TextMind.py

#!/usr/bin/env python3

import sys, re, os, json
from collections import Counter
from pathlib import Path
from datetime import datetime

config = {
    'positive_words': {'good', 'great', 'excellent', 'amazing', 'wonderful', 'fantastic', 'happy', 'joy', 'love', 'brilliant', 'perfect', 'beautiful', 'best', 'success', 'win', 'positive', 'hope', 'improve', 'benefit', 'advantage'},
    'negative_words': {'bad', 'terrible', 'awful', 'horrible', 'poor', 'worst', 'hate', 'sad', 'anger', 'fail', 'problem', 'issue', 'wrong', 'error', 'difficult', 'crisis', 'threat', 'danger', 'risk', 'decline', 'worse', 'inferior'},
    'stop_words': {'the', 'a', 'an', 'and', 'or', 'but', 'in', 'on', 'at', 'to', 'for', 'of', 'with', 'by', 'from', 'as', 'is', 'was', 'are', 'were', 'been', 'be', 'have', 'has', 'had', 'do', 'does', 'did', 'will', 'would', 'could', 'should', 'may', 'might', 'can', 'this', 'that', 'these', 'those', 'i', 'you', 'he', 'she', 'it', 'we', 'they', 'what', 'which'},
    'subjective_words': {'believe', 'think', 'feel', 'opinion', 'suggest', 'claim'}, 'objective_words': {'data', 'fact', 'study', 'research', 'evidence', 'result'},
    'colors': {'red': '\033[91m', 'green': '\033[92m', 'yellow': '\033[93m', 'blue': '\033[94m', 'cyan': '\033[96m', 'white': '\033[97m', 'reset': '\033[0m', 'bold': '\033[1m'}
}

data = {'raw': '', 'sentences': [], 'words': [], 'clean_words': []}
results = {}

args = sys.argv[1:] if len(sys.argv) > 1 else []

def check_install(pkg_import, pkg_name=None):
    pkg_name = pkg_name or pkg_import
    try:
        __import__(pkg_import)
        return True
    except ImportError:
        print(f"{config['colors']['yellow']}Installing {pkg_name}...{config['colors']['reset']}")
        os.system(f"{sys.executable} -m pip install {pkg_name} -q")
        return True

def load_text(path):
    if not Path(path).exists():
        print(f"{config['colors']['red']}Error: File not found{config['colors']['reset']}")
        sys.exit(1)
    temp = Path(path).suffix.lower()
    if temp in ['.txt', '.md']:
        with open(path, 'r', encoding='utf-8') as file_handle:
            return file_handle.read()
    elif temp == '.pdf':
        check_install('PyPDF2')
        import PyPDF2
        temp2 = ''
        with open(path, 'rb') as file_handle:
            temp3 = PyPDF2.PdfReader(file_handle)
            for temp4 in range(len(temp3.pages)):
                temp2 += temp3.pages[temp4].extract_text() + '\n'
        # Format multi-column text
        temp2 = re.sub(r'\n\s*\n\s*\n+', '\n\n', temp2)
        temp2 = re.sub(r'[ \t]+', ' ', temp2)
        temp2 = re.sub(r'(\w+)-\s*\n\s*(\w+)', r'\1\2', temp2)
        return temp2
    elif temp == '.docx':
        check_install('docx', 'python-docx')
        import docx
        temp2 = docx.Document(path)
        return '\n'.join([temp3.text for temp3 in temp2.paragraphs])
    print(f"{config['colors']['red']}Unsupported format{config['colors']['reset']}")
    sys.exit(1)

def preprocess():
    data['sentences'] = [temp.strip() for temp in re.split(r'[.!?]+', data['raw']) if temp.strip()]
    data['words'] = re.findall(r'\b[a-z]+\b', data['raw'].lower())
    data['clean_words'] = [temp for temp in data['words'] if temp not in config['stop_words'] and len(temp) > 2]

def analyze():
    # Word intelligence
    temp = Counter(data['clean_words'])
    results['word_freq'] = temp.most_common(10)
    results['word_diversity'] = round(len(set(data['words'])) / len(data['words']) * 100, 2) if data['words'] else 0
    results['rare_words'] = [w for w, f in temp.items() if f == 1 and len(w) > 6][:5]
    # Sentence ranking and summaries
    temp2 = []
    for temp3 in data['sentences']:
        temp4 = sum([temp.get(temp5.lower(), 0) for temp5 in re.findall(r'\b\w+\b', temp3)])
        temp5 = data['sentences'].index(temp3) / len(data['sentences'])
        temp4 = temp4 * (1 + (0.3 if temp5 < 0.2 or temp5 > 0.8 else 0))
        temp2.append((temp3, temp4, data['sentences'].index(temp3)))
    results['top_sentences'] = sorted(temp2, key=lambda t: t[1], reverse=True)[:5]
    # Small summary (15%)
    temp6 = max(2, int(len(data['sentences']) * 0.15))
    temp7 = sorted(temp2, key=lambda t: t[1], reverse=True)[:temp6]
    temp7 = sorted(temp7, key=lambda t: t[2])
    results['summary_small'] = ' '.join([t[0] for t in temp7])
    # Detailed summary (40%)
    temp8 = max(4, int(len(data['sentences']) * 0.4))
    temp9 = sorted(temp2, key=lambda t: t[1], reverse=True)[:temp8]
    temp9 = sorted(temp9, key=lambda t: t[2])
    temp10 = len(temp9) // 3
    temp11 = []
    for i in range(0, len(temp9), temp10):
        temp11.append(' '.join([t[0] for t in temp9[i:i + temp10]]))
    results['summary_detailed'] = '\n\n'.join(temp11[:3])
    # Tone analysis
    temp12 = {'positive': 0, 'negative': 0, 'neutral': 0}
    temp13 = []
    for temp14 in data['sentences']:
        temp15 = temp14.lower()
        temp16 = sum(1 for w in config['positive_words'] if w in temp15)
        temp17 = sum(1 for w in config['negative_words'] if w in temp15)
        if temp16 > temp17:
            temp12['positive'] += 1
            temp13.append(('positive', temp14))
        elif temp17 > temp16:
            temp12['negative'] += 1
            temp13.append(('negative', temp14))
        else:
            temp12['neutral'] += 1
            temp13.append(('neutral', temp14))
    results['tone_counts'] = temp12
    results['tone_sentences'] = temp13
    # Bias analysis
    temp18 = sum(1 for w in data['words'] if w in config['subjective_words'])
    temp19 = sum(1 for w in data['words'] if w in config['objective_words'])
    temp20 = temp18 + temp19
    results['subjectivity'] = round(temp18 / temp20 * 100, 2) if temp20 > 0 else 0
    results['objectivity'] = round(temp19 / temp20 * 100, 2) if temp20 > 0 else 0
    # Semantic fingerprint & topic distribution
    results['features'] = {
        'lexical_density': len(data['clean_words']) / len(data['words']) if data['words'] else 0,
        'avg_word_length': sum(len(w) for w in data['clean_words']) / len(data['clean_words']) if data['clean_words'] else 0,
        'complexity': len(data['words']) / len(data['sentences']) if data['sentences'] else 0,
        'vocabulary_richness': len(set(data['clean_words'])) / len(data['clean_words']) if data['clean_words'] else 0
    }
    # Topic distribution
    temp21 = {'emotion': 0, 'action': 0, 'abstract': 0, 'technical': 0, 'social': 0}
    temp22 = {'love', 'hate', 'happy', 'sad', 'fear', 'anger', 'joy', 'hope', 'feel'}
    temp23 = {'do', 'make', 'create', 'build', 'move', 'act', 'run', 'work', 'start'}
    temp24 = {'idea', 'concept', 'theory', 'principle', 'thought', 'belief', 'value'}
    temp25 = {'data', 'system', 'process', 'function', 'method', 'algorithm', 'analysis'}
    temp26 = {'people', 'society', 'community', 'group', 'team', 'family', 'human'}
    for w in data['clean_words']:
        if w in temp22: temp21['emotion'] += 1
        elif w in temp23: temp21['action'] += 1
        elif w in temp24: temp21['abstract'] += 1
        elif w in temp25: temp21['technical'] += 1
        elif w in temp26: temp21['social'] += 1
    temp27 = sum(temp21.values())
    results['topics'] = {k: round(v/temp27*100, 2) if temp27 > 0 else 0 for k, v in temp21.items()}
    # Stats
    results['stats'] = {
        'sentences': len(data['sentences']),
        'words': len(data['words']),
        'unique': len(set(data['words'])),
        'read_time': round(len(data['words']) / 300, 2),
        'speak_time': round(len(data['words']) / 150, 2),
        'avg_length': round(len(data['words']) / len(data['sentences']), 2) if data['sentences'] else 0
    }

def keyword_context(keyword):
    temp = []
    temp2 = r'\b' + re.escape(keyword.lower()) + r'\b'
    for temp3 in data['sentences']:
        if re.search(temp2, temp3.lower()):
            temp4 = re.sub(temp2, lambda m: f"{config['colors']['yellow']}{config['colors']['bold']}{temp3[m.start():m.end()]}{config['colors']['reset']}", temp3, flags=re.IGNORECASE)
            temp.append(temp4)
    return temp

def export_json():
    temp = Path(args[0]).stem
    temp2 = f"textmind_{temp}_{datetime.now().strftime('%Y%m%d_%H%M%S')}.json"
    temp3 = results['tone_counts']
    temp4 = sum(temp3.values())
    temp5 = {
        'metadata': {'source': args[0], 'generated': datetime.now().isoformat(), 'version': '1.0'},
        'statistics': results['stats'],
        'summaries': { 'small': results['summary_small'], 'detailed': results['summary_detailed'], 'top_sentences': [{'sentence': s, 'score': sc} for s, sc, _ in results['top_sentences']] },
        'sentiment_analysis': { 'counts': results['tone_counts'], 'percentages': {k: round(v/temp4*100, 2) for k, v in temp3.items()} if temp4 > 0 else {}, 'all_sentences': [{'tone': t, 'sentence': s} for t, s in results['tone_sentences']], 'positive_sentences': [s for t, s in results['tone_sentences'] if t == 'positive'], 'negative_sentences': [s for t, s in results['tone_sentences'] if t == 'negative'], 'neutral_sentences': [s for t, s in results['tone_sentences'] if t == 'neutral'] },
        'debate_analysis': { 'balance_positive_percent': round(temp3['positive']/(temp3['positive']+temp3['negative'])*100, 2) if (temp3['positive']+temp3['negative']) > 0 else 0, 'balance_negative_percent': round(temp3['negative']/(temp3['positive']+temp3['negative'])*100, 2) if (temp3['positive']+temp3['negative']) > 0 else 0, 'positive_arguments': [s for t, s in results['tone_sentences'] if t == 'positive'], 'negative_arguments': [s for t, s in results['tone_sentences'] if t == 'negative'] },
        'word_intelligence': { 'top_keywords': [{'word': w, 'frequency': f} for w, f in results['word_freq']], 'word_diversity_percent': results['word_diversity'] },
        'bias_analysis': { 'subjectivity_percent': results['subjectivity'], 'objectivity_percent': results['objectivity'], 'classification': 'subjective' if results['subjectivity'] > 60 else 'objective' if results['objectivity'] > 60 else 'balanced' },
        'linguistic_features': results['features']
    }
    with open(temp2, 'w', encoding='utf-8') as file_handle:
        json.dump(temp5, file_handle, indent=2, ensure_ascii=False)
    pc(f"{config['colors']['green']}✅ Exported: {temp2}{config['colors']['reset']}")
    pc(f"  Contains: {len(temp5)} sections with complete analysis", 'cyan')

def export_ml():
    temp = Path(args[0]).stem
    temp2 = f"textmind_ml_{temp}_{datetime.now().strftime('%Y%m%d_%H%M%S')}.json"
    temp3 = results['tone_counts']
    temp4 = sum(temp3.values())
    # Calculate additional ML metrics
    temp5 = results['topics']
    temp6 = max(temp5.values()) if temp5 else 0
    temp7 = max(temp5, key=temp5.get) if temp5 else 'general'
    temp8 = {
        'metadata': { 'document_id': temp, 'timestamp': datetime.now().isoformat(), 'source_file': args[0], 'processing_engine': 'TextMind v1.0' },
        'feature_vectors': {
            'linguistic_features': { 'lexical_density': round(results['features']['lexical_density'], 4), 'avg_word_length': round(results['features']['avg_word_length'], 4), 'avg_word_length_normalized': round(results['features']['avg_word_length'] / 15, 4), 'sentence_complexity': round(results['features']['complexity'], 4), 'sentence_complexity_normalized': round(min(results['features']['complexity'] / 50, 1.0), 4), 'vocabulary_richness': round(results['features']['vocabulary_richness'], 4), 'word_diversity': round(results['word_diversity'] / 100, 4) },
            'sentiment_features': { 'positive_ratio': round(temp3['positive'] / temp4, 4) if temp4 > 0 else 0, 'negative_ratio': round(temp3['negative'] / temp4, 4) if temp4 > 0 else 0, 'neutral_ratio': round(temp3['neutral'] / temp4, 4) if temp4 > 0 else 0, 'sentiment_polarity': round((temp3['positive'] - temp3['negative']) / temp4, 4) if temp4 > 0 else 0, 'sentiment_strength': round((temp3['positive'] + temp3['negative']) / temp4, 4) if temp4 > 0 else 0 },
            'bias_features': { 'subjectivity_score': round(results['subjectivity'] / 100, 4), 'objectivity_score': round(results['objectivity'] / 100, 4), 'bias_ratio': round(results['subjectivity'] / (results['objectivity'] + 1), 4) },
            'topic_features': { 'emotion_score': round(temp5['emotion'] / 100, 4), 'action_score': round(temp5['action'] / 100, 4), 'abstract_score': round(temp5['abstract'] / 100, 4), 'technical_score': round(temp5['technical'] / 100, 4), 'social_score': round(temp5['social'] / 100, 4), 'dominant_topic_score': round(temp6 / 100, 4) },
            'document_features': { 'length_words_normalized': round(min(results['stats']['words'] / 5000, 1.0), 4), 'length_sentences_normalized': round(min(results['stats']['sentences'] / 500, 1.0), 4), 'unique_word_ratio': round(results['stats']['unique'] / results['stats']['words'], 4) if results['stats']['words'] > 0 else 0, 'reading_difficulty': round(results['stats']['avg_length'] / 30, 4) },
            'debate_features': { 'positive_argument_ratio': round(temp3['positive'] / (temp3['positive'] + temp3['negative']), 4) if (temp3['positive'] + temp3['negative']) > 0 else 0.5, 'negative_argument_ratio': round(temp3['negative'] / (temp3['positive'] + temp3['negative']), 4) if (temp3['positive'] + temp3['negative']) > 0 else 0.5, 'argumentative_balance': round(1 - abs(temp3['positive'] - temp3['negative']) / (temp3['positive'] + temp3['negative']), 4) if (temp3['positive'] + temp3['negative']) > 0 else 1.0 }
        },
        'categorical_labels': { 'sentiment_primary': 'positive' if temp3['positive'] > temp3['negative'] else 'negative' if temp3['negative'] > temp3['positive'] else 'neutral', 'sentiment_strength': 'strong' if max(temp3.values()) / temp4 > 0.6 else 'moderate' if max(temp3.values()) / temp4 > 0.4 else 'weak', 'writing_style': 'subjective' if results['subjectivity'] > 60 else 'objective' if results['objectivity'] > 60 else 'balanced', 'debate_position': 'pro' if temp3['positive'] > temp3['negative'] else 'con' if temp3['negative'] > temp3['positive'] else 'neutral', 'complexity_level': 'high' if results['features']['complexity'] > 25 else 'medium' if results['features']['complexity'] > 15 else 'low', 'dominant_topic': temp7, 'content_type': 'technical' if temp5['technical'] > 20 else 'emotional' if temp5['emotion'] > 20 else 'analytical' if temp5['abstract'] > 20 else 'general' },
        'training_datasets': {
            'text_samples': { 'summary_short': results['summary_small'], 'summary_long': results['summary_detailed'], 'all_sentences': data['sentences'] },
            'labeled_sentences': { 'positive': [s for t, s in results['tone_sentences'] if t == 'positive'], 'negative': [s for t, s in results['tone_sentences'] if t == 'negative'], 'neutral': [s for t, s in results['tone_sentences'] if t == 'neutral'] },
            'keywords': { 'top_10': [w for w, f in results['word_freq']], 'top_10_with_freq': [{'word': w, 'frequency': f, 'normalized_freq': round(f/results['stats']['words'], 4)} for w, f in results['word_freq']], 'rare_impactful': results.get('rare_words', []) }
        },
        'raw_statistics': { 'word_count': results['stats']['words'], 'sentence_count': results['stats']['sentences'], 'unique_word_count': results['stats']['unique'], 'avg_sentence_length': results['stats']['avg_length'], 'reading_time_minutes': results['stats']['read_time'], 'speaking_time_minutes': results['stats']['speak_time'], 'tone_distribution': results['tone_counts'], 'topic_distribution': results['topics'] },
        'ml_usage_guide': {
            'recommended_models': { 'classification': ['RandomForest', 'SVM', 'XGBoost', 'Neural Networks'], 'regression': ['Linear Regression', 'Ridge', 'Lasso', 'Gradient Boosting'], 'clustering': ['KMeans', 'DBSCAN', 'Hierarchical'], 'deep_learning': ['LSTM', 'BERT', 'Transformers'] },
            'feature_engineering_tips': [ 'Use feature_vectors for direct ML input', 'Combine linguistic + sentiment features for classification', 'Use labeled_sentences for supervised learning', 'Topic features good for content categorization', 'Debate features useful for argument mining' ],
            'use_cases': [ 'Sentiment Analysis: Use sentiment_features + labeled_sentences', 'Content Classification: Use topic_features + categorical_labels', 'Readability Scoring: Use linguistic_features + complexity_level', 'Debate Analysis: Use debate_features + debate_position', 'Style Transfer: Use writing_style + bias_features', 'Text Quality: Use vocabulary_richness + word_diversity' ],
            'feature_vector_format': 'All features normalized to [0, 1] range for direct ML consumption',
            'total_features': 27
        }
    }
    with open(temp2, 'w', encoding='utf-8') as file_handle:
        json.dump(temp8, file_handle, indent=2, ensure_ascii=False)
    
    pc(f"\n{config['colors']['green']}✅ ML DATASET EXPORTED SUCCESSFULLY{config['colors']['reset']}", 'green', True)
    pc(f"{'='*60}", 'green')
    pc(f"📁 File: {temp2}", 'cyan')
    pc(f"📊 Features: {temp8['ml_usage_guide']['total_features']} normalized vectors", 'cyan')
    pc(f"🏷️  Labels: {len(temp8['categorical_labels'])} categorical + {len(temp8['feature_vectors'])} feature groups", 'cyan')
    pc(f"📝 Training Samples: {len(data['sentences'])} sentences", 'cyan')
    pc(f"{'='*60}", 'green')
    pc("\n🎯 QUICK STATS:", 'yellow', True)
    pc(f"  Sentiment: {temp8['categorical_labels']['sentiment_primary'].upper()} ({temp8['categorical_labels']['sentiment_strength']})", 'white')
    pc(f"  Style: {temp8['categorical_labels']['writing_style'].upper()}", 'white')
    pc(f"  Complexity: {temp8['categorical_labels']['complexity_level'].upper()}", 'white')
    pc(f"  Dominant Topic: {temp8['categorical_labels']['dominant_topic'].upper()}", 'white')
    pc("\n💡 READY FOR:", 'yellow', True)
    pc("  ✓ scikit-learn (Classification, Regression, Clustering)", 'green')
    pc("  ✓ TensorFlow/PyTorch (Deep Learning)", 'green')
    pc("  ✓ XGBoost/LightGBM (Gradient Boosting)", 'green')
    pc("  ✓ BERT/Transformers (NLP Tasks)", 'green')

def pc(text, color='white', bold=False):
    temp = config['colors'].get(color, config['colors']['white'])
    temp2 = config['colors']['bold'] if bold else ''
    print(f"{temp2}{temp}{text}{config['colors']['reset']}")

def show_help():
    pc("\n╔════════════════════════════════════════════════════════════╗", 'cyan', True)
    pc("║           TEXTMIND - Text Intelligence System              ║", 'cyan', True)
    pc("╚════════════════════════════════════════════════════════════╝", 'cyan', True)
    print("\nUSAGE: python textmind.py <file> [options]")
    print("\nFORMATS: .txt, .md, .pdf, .docx")
    print("\nOPTIONS:")
    print("  --summary         Small summary (1 paragraph, default)")
    print("  --summary-detailed Detailed summary (2-3 paragraphs)")
    print("  --report          Full intelligence report")
    print("  --tone            Sentiment analysis")
    print("  --positive        List positive sentences")
    print("  --negative        List negative sentences")
    print("  --debate          Pro vs con analysis")
    print("  --context <word>  Find keyword with highlighting")
    print("  --export-json     Export complete JSON report")
    print("  --export-ml       Export ML-ready dataset")
    print("  --help            Show this help")

def main():
    if not args or '--help' in args:
        show_help()
        sys.exit(0)
    temp = args[0]
    pc("\n🧠 TextMind - Initializing...", 'cyan', True)
    print(f"Loading: {temp}")
    data['raw'] = load_text(temp)
    preprocess()
    if not data['sentences']:
        pc("\n❌ Error: No readable text found", 'red', True)
        sys.exit(1)
    analyze()
    if '--export-json' in args:
        export_json()
    elif '--export-ml' in args:
        export_ml()
    elif '--context' in args:
        try:
            temp2 = args[args.index('--context') + 1]
            temp3 = keyword_context(temp2)
            pc(f"\n{'='*60}", 'cyan')
            pc(f"  KEYWORD CONTEXT: '{temp2}'", 'cyan', True)
            pc(f"{'='*60}", 'cyan')
            if temp3:
                pc(f"\nFound {len(temp3)} occurrence(s):\n", 'cyan')
                for i, sent in enumerate(temp3, 1):
                    print(f"{i}. {sent}\n")
            else:
                pc(f"\n❌ Keyword '{temp2}' not found", 'red')
        except (IndexError, ValueError):
            pc("\n❌ Error: Specify keyword after --context", 'red')
    elif '--report' in args:
        pc(f"\n{'='*32}", 'cyan', True)
        pc("║ TEXTMIND INTELLIGENCE REPORT ║", 'cyan', True)
        pc(f"{'='*32}", 'cyan', True)
        pc("\n📊 DOCUMENT STATISTICS", 'blue', True)
        pc("─" * 70, 'blue')
        print(f"  📄 Total Sentences: {results['stats']['sentences']}")
        print(f"  📝 Total Words: {results['stats']['words']}")
        print(f"  🔤 Unique Words: {results['stats']['unique']}")
        print(f"  📈 Word Diversity: {results['word_diversity']}%")
        print(f"  📏 Avg Sentence Length: {results['stats']['avg_length']} words")
        print(f"  ⏱️  Reading Time: {results['stats']['read_time']} minutes")
        print(f"  🗣️  Speaking Time: {results['stats']['speak_time']} minutes")
        pc("\n🎯 SENTIMENT ANALYSIS", 'blue', True)
        pc("─" * 70, 'blue')
        temp = results['tone_counts']
        temp2 = sum(temp.values())
        for k, v in temp.items():
            temp3 = round(v/temp2*100, 1) if temp2 > 0 else 0
            temp4 = 'green' if k == 'positive' else 'red' if k == 'negative' else 'yellow'
            temp5 = '█' * int(temp3/5)
            pc(f"  {k.capitalize():10s} [{temp5:20s}] {v:3d} ({temp3}%)", temp4)
        pc("\n💡 WORD INTELLIGENCE", 'blue', True)
        pc("─" * 70, 'blue')
        pc("  Top 10 Keywords:", 'cyan')
        for i, (w, f) in enumerate(results['word_freq'], 1):
            temp6 = '▓' * min(int(f/2), 20)
            pc(f"    {i:2d}. {w:15s} {temp6} {f}", 'white')
        if results.get('rare_words'):
            pc("\n  🔍 Hidden Gems (rare impactful words):", 'magenta')
            for w in results['rare_words']:
                pc(f"    • {w}", 'magenta')
        pc("\n📄 DOCUMENT SUMMARY", 'blue', True)
        pc("─" * 70, 'blue')
        pc("  (Extractive summary - maintains narrative flow)\n", 'cyan')
        print(f"  {results['summary_small']}")
        pc("\n🎭 BIAS ANALYSIS", 'blue', True)
        pc("─" * 70, 'blue')
        print(f"  📊 Subjectivity: {results['subjectivity']}%")
        print(f"  📊 Objectivity: {results['objectivity']}%")
        if results['subjectivity'] > 60:
            pc("  → Writing Style: HIGHLY SUBJECTIVE", 'yellow')
        elif results['objectivity'] > 60:
            pc("  → Writing Style: HIGHLY OBJECTIVE", 'green')
        else:
            pc("  → Writing Style: BALANCED PERSPECTIVE", 'cyan')
        pc("\n🔬 SEMANTIC FINGERPRINT", 'blue', True)
        pc("─" * 70, 'blue')
        temp7 = results['features']
        print(f"  • Lexical Density: {temp7['lexical_density']:.2%} (meaningful word ratio)")
        print(f"  • Avg Word Length: {temp7['avg_word_length']:.1f} characters")
        print(f"  • Sentence Complexity: {temp7['complexity']:.1f} words/sentence")
        print(f"  • Vocabulary Richness: {temp7['vocabulary_richness']:.2%} (unique ratio)")
        pc("\n📊 TOPIC DISTRIBUTION", 'blue', True)
        pc("─" * 70, 'blue')
        temp8 = results['topics']
        for k, v in sorted(temp8.items(), key=lambda x: x[1], reverse=True):
            if v > 0:
                temp9 = '█' * int(v/5)
                pc(f"  {k.capitalize():12s} [{temp9:20s}] {v:.1f}%", 'cyan')
        pc(f"\n{'='*70}", 'cyan', True)
    elif '--summary-detailed' in args:
        pc(f"\n{'='*60}", 'cyan')
        pc("  DETAILED SUMMARY", 'cyan', True)
        pc(f"{'='*60}", 'cyan')
        print(f"\n{results['summary_detailed']}")
    elif '--tone' in args:
        pc(f"\n{'='*60}", 'cyan')
        pc("  TONE ANALYSIS", 'cyan', True)
        pc(f"{'='*60}", 'cyan')
        for k, v in results['tone_counts'].items():
            temp4 = 'green' if k == 'positive' else 'red' if k == 'negative' else 'yellow'
            pc(f"{k.capitalize()}: {v} sentences", temp4)
    elif '--positive' in args:
        pc(f"\n{'='*60}", 'cyan')
        pc("  POSITIVE SENTENCES", 'cyan', True)
        pc(f"{'='*60}", 'cyan')
        for tone, sent in results['tone_sentences']:
            if tone == 'positive':
                pc(f"✓ {sent}", 'green')
    elif '--negative' in args:
        pc(f"\n{'='*60}", 'cyan')
        pc("  NEGATIVE SENTENCES", 'cyan', True)
        pc(f"{'='*60}", 'cyan')
        for tone, sent in results['tone_sentences']:
            if tone == 'negative':
                pc(f"✗ {sent}", 'red')
    elif '--debate' in args:
        pc(f"\n{'='*60}", 'cyan')
        pc("  DEBATE ANALYSIS", 'cyan', True)
        pc(f"{'='*60}", 'cyan')
        temp5 = results['tone_counts']
        temp6 = temp5['positive'] + temp5['negative']
        if temp6 > 0:
            pc(f"\n📊 Balance: {round(temp5['positive']/temp6*100,1)}% Positive / {round(temp5['negative']/temp6*100,1)}% Negative", 'cyan', True)
            temp7 = [s for t, s in results['tone_sentences'] if t == 'positive']
            temp8 = [s for t, s in results['tone_sentences'] if t == 'negative']
            pc(f"\n✓ POSITIVE ARGUMENTS ({len(temp7)} total):", 'green', True)
            for i, sent in enumerate(temp7, 1):
                print(f"  {i}. {sent}")
            pc(f"\n✗ NEGATIVE ARGUMENTS ({len(temp8)} total):", 'red', True)
            for i, sent in enumerate(temp8, 1):
                print(f"  {i}. {sent}")
    else:
        pc(f"\n{'='*60}", 'cyan')
        pc("  DOCUMENT SUMMARY", 'cyan', True)
        pc(f"{'='*60}", 'cyan')
        print(f"\n{results['summary_small']}")
        pc("\n💡 Tip: Use --summary-detailed or --report for more", 'yellow')

if __name__ == '__main__':
    main()