source/pdfimport.cc

   1 /* This file is part of Shapes.
   2  *
   3  * Shapes is free software: you can redistribute it and/or modify
   4  * it under the terms of the GNU General Public License as published by
   5  * the Free Software Foundation, either version 3 of the License, or
   6  * any later version.
   7  *
   8  * Shapes is distributed in the hope that it will be useful,
   9  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  10  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  11  * GNU General Public License for more details.
  12  *
  13  * You should have received a copy of the GNU General Public License
  14  * along with Shapes.  If not, see <http://www.gnu.org/licenses/>.
  15  *
  16  * Copyright 2008 Henrik Tidefelt
  17  */
  18
  19 #include "pdfimport.h"
  20 #include "globals.h"
  21 #include "shapesexceptions.h"
  22 #include "pathtypes.h"
  23 #include "drawabletypes.h"
  24 #include "texlabelmanager.h"
  25
  26 using namespace Shapes;
  27 using namespace SimplePDF;
  28
  29 Kernel::Import::Import( )
  30 { }
  31
  32 RefCountPtr< const std::vector< RefCountPtr< const Shapes::Lang::XObject > > >
  33 Kernel::Import::addPagesAsXObjects( RefCountPtr< PDF_in > pdfi )
  34 {
  35         using namespace Shapes;
  36
  37         importSources_.push_back( pdfi ); // Keep the source alive so that it can be used when finally producing output
  38         IndirectRemapType indirectRemap;
  39         std::vector< RefCountPtr< const Lang::XObject > > * res = new std::vector< RefCountPtr< const Lang::XObject > >;
  40         for( size_t pageNumber( 0 ); pageNumber < pdfi->getPageCount( ); ++pageNumber )
  41                 {
  42                         RefCountPtr< PDF_Dictionary > pageDic( pdfi->getPage( pageNumber ) );
  43                         RefCountPtr< PDF_Vector > cropBox = down_cast_follow< PDF_Vector >( pageDic->getInheritable( "CropBox" ) );
  44                         if( cropBox == NullPtr< PDF_Vector >( ) )
  45                                 {
  46                                         cropBox = down_cast_follow< PDF_Vector >( pageDic->getInheritable( "MediaBox" ) );
  47                                 }
  48                         if( cropBox == NullPtr< PDF_Vector >( ) )
  49                                 {
  50                                         throw Exceptions::InternalError( strrefdup( "Failed to find crop box of imported page.  (Searched to the page tree root.)" ) );
  51                                 }
  52                         RefCountPtr< PDF_Stream_in > original( pdfi->follow< PDF_Stream_in >( (*pageDic)[ "Contents" ] ) );
  53
  54 //                       RefCountPtr< PDF_Dictionary > xobjectDic( pdfi->follow< PDF_Dictionary >( (*pdfi->follow< PDF_Dictionary >( (*pageDic)[ "Resources" ] ))[ "XObject" ] ) );
  55 //                       if( xobjectDic->dic.size( ) != 1 )
  56 //                               {
  57 //                                       throw Exceptions::InternalError( strrefdup( "Expected exactly 1 XObject on a TeX label input page" ) );
  58 //                               }
  59 //                       RefCountPtr< PDF_Stream_in > original( pdfi->follow< PDF_Stream_in >( xobjectDic->dic.begin( )->second ) );
  60                         RefCountPtr< PDF_Stream_in > newObj( new PDF_Stream_in( original->is, original->dataStart ) );
  61                         (*newObj)[ "Subtype" ] = newName( "Form" );
  62                         (*newObj)[ "FormType" ] = newInt( 1 );
  63                         (*newObj)[ "BBox" ] = cropBox;                                                   // ->rectangleIntersection( down_cast_follow< PDF_Vector >( (*original)[ "BBox" ] ) );
  64                         (*newObj)[ "Matrix" ] = RefCountPtr< PDF_Object >( new PDF_Vector( 1, 0, 0, 1, 0, 0 ) );
  65                         (*newObj)[ "Filter" ] = (*original)[ "Filter" ];
  66                         (*newObj)[ "Length" ] = RefCountPtr< PDF_Object >( new PDF_Int( pdfi->follow< PDF_Int >( (*original)[ "Length" ] )->value( ) ) );
  67                         (*newObj)[ "Resources" ] = deepCopy( (*pageDic)[ "Resources" ], & Kernel::theIndirectObjectCount, & indirectRemap );
  68
  69                         Concrete::Length xmin;
  70                         Concrete::Length xmax;
  71                         Concrete::Length ymin;
  72                         Concrete::Length ymax;
  73                         try
  74                                 {
  75                                         RefCountPtr< PDF_Vector > bboxTyped = pdfi->follow< PDF_Vector >( (*newObj)[ "BBox" ] );
  76                                         if( bboxTyped->vec.size( ) != 4 )
  77                                                 {
  78                                                         throw Exceptions::InternalError( strrefdup( "The bbox of the imported page was not of size 4." ) );
  79                                                 }
  80                                         xmin = Concrete::Length( pdfi->follow< PDF_Float >( bboxTyped->vec[ 0 ] )->value( ) );
  81                                         xmax = Concrete::Length( pdfi->follow< PDF_Float >( bboxTyped->vec[ 2 ] )->value( ) );
  82                                         ymin = Concrete::Length( pdfi->follow< PDF_Float >( bboxTyped->vec[ 1 ] )->value( ) );
  83                                         ymax = Concrete::Length( pdfi->follow< PDF_Float >( bboxTyped->vec[ 3 ] )->value( ) );
  84                                 }
  85                         catch( Exceptions::Exception & ball )
  86                                 {
  87                                         throw;
  88                                 }
  89                         catch( const char * ball )
  90                                 {
  91                                         std::ostringstream oss;
  92                                         oss << "An error occurred while evaluating the bbox of the imported page: " << ball ;
  93                                         throw Exceptions::InternalError( strrefdup( oss.str( ).c_str( ) ) );
  94                                 }
  95                         catch( ... )
  96                                 {
  97                                         throw Exceptions::InternalError( strrefdup( "An error occurred while evaluating the bbox of the imported page." ) );
  98                                 }
  99                         Lang::ElementaryPath2D * bboxpath = new Lang::ElementaryPath2D;
 100                         bboxpath->push_back( new Concrete::PathPoint2D( xmin, ymin ) );
 101                         bboxpath->push_back( new Concrete::PathPoint2D( xmin, ymax ) );
 102                         bboxpath->push_back( new Concrete::PathPoint2D( xmax, ymax ) );
 103                         bboxpath->push_back( new Concrete::PathPoint2D( xmax, ymin ) );
 104                         bboxpath->close( );
 105
 106                         RefCountPtr< PDF_Object > indirection = SimplePDF::indirect( newObj, & Kernel::theIndirectObjectCount );
 107                         RefCountPtr< const Lang::ElementaryPath2D > bbox = RefCountPtr< const Lang::ElementaryPath2D >( bboxpath );
 108                         res->push_back( RefCountPtr< Lang::XObject >( new Lang::XObject( indirection, bbox, bbox, Kernel::THE_DEFAULT_STATE ) ) );
 109                 }
 110         return RefCountPtr< const std::vector< RefCountPtr< const Lang::XObject > > >( res );
 111 }
 112
 113 void
 114 Kernel::Import::importBtexEtexThings( RefCountPtr< PDF_in > pdfi, Shapes::Kernel::TeXLabelManager::MapType * dstMap, const std::string & setupCodeHash )
 115 {
 116         using namespace Shapes;
 117
 118         importSources_.push_back( pdfi ); // Keep the source alive so that it can be used when finally producing output
 119         IndirectRemapType indirectRemap;
 120
 121         PDF_in::PageIterator theEnd = pdfi->endPages( );
 122         for( PDF_in::PageIterator i = pdfi->beginPages( ); i != theEnd; ++i )
 123                 {
 124                         RefCountPtr< PDF_Dictionary > xobjectDic( pdfi->follow< PDF_Dictionary >( (*pdfi->follow< PDF_Dictionary >( (**i)[ "Resources" ] ))[ "XObject" ] ) );
 125                         if( xobjectDic->dic.size( ) != 1 )
 126                                 {
 127                                         throw Exceptions::InternalError( strrefdup( "Expected exactly 1 XObject on a TeX label input page" ) );
 128                                 }
 129                         RefCountPtr< PDF_Stream_in > original( pdfi->follow< PDF_Stream_in >( xobjectDic->dic.begin( )->second ) );
 130
 131                         std::string texStr;
 132                         {
 133                                 RefCountPtr< PDF_Stream_in > texStream( pdfi->follow< PDF_Stream_in >( (*original)[ "TeXsrc" ] ) );
 134                                 std::ostringstream tmp;
 135                                 texStream->writeDataDefilteredTo( tmp );
 136                                 texStr = tmp.str( );
 137                         }
 138                         if( i == pdfi->beginPages( ) )
 139                                 {
 140                                         /* This page only contains information about the TeX context */
 141                                         if( texStr != setupCodeHash )
 142                                                 {
 143                                                         throw Exceptions::TeXSetupHasChanged( );
 144                                                 }
 145                                         continue;
 146                                 }
 147                         Concrete::Length height( pdfNameToDouble( (*original)[ "TeXht" ] ) );
 148                         Concrete::Length depth( pdfNameToDouble( (*original)[ "TeXdp" ] ) );
 149                         Concrete::Length width( pdfNameToDouble( (*original)[ "TeXwd" ] ) );
 150
 151                         const Concrete::Length bboxAddY = 0.08 * ( depth + height );
 152                         const Concrete::Length xmin = Concrete::ZERO_LENGTH;
 153                         const Concrete::Length xmax = width;
 154                         const Concrete::Length ymin = -depth - bboxAddY;
 155                         const Concrete::Length ymax = height + bboxAddY;
 156
 157                         RefCountPtr< PDF_Stream_in > newObj( new PDF_Stream_in( original->is, original->dataStart ) );
 158                         (*newObj)[ "Subtype" ] = newName( "Form" );
 159                         (*newObj)[ "FormType" ] = newInt( 1 );
 160                         (*newObj)[ "BBox" ] = RefCountPtr< PDF_Object >( new PDF_Vector( 0.0,
 161                                                                                                                                                                                                                                                                                          - bboxAddY.offtype< 1, 0 >( ),
 162                                                                                                                                                                                                                                                                                          ( xmax - xmin ).offtype< 1, 0 >( ),
 163                                                                                                                                                                                                                                                                                          ( ymax - ymin ).offtype< 1, 0 >( ) ) );
 164                         //                      (*newObj)[ "BBox" ] = (*original)[ "BBox" ];  /* This box is too small! */
 165                         (*newObj)[ "Matrix" ] = RefCountPtr< PDF_Object >( new PDF_Vector( 1, 0, 0, 1, 0, - depth.offtype< 1, 0 >( ) ) );
 166                         (*newObj)[ "Filter" ] = (*original)[ "Filter" ];
 167                         (*newObj)[ "Length" ] = RefCountPtr< PDF_Object >( new PDF_Int( pdfi->follow< PDF_Int >( (*original)[ "Length" ] )->value( ) ) );
 168                         (*newObj)[ "Resources" ] = deepCopy( (*original)[ "Resources" ], & Kernel::theIndirectObjectCount, & indirectRemap );
 169
 170                         Lang::ElementaryPath2D * bboxpath = new Lang::ElementaryPath2D;
 171                         bboxpath->push_back( new Concrete::PathPoint2D( xmin, ymin ) );
 172                         bboxpath->push_back( new Concrete::PathPoint2D( xmin, ymax ) );
 173                         bboxpath->push_back( new Concrete::PathPoint2D( xmax, ymax ) );
 174                         bboxpath->push_back( new Concrete::PathPoint2D( xmax, ymin ) );
 175                         bboxpath->close( );
 176
 177                         if( dstMap->find( texStr ) != dstMap->end( ) )
 178                                 {
 179                                         throw Exceptions::InternalError( strrefdup( "Multiply generated TeX label: " + texStr ) );
 180                                 }
 181
 182                         RefCountPtr< PDF_Object > indirection = SimplePDF::indirect( newObj, & Kernel::theIndirectObjectCount );
 183                         dstMap->insert( Shapes::Kernel::TeXLabelManager::MapType::value_type( texStr, RefCountPtr< const Lang::XObject >( new Lang::XObject( indirection, RefCountPtr< const Lang::ElementaryPath2D >( bboxpath ) ) ) ) );
 184                 }
 185 }
 186
 187 void
 188 Kernel::Import::free( )
 189 {
 190         while( ! importSources_.empty( ) )
 191                 {
 192                         importSources_.pop_back( );
 193                 }
 194 }
 195
 196 double
 197 Kernel::Import::pdfNameToDouble( RefCountPtr< PDF_Object > nameObject )
 198 {
 199         RefCountPtr< PDF_Name > name( nameObject.down_cast< PDF_Name >( ) );
 200         if( name == NullPtr< PDF_Name >( ) )
 201                 {
 202                         throw( "PDF_out::pdfNameToDouble: The object was not a name" );
 203                 }
 204         const double pointToBigPointFactor = 72 / 72.27;
 205         char * end;
 206         double res = pointToBigPointFactor * strtod( name->name( ).c_str( ), & end );
 207         if( strcmp( end, "pt" ) != 0 )
 208                 {
 209                         throw( "PDF_out::pdfNameToDouble: Expected \"pt\" to follow length number" );
 210                 }
 211         return res;
 212 }
 213