sdext/source/pdfimport/pdfiadaptor.cxx

   1 /*************************************************************************
   2  *
   3  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
   4  *
   5  * Copyright 2008 by Sun Microsystems, Inc.
   6  *
   7  * OpenOffice.org - a multi-platform office productivity suite
   8  *
   9  * $RCSfile: pdfiadaptor.cxx,v $
  10  *
  11  * $Revision: 1.2 $
  12  *
  13  * This file is part of OpenOffice.org.
  14  *
  15  * OpenOffice.org is free software: you can redistribute it and/or modify
  16  * it under the terms of the GNU Lesser General Public License version 3
  17  * only, as published by the Free Software Foundation.
  18  *
  19  * OpenOffice.org is distributed in the hope that it will be useful,
  20  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  21  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  22  * GNU Lesser General Public License version 3 for more details
  23  * (a copy is included in the LICENSE file that accompanied this code).
  24  *
  25  * You should have received a copy of the GNU Lesser General Public License
  26  * version 3 along with OpenOffice.org.  If not, see
  27  * <http://www.openoffice.org/license.html>
  28  * for a copy of the LGPLv3 License.
  29  *
  30  ************************************************************************/
  31
  32 // MARKER(update_precomp.py): autogen include statement, do not remove
  33 #include "precompiled_sdext.hxx"
  34
  35 #include "pdfiadaptor.hxx"
  36 #include "filterdet.hxx"
  37 #include "saxemitter.hxx"
  38 #include "odfemitter.hxx"
  39 #include "inc/wrapper.hxx"
  40 #include "inc/contentsink.hxx"
  41 #include "tree/pdfiprocessor.hxx"
  42
  43 #include <osl/file.h>
  44 #include <osl/thread.h>
  45 #include <osl/diagnose.h>
  46 #include <cppuhelper/factory.hxx>
  47 #include <cppuhelper/implementationentry.hxx>
  48 #include <com/sun/star/lang/XMultiComponentFactory.hpp>
  49 #include <com/sun/star/uno/RuntimeException.hpp>
  50 #include <com/sun/star/io/XInputStream.hpp>
  51 #include <com/sun/star/frame/XLoadable.hpp>
  52 #include <com/sun/star/xml/sax/XDocumentHandler.hpp>
  53 #include <com/sun/star/io/XSeekable.hpp>
  54
  55
  56 #include <boost/shared_ptr.hpp>
  57
  58 using namespace com::sun::star;
  59
  60
  61 namespace pdfi
  62 {
  63
  64 PDFIHybridAdaptor::PDFIHybridAdaptor( const uno::Reference< uno::XComponentContext >& xContext ) :
  65     PDFIHybridAdaptorBase( m_aMutex ),
  66     m_xContext( xContext ),
  67     m_xModel()
  68 {
  69 }
  70
  71 // XFilter
  72 sal_Bool SAL_CALL PDFIHybridAdaptor::filter( const uno::Sequence< beans::PropertyValue >& rFilterData ) throw( uno::RuntimeException )
  73 {
  74     sal_Bool bRet = sal_False;
  75     if( m_xModel.is() )
  76     {
  77         uno::Reference< io::XStream > xSubStream;
  78         rtl::OUString aPwd;
  79         const beans::PropertyValue* pAttribs = rFilterData.getConstArray();
  80         sal_Int32 nAttribs = rFilterData.getLength();
  81         sal_Int32 nPwPos = -1;
  82         for( sal_Int32 i = 0; i < nAttribs; i++ )
  83         {
  84             #if OSL_DEBUG_LEVEL > 1
  85             rtl::OUString aVal( RTL_CONSTASCII_USTRINGPARAM( "<no string>" ) );
  86             pAttribs[i].Value >>= aVal;
  87             OSL_TRACE( "filter: Attrib: %s = %s\n",
  88                        rtl::OUStringToOString( pAttribs[i].Name, RTL_TEXTENCODING_UTF8 ).getStr(),
  89                        rtl::OUStringToOString( aVal, RTL_TEXTENCODING_UTF8 ).getStr() );
  90             #endif
  91             if( pAttribs[i].Name.equalsAscii( "EmbeddedSubstream" ) )
  92                 pAttribs[i].Value >>= xSubStream;
  93             else if( pAttribs[i].Name.equalsAscii( "Password" ) )
  94             {
  95                 nPwPos = i;
  96                 pAttribs[i].Value >>= aPwd;
  97             }
  98         }
  99         bool bAddPwdProp = false;
 100         if( ! xSubStream.is() )
 101         {
 102             uno::Reference< io::XInputStream > xInput;
 103             for( sal_Int32 i = 0; i < nAttribs; i++ )
 104             {
 105                 if( pAttribs[i].Name.equalsAscii( "InputStream" ) )
 106                 {
 107                     pAttribs[i].Value >>= xInput;
 108                     break;
 109                 }
 110             }
 111             if( xInput.is() )
 112             {
 113                 // TODO(P2): extracting hybrid substream twice - once during detection, second time here
 114                 uno::Reference< io::XSeekable > xSeek( xInput, uno::UNO_QUERY );
 115                 if( xSeek.is() )
 116                     xSeek->seek( 0 );
 117                 oslFileHandle aFile = NULL;
 118                 sal_uInt64 nWritten = 0;
 119                 rtl::OUString aURL;
 120                 if( osl_createTempFile( NULL, &aFile, &aURL.pData ) == osl_File_E_None )
 121                 {
 122                     OSL_TRACE( "created temp file %s\n", rtl::OUStringToOString( aURL, RTL_TEXTENCODING_UTF8 ).getStr() );
 123                     const sal_Int32 nBufSize = 4096;
 124                     uno::Sequence<sal_Int8> aBuf(nBufSize);
 125                     // copy the bytes
 126                     sal_Int32 nBytes;
 127                     do
 128                     {
 129                         nBytes = xInput->readBytes( aBuf, nBufSize );
 130                         if( nBytes > 0 )
 131                         {
 132                             osl_writeFile( aFile, aBuf.getConstArray(), nBytes, &nWritten );
 133                             if( static_cast<sal_Int32>(nWritten) != nBytes )
 134                             {
 135                                 xInput.clear();
 136                                 break;
 137                             }
 138                         }
 139                     } while( nBytes == nBufSize );
 140                     osl_closeFile( aFile );
 141                     if( xInput.is() )
 142                     {
 143                         rtl::OUString aEmbedMimetype;
 144                         rtl::OUString aOrgPwd( aPwd );
 145                         xSubStream = getAdditionalStream( aURL, aEmbedMimetype, aPwd, m_xContext, rFilterData, true );
 146                         if( aOrgPwd != aPwd )
 147                             bAddPwdProp = true;
 148                     }
 149                     osl_removeFile( aURL.pData );
 150                 }
 151                 else
 152                     xSubStream.clear();
 153             }
 154         }
 155         if( xSubStream.is() )
 156         {
 157             uno::Sequence< uno::Any > aArgs( 2 );
 158             aArgs[0] <<= m_xModel;
 159             aArgs[1] <<= xSubStream;
 160
 161             OSL_TRACE( "try to instantiate subfilter\n" );
 162             uno::Reference< document::XFilter > xSubFilter;
 163             try {
 164                 xSubFilter = uno::Reference<document::XFilter>(
 165                     m_xContext->getServiceManager()->createInstanceWithArgumentsAndContext(
 166                         rtl::OUString( RTL_CONSTASCII_USTRINGPARAM( "com.sun.star.document.OwnSubFilter" ) ),
 167                         aArgs,
 168                         m_xContext ),
 169                     uno::UNO_QUERY );
 170             }
 171             catch(uno::Exception& e)
 172             {
 173                 (void)e;
 174                 OSL_TRACE( "subfilter exception: %s\n",
 175                            OUStringToOString( e.Message, RTL_TEXTENCODING_UTF8 ).getStr() );
 176             }
 177
 178             OSL_TRACE( "subfilter: %p\n", xSubFilter.get() );
 179             if( xSubFilter.is() )
 180             {
 181                 if( bAddPwdProp )
 182                 {
 183                     uno::Sequence<beans::PropertyValue> aFilterData( rFilterData );
 184                     if( nPwPos == -1 )
 185                     {
 186                         nPwPos = aFilterData.getLength();
 187                         aFilterData.realloc( nPwPos+1 );
 188                         aFilterData[nPwPos].Name = rtl::OUString(
 189                             RTL_CONSTASCII_USTRINGPARAM( "Password" ) );
 190                     }
 191                     aFilterData[nPwPos].Value <<= aPwd;
 192                     bRet = xSubFilter->filter( aFilterData );
 193                 }
 194                 else
 195                     bRet = xSubFilter->filter( rFilterData );
 196             }
 197         }
 198         #if OSL_DEBUG_LEVEL > 1
 199         else
 200             OSL_TRACE( "PDFIAdaptor::filter: no embedded substream set\n" );
 201         #endif
 202     }
 203     #if OSL_DEBUG_LEVEL > 1
 204     else
 205         OSL_TRACE( "PDFIAdaptor::filter: no model set\n" );
 206     #endif
 207
 208     return bRet;
 209 }
 210
 211 void SAL_CALL PDFIHybridAdaptor::cancel() throw()
 212 {
 213 }
 214
 215 //XImporter
 216 void SAL_CALL PDFIHybridAdaptor::setTargetDocument( const uno::Reference< lang::XComponent >& xDocument ) throw( lang::IllegalArgumentException )
 217 {
 218     OSL_TRACE( "PDFIAdaptor::setTargetDocument\n" );
 219     m_xModel = uno::Reference< frame::XModel >( xDocument, uno::UNO_QUERY );
 220     if( xDocument.is() && ! m_xModel.is() )
 221         throw lang::IllegalArgumentException();
 222 }
 223
 224 //---------------------------------------------------------------------------------------
 225
 226 PDFIRawAdaptor::PDFIRawAdaptor( const uno::Reference< uno::XComponentContext >& xContext ) :
 227     PDFIAdaptorBase( m_aMutex ),
 228     m_xContext( xContext ),
 229     m_xModel(),
 230     m_pVisitorFactory(),
 231     m_bEnableToplevelText(false)
 232 {
 233 }
 234
 235 void PDFIRawAdaptor::setTreeVisitorFactory(const TreeVisitorFactorySharedPtr& rVisitorFactory)
 236 {
 237     m_pVisitorFactory = rVisitorFactory;
 238 }
 239
 240 bool PDFIRawAdaptor::parse( const uno::Reference<io::XInputStream>&       xInput,
 241                             const uno::Reference<task::XInteractionHandler>& xIHdl,
 242                             const rtl::OUString&                          rPwd,
 243                             const uno::Reference<task::XStatusIndicator>& xStatus,
 244                             const XmlEmitterSharedPtr&                    rEmitter,
 245                             const rtl::OUString&                          rURL )
 246 {
 247     // container for metaformat
 248     boost::shared_ptr<PDFIProcessor> pSink(
 249         new PDFIProcessor(xStatus));
 250
 251     // TEMP! TEMP!
 252     if( m_bEnableToplevelText )
 253         pSink->enableToplevelText();
 254
 255     bool bSuccess=false;
 256
 257     if( xInput.is() && (!rURL.getLength() || rURL.compareToAscii( "file:", 5 ) != 0) )
 258         bSuccess = xpdf_ImportFromStream( xInput, pSink, xIHdl, rPwd, m_xContext );
 259     else
 260         bSuccess = xpdf_ImportFromFile( rURL, pSink, xIHdl, rPwd, m_xContext );
 261
 262     if( bSuccess )
 263         pSink->emit(*rEmitter,*m_pVisitorFactory);
 264
 265     return bSuccess;
 266 }
 267
 268 bool PDFIRawAdaptor::odfConvert( const rtl::OUString&                          rURL,
 269                                  const uno::Reference<io::XOutputStream>&      xOutput,
 270                                  const uno::Reference<task::XStatusIndicator>& xStatus )
 271 {
 272     XmlEmitterSharedPtr pEmitter = createOdfEmitter(xOutput);
 273     const bool bSuccess = parse(uno::Reference<io::XInputStream>(),
 274                                 uno::Reference<task::XInteractionHandler>(),
 275                                 rtl::OUString(),
 276                                 xStatus,pEmitter,rURL);
 277
 278     // tell input stream that it is no longer needed
 279     xOutput->closeOutput();
 280
 281     return bSuccess;
 282 }
 283
 284 // XImportFilter
 285 sal_Bool SAL_CALL PDFIRawAdaptor::importer( const uno::Sequence< beans::PropertyValue >&        rSourceData,
 286                                             const uno::Reference< xml::sax::XDocumentHandler >& rHdl,
 287                                             const uno::Sequence< rtl::OUString >&               /*rUserData*/ ) throw( uno::RuntimeException )
 288 {
 289     // get the InputStream carrying the PDF content
 290     uno::Reference< io::XInputStream > xInput;
 291     uno::Reference< task::XStatusIndicator > xStatus;
 292     uno::Reference< task::XInteractionHandler > xInteractionHandler;
 293     rtl::OUString aURL;
 294     rtl::OUString aPwd;
 295     const beans::PropertyValue* pAttribs = rSourceData.getConstArray();
 296     sal_Int32 nAttribs = rSourceData.getLength();
 297     for( sal_Int32 i = 0; i < nAttribs; i++, pAttribs++ )
 298     {
 299         OSL_TRACE("importer Attrib: %s\n", OUStringToOString( pAttribs->Name, RTL_TEXTENCODING_UTF8 ).getStr() );
 300         if( pAttribs->Name.equalsAscii( "InputStream" ) )
 301             pAttribs->Value >>= xInput;
 302         else if( pAttribs->Name.equalsAscii( "URL" ) )
 303             pAttribs->Value >>= aURL;
 304         else if( pAttribs->Name.equalsAscii( "StatusIndicator" ) )
 305             pAttribs->Value >>= xStatus;
 306         else if( pAttribs->Name.equalsAscii( "InteractionHandler" ) )
 307             pAttribs->Value >>= xInteractionHandler;
 308         else if( pAttribs->Name.equalsAscii( "Password" ) )
 309             pAttribs->Value >>= aPwd;
 310     }
 311     if( !xInput.is() )
 312         return sal_False;
 313
 314     XmlEmitterSharedPtr pEmitter = createSaxEmitter(rHdl);
 315     const bool bSuccess = parse(xInput,xInteractionHandler, aPwd, xStatus,pEmitter,aURL);
 316
 317     // tell input stream that it is no longer needed
 318     xInput->closeInput();
 319     xInput.clear();
 320
 321     return bSuccess;
 322 }
 323
 324 //XImporter
 325 void SAL_CALL PDFIRawAdaptor::setTargetDocument( const uno::Reference< lang::XComponent >& xDocument ) throw( lang::IllegalArgumentException )
 326 {
 327     OSL_TRACE( "PDFIAdaptor::setTargetDocument\n" );
 328     m_xModel = uno::Reference< frame::XModel >( xDocument, uno::UNO_QUERY );
 329     if( xDocument.is() && ! m_xModel.is() )
 330         throw lang::IllegalArgumentException();
 331 }
 332
 333 }