TextExtractor.h

Go to the documentation of this file.

//---------------------------------------------------------------------------------------
// Copyright (c) 2001-2024 by Apryse Software Inc. All Rights Reserved.
// Consult legal.txt regarding legal and license information.
//---------------------------------------------------------------------------------------
#ifndef PDFTRON_H_CPPPDFTextExtractor
#define PDFTRON_H_CPPPDFTextExtractor

#include <PDF/Page.h>
#include <PDF/Rect.h>
#include <Common/UString.h>
#include <C/PDF/TRN_TextExtractor.h>
#include <vector>
#include <PDF/Highlights.h>
#include<PDF/OCG/Context.h>

namespace pdftron { 
    namespace PDF {

class Style;
class Word;
class Line;

struct CharRange
{
    int index; // character index
    int length; // character length
};

class TextExtractor 
{
public:
    typedef pdftron::PDF::Style Style;
    typedef pdftron::PDF::Word Word;
    typedef pdftron::PDF::Line Line;

     TextExtractor();
     ~TextExtractor();

     enum ProcessingFlags
    {
        // Disables expanding of ligatures using a predefined mapping. 
        // Default ligatures are: fi, ff, fl, ffi, ffl, ch, cl, ct, ll, 
        // ss, fs, st, oe, OE. 
        e_no_ligature_exp = 1, 

        // Disables removing duplicated text that is frequently used to 
        // achieve visual effects of drop shadow and fake bold. 
        e_no_dup_remove = 2,

        // Treat punctuation (e.g. full stop, comma, semicolon, etc.) as 
        // word break characters. 
        e_punct_break = 4,

        // Enables removal of text that is obscured by images or 
        // rectangles. Since this option has small performance penalty 
        // on performance of text extraction, by default it is not 
        // enabled.
        e_remove_hidden_text = 8, 

        // Enables removing text that uses rendering mode 3 (i.e. invisible text).
        // Invisible text is usually used in 'PDF Searchable Images' (i.e. scanned 
        // pages with a corresponding OCR text). As a result, invisible text 
        // will be extracted by default.
        e_no_invisible_text = 16,

        // Enables removal of text that is marked as part of a Watermark layer
        e_no_watermarks = 128,

        // Use Z-order as reading order for text
        e_extract_using_zorder = 256
    };

     void Begin(Page page, const Rect* clip_ptr = 0, UInt32 flags = 0);

    void SetOCGContext(OCG::Context* ctx);
    
     int GetWordCount();
     
     void SetRightToLeftLanguage(bool rtl);
     bool GetRightToLeftLanguage();
     UString GetAsText(bool dehyphen = true);

#ifndef SWIG
     void GetAsText(UString& out_str, bool dehyphen = true);
#endif
    
    UString GetTextUnderAnnot(const Annot& annot);
    
#ifndef SWIG
    void GetTextUnderAnnot(UString& out_str, const Annot& annot);
#endif


    enum XMLOutputFlags 
    {
        // Output words as XML elements instead of inline text.
        e_words_as_elements = 1, 

        // Include bounding box information for each XML element. 
        // The bounding box information will be stored as 'bbox' attribute.
        e_output_bbox = 2, 

        // Include font and styling information.
        e_output_style_info = 4
    };

     UString GetAsXML(UInt32 xml_output_flags = 0);

#ifndef SWIG
     void GetAsXML(UString& out_xml, UInt32 xml_output_flags = 0);
#endif

     Highlights GetHighlights(const std::vector<CharRange>& char_ranges);

#ifndef SWIG

     Highlights GetHighlights(const CharRange* char_ranges, size_t char_ranges_count);
#endif

     int GetNumLines();


     Line GetFirstLine();

     void Destroy();

     // @cond PRIVATE_DOC
private:
    TRN_TextExtractor mp_extractor;

    // TextExtractor should not be copied
    TextExtractor(const TextExtractor& other);
    TextExtractor& operator= (const TextExtractor&);
    // @endcond
};

class Style 
{
public:

    SDF::Obj GetFont();

    UString GetFontName();

    double GetFontSize();

    int GetWeight();

    bool IsItalic();

    bool IsSerif();

    std::vector<int> GetColor();

#ifndef SWIG
    void GetColor(UInt8 rgb[3]);
#endif

    bool operator== (const Style& s) const;
    bool operator!= (const Style& s) const;

    Style();

    // @cond PRIVATE_DOC 
    #ifndef SWIGHIDDEN
    Style(const Style& s);
    Style(TRN_TextExtractorStyle impl);
    TRN_TextExtractorStyle mp_style;
    #endif
    // @endcond
};

class Word 
{
public:
        int GetNumGlyphs();

    Rect GetBBox();

#ifndef SWIG
    void GetBBox(double out_bbox[4]);
#endif

    std::vector<double> GetQuad();

#ifndef SWIG
    void GetQuad(double out_quad[8]);
#endif

    std::vector<double> GetGlyphQuad(int glyph_idx);

#ifndef SWIG
    void GetGlyphQuad(int glyph_idx, double out_quad[8]);
#endif

    Style GetCharStyle(int char_idx);

    Style GetStyle();

    int GetStringLen();

#ifdef SWIG
    UString GetString();
#else
    const Unicode* GetString();
#endif

    Word GetNextWord();

    int GetCurrentNum();

    bool IsValid();

    bool operator== (const Word&) const;
    bool operator!= (const Word&) const;
    Word();

    // @cond PRIVATE_DOC 
    #ifndef SWIGHIDDEN
    Word(TRN_TextExtractorWord impl);
    TRN_TextExtractorWord mp_word;
    #endif
    // @endcond
};

class Line {
public: 

    int GetNumWords();

    bool IsSimpleLine();

#ifdef SWIG
    Rect GetBBox();
#else
    const double* GetBBox();
#endif

    std::vector<double> GetQuad();

#ifndef SWIG

    void GetQuad(double out_quad[8]);
#endif

    Word GetFirstWord();

    Word GetWord(int word_idx);

    Line GetNextLine();

    int GetCurrentNum();

    Style GetStyle();

    int GetParagraphID();

    int GetFlowID();

    bool EndsWithHyphen();

    bool IsValid();

    bool operator== (const Line&) const;
    bool operator!= (const Line&) const;
    Line();

    // @cond PRIVATE_DOC 
    #ifndef SWIGHIDDEN
    Line(TRN_TextExtractorLine impl);
    TRN_TextExtractorLine mp_line;
    #endif
    // @endcond
};




#include <Impl/TextExtractor.inl>

    };  // namespace PDF
};  // namespace pdftron

#endif // PDFTRON_H_CPPPDFTextExtractor