pdfium_searchify.cc | Explore in Territory

// Copyright 2024 The Chromium Authors
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.

#include "pdf/pdfium/pdfium_searchify.h"

#include <math.h>
#include <stdint.h>

#include <algorithm>
#include <array>
#include <string>
#include <utility>
#include <vector>

#include "base/check.h"
#include "base/check_op.h"
#include "base/compiler_specific.h"
#include "base/containers/span.h"
#include "base/functional/callback.h"
#include "base/numerics/angle_conversions.h"
#include "base/strings/utf_string_conversions.h"
#include "pdf/pdfium/pdfium_engine.h"
#include "pdf/pdfium/pdfium_mem_buffer_file_write.h"
#include "pdf/pdfium/pdfium_ocr.h"
#include "pdf/pdfium/pdfium_searchify_font.h"
#include "services/screen_ai/public/mojom/screen_ai_service.mojom.h"
#include "third_party/pdfium/public/cpp/fpdf_scopers.h"
#include "third_party/pdfium/public/fpdf_edit.h"
#include "third_party/pdfium/public/fpdf_save.h"
#include "third_party/pdfium/public/fpdfview.h"
#include "third_party/skia/include/core/SkBitmap.h"
#include "third_party/skia/include/core/SkImageInfo.h"
#include "third_party/skia/include/core/SkPixmap.h"
#include "ui/gfx/codec/jpeg_codec.h"
#include "ui/gfx/geometry/point_f.h"
#include "ui/gfx/geometry/rect.h"
#include "ui/gfx/geometry/size.h"
#include "ui/gfx/geometry/size_f.h"

namespace chrome_pdf {

namespace {

std::vector<uint32_t> Utf8ToCharcodes(const std::string& string) { … }

// The coordinate systems between OCR and PDF are different. OCR's origin is at
// top-left, so we need to convert them to PDF's bottom-left.
SearchifyBoundingBoxOrigin ConvertToPdfOrigin(const gfx::Rect& rect,
                                              float angle,
                                              float coordinate_system_height) { … }

// Project the text object's origin to the baseline's origin.
SearchifyBoundingBoxOrigin ProjectToBaseline(
    const gfx::PointF& origin_point,
    const SearchifyBoundingBoxOrigin& baseline_origin) { … }

gfx::SizeF GetRenderedImageSize(FPDF_PAGEOBJECT image) { … }

bool CalculateImageWithoutScalingMatrix(FPDF_PAGEOBJECT image,
                                        const gfx::SizeF& rendered_size,
                                        FS_MATRIX& image_matrix) { … }

// Returns the transformation matrix needed to move a word to where it is
// positioned on the image.
FS_MATRIX CalculateWordMoveMatrix(const SearchifyBoundingBoxOrigin& word_origin,
                                  int word_bounding_box_width,
                                  bool word_is_rtl) { … }

void AddWordOnImage(FPDF_DOCUMENT document,
                    FPDF_PAGE page,
                    FPDF_FONT font,
                    const screen_ai::mojom::WordBoxPtr& word,
                    base::span<const FS_MATRIX> transform_matrices) { … }

void AddTextOnImage(FPDF_DOCUMENT document,
                    FPDF_PAGE page,
                    FPDF_FONT font,
                    FPDF_PAGEOBJECT image,
                    screen_ai::mojom::VisualAnnotationPtr annotation,
                    const gfx::Size& image_pixel_size) { … }

ScopedFPDFFont CreateFont(FPDF_DOCUMENT document) { … }

int GetBlockForJpeg(void* param,
                    unsigned long pos,
                    unsigned char* buf,
                    unsigned long size) { … }

}  // namespace

std::vector<uint8_t> PDFiumSearchify(
    base::span<const uint8_t> pdf_buffer,
    base::RepeatingCallback<screen_ai::mojom::VisualAnnotationPtr(
        const SkBitmap& bitmap)> perform_ocr_callback) { … }

SearchifyBoundingBoxOrigin ConvertToPdfOriginForTesting(
    const gfx::Rect& rect,
    float angle,
    float coordinate_system_height) { … }

FS_MATRIX CalculateWordMoveMatrixForTesting(
    const SearchifyBoundingBoxOrigin& origin,
    int word_bounding_box_width,
    bool word_is_rtl) { … }

PdfiumProgressiveSearchifier::ScopedSdkInitializer::ScopedSdkInitializer() { … }

PdfiumProgressiveSearchifier::ScopedSdkInitializer::~ScopedSdkInitializer() { … }

PdfiumProgressiveSearchifier::PdfiumProgressiveSearchifier()
    : … { … }

PdfiumProgressiveSearchifier::~PdfiumProgressiveSearchifier() = default;

// TODO(chuhsuan): Return bool instead of crashing on error.
void PdfiumProgressiveSearchifier::AddPage(
    const SkBitmap& bitmap,
    uint32_t page_index,
    screen_ai::mojom::VisualAnnotationPtr annotation) { … }

void PdfiumProgressiveSearchifier::DeletePage(uint32_t page_index) { … }

std::vector<uint8_t> PdfiumProgressiveSearchifier::Save() { … }

}  // namespace chrome_pdf
chromium/pdf/pdfium/pdfium_searchify.cc
