src/tools/sentiment_analyzer.py

"""Sentiment analysis tool using word dictionaries."""

import re
from typing import Dict, Any

from .base_tool import BaseTool


class SentimentAnalyzer(BaseTool):
    """Analyzes sentiment using positive and negative word dictionaries."""

    def __init__(self):
        super().__init__()
        # Positive words
        self.positive_words = {
            'good', 'great', 'excellent', 'wonderful', 'fantastic', 'amazing',
            'love', 'like', 'best', 'happy', 'joy', 'pleased', 'perfect',
            'beautiful', 'brilliant', 'awesome', 'outstanding', 'superb',
            'delightful', 'pleasant', 'positive', 'successful', 'benefit',
            'advantage', 'improve', 'better', 'enjoy', 'excited', 'glad'
        }

        # Negative words
        self.negative_words = {
            'bad', 'terrible', 'horrible', 'awful', 'poor', 'worst', 'hate',
            'dislike', 'sad', 'unhappy', 'disappointed', 'disappointing',
            'ugly', 'disgusting', 'negative', 'problem', 'issue', 'fail',
            'failure', 'difficult', 'hard', 'wrong', 'error', 'unfortunately',
            'worse', 'boring', 'annoying', 'frustrating', 'concerned', 'worry'
        }

    @property
    def description(self) -> str:
        return (
            "Analyzes the sentiment (positive/negative/neutral) of the text. "
            "Returns a sentiment score between -1 (very negative) and 1 (very positive), "
            "along with a sentiment label. Use this when you need to understand "
            "the emotional tone or opinion expressed in the text."
        )

    def run(self, text: str) -> Dict[str, Any]:
        """Analyze sentiment of the text.

        Args:
            text: Input text to analyze

        Returns:
            Dictionary with sentiment score and label
        """
        # Tokenize
        words = re.findall(r'\b[a-zA-Z]+\b', text.lower())

        if not words:
            return {
                "sentiment_score": 0.0,
                "sentiment_label": "neutral",
                "positive_words_found": [],
                "negative_words_found": []
            }

        # Count positive and negative words
        positive_found = [w for w in words if w in self.positive_words]
        negative_found = [w for w in words if w in self.negative_words]

        pos_count = len(positive_found)
        neg_count = len(negative_found)

        # Calculate sentiment score
        total_sentiment_words = pos_count + neg_count

        if total_sentiment_words == 0:
            score = 0.0
            label = "neutral"
        else:
            # Score ranges from -1 to 1
            score = (pos_count - neg_count) / len(words)

            # Determine label
            if score > 0.02:
                label = "positive"
            elif score < -0.02:
                label = "negative"
            else:
                label = "neutral"

        return {
            "sentiment_score": round(score, 4),
            "sentiment_label": label,
            "positive_words_count": pos_count,
            "negative_words_count": neg_count,
            "positive_words_found": list(set(positive_found))[:5],
            "negative_words_found": list(set(negative_found))[:5]
        }