web/lib/constants.ts

// Augmentation feature definitions
export const augmentationFeatures = [
    { id: 'rotation', name: 'Rotation/Tilt', min: 0, max: 30, default: 5 },
    { id: 'skew', name: 'Skew/Perspective', min: 0, max: 20, default: 3 },
    { id: 'gaussian_blur', name: 'Gaussian Blur', min: 0, max: 5, default: 1 },
    { id: 'motion_blur', name: 'Motion Blur', min: 0, max: 10, default: 3 },
    { id: 'gaussian_noise', name: 'Gaussian Noise', min: 0, max: 50, default: 10 },
    { id: 'salt_pepper', name: 'Salt & Pepper', min: 0, max: 10, default: 1 },
    { id: 'brightness', name: 'Brightness Var.', min: 0, max: 50, default: 15 },
    { id: 'contrast', name: 'Contrast Var.', min: 0, max: 100, default: 20 },
    { id: 'jpeg_quality', name: 'JPEG Artifacts', min: 0, max: 100, default: 70 },
    { id: 'resolution', name: 'Resolution Loss', min: 0, max: 80, default: 30 },
    { id: 'paper_texture', name: 'Paper Texture', min: 0, max: 100, default: 50 },
    { id: 'shadow', name: 'Shadow/Lighting', min: 0, max: 100, default: 30 },
    { id: 'ink_bleed', name: 'Ink Bleed', min: 0, max: 100, default: 20 },
]

export interface TextFileInfo {
    name: string
    wordCount: number
    lineCount: number
    charCount: number
    maxSamples: {
        character: number
        word: number
        line: number
        sentence: number
        ngram: number
    }
    parsedData: {
        characters: string[]
        words: string[]
        lines: string[]
        sentences: string[]
        ngrams: string[]
    }
}

export interface AppConfig {
    dataset: {
        size: number
        seed: number
    }
    input: {
        segmentation: string
        file?: string
    }
    image: {
        width: number
        height: number
        background: string
        backgroundStyle: string
        backgroundMode: 'single' | 'mix'
        backgroundPercentages: Record<string, number>
        customBackgrounds: { name: string; dataUrl: string; percentage: number }[]
        textColor: string
        direction: string
    }
    fonts: {
        distribution: { family: string; percentage: number; name?: string; dataUrl?: string }[]
    }
    augmentation: {
        enabled: boolean
        applyPercentage: number
        preset: string
        customMode: boolean
        values: Record<string, number>
    }
    output: {
        formats: string[]
    }
}

export function getDefaultConfig(): AppConfig {
    return {
        dataset: {
            size: 100,
            seed: 42,
        },
        input: {
            segmentation: 'word',
        },
        image: {
            width: 256,
            height: 64,
            background: '#FFFFFF',
            backgroundStyle: 'clean_white',
            backgroundMode: 'single',
            backgroundPercentages: {
                clean_white: 100, aged_paper: 0, book_page: 0, newspaper: 0,
                notebook: 0, parchment: 0, weathered: 0, coffee_stain: 0,
                old_book: 0, recycled: 0, cream: 0, ivory: 0

            },
            customBackgrounds: [],
            textColor: '#000000',
            direction: 'rtl',
        },
        fonts: {
            distribution: [],
        },
        augmentation: {
            enabled: true,
            applyPercentage: 70,
            preset: 'balanced',
            customMode: false,
            values: Object.fromEntries(augmentationFeatures.map(f => [f.id, f.default])),
        },
        output: {
            formats: ['crnn', 'trocr'],
        },
    }
}