tesseract::TessPDFRenderer Class Reference

#include <renderer.h>

Inheritance diagram for tesseract::TessPDFRenderer:

Public Member Functions
	TessPDFRenderer (const char *outputbase, const char *datadir, bool textonly=false)
Public Member Functions inherited from tesseract::TessResultRenderer
virtual	~TessResultRenderer ()
void	insert (TessResultRenderer *next)
TessResultRenderer *	next ()
bool	BeginDocument (const char *title)
bool	AddImage (TessBaseAPI *api)
bool	EndDocument ()
const char *	file_extension () const
const char *	title () const
bool	happy ()
int	imagenum () const

Protected Member Functions
bool	BeginDocumentHandler () override
bool	AddImageHandler (TessBaseAPI *api) override
bool	EndDocumentHandler () override
Protected Member Functions inherited from tesseract::TessResultRenderer
	TessResultRenderer (const char *outputbase, const char *extension)
void	AppendString (const char *s)
void	AppendData (const char *s, int len)

Detailed Description

Renders tesseract output into searchable PDF

Definition at line 215 of file renderer.h.

Constructor & Destructor Documentation

TessPDFRenderer()

tesseract::TessPDFRenderer::TessPDFRenderer	(	const char *	outputbase,
		const char *	datadir,
		bool	textonly = false
	)

Definition at line 178 of file pdfrenderer.cpp.

    : TessResultRenderer(outputbase, "pdf"),
      datadir_(datadir) {
  obj_  = 0;
  textonly_ = textonly;
  offsets_.push_back(0);

Member Function Documentation

AddImageHandler()

bool tesseract::TessPDFRenderer::AddImageHandler ( TessBaseAPI *  api )

overrideprotectedvirtual

Implements tesseract::TessResultRenderer.

Definition at line 796 of file pdfrenderer.cpp.

                                                      {
  Pix *pix = api->GetInputImage();
  const char* filename = api->GetInputName();
  int ppi = api->GetSourceYResolution();
  if (!pix || ppi <= 0)
    return false;
  double width = pixGetWidth(pix) * 72.0 / ppi;
  double height = pixGetHeight(pix) * 72.0 / ppi;
 
  std::stringstream xobject;
  // Use "C" locale (needed for int values larger than 999).
  xobject.imbue(std::locale::classic());
  if (!textonly_) {
    xobject << "/XObject << /Im1 " << (obj_ + 2) << " 0 R >>\n";
  }
 
  // PAGE
  std::stringstream stream;
  // Use "C" locale (needed for double values width and height).
  stream.imbue(std::locale::classic());
  stream.precision(2);
  stream << std::fixed <<
    obj_ << " 0 obj\n"
    "<<\n"
    "  /Type /Page\n"
    "  /Parent 2 0 R\n" // Pages object
    "  /MediaBox [0 0 " << width << " " << height << "]\n"
    "  /Contents " << (obj_ + 1) << " 0 R\n" // Contents object
    "  /Resources\n"
    "  <<\n"
    "    " << xobject.str() << // Image object
    "    /ProcSet [ /PDF /Text /ImageB /ImageI /ImageC ]\n"
    "    /Font << /f-0-0 3 0 R >>\n" // Type0 Font
    "  >>\n"
    ">>\n"
    "endobj\n";
  pages_.push_back(obj_);
  AppendPDFObject(stream.str().c_str());
 
  // CONTENTS
  const std::unique_ptr<char[]> pdftext(GetPDFTextObjects(api, width, height));
  const size_t pdftext_len = strlen(pdftext.get());
  size_t len;
  unsigned char *comp_pdftext = zlibCompress(
      reinterpret_cast<unsigned char *>(pdftext.get()), pdftext_len, &len);
  long comp_pdftext_len = len;
  stream.str("");
  stream <<
    obj_ << " 0 obj\n"
    "<<\n"
    "  /Length " << comp_pdftext_len << " /Filter /FlateDecode\n"
    ">>\n"
    "stream\n";
  AppendString(stream.str().c_str());
  long objsize = stream.str().size();
  AppendData(reinterpret_cast<char *>(comp_pdftext), comp_pdftext_len);
  objsize += comp_pdftext_len;
  lept_free(comp_pdftext);
  const char *b2 =
      "endstream\n"
      "endobj\n";
  AppendString(b2);
  objsize += strlen(b2);
  AppendPDFObjectDIY(objsize);
 
  if (!textonly_) {
    char *pdf_object = nullptr;
    int jpg_quality;
    api->GetIntVariable("jpg_quality", &jpg_quality);
    if (!imageToPDFObj(pix, filename, obj_, &pdf_object, &objsize,
                       jpg_quality)) {
      return false;
    }
    AppendData(pdf_object, objsize);
    AppendPDFObjectDIY(objsize);
    delete[] pdf_object;
  }
  return true;

BeginDocumentHandler()

bool tesseract::TessPDFRenderer::BeginDocumentHandler ( )

overrideprotectedvirtual

Reimplemented from tesseract::TessResultRenderer.

Definition at line 493 of file pdfrenderer.cpp.

                                           {
  AppendPDFObject("%PDF-1.5\n%\xDE\xAD\xBE\xEB\n");
 
  // CATALOG
  AppendPDFObject("1 0 obj\n"
                  "<<\n"
                  "  /Type /Catalog\n"
                  "  /Pages 2 0 R\n"
                  ">>\nendobj\n");
 
  // We are reserving object #2 for the /Pages
  // object, which I am going to create and write
  // at the end of the PDF file.
  AppendPDFObject("");
 
  // TYPE0 FONT
  AppendPDFObject("3 0 obj\n"
                  "<<\n"
                  "  /BaseFont /GlyphLessFont\n"
                  "  /DescendantFonts [ 4 0 R ]\n" // CIDFontType2 font
                  "  /Encoding /Identity-H\n"
                  "  /Subtype /Type0\n"
                  "  /ToUnicode 6 0 R\n" // ToUnicode
                  "  /Type /Font\n"
                  ">>\n"
                  "endobj\n");
 
  // CIDFONTTYPE2
  std::stringstream stream;
  // Use "C" locale (needed for int values larger than 999).
  stream.imbue(std::locale::classic());
  stream <<
    "4 0 obj\n"
    "<<\n"
    "  /BaseFont /GlyphLessFont\n"
    "  /CIDToGIDMap 5 0 R\n" // CIDToGIDMap
    "  /CIDSystemInfo\n"
    "  <<\n"
    "     /Ordering (Identity)\n"
    "     /Registry (Adobe)\n"
    "     /Supplement 0\n"
    "  >>\n"
    "  /FontDescriptor 7 0 R\n" // Font descriptor
    "  /Subtype /CIDFontType2\n"
    "  /Type /Font\n"
    "  /DW " << (1000 / kCharWidth) << "\n"
    ">>\n"
    "endobj\n";
  AppendPDFObject(stream.str().c_str());
 
  // CIDTOGIDMAP
  const int kCIDToGIDMapSize = 2 * (1 << 16);
  const std::unique_ptr<unsigned char[]> cidtogidmap(
      new unsigned char[kCIDToGIDMapSize]);
  for (int i = 0; i < kCIDToGIDMapSize; i++) {
    cidtogidmap[i] = (i % 2) ? 1 : 0;
  }
  size_t len;
  unsigned char *comp = zlibCompress(cidtogidmap.get(), kCIDToGIDMapSize, &len);
  stream.str("");
  stream <<
    "5 0 obj\n"
    "<<\n"
    "  /Length " << len << " /Filter /FlateDecode\n"
    ">>\n"
    "stream\n";
  AppendString(stream.str().c_str());
  long objsize = stream.str().size();
  AppendData(reinterpret_cast<char *>(comp), len);
  objsize += len;
  lept_free(comp);
  const char *endstream_endobj =
      "endstream\n"
      "endobj\n";
  AppendString(endstream_endobj);
  objsize += strlen(endstream_endobj);
  AppendPDFObjectDIY(objsize);
 
  const char stream2[] =
      "/CIDInit /ProcSet findresource begin\n"
      "12 dict begin\n"
      "begincmap\n"
      "/CIDSystemInfo\n"
      "<<\n"
      "  /Registry (Adobe)\n"
      "  /Ordering (UCS)\n"
      "  /Supplement 0\n"
      ">> def\n"
      "/CMapName /Adobe-Identify-UCS def\n"
      "/CMapType 2 def\n"
      "1 begincodespacerange\n"
      "<0000> <FFFF>\n"
      "endcodespacerange\n"
      "1 beginbfrange\n"
      "<0000> <FFFF> <0000>\n"
      "endbfrange\n"
      "endcmap\n"
      "CMapName currentdict /CMap defineresource pop\n"
      "end\n"
      "end\n";
 
  // TOUNICODE
  stream.str("");
  stream <<
    "6 0 obj\n"
    "<< /Length " << (sizeof(stream2) - 1) << " >>\n"
    "stream\n" << stream2 <<
    "endstream\n"
    "endobj\n";
  AppendPDFObject(stream.str().c_str());
 
  // FONT DESCRIPTOR
  stream.str("");
  stream <<
    "7 0 obj\n"
    "<<\n"
    "  /Ascent 1000\n"
    "  /CapHeight 1000\n"
    "  /Descent -1\n"       // Spec says must be negative
    "  /Flags 5\n"          // FixedPitch + Symbolic
    "  /FontBBox  [ 0 0 " << (1000 / kCharWidth) << " 1000 ]\n"
    "  /FontFile2 8 0 R\n"
    "  /FontName /GlyphLessFont\n"
    "  /ItalicAngle 0\n"
    "  /StemV 80\n"
    "  /Type /FontDescriptor\n"
    ">>\n"
    "endobj\n";
  AppendPDFObject(stream.str().c_str());
 
  stream.str("");
  stream << datadir_.c_str() << "/pdf.ttf";
  FILE *fp = fopen(stream.str().c_str(), "rb");
  if (!fp) {
    tprintf("Cannot open file \"%s\"!\n", stream.str().c_str());
    return false;
  }
  fseek(fp, 0, SEEK_END);
  auto size = std::ftell(fp);
  if (size < 0) {
    fclose(fp);
    return false;
  }
  fseek(fp, 0, SEEK_SET);
  const std::unique_ptr<char[]> buffer(new char[size]);
  if (!tesseract::DeSerialize(fp, buffer.get(), size)) {
    fclose(fp);
    return false;
  }
  fclose(fp);
  // FONTFILE2
  stream.str("");
  stream <<
    "8 0 obj\n"
    "<<\n"
    "  /Length " << size << "\n"
    "  /Length1 " << size << "\n"
    ">>\n"
    "stream\n";
  AppendString(stream.str().c_str());
  objsize  = stream.str().size();
  AppendData(buffer.get(), size);
  objsize += size;
  AppendString(endstream_endobj);
  objsize += strlen(endstream_endobj);
  AppendPDFObjectDIY(objsize);
  return true;

EndDocumentHandler()

bool tesseract::TessPDFRenderer::EndDocumentHandler ( )

overrideprotectedvirtual

Reimplemented from tesseract::TessResultRenderer.

Definition at line 877 of file pdfrenderer.cpp.

                                         {
  // We reserved the /Pages object number early, so that the /Page
  // objects could refer to their parent. We finally have enough
  // information to go fill it in. Using lower level calls to manipulate
  // the offset record in two spots, because we are placing objects
  // out of order in the file.
 
  // PAGES
  const long int kPagesObjectNumber = 2;
  offsets_[kPagesObjectNumber] = offsets_.back();  // manipulation #1
  std::stringstream stream;
  // Use "C" locale (needed for int values larger than 999).
  stream.imbue(std::locale::classic());
  stream << kPagesObjectNumber << " 0 obj\n<<\n  /Type /Pages\n  /Kids [ ";
  AppendString(stream.str().c_str());
  size_t pages_objsize  = stream.str().size();
  for (size_t i = 0; i < pages_.unsigned_size(); i++) {
    stream.str("");
    stream << pages_[i] << " 0 R ";
    AppendString(stream.str().c_str());
    pages_objsize += stream.str().size();
  }
  stream.str("");
  stream << "]\n  /Count " << pages_.size() << "\n>>\nendobj\n";
  AppendString(stream.str().c_str());
  pages_objsize += stream.str().size();
  offsets_.back() += pages_objsize;    // manipulation #2
 
  // INFO
  STRING utf16_title = "FEFF";  // byte_order_marker
  std::vector<char32> unicodes = UNICHAR::UTF8ToUTF32(title());
  char utf16[kMaxBytesPerCodepoint];
  for (char32 code : unicodes) {
    if (CodepointToUtf16be(code, utf16)) {
      utf16_title += utf16;
    }
  }
 
  char* datestr = l_getFormattedDate();
  stream.str("");
  stream
    << obj_ << " 0 obj\n"
       "<<\n"
       "  /Producer (Tesseract " << tesseract::TessBaseAPI::Version() << ")\n"
       "  /CreationDate (D:" << datestr << ")\n"
       "  /Title <" << utf16_title.c_str() << ">\n"
       ">>\n"
       "endobj\n";
  lept_free(datestr);
  AppendPDFObject(stream.str().c_str());
  stream.str("");
  stream << "xref\n0 " << obj_ << "\n0000000000 65535 f \n";
  AppendString(stream.str().c_str());
  for (int i = 1; i < obj_; i++) {
    stream.str("");
    stream.width(10);
    stream.fill('0');
    stream << offsets_[i] << " 00000 n \n";
    AppendString(stream.str().c_str());
  }
  stream.str("");
  stream
    << "trailer\n<<\n  /Size " << obj_ << "\n"
    "  /Root 1 0 R\n" // catalog
    "  /Info " << (obj_ - 1) << " 0 R\n" // info
    ">>\nstartxref\n" << offsets_.back() << "\n%%EOF\n";
  AppendString(stream.str().c_str());
  return true;

---

The documentation for this class was generated from the following files:

• /tesseract/include/tesseract/renderer.h
• /tesseract/src/api/pdfrenderer.cpp