Component Reference - Tokenizador

Component Overview

Tokenizador consists of four main components, each with specific responsibilities:

TokenAnalyzer

Application orchestrator that coordinates all other components and manages the application lifecycle

TokenizationService

Tokenization engine that integrates with tiktoken and provides fallback tokenization

UIController

UI manager that handles all DOM manipulation, user input, and visual updates

StatisticsCalculator

Computation engine that calculates metrics, costs, and context utilization

TokenAnalyzer

The TokenAnalyzer class is the main application orchestrator. It creates and coordinates all other components, handling the overall application lifecycle and user workflows.

Responsibilities

Initialize all services and controllers
Coordinate communication between components
Handle user actions (text change, model change, clear)
Manage error handling and warnings
Provide export functionality

Component Diagram

Key Methods

constructor()

Creates instances of all three core components and calls init().

constructor() {
    this.tokenizationService = new TokenizationService();
    this.uiController = new UIController();
    this.statisticsCalculator = new StatisticsCalculator();
    this.init();
}

init()

Initializes the application by setting up event handlers and waiting for tiktoken to load.

async init() {
    // Configure event handlers
    this.uiController.setEventHandlers({
        onTextChange: () => this.handleTextChange(),
        onModelChange: () => this.handleModelChange(),
        onClear: () => this.handleClear()
    });

    // Wait for tiktoken initialization
    await this.tokenizationService.waitForInitialization();

    // Trigger initial model change
    this.uiController.triggerModelChange();
}

performRealTimeAnalysis()

The core analysis workflow that tokenizes text, calculates statistics, and updates the UI.

async performRealTimeAnalysis() {
    const text = this.uiController.getTextInput().trim();
    const selectedModel = this.uiController.getSelectedModel();

    if (!text) {
        this.resetDisplays();
        return;
    }

    try {
        this.uiController.showLoading();

        // Tokenize text
        const tokenResult = await this.tokenizationService
            .tokenizeText(text, selectedModel);
        
        // Calculate statistics
        const statistics = this.statisticsCalculator
            .calculateStatistics(text, tokenResult, selectedModel);

        // Update UI
        this.updateDisplays(tokenResult, statistics);

        // Check for warnings
        this.showContextWarnings(statistics, selectedModel);
    } catch (error) {
        console.error('Error durante el análisis:', error);
        this.showError('Error al analizar el texto.');
    }
}

compareModels()

Compares tokenization across multiple models.

async compareModels(modelIds = ['gpt-4o', 'claude-3.5-sonnet', 'llama-3.1-70b']) {
    const text = this.uiController.getTextInput();
    
    if (!text.trim()) {
        throw new Error('No hay texto para analizar');
    }

    return await this.statisticsCalculator.compareModels(
        text, 
        modelIds, 
        this.tokenizationService
    );
}

exportResults()

Exports analysis results in JSON, CSV, or TXT format.

async exportResults(format = 'json') {
    const state = this.getState();
    const text = state.currentText;
    
    if (!text.trim()) {
        throw new Error('No hay datos para exportar');
    }

    const tokenResult = await this.tokenizationService
        .tokenizeText(text, state.selectedModel);
    const statistics = this.statisticsCalculator
        .calculateStatistics(text, tokenResult, state.selectedModel);

    const data = {
        timestamp: new Date().toISOString(),
        model: state.selectedModel,
        text: text,
        statistics: statistics,
        tokens: tokenResult.tokens
    };

    switch (format.toLowerCase()) {
        case 'json':
            return JSON.stringify(data, null, 2);
        case 'csv':
            return this.exportToCSV(data);
        case 'txt':
            return this.exportToText(data);
    }
}

Usage Example

// Application initialization (happens automatically on page load)
const app = new TokenAnalyzer();

// Get current application state
const state = app.getState();
console.log(state);
// {
//   currentText: "Hello world",
//   selectedModel: "gpt-4o",
//   isInitialized: true,
//   availableModels: ["gpt-4o", "claude-3.5-sonnet", ...]
// }

// Compare multiple models
const comparison = await app.compareModels([
    'gpt-4o', 
    'claude-3.5-sonnet', 
    'llama-3.1-70b'
]);

// Export results
const jsonData = await app.exportResults('json');
const csvData = await app.exportResults('csv');

TokenizationService

The TokenizationService handles all tokenization logic, including integration with the tiktoken library and providing fallback tokenization when tiktoken is unavailable.

Responsibilities

Initialize and manage the tiktoken encoder
Tokenize text using the appropriate encoding
Create token objects with type classification
Provide fallback tokenization when tiktoken fails
Apply model-specific token ratios

Component Diagram

Key Methods

initializeTokenizer()

Asynchronously loads and initializes the tiktoken library.

async initializeTokenizer() {
    try {
        // Wait for tiktoken to be available
        let attempts = 0;
        const maxAttempts = 20;
        
        while (attempts < maxAttempts) {
            if (typeof window !== 'undefined' && window.tiktokenLoaded) {
                break;
            }
            await new Promise(resolve => setTimeout(resolve, 500));
            attempts++;
        }
        
        // Initialize encoder
        let tiktokenLib = tiktoken || window.tiktoken;
        
        if (tiktokenLib && typeof tiktokenLib.get_encoding === 'function') {
            this.encoder = tiktokenLib.get_encoding('cl100k_base');
            this.isInitialized = true;
            this.isRealTiktoken = !window.tiktokenFallback;
            
            console.log(this.isRealTiktoken 
                ? 'Tiktoken REAL inicializado correctamente' 
                : 'Tiktoken RESPALDO inicializado correctamente');
        }
    } catch (error) {
        console.error('Error durante la inicialización:', error);
        this.isInitialized = false;
    }
}

tokenizeText()

The main tokenization method that processes text using tiktoken or fallback.

async tokenizeText(text, modelId) {
    if (!text.trim()) {
        return { tokens: [], count: 0 };
    }

    await this.waitForInitialization();

    const modelInfo = MODELS_DATA[modelId];
    let tokenCount = 0;
    let tokens = [];

    if (this.isInitialized && this.encoder && 
        modelInfo.encoding === 'cl100k_base') {
        try {
            // Use tiktoken for precise tokenization
            const encoded = this.encoder.encode(text);
            tokenCount = encoded.length;
            
            // Apply model-specific ratio
            if (modelInfo.tokenRatio) {
                tokenCount = Math.round(tokenCount * modelInfo.tokenRatio);
            }

            // Create token objects
            tokens = this.createTokensFromEncoding(text, encoded, modelId);
        } catch (error) {
            console.error('Error de tokenización tiktoken:', error);
            return this.fallbackTokenization(text, modelId);
        }
    } else {
        // Use fallback tokenization
        return this.fallbackTokenization(text, modelId);
    }

    return { tokens, count: tokenCount };
}

createTokensFromEncoding()

Creates visual token objects from tiktoken encoding with type classification.

createTokensFromEncoding(text, encoded, modelId) {
    const tokens = [];
    const isApproximate = !this.isRealTiktoken;
    
    for (let i = 0; i < encoded.length; i++) {
        const tokenId = encoded[i];
        const tokenText = this.encoder.decode([tokenId]);
        
        // Classify token type
        let type = 'palabra';
        if (/^\s+$/.test(tokenText)) {
            type = 'espacio_en_blanco';
        } else if (/^\d+$/.test(tokenText.trim())) {
            type = 'number';
        } else if (/^[.,!?;:'"()\[\]{}]+$/.test(tokenText.trim())) {
            type = 'punctuation';
        } else if (tokenText.startsWith(' ')) {
            type = 'palabra_con_espacio';
        } else if (tokenText.length <= 2 && !/^\s+$/.test(tokenText)) {
            type = 'subword';
        }

        tokens.push({
            text: tokenText,
            type: type,
            id: `token_${i}`,
            tokenId: tokenId,
            index: i,
            isApproximate: isApproximate
        });
    }

    return tokens;
}

Token Types:

palabra - Regular word token
espacio_en_blanco - Whitespace
number - Numeric token
punctuation - Punctuation mark
palabra_con_espacio - Word with leading space
subword - Subword/partial word token

fallbackTokenization()

Provides approximate tokenization when tiktoken is unavailable.

fallbackTokenization(text, modelId) {
    console.log('⚠️ Usando tokenización de respaldo');
    
    const tokens = [];
    let tokenIndex = 0;
    
    // Split into segments
    const segments = text.match(/\S+|\s+/g) || [];
    
    for (const segment of segments) {
        if (/\s/.test(segment)) {
            // Whitespace token
            tokens.push({
                text: segment,
                type: 'espacio_en_blanco',
                id: `token_${tokenIndex}`,
                tokenId: this.createDeterministicId(segment, tokenIndex),
                index: tokenIndex,
                isApproximate: true
            });
            tokenIndex++;
        } else {
            // Split word into tokens
            const wordTokens = this.splitWordIntoTokens(segment, tokenIndex);
            tokens.push(...wordTokens);
            tokenIndex += wordTokens.length;
        }
    }

    return { tokens, count: tokens.length };
}

Token Structure

Each token object returned by the service has this structure:

interface Token {
  text: string;           // The actual text of the token
  type: string;           // Token classification (palabra, number, etc.)
  id: string;             // Unique identifier for UI rendering
  tokenId: number;        // Numeric token ID from tiktoken
  index: number;          // Position in token array
  isApproximate: boolean; // Whether this is a real or approximate token
}

Usage Example

const service = new TokenizationService();

// Wait for initialization
await service.waitForInitialization();

// Tokenize text
const result = await service.tokenizeText("Hello world", "gpt-4o");
console.log(result);
// {
//   tokens: [
//     { text: "Hello", type: "palabra", id: "token_0", tokenId: 9906, ... },
//     { text: " world", type: "palabra_con_espacio", id: "token_1", tokenId: 1917, ... }
//   ],
//   count: 2
// }

// Get tokenizer name
const tokenizerName = service.getTokenizerName('cl100k_base');
console.log(tokenizerName); // "Tokenizador GPT-4"

// Get algorithm name for a model
const algorithm = service.getAlgorithmName('claude-3.5-sonnet');
console.log(algorithm); // "Tokenización Claude (~20% más tokens)"

UIController

The UIController manages all user interface interactions, DOM manipulation, and visual updates. It serves as the presentation layer of the application.

Responsibilities

Initialize and cache DOM element references
Handle user input events (keyboard, mouse)
Update statistics displays
Render token visualizations
Manage UI state (loading, empty, error)
Provide visual feedback (highlighting, scrolling)

Component Diagram

Key Methods

initializeElements()

Caches all DOM element references for performance.

initializeElements() {
    return {
        textInput: document.getElementById('text-input'),
        modelSelect: document.getElementById('model-select'),
        clearBtn: document.getElementById('clear-btn'),
        tokensContainer: document.getElementById('tokens-container'),
        tokensArray: document.getElementById('tokens-array'),
        modelDetailsLink: document.getElementById('model-details-link'),
        
        // Statistics elements
        tokenCount: document.getElementById('token-count'),
        charCount: document.getElementById('char-count'),
        wordCount: document.getElementById('word-count'),
        costEstimate: document.getElementById('cost-estimate'),
        
        // Model info elements
        selectedModelName: document.getElementById('selected-model-name'),
        contextLimit: document.getElementById('context-limit'),
        tokenizationType: document.getElementById('tokenization-type'),
        costPer1k: document.getElementById('cost-per-1k'),
        activeAlgorithm: document.getElementById('active-algorithm')
    };
}

Caching element references improves performance by avoiding repeated DOM queries.

bindEvents()

Attaches event listeners to DOM elements.

bindEvents() {
    // Real-time analysis on input
    if (this.elements.textInput) {
        this.elements.textInput.addEventListener('input', () => {
            this.onTextChange?.();
        });
    }

    // Model selection change
    if (this.elements.modelSelect) {
        this.elements.modelSelect.addEventListener('change', () => {
            this.onModelChange?.();
        });
    }

    // Clear button
    if (this.elements.clearBtn) {
        this.elements.clearBtn.addEventListener('click', () => {
            this.onClear?.();
        });
    }

    // Keyboard shortcuts
    document.addEventListener('keydown', (e) => {
        if (e.ctrlKey && e.key === 'k') {
            e.preventDefault();
            this.elements.textInput?.focus();
        }
    });
}

updateStatistics()

Updates all statistics displays with formatted values.

updateStatistics(stats = {}) {
    const updates = {
        tokenCount: (stats.tokenCount || 0).toLocaleString(),
        charCount: (stats.charCount || 0).toLocaleString(),
        wordCount: (stats.wordCount || 0).toLocaleString(),
        costEstimate: `$${(stats.costEstimate || 0).toFixed(6)}`
    };

    Object.entries(updates).forEach(([key, value]) => {
        if (this.elements[key]) {
            this.elements[key].textContent = value;
        }
    });
}

updateTokenVisualization()

Renders color-coded token spans in the main visualization area.

updateTokenVisualization(tokens = []) {
    if (!this.elements.tokensContainer) return;

    if (!tokens.length) {
        this.elements.tokensContainer.innerHTML = 
            '<p style="color: var(--text-secondary); font-style: italic;">'
            + 'Escribe algo para ver los tokens...</p>';
        return;
    }

    this.elements.tokensContainer.innerHTML = '';
    tokens.forEach(token => {
        const tokenElement = document.createElement('span');
        tokenElement.className = `token ${token.type}`;
        tokenElement.textContent = token.text;
        
        if (token.isApproximate) {
            tokenElement.classList.add('approximate-id');
        }
        
        tokenElement.addEventListener('click', () => {
            this.highlightTokenInList(token.id);
        });

        this.elements.tokensContainer.appendChild(tokenElement);
    });
}

updateTokensList()

Renders the detailed token list with IDs and types.

updateTokensList(tokens = []) {
    if (!this.elements.tokensArray) return;

    if (!tokens.length) {
        this.elements.tokensArray.innerHTML = 
            '<p style="color: var(--text-secondary); font-style: italic;">'
            + 'Los tokens aparecerán aquí...</p>';
        return;
    }

    // Check for approximate tokens
    const hasApproximateTokens = tokens.some(token => token.isApproximate);
    
    // Add precision info header
    let headerHtml = '';
    if (hasApproximateTokens) {
        headerHtml = `
            <div class="tokens-precision-info approximate">
                <i class="fas fa-exclamation-triangle"></i>
                <span>IDs Aproximados - tiktoken no disponible</span>
            </div>
        `;
    } else {
        headerHtml = `
            <div class="tokens-precision-info accurate">
                <i class="fas fa-check-circle"></i>
                <span>IDs Reales de tiktoken</span>
            </div>
        `;
    }

    this.elements.tokensArray.innerHTML = headerHtml;
    
    tokens.forEach((token, index) => {
        const tokenItem = document.createElement('div');
        tokenItem.className = 'token-item';
        tokenItem.id = `token-item-${token.id}`;
        
        const idType = token.isApproximate ? '≈' : '=';
        const idDescription = token.isApproximate ? 'Aproximado' : 'Tiktoken Real';
        
        if (token.isApproximate) {
            tokenItem.classList.add('approximate-token');
        }
        
        tokenItem.innerHTML = `
            <span class="token-text">"${token.text}"</span>
            <span class="token-id">
                #${index + 1} (${token.type})
                → ID ${idType} ${token.tokenId} (${idDescription})
            </span>
        `;

        this.elements.tokensArray.appendChild(tokenItem);
    });
}

highlightTokenInList()

Highlights a specific token in the list and scrolls it into view.

highlightTokenInList(tokenId) {
    // Remove previous highlights
    document.querySelectorAll('.token-item').forEach(item => {
        item.style.background = '';
        item.style.color = '';
    });

    // Highlight selected token
    const tokenItem = document.getElementById(`token-item-${tokenId}`);
    if (tokenItem) {
        tokenItem.style.background = 'var(--primary-color)';
        tokenItem.style.color = 'white';
        tokenItem.scrollIntoView({ behavior: 'smooth', block: 'center' });
        
        // Remove highlight after 2 seconds
        setTimeout(() => {
            tokenItem.style.background = '';
            tokenItem.style.color = '';
        }, 2000);
    }
}

Event Handler Pattern

The UIController uses a callback pattern for loose coupling:

// Set event handlers
uiController.setEventHandlers({
    onTextChange: () => console.log('Text changed'),
    onModelChange: () => console.log('Model changed'),
    onClear: () => console.log('Clear clicked')
});

// Internally, the controller calls these when events occur
this.elements.textInput.addEventListener('input', () => {
    this.onTextChange?.(); // Safe call using optional chaining
});

Usage Example

const uiController = new UIController();

// Set event handlers
uiController.setEventHandlers({
    onTextChange: async () => {
        const text = uiController.getTextInput();
        console.log('User typed:', text);
    },
    onModelChange: () => {
        const model = uiController.getSelectedModel();
        console.log('Model changed to:', model);
    },
    onClear: () => {
        console.log('Clear button clicked');
    }
});

// Get user input
const text = uiController.getTextInput();
const model = uiController.getSelectedModel();

// Update displays
uiController.updateStatistics({
    tokenCount: 42,
    charCount: 100,
    wordCount: 15,
    costEstimate: 0.000105
});

// Show loading state
uiController.showLoading();

// Clear input
uiController.clearTextInput();

StatisticsCalculator

The StatisticsCalculator performs all statistical computations including token counts, costs, context utilization, and efficiency metrics.

Responsibilities

Calculate comprehensive text statistics
Estimate API costs based on token counts
Compute context window utilization
Generate context limit warnings
Compare statistics across models
Calculate efficiency metrics

Component Diagram

Key Methods

calculateStatistics()

Computes all statistics for a given text and model.

calculateStatistics(text, tokenResult, modelId) {
    const modelInfo = MODELS_DATA[modelId];
    
    if (!modelInfo) {
        throw new Error(`Modelo desconocido: ${modelId}`);
    }

    const tokenCount = tokenResult.count || 0;
    const charCount = text.length;
    const wordCount = this.countWords(text);
    const costEstimate = this.calculateCost(tokenCount, modelInfo);

    return {
        tokenCount,
        charCount,
        wordCount,
        costEstimate,
        contextUtilization: this.calculateContextUtilization(
            tokenCount, 
            modelInfo.contextLimit
        ),
        tokensPerWord: wordCount > 0 ? (tokenCount / wordCount) : 0,
        inputCostPer1M: modelInfo.inputCost,
        outputCostPer1M: modelInfo.outputCost
    };
}

Returns:

{
  tokenCount: number;         // Total tokens
  charCount: number;          // Total characters
  wordCount: number;          // Total words
  costEstimate: number;       // Estimated cost in USD
  contextUtilization: number; // Percentage (0-100)
  tokensPerWord: number;      // Average ratio
  inputCostPer1M: number;     // Cost per 1M input tokens
  outputCostPer1M: number;    // Cost per 1M output tokens
}

calculateCost()

Calculates the estimated API cost based on token count.

calculateCost(tokenCount, modelInfo) {
    if (!modelInfo || !modelInfo.inputCost) {
        return 0;
    }
    // Calculate cost based on input tokens (cost per 1M tokens)
    return (tokenCount / 1000000) * modelInfo.inputCost;
}

Cost calculation uses input pricing since the text being analyzed represents input to the model. Output costs are shown separately for reference.

calculateContextUtilization()

Computes what percentage of the model’s context window is being used.

calculateContextUtilization(tokenCount, contextLimit) {
    return contextLimit > 0 
        ? Math.min((tokenCount / contextLimit) * 100, 100) 
        : 0;
}

getContextWarning()

Returns appropriate warning messages based on context utilization.

getContextWarning(tokenCount, modelId) {
    const modelInfo = MODELS_DATA[modelId];
    
    if (!modelInfo) return null;

    const utilization = this.calculateContextUtilization(
        tokenCount, 
        modelInfo.contextLimit
    );
    
    if (utilization >= 100) {
        return `⚠️ Texto excede el límite de contexto del modelo `
            + `(${modelInfo.contextLimit.toLocaleString()} tokens)`;
    } else if (utilization >= 90) {
        return `⚠️ Cerca del límite de contexto `
            + `(${utilization.toFixed(1)}% utilizado)`;
    } else if (utilization >= 75) {
        return `ℹ️ Alto uso del contexto `
            + `(${utilization.toFixed(1)}% utilizado)`;
    }
    
    return null;
}

Warning Thresholds:

100%+: Critical - exceeds context limit
90-99%: Warning - approaching limit
75-89%: Info - high utilization
Under 75%: No warning

compareModels()

Compares statistics across multiple models for the same text.

async compareModels(text, modelIds, tokenizationService) {
    const comparisons = [];

    for (const modelId of modelIds) {
        try {
            const tokenResult = await tokenizationService
                .tokenizeText(text, modelId);
            const stats = this.calculateStatistics(
                text, 
                tokenResult, 
                modelId
            );
            
            comparisons.push({
                modelId,
                company: MODELS_DATA[modelId]?.company || 'Desconocido',
                stats,
                formatted: this.formatStatistics(stats)
            });
        } catch (error) {
            console.warn(`Error comparando modelo ${modelId}:`, error);
        }
    }

    // Sort by cost (cheapest first)
    return comparisons.sort((a, b) => 
        a.stats.costEstimate - b.stats.costEstimate
    );
}

getEfficiencyMetrics()

Calculates efficiency metrics for deeper analysis.

getEfficiencyMetrics(stats) {
    return {
        // Cost per thousand tokens
        costEfficiency: stats.tokenCount > 0 
            ? (stats.costEstimate / stats.tokenCount * 1000) 
            : 0,
        
        // Tokens per character (compression ratio)
        compressionRatio: stats.charCount > 0 
            ? (stats.tokenCount / stats.charCount) 
            : 0,
        
        // Tokens per word (verbosity index)
        verbosityIndex: stats.wordCount > 0 
            ? (stats.tokenCount / stats.wordCount) 
            : 0
    };
}

Statistics Object Structure

A complete statistics object contains:

interface Statistics {
  // Basic counts
  tokenCount: number;         // Total tokens in text
  charCount: number;          // Total characters
  wordCount: number;          // Total words
  
  // Cost analysis
  costEstimate: number;       // Estimated cost in USD
  inputCostPer1M: number;     // Input cost per 1M tokens
  outputCostPer1M: number;    // Output cost per 1M tokens
  
  // Context analysis
  contextUtilization: number; // Percentage of context used (0-100)
  tokensPerWord: number;      // Average tokens per word
}

Usage Example

const calculator = new StatisticsCalculator();

// Calculate statistics
const stats = calculator.calculateStatistics(
    "Hello world",
    { tokens: [...], count: 2 },
    "gpt-4o"
);

console.log(stats);
// {
//   tokenCount: 2,
//   charCount: 11,
//   wordCount: 2,
//   costEstimate: 0.000005,
//   contextUtilization: 0.0016,
//   tokensPerWord: 1.0,
//   inputCostPer1M: 2.50,
//   outputCostPer1M: 10.00
// }

// Check for warnings
const warning = calculator.getContextWarning(150000, "gpt-4o");
if (warning) {
    console.log(warning); // "⚠️ Cerca del límite de contexto (117.2% utilizado)"
}

// Get efficiency metrics
const efficiency = calculator.getEfficiencyMetrics(stats);
console.log(efficiency);
// {
//   costEfficiency: 0.0025,
//   compressionRatio: 0.182,
//   verbosityIndex: 1.0
// }

// Compare models
const comparison = await calculator.compareModels(
    "Hello world",
    ["gpt-4o", "claude-3.5-sonnet", "llama-3.1-70b"],
    tokenizationService
);

console.log(comparison);
// [
//   { modelId: "llama-3.1-70b", company: "Meta", stats: {...}, ... },
//   { modelId: "gpt-4o", company: "OpenAI", stats: {...}, ... },
//   { modelId: "claude-3.5-sonnet", company: "Anthropic", stats: {...}, ... }
// ]

Component Interactions

Here’s a comprehensive example showing how all components work together:

// 1. Application initialization
class TokenAnalyzer {
    constructor() {
        // Create all components
        this.tokenizationService = new TokenizationService();
        this.uiController = new UIController();
        this.statisticsCalculator = new StatisticsCalculator();
        this.init();
    }

    async init() {
        // 2. Set up event handlers
        this.uiController.setEventHandlers({
            onTextChange: () => this.handleTextChange(),
            onModelChange: () => this.handleModelChange(),
            onClear: () => this.handleClear()
        });

        // 3. Wait for services to initialize
        await this.tokenizationService.waitForInitialization();

        // 4. Trigger initial UI update
        this.uiController.triggerModelChange();
    }

    // 5. Handle user typing
    async handleTextChange() {
        await this.performRealTimeAnalysis();
    }

    // 6. Perform analysis
    async performRealTimeAnalysis() {
        // Get input from UI
        const text = this.uiController.getTextInput();
        const model = this.uiController.getSelectedModel();

        if (!text) return;

        // Tokenize via service
        const tokenResult = await this.tokenizationService
            .tokenizeText(text, model);
        
        // Calculate stats
        const statistics = this.statisticsCalculator
            .calculateStatistics(text, tokenResult, model);

        // Update UI
        this.uiController.updateStatistics(statistics);
        this.uiController.updateTokenVisualization(tokenResult.tokens);
        this.uiController.updateTokensList(tokenResult.tokens);

        // Check warnings
        const warning = this.statisticsCalculator
            .getContextWarning(statistics.tokenCount, model);
        if (warning) {
            console.warn(warning);
        }
    }
}

// 7. Initialize application
const app = new TokenAnalyzer();

Summary

TokenAnalyzer

Role: Application OrchestratorKey Features:

Coordinates all components
Handles user workflows
Manages application lifecycle
Provides export functionality

TokenizationService

Role: Tokenization EngineKey Features:

Integrates with tiktoken
Classifies token types
Provides fallback tokenization
Applies model-specific ratios

UIController

Role: Presentation ManagerKey Features:

Manages DOM interactions
Handles user input
Updates visualizations
Provides visual feedback

StatisticsCalculator

Role: Computation EngineKey Features:

Calculates comprehensive stats
Estimates costs
Compares models
Generates warnings

Next Steps

Architecture Overview

Return to the architecture overview

API Reference

Explore the complete API documentation

How to Use

Learn how to use the application

Contributing

Contribute to Tokenizador

Get Started

Guides

Architecture

​Component Overview

TokenAnalyzer

TokenizationService

UIController

StatisticsCalculator

​TokenAnalyzer

​Responsibilities

​Component Diagram

​Key Methods

​Usage Example

​TokenizationService

​Responsibilities

​Component Diagram

​Key Methods

​Token Structure

​Usage Example

​UIController

​Responsibilities

​Component Diagram

​Key Methods

​Event Handler Pattern

​Usage Example

​StatisticsCalculator

​Responsibilities

​Component Diagram

​Key Methods

​Statistics Object Structure

​Usage Example

​Component Interactions

​Summary

TokenAnalyzer

TokenizationService

UIController

StatisticsCalculator

​Next Steps

Architecture Overview

API Reference

How to Use

Contributing

Build docs developers (and LLMs) love

Component Overview

TokenAnalyzer

Responsibilities

Component Diagram

Key Methods

Usage Example

TokenizationService

Responsibilities

Component Diagram

Key Methods

Token Structure

Usage Example

UIController

Responsibilities

Component Diagram

Key Methods

Event Handler Pattern

Usage Example

StatisticsCalculator

Responsibilities

Component Diagram

Key Methods

Statistics Object Structure

Usage Example

Component Interactions

Summary

Next Steps