pdf/engine/include/pdfengine/document/raw_ocr_document.hpp

#pragma once

#include <string>
#include <vector>
#include <cstdint>

namespace pdfengine::document {

struct RawOCRWord {
    std::string text;
    double x = 0.0;
    double y = 0.0;
    double width = 0.0;
    double height = 0.0;
    double confidence = 0.0;
    std::vector<std::pair<double, double>> polygon;
};

struct RawOCRLine {
    std::string text;
    double x = 0.0;
    double y = 0.0;
    double width = 0.0;
    double height = 0.0;
    double baselineY = 0.0;
    double confidence = 0.0;
    std::vector<RawOCRWord> words;
    std::string fontName;
    std::string fontId;
    std::string fontFace;
    int fontWeight = 400;
    std::string fontStyle = "normal";
    double fontSize = 0.0;
    double lineSpacing = 1.2;
    double letterSpacing = 0.0;
    bool isBold = false;
    bool isItalic = false;
    bool isEmbeddedFont = false;
    bool isPredictedFont = true;
};

struct RawOCRPage {
    int pageIndex = 0;
    double imageWidth = 0.0;
    double imageHeight = 0.0;
    double pageWidth = 0.0;
    double pageHeight = 0.0;
    std::vector<RawOCRLine> lines;
    double processTimeMs = 0.0;
};

} // namespace pdfengine::document