app.py

import gradio as gr
import re
from collections import Counter
from datetime import datetime
import emoji
from transformers import pipeline
import logging
from typing import Tuple, List, Optional

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

class CommentAnalyzer:
    def __init__(self):
        """Initialize the analyzer with sentiment model and compile regex patterns"""
        try:
            self.sentiment_model = pipeline("sentiment-analysis")
        except Exception as e:
            logger.error(f"Failed to load sentiment model: {e}")
            raise

        # Compile regex patterns for better performance
        self.mention_pattern = re.compile(r'@[\w\.]+')
        self.comment_pattern = re.compile(
            r'Фото профиля\s+(.+?)\s+'  # Username
            r'((?:(?!Фото профиля).)+?)\s+'  # Comment text
            r'(\d+)?\s*(?:нравится|like[s]?)?\s*'  # Likes count
            r'(\d+)\s*(?:н|w)'  # Week number
            , re.DOTALL
        )

    def clean_text(self, text: str) -> str:
        """Clean text by removing extra whitespace and normalizing line breaks"""
        return ' '.join(text.split())

    def count_emojis(self, text: str) -> int:
        """Count the number of emoji characters in text"""
        return len([c for c in text if c in emoji.EMOJI_DATA])

    def extract_mentions(self, text: str) -> List[str]:
        """Extract @mentions from text"""
        return self.mention_pattern.findall(text)

    def analyze_sentiment(self, text: str) -> str:
        """Analyze text sentiment using the loaded model"""
        try:
            result = self.sentiment_model(text[:512])  # Limit text length for model
            sentiment = result[0]['label']
            if sentiment == 'POSITIVE':
                return 'positive'
            elif sentiment == 'NEGATIVE':
                return 'negative'
            return 'neutral'
        except Exception as e:
            logger.warning(f"Sentiment analysis failed: {e}")
            return 'neutral'

    def extract_comment_data(self, comment_block: str) -> Tuple[Optional[str], Optional[str], int, int]:
        """
        Extract structured data from a comment block
        Returns: (username, comment_text, likes_count, week_number)
        """
        match = self.comment_pattern.search(comment_block)
        if not match:
            return None, None, 0, 0
            
        username, comment, likes, week = match.groups()
        return (
            username.strip(),
            self.clean_text(comment),
            int(likes or 0),
            int(week or 0)
        )

    def analyze_post(self, 
                    content_type: str,
                    link_to_post: str,
                    post_likes: int,
                    post_date: str,
                    description: str,
                    comment_count: int,
                    all_comments: str) -> Tuple[str, str, str, str, str]:
        """
        Analyze Instagram post comments and generate comprehensive analytics
        
        Args:
            content_type: Type of content ("Photo" or "Video")
            link_to_post: URL of the post
            post_likes: Number of likes on the post
            post_date: Date of post publication
            description: Post description/caption
            comment_count: Total number of comments
            all_comments: Raw text containing all comments
            
        Returns:
            Tuple containing:
            - Analytics summary
            - List of usernames
            - List of comments
            - Chronological list of likes
            - Total likes count
        """
        try:
            # Split comments into blocks
            comments_blocks = [block for block in re.split(r'(?=Фото профиля)', all_comments) if block.strip()]
            
            # Initialize data containers
            data = {
                'usernames': [],
                'comments': [],
                'likes': [],
                'weeks': [],
                'emojis': 0,
                'mentions': [],
                'sentiments': [],
                'lengths': []
            }
            
            # Process each comment block
            for block in comments_blocks:
                username, comment, like_count, week = self.extract_comment_data(block)
                if username and comment:
                    data['usernames'].append(username)
                    data['comments'].append(comment)
                    data['likes'].append(like_count)
                    data['weeks'].append(week)
                    
                    # Collect metrics
                    data['emojis'] += self.count_emojis(comment)
                    data['mentions'].extend(self.extract_mentions(comment))
                    data['sentiments'].append(self.analyze_sentiment(comment))
                    data['lengths'].append(len(comment))
            
            # Calculate analytics
            total_comments = len(data['comments'])
            if total_comments == 0:
                raise ValueError("No valid comments found in input")
                
            analytics = {
                'avg_length': sum(data['lengths']) / total_comments,
                'sentiment_dist': Counter(data['sentiments']),
                'active_users': Counter(data['usernames']).most_common(5),
                'top_mentions': Counter(data['mentions']).most_common(5),
                'avg_likes': sum(data['likes']) / total_comments,
                'weeks_range': (min(data['weeks']), max(data['weeks'])),
                'total_likes': sum(data['likes'])
            }
            
            # Generate summary
            summary = self._format_analytics_summary(
                content_type, link_to_post, data, analytics, total_comments
            )
            
            return (
                summary,
                '\n'.join(data['usernames']),
                '\n'.join(data['comments']),
                '\n'.join(map(str, data['likes'])),
                str(analytics['total_likes'])
            )
            
        except Exception as e:
            logger.error(f"Error analyzing post: {e}", exc_info=True)
            return (f"Error during analysis: {str(e)}", "", "", "", "0")

    def _format_analytics_summary(self, content_type, link, data, analytics, total_comments):
        """Format analytics data into a readable summary"""
        return f"""
Content Type: {content_type}
Link to Post: {link}

ОСНОВНАЯ СТАТИСТИКА:
- Всего комментариев: {total_comments}
- Всего лайков на комментариях: {analytics['total_likes']}
- Среднее количество лайков: {analytics['avg_likes']:.1f}
- Период активности: {analytics['weeks_range'][0]}-{analytics['weeks_range'][1]} недель

АНАЛИЗ КОНТЕНТА:
- Средняя длина комментария: {analytics['avg_length']:.1f} символов
- Всего эмодзи использовано: {data['emojis']}
- Тональность комментариев:
  * Позитивных: {analytics['sentiment_dist']['positive']}
  * Нейтральных: {analytics['sentiment_dist']['neutral']}
  * Негативных: {analytics['sentiment_dist']['negative']}

АКТИВНОСТЬ ПОЛЬЗОВАТЕЛЕЙ:
Самые активные комментаторы:
{chr(10).join(f"- {user}: {count} комментариев" for user, count in analytics['active_users'])}

Самые упоминаемые пользователи:
{chr(10).join(f"- {user}: {count} упоминаний" for user, count in analytics['top_mentions'] if user)}

ВОВЛЕЧЕННОСТЬ:
- Процент комментариев с лайками: {(sum(1 for l in data['likes'] if l > 0) / total_comments * 100):.1f}%
- Процент комментариев с эмодзи: {(sum(1 for c in data['comments'] if self.count_emojis(c) > 0) / total_comments * 100):.1f}%
"""

def create_interface():
    """Create and configure the Gradio interface"""
    analyzer = CommentAnalyzer()
    
    iface = gr.Interface(
        fn=analyzer.analyze_post,
        inputs=[
            gr.Radio(
                choices=["Photo", "Video"],
                label="Content Type",
                value="Photo"
            ),
            gr.Textbox(
                label="Link to Post",
                placeholder="Введите ссылку на пост"
            ),
            gr.Number(
                label="Likes",
                value=0
            ),
            gr.Textbox(
                label="Post Date",
                placeholder="Введите дату публикации"
            ),
            gr.Textbox(
                label="Description",
                placeholder="Введите описание поста",
                lines=3
            ),
            gr.Number(
                label="Total Comment Count",
                value=0
            ),
            gr.Textbox(
                label="All Comments",
                placeholder="Вставьте комментарии",
                lines=10
            )
        ],
        outputs=[
            gr.Textbox(label="Analytics Summary", lines=20),
            gr.Textbox(label="Usernames (Output 1)", lines=5),
            gr.Textbox(label="Comments (Output 2)", lines=5),
            gr.Textbox(label="Likes Chronology (Output 3)", lines=5),
            gr.Textbox(label="Total Likes on Comments (Output 4)")
        ],
        title="Instagram Comment Analyzer Pro",
        description="Расширенный анализатор комментариев Instagram с детальной аналитикой"
    )
    return iface

if __name__ == "__main__":
    iface = create_interface()
    iface.launch()