β€’ **Time:** {processing_time:.1f}s β€’ **Length:** {prompt_length} chars

**πŸ“Š CINEMATOGRAPHY ANALYSIS:**
**Scene Type:** {scene_type.replace('_', ' ').title()}
**Professional Context:** {'βœ… Applied' if has_cinema_context else '❌ Not Applied'}

**🎯 OPTIMIZATIONS APPLIED:**
βœ… Professional camera configuration
βœ… Cinematography lighting setup  
βœ… Technical specifications
βœ… Multi-engine compatibility
βœ… Cinema-quality enhancement

**⚑ Powered by Pariente AI for MIA TV Series**"""
    
    return report


def safe_execute(func, *args, **kwargs) -> Tuple[bool, Any]:
    """
    Safely execute a function with error handling
    
    Args:
        func: Function to execute
        *args: Function arguments
        **kwargs: Function keyword arguments
        
    Returns:
        Tuple of (success: bool, result: Any)
    """
    try:
        result = func(*args, **kwargs)
        return True, result
    except Exception as e:
        logger.error(f"Safe execution failed for {func.__name__}: {e}")
        return False, str(e)


def truncate_text(text: str, max_length: int = 100) -> str:
    """
    Truncate text to specified length with ellipsis
    
    Args:
        text: Text to truncate
        max_length: Maximum length
        
    Returns:
        Truncated text
    """
    if not text or len(text) <= max_length:
        return text
    
    return text[:max_length-3] + "..."


def enhance_prompt_with_cinematography_knowledge(original_prompt: str, scene_type: str = "default") -> str:
    """
    Enhance prompt with professional cinematography knowledge
    
    Args:
        original_prompt: Base prompt text
        scene_type: Detected scene type
        
    Returns:
        Enhanced prompt with cinematography context
    """
    try:
        # Import here to avoid circular imports
        from professional_photography import enhance_flux_prompt_with_professional_knowledge
        
        # Apply professional cinematography enhancement
        enhanced = enhance_flux_prompt_with_professional_knowledge(original_prompt)
        
        logger.info(f"Enhanced prompt with cinematography knowledge for {scene_type} scene")
        return enhanced
        
    except ImportError:
        logger.warning("Professional photography module not available")
        return original_prompt
    except Exception as e:
        logger.warning(f"Cinematography enhancement failed: {e}")
        return original_prompt


# Export main functions
__all__ = [
    "setup_logging",
    "optimize_image", 
    "validate_image",
    "clean_memory",
    "apply_flux_rules",
    "calculate_prompt_score",
    "calculate_professional_enhanced_score",
    "get_score_grade",
    "format_analysis_report",
    "safe_execute",
    "truncate_text",
    "enhance_prompt_with_cinematography_knowledge",
    "detect_scene_type_from_analysis"
]