diodiogod-TTS-Audio-Suite/utils/timing/reporting.py

"""
SRT Reporting - Handles report generation and SRT output formatting
Extracted from the massive SRT TTS node for better maintainability
"""

from typing import Dict, Any, Optional, List, Tuple
from datetime import datetime


class SRTReportGenerator:
    """
    Handles generation of timing reports and adjusted SRT content
    """
    
    def __init__(self):
        """Initialize report generator"""
        pass
    
    def _get_original_overlapping_segments(self, subtitles: List) -> set:
        """Get set of segment indices that overlap with next segments in original SRT."""
        overlapping_segments = set()
        for i in range(len(subtitles) - 1):
            current = subtitles[i]
            next_sub = subtitles[i + 1]
            if current.end_time > next_sub.start_time:
                overlapping_segments.add(i)
        return overlapping_segments
    
    def generate_timing_report(self, subtitles: List, adjustments: List[Dict], timing_mode: str, has_original_overlaps: bool = False, mode_switched: bool = False, original_mode: str = None, stretch_method: str = None) -> str:
        """Generate detailed timing report with original vs generated overlap distinction"""
        import numpy as np
        
        # Get original overlapping segments if needed
        original_overlapping_segments = set()
        if has_original_overlaps:
            original_overlapping_segments = self._get_original_overlapping_segments(subtitles)
        
        # Prepare timing mode display info
        mode_display = timing_mode
        if mode_switched and original_mode:
            mode_display = f"{timing_mode} (switched from {original_mode} due to overlapping subtitles)"
        
        # Handle edge case where no subtitles were processed (immediate interruption)
        if not subtitles:
            return f"""SRT Timing Report ({mode_display} mode)
{'=' * 50}
Total subtitles: 0 (interrupted immediately)
Total duration: 0.000s

No segments were processed due to immediate interruption.
"""
        
        report_lines = [
            f"SRT Timing Report ({mode_display} mode)",
            "=" * 50,
            f"Total subtitles: {len(subtitles)}",
            f"Total duration: {subtitles[-1].end_time:.3f}s",
        ]

        # Add stretcher method info if available
        if stretch_method and timing_mode == "stretch_to_fit":
            report_lines.append(f"Time stretcher: {stretch_method}")

        report_lines.extend(["", "Per-subtitle analysis:"])
        
        if timing_mode == "smart_natural":
            # For smart_natural mode, iterate directly over the detailed adjustments report
            for adj in adjustments:
                report_lines.append(
                    f"  {adj['sequence']:2d}. Original SRT: {adj['original_srt_start']:6.2f}-{adj['original_srt_end']:6.2f}s "
                    f"(Target: {adj['original_srt_duration']:.2f}s)"
                )
                report_lines.append(f"      Natural Audio: {adj['natural_audio_duration']:.3f}s")
                
                for action in adj['actions']:
                    report_lines.append(f"      - {action}")
                
                report_lines.append(
                    f"      Final Audio Duration: {adj['final_segment_duration']:.3f}s "
                    f"(Final SRT: {adj['final_srt_start']:6.2f}-{adj['final_srt_end']:6.2f}s)"
                )
                # Find the corresponding subtitle to get its text
                # This assumes subtitles are sorted by sequence or index
                original_subtitle_text = next((s.text for s in subtitles if s.sequence == adj['sequence']), "N/A")
                report_lines.append(f"      Text: {original_subtitle_text[:60]}{'...' if len(original_subtitle_text) > 60 else ''}")
        elif timing_mode == "concatenate":
            # For concatenate mode, show original vs new timings
            for adj in adjustments:
                timing_change = adj.get('timing_change', 0.0)
                timing_change_info = ""
                if abs(timing_change) > 0.01:  # Only show significant changes
                    if timing_change > 0:
                        timing_change_info = f" [timing +{timing_change:.2f}s later]"
                    else:
                        timing_change_info = f" [timing {timing_change:.2f}s earlier]"
                
                report_lines.append(
                    f"  {adj['sequence']:2d}. Original SRT: {adj['original_srt_start']:6.2f}-{adj['original_srt_end']:6.2f}s "
                    f"→ New: {adj['start_time']:6.2f}-{adj['end_time']:6.2f}s ({adj['natural_duration']:.2f}s){timing_change_info}"
                )
                report_lines.append(f"      Text: {adj['original_text'][:60]}{'...' if len(adj['original_text']) > 60 else ''}")
        else:
            # For other modes, iterate using zip with original subtitles
            for i, (subtitle, adj) in enumerate(zip(subtitles, adjustments)):
                if timing_mode == "pad_with_silence":
                    # For pad_with_silence mode, show overlap/gap information with distinction
                    timing_info = ""
                    if adj['natural_duration'] > subtitle.duration:
                        # Audio is longer than SRT slot - check if it's a real overlap or gap extension
                        overlap = adj['natural_duration'] - subtitle.duration
                        audio_end_time = subtitle.start_time + adj['natural_duration']
                        
                        # Check if there's a next segment and if audio truly overlaps with it
                        has_real_overlap = False
                        if i < len(subtitles) - 1:
                            next_subtitle = subtitles[i + 1]
                            if audio_end_time > next_subtitle.start_time:
                                has_real_overlap = True
                                real_overlap = audio_end_time - next_subtitle.start_time
                        
                        if has_real_overlap:
                            # True overlap - audio collides with next segment
                            if i in original_overlapping_segments:
                                timing_info = f" 🔁✔️ [ORIGINAL+REAL OVERLAP: +{real_overlap:.2f}s into next]"
                            else:
                                timing_info = f" 🔁 [REAL OVERLAP: +{real_overlap:.2f}s into next segment]"
                        else:
                            # Gap extension - audio extends into silence, not a real problem
                            if i in original_overlapping_segments:
                                timing_info = f" 🔁✔️ [ORIGINAL OVERLAP] [gap extension: +{overlap:.2f}s into silence]"
                            else:
                                timing_info = f" [gap extension: +{overlap:.2f}s into silence]"
                    elif i in original_overlapping_segments:
                        # Only original overlap (generation fits within SRT timing)
                        timing_info = f" 🔁✔️ [ORIGINAL OVERLAP]"
                    elif i < len(subtitles) - 1:
                        # Check for silence gap to next subtitle
                        next_subtitle = subtitles[i + 1]
                        gap_duration = next_subtitle.start_time - subtitle.end_time
                        if gap_duration > 0:
                            timing_info = f" [+{gap_duration:.2f}s silence available]"
                    
                    report_lines.append(
                        f"  {subtitle.sequence:2d}. {subtitle.start_time:6.2f}-{subtitle.end_time:6.2f}s "
                        f"({subtitle.duration:.2f}s target, {adj['natural_duration']:.2f}s natural){timing_info}"
                    )
                else:
                    # For other modes (e.g., stretch_to_fit), show stretch information and original overlaps
                    stretch_info = ""
                    if adj['needs_stretching']:
                        # Add emoji for compress/expand modes
                        emoji = ""
                        if adj['stretch_type'] == 'compress':
                            emoji = "🐰 "
                        elif adj['stretch_type'] == 'expand':
                            emoji = "🐌 "
                        stretch_info = f" [{emoji}{adj['stretch_type']} {adj['stretch_factor']:.2f}x]"
                    
                    # Add original overlap indicator for stretch_to_fit mode
                    overlap_info = ""
                    if i in original_overlapping_segments:
                        overlap_info = f" 🔁✔️ [ORIGINAL OVERLAP - expected]"
                    
                    report_lines.append(
                        f"  {subtitle.sequence:2d}. {subtitle.start_time:6.2f}-{subtitle.end_time:6.2f}s "
                        f"({subtitle.duration:.2f}s target, {adj['natural_duration']:.2f}s natural){stretch_info}{overlap_info}"
                    )
                
                report_lines.append(f"      Text: {subtitle.text[:60]}{'...' if len(subtitle.text) > 60 else ''}")
        
        # Summary statistics
        if timing_mode == "pad_with_silence":
            total_gaps = 0
            total_gap_duration = 0
            total_real_overlaps = 0
            total_real_overlap_duration = 0
            total_gap_extensions = 0
            total_gap_extension_duration = 0
            total_original_overlaps = len(original_overlapping_segments)
            total_combined_overlaps = 0
            
            for i, (subtitle, adj) in enumerate(zip(subtitles, adjustments)):
                if adj['natural_duration'] > subtitle.duration:
                    extend_duration = adj['natural_duration'] - subtitle.duration
                    audio_end_time = subtitle.start_time + adj['natural_duration']
                    
                    # Check if there's a real overlap with next segment
                    has_real_overlap = False
                    real_overlap_duration = 0
                    if i < len(subtitles) - 1:
                        next_subtitle = subtitles[i + 1]
                        if audio_end_time > next_subtitle.start_time:
                            has_real_overlap = True
                            real_overlap_duration = audio_end_time - next_subtitle.start_time
                    
                    if has_real_overlap:
                        # True overlap with next segment
                        if i in original_overlapping_segments:
                            total_combined_overlaps += 1
                        else:
                            total_real_overlaps += 1
                        total_real_overlap_duration += real_overlap_duration
                    else:
                        # Gap extension into silence
                        total_gap_extensions += 1
                        total_gap_extension_duration += extend_duration
                        
                elif i < len(subtitles) - 1:
                    gap_duration = subtitles[i + 1].start_time - subtitles[i].end_time
                    if gap_duration > 0:
                        total_gaps += 1
                        total_gap_duration += gap_duration
            
            summary_lines = [
                "",
                "Summary:",
                f"  Audio preserved at natural timing (no stretching)"
            ]
            
            if total_original_overlaps > 0:
                summary_lines.append(f"  Original SRT overlaps: {total_original_overlaps} segments (expected from subtitle file)")
                
            if total_real_overlaps > 0:
                summary_lines.append(f"  🔁 Real overlaps: {total_real_overlaps} segments, +{total_real_overlap_duration:.2f}s total (audio collides with next segment)")
                
            if total_combined_overlaps > 0:
                summary_lines.append(f"  Combined overlaps: {total_combined_overlaps} segments (both original + real generation overlaps)")
            
            if total_gap_extensions > 0:
                summary_lines.append(f"  Gap extensions: {total_gap_extensions} segments, +{total_gap_extension_duration:.2f}s total (extending into silence - not problematic)")
                
            if total_gaps > 0:
                summary_lines.append(f"  Silence gaps available: {total_gaps} gaps, {total_gap_duration:.2f}s total silence")
            
            if total_real_overlaps == 0 and total_gap_extensions == 0 and total_gaps == 0 and total_original_overlaps == 0:
                summary_lines.append(f"  Perfect timing match - no gaps or overlaps")
            
            report_lines.extend(summary_lines)
        elif timing_mode == "concatenate":
            # Summary for concatenate mode
            total_original_duration = sum(adj.get('original_srt_duration', 0) for adj in adjustments)
            total_new_duration = sum(adj.get('natural_duration', 0) for adj in adjustments)
            duration_difference = total_new_duration - total_original_duration
            
            segments_earlier = sum(1 for adj in adjustments if adj.get('timing_change', 0) < -0.01)
            segments_later = sum(1 for adj in adjustments if adj.get('timing_change', 0) > 0.01)
            segments_unchanged = len(adjustments) - segments_earlier - segments_later
            
            summary_lines = [
                "",
                "Summary (Concatenate Mode):",
                f"  Original SRT duration: {total_original_duration:.2f}s",
                f"  Natural audio duration: {total_new_duration:.2f}s",
                f"  Duration change: {duration_difference:+.2f}s",
                f"  Segments with earlier timing: {segments_earlier}",
                f"  Segments with later timing: {segments_later}",
                f"  Segments with similar timing: {segments_unchanged}",
                f"  Audio quality: Natural (no stretching or padding applied)"
            ]
            report_lines.extend(summary_lines)
        elif timing_mode == "smart_natural":
            # Define the same threshold used in smart timing processing
            INSIGNIFICANT_TRUNCATION_THRESHOLD = 0.05
            
            total_shifted = sum(1 for adj in adjustments if adj['next_segment_shifted_by'] > 0)
            total_stretched = sum(1 for adj in adjustments if abs(adj['stretch_factor_applied'] - 1.0) > 0.01)
            total_padded = sum(1 for adj in adjustments if adj['padding_added'] > 0)
            # Only count significant truncations in the summary
            total_truncated = sum(1 for adj in adjustments if adj['truncated_by'] > INSIGNIFICANT_TRUNCATION_THRESHOLD)
            
            summary_lines = [
                "",
                "Summary (Smart Natural Mode):",
                f"  Segments with next segment shifted: {total_shifted}/{len(adjustments)}",
                f"  Segments with audio stretched/shrunk: {total_stretched}/{len(adjustments)}",
                f"  Segments with silence padded: {total_padded}/{len(adjustments)}",
                f"  Segments truncated: {total_truncated}/{len(adjustments)}",
            ]
            report_lines.extend(summary_lines)
        else:
            total_stretch_needed = sum(1 for adj in adjustments if adj['needs_stretching'])
            avg_stretch = np.mean([adj['stretch_factor'] for adj in adjustments])
            total_original_overlaps = len(original_overlapping_segments)
            
            summary_lines = [
                "",
                "Summary:",
                f"  Segments needing time adjustment: {total_stretch_needed}/{len(adjustments)}",
                f"  Average stretch factor: {avg_stretch:.2f}x",
            ]
            
            if total_original_overlaps > 0:
                summary_lines.append(f"  Original SRT overlaps: {total_original_overlaps} segments (expected behavior - segments will overlap as intended)")
            
            report_lines.extend(summary_lines)
        
        return "\n".join(report_lines)
    
    def generate_adjusted_srt_string(self, subtitles: List, adjustments: List[Dict], timing_mode: str) -> str:
        """Generate adjusted SRT string from final timings - EXACT ORIGINAL LOGIC"""
        if not adjustments:
            return "# No subtitles processed\n"
        
        srt_lines = []
        for i, adj in enumerate(adjustments):
            def format_time(seconds):
                hours = int(seconds // 3600)
                minutes = int((seconds % 3600) // 60)
                seconds_int = int(seconds % 60)
                milliseconds = int((seconds - int(seconds)) * 1000)
                return f"{hours:02}:{minutes:02}:{seconds_int:02},{milliseconds:03}"
            
            # Determine timing based on mode
            if timing_mode == "smart_natural":
                start_time = adj.get('final_srt_start', adj.get('original_srt_start', 0))
                end_time = start_time + adj.get('final_segment_duration', adj.get('natural_audio_duration', 1))
            elif timing_mode == "concatenate":
                # Use the recalculated timings from concatenation
                start_time = adj.get('start_time', 0)
                end_time = adj.get('end_time', adj.get('start_time', 0) + adj.get('natural_duration', 1))
            else:
                start_time = adj.get('start_time', subtitles[i].start_time if i < len(subtitles) else 0)
                end_time = adj.get('end_time', subtitles[i].end_time if i < len(subtitles) else 1)
            
            start_time_str = format_time(start_time)
            end_time_str = format_time(end_time)
            
            # Get text
            text = adj.get('original_text', subtitles[i].text if i < len(subtitles) else f"Subtitle {i+1}")
            
            srt_lines.append(str(adj.get('sequence', i+1)))
            srt_lines.append(f"{start_time_str} --> {end_time_str}")
            srt_lines.append(text)
            srt_lines.append("")
        
        return "\n".join(srt_lines)
    
    def generate_generation_info(self, total_duration: float, segment_count: int, 
                               timing_mode: str, cache_status: str, 
                               model_source: str) -> str:
        """
        Generate concise generation info string
        
        Args:
            total_duration: Total audio duration in seconds
            segment_count: Number of segments processed
            timing_mode: Timing mode used
            cache_status: Cache status ("cached", "generated", "mixed")
            model_source: Model source information
            
        Returns:
            Formatted generation info string
        """
        return (f"Generated {total_duration:.1f}s SRT-timed audio from {segment_count} subtitles "
                f"using {timing_mode} mode ({cache_status} segments, {model_source} models)")
    
    def _generate_smart_natural_analysis(self, adjustments: List[Dict]) -> List[str]:
        """Generate analysis for smart natural timing mode"""
        analysis_lines = []
        
        for adj in adjustments:
            sequence = adj.get('sequence', adj.get('segment_index', 0) + 1)
            original_start = adj.get('original_srt_start', 0)
            original_end = adj.get('original_srt_end', 0)
            original_duration = adj.get('original_srt_duration', original_end - original_start)
            
            analysis_lines.append(
                f"  {sequence:2d}. Original SRT: {original_start:6.2f}-{original_end:6.2f}s "
                f"(Target: {original_duration:.2f}s)"
            )
            
            # Add actions taken
            actions = adj.get('actions', ['No actions recorded'])
            for action in actions:
                analysis_lines.append(f"      - {action}")
            
            # Add timing deviation info if available
            if 'timing_deviation' in adj:
                deviation = adj['timing_deviation']
                within_tolerance = adj.get('within_tolerance', True)
                status = "✓" if within_tolerance else "⚠"
                analysis_lines.append(f"      {status} Timing deviation: {deviation:.3f}s")
            
            # Add text preview
            text = adj.get('original_text', 'No text available')
            preview = text[:60] + '...' if len(text) > 60 else text
            analysis_lines.append(f"      Text: {preview}")
            analysis_lines.append("")  # Empty line between entries
        
        return analysis_lines
    
    def _generate_standard_analysis(self, subtitles: List, adjustments: List[Dict]) -> List[str]:
        """Generate analysis for standard timing modes"""
        analysis_lines = []
        
        for i, adj in enumerate(adjustments):
            if i < len(subtitles):
                subtitle = subtitles[i]
                sequence = subtitle.sequence
                start_time = subtitle.start_time
                end_time = subtitle.end_time
                duration = subtitle.duration
                text = subtitle.text
            else:
                sequence = adj.get('sequence', i + 1)
                start_time = adj.get('start_time', 0)
                end_time = adj.get('end_time', 1)
                duration = end_time - start_time
                text = adj.get('original_text', f'Subtitle {i+1}')
            
            # Format stretch information
            stretch_info = ""
            if adj.get('needs_stretching'):
                stretch_type = adj.get('stretch_type', 'unknown')
                stretch_factor = adj.get('stretch_factor', 1.0)
                # Add emoji for compress/expand modes
                emoji = ""
                if stretch_type == 'compress':
                    emoji = "🐰 "
                elif stretch_type == 'expand':
                    emoji = "🐌 "
                stretch_info = f" [{emoji}{stretch_type} {stretch_factor:.2f}x]"
            
            natural_duration = adj.get('natural_duration', duration)
            
            analysis_lines.append(
                f"  {sequence:2d}. {start_time:6.2f}-{end_time:6.2f}s "
                f"({duration:.2f}s target, {natural_duration:.2f}s natural){stretch_info}"
            )
            
            # Add text preview
            preview = text[:60] + '...' if len(text) > 60 else text
            analysis_lines.append(f"      Text: {preview}")
        
        return analysis_lines
    
    def _generate_summary_statistics(self, adjustments: List[Dict], timing_mode: str) -> List[str]:
        """Generate summary statistics from adjustments"""
        stats_lines = []
        
        # Count different types of adjustments
        stretched_count = sum(1 for adj in adjustments if adj.get('needs_stretching', False))
        no_change_count = len(adjustments) - stretched_count
        
        if timing_mode == "smart_natural":
            # Smart natural specific stats
            shifted_count = sum(1 for adj in adjustments if adj.get('next_segment_shifted_by', 0) > 0)
            padded_count = sum(1 for adj in adjustments if adj.get('padding_added', 0) > 0)
            
            stats_lines.extend([
                f"- Segments using natural timing: {no_change_count}",
                f"- Segments requiring adjustment: {stretched_count}",
                f"- Segments causing shifts: {shifted_count}",
                f"- Segments with padding: {padded_count}"
            ])
            
            # Calculate average timing deviation
            deviations = [adj.get('timing_deviation', 0) for adj in adjustments if 'timing_deviation' in adj]
            if deviations:
                avg_deviation = sum(deviations) / len(deviations)
                max_deviation = max(deviations)
                stats_lines.append(f"- Average timing deviation: {avg_deviation:.3f}s")
                stats_lines.append(f"- Maximum timing deviation: {max_deviation:.3f}s")
        
        else:
            # Standard timing mode stats
            if stretched_count > 0:
                compress_count = sum(1 for adj in adjustments 
                                   if adj.get('stretch_type') == 'compress')
                expand_count = sum(1 for adj in adjustments 
                                 if adj.get('stretch_type') == 'expand')
                
                stats_lines.extend([
                    f"- Segments requiring stretching: {stretched_count}",
                    f"  - Compressed: {compress_count}",
                    f"  - Extended: {expand_count}",
                    f"- Segments with natural timing: {no_change_count}"
                ])
            else:
                stats_lines.append(f"- All {len(adjustments)} segments used natural timing")
        
        return stats_lines
    
    def _format_srt_timestamp(self, seconds: float) -> str:
        """
        Format seconds as SRT timestamp (HH:MM:SS,mmm)
        
        Args:
            seconds: Time in seconds
            
        Returns:
            Formatted timestamp string
        """
        hours = int(seconds // 3600)
        minutes = int((seconds % 3600) // 60)
        seconds_int = int(seconds % 60)
        milliseconds = int((seconds - int(seconds)) * 1000)
        
        return f"{hours:02}:{minutes:02}:{seconds_int:02},{milliseconds:03}"
    
    def generate_debug_report(self, subtitles: List, adjustments: List[Dict], 
                            audio_segments: List = None) -> str:
        """
        Generate detailed debug report for troubleshooting
        
        Args:
            subtitles: List of SRTSubtitle objects
            adjustments: List of timing adjustments
            audio_segments: Optional list of audio segments for analysis
            
        Returns:
            Detailed debug report string
        """
        debug_lines = [
            "SRT Processing Debug Report",
            "=" * 50,
            f"Generated: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}",
            "",
            f"Input subtitles: {len(subtitles) if subtitles else 0}",
            f"Generated adjustments: {len(adjustments)}",
            f"Audio segments: {len(audio_segments) if audio_segments else 'N/A'}",
            ""
        ]
        
        # Add detailed adjustment analysis
        if adjustments:
            debug_lines.append("Detailed Adjustment Analysis:")
            for i, adj in enumerate(adjustments):
                debug_lines.append(f"Adjustment {i+1}:")
                for key, value in adj.items():
                    if isinstance(value, float):
                        debug_lines.append(f"  {key}: {value:.6f}")
                    elif isinstance(value, list):
                        debug_lines.append(f"  {key}: {len(value)} items")
                    else:
                        debug_lines.append(f"  {key}: {value}")
                debug_lines.append("")
        
        return "\n".join(debug_lines)