sdext/source/pdfimport/inc/pdfparse.hxx

   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: pdfparse.hxx,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 #ifndef INCLUDED_PDFI_PDFPARSE_HXX
  33 #define INCLUDED_PDFI_PDFPARSE_HXX
  34
  35 #include <sal/types.h>
  36 #include <rtl/ustring.hxx>
  37 #include <rtl/string.hxx>
  38
  39 #include <vector>
  40 #include <hash_map>
  41
  42 namespace pdfparse
  43 {
  44
  45 struct EmitImplData;
  46 struct PDFContainer;
  47 class EmitContext
  48 {
  49     public:
  50     virtual bool write( const void* pBuf, unsigned int nLen ) = 0;
  51     virtual unsigned int getCurPos() = 0;
  52     virtual bool copyOrigBytes( unsigned int nOrigOffset, unsigned int nLen ) = 0;
  53     virtual unsigned int readOrigBytes( unsigned int nOrigOffset, unsigned int nLen, void* pBuf ) = 0;
  54
  55     EmitContext( const PDFContainer* pTop = NULL );
  56     virtual ~EmitContext();
  57
  58     // set this to deflate contained streams
  59     bool m_bDeflate;
  60     // set this to decrypt the PDF file
  61     bool m_bDecrypt;
  62
  63     private:
  64     friend struct PDFEntry;
  65     EmitImplData* m_pImplData;
  66 };
  67
  68 struct PDFEntry
  69 {
  70     PDFEntry() {}
  71     virtual ~PDFEntry();
  72
  73     virtual bool emit( EmitContext& rWriteContext ) const = 0;
  74     virtual PDFEntry* clone() const = 0;
  75
  76     protected:
  77     EmitImplData* getEmitData( EmitContext& rContext ) const;
  78     void setEmitData( EmitContext& rContext, EmitImplData* pNewEmitData ) const;
  79 };
  80
  81 struct PDFComment : public PDFEntry
  82 {
  83     rtl::OString  m_aComment;
  84
  85     PDFComment( const rtl::OString& rComment )
  86     : PDFEntry(), m_aComment( rComment ) {}
  87     virtual ~PDFComment();
  88     virtual bool emit( EmitContext& rWriteContext ) const;
  89     virtual PDFEntry* clone() const;
  90 };
  91
  92 struct PDFValue : public PDFEntry
  93 {
  94     // abstract base class for simple values
  95     PDFValue() : PDFEntry() {}
  96     virtual ~PDFValue();
  97 };
  98
  99 struct PDFName : public PDFValue
 100 {
 101     rtl::OString  m_aName;
 102
 103     PDFName( const rtl::OString& rName )
 104     : PDFValue(), m_aName( rName ) {}
 105     virtual ~PDFName();
 106     virtual bool emit( EmitContext& rWriteContext ) const;
 107     virtual PDFEntry* clone() const;
 108
 109     rtl::OUString getFilteredName() const;
 110 };
 111
 112 struct PDFString : public PDFValue
 113 {
 114     rtl::OString  m_aString;
 115
 116     PDFString( const rtl::OString& rString )
 117     : PDFValue(), m_aString( rString ) {}
 118     virtual ~PDFString();
 119     virtual bool emit( EmitContext& rWriteContext ) const;
 120     virtual PDFEntry* clone() const;
 121
 122     rtl::OString getFilteredString() const;
 123 };
 124
 125 struct PDFNumber : public PDFValue
 126 {
 127     double m_fValue;
 128
 129     PDFNumber( double fVal )
 130     : PDFValue(), m_fValue( fVal ) {}
 131     virtual ~PDFNumber();
 132     virtual bool emit( EmitContext& rWriteContext ) const;
 133     virtual PDFEntry* clone() const;
 134 };
 135
 136 struct PDFBool : public PDFValue
 137 {
 138     bool m_bValue;
 139
 140     PDFBool( bool bVal )
 141     : PDFValue(), m_bValue( bVal ) {}
 142     virtual ~PDFBool();
 143     virtual bool emit( EmitContext& rWriteContext ) const;
 144     virtual PDFEntry* clone() const;
 145 };
 146
 147 struct PDFObjectRef : public PDFValue
 148 {
 149     unsigned int    m_nNumber;
 150     unsigned int    m_nGeneration;
 151
 152     PDFObjectRef( unsigned int nNr, unsigned int nGen )
 153     : PDFValue(), m_nNumber( nNr ), m_nGeneration( nGen ) {}
 154     virtual ~PDFObjectRef();
 155     virtual bool emit( EmitContext& rWriteContext ) const;
 156     virtual PDFEntry* clone() const;
 157 };
 158
 159 struct PDFNull : public PDFValue
 160 {
 161     PDFNull() {}
 162     virtual ~PDFNull();
 163     virtual bool emit( EmitContext& rWriteContext ) const;
 164     virtual PDFEntry* clone() const;
 165 };
 166
 167 struct PDFObject;
 168 struct PDFContainer : public PDFEntry
 169 {
 170     sal_Int32              m_nOffset;
 171     std::vector<PDFEntry*> m_aSubElements;
 172
 173     // this is an abstract base class for identifying
 174     // entries that can contain sub elements besides comments
 175     PDFContainer() : PDFEntry(), m_nOffset( 0 ) {}
 176     virtual ~PDFContainer();
 177     virtual bool emitSubElements( EmitContext& rWriteContext ) const;
 178     virtual void cloneSubElements( std::vector<PDFEntry*>& rNewSubElements ) const;
 179
 180     PDFObject* findObject( unsigned int nNumber, unsigned int nGeneration ) const;
 181     PDFObject* findObject( PDFObjectRef* pRef ) const
 182     { return findObject( pRef->m_nNumber, pRef->m_nGeneration ); }
 183 };
 184
 185 struct PDFArray : public PDFContainer
 186 {
 187     PDFArray() {}
 188     virtual ~PDFArray();
 189     virtual bool emit( EmitContext& rWriteContext ) const;
 190     virtual PDFEntry* clone() const;
 191 };
 192
 193 struct PDFDict : public PDFContainer
 194 {
 195     typedef std::hash_map<rtl::OString,PDFEntry*,rtl::OStringHash> Map;
 196     Map m_aMap;
 197
 198     PDFDict() {}
 199     virtual ~PDFDict();
 200     virtual bool emit( EmitContext& rWriteContext ) const;
 201     virtual PDFEntry* clone() const;
 202
 203     // inserting a value of NULL will remove rName and the previous value
 204     // from the dictionary
 205     void insertValue( const rtl::OString& rName, PDFEntry* pValue );
 206     // removes a name/value pair from the dict
 207     void eraseValue( const rtl::OString& rName );
 208     // builds new map as of sub elements
 209     // returns NULL if successfull, else the first offending element
 210     PDFEntry* buildMap();
 211 };
 212
 213 struct PDFStream : public PDFEntry
 214 {
 215     unsigned int    m_nBeginOffset;
 216     unsigned int    m_nEndOffset; // offset of the byte after the stream
 217     PDFDict*        m_pDict;
 218
 219     PDFStream( unsigned int nBegin, unsigned int nEnd, PDFDict* pStreamDict )
 220     : PDFEntry(), m_nBeginOffset( nBegin ), m_nEndOffset( nEnd ), m_pDict( pStreamDict ) {}
 221     virtual ~PDFStream();
 222     virtual bool emit( EmitContext& rWriteContext ) const;
 223     virtual PDFEntry* clone() const;
 224
 225     unsigned int getDictLength( const PDFContainer* pObjectContainer = NULL ) const; // get contents of the "Length" entry of the dict
 226 };
 227
 228 struct PDFTrailer : public PDFContainer
 229 {
 230     PDFDict*        m_pDict;
 231
 232     PDFTrailer() : PDFContainer(), m_pDict( NULL ) {}
 233     virtual ~PDFTrailer();
 234     virtual bool emit( EmitContext& rWriteContext ) const;
 235     virtual PDFEntry* clone() const;
 236 };
 237
 238 struct PDFFileImplData;
 239 struct PDFFile : public PDFContainer
 240 {
 241     private:
 242     mutable PDFFileImplData*    m_pData;
 243     PDFFileImplData*            impl_getData() const;
 244     public:
 245     unsigned int        m_nMajor;           // PDF major
 246     unsigned int        m_nMinor;           // PDF minor
 247
 248     PDFFile()
 249     : PDFContainer(),
 250       m_pData( NULL ),
 251       m_nMajor( 0 ), m_nMinor( 0 )
 252     {}
 253     virtual ~PDFFile();
 254
 255     virtual bool emit( EmitContext& rWriteContext ) const;
 256     virtual PDFEntry* clone() const;
 257
 258     bool isEncrypted() const;
 259     // this method checks whether rPwd is compatible with
 260     // either user or owner password and sets up decrypt data in that case
 261     // returns true if decryption can be done
 262     bool setupDecryptionData( const rtl::OString& rPwd ) const;
 263
 264     bool decrypt( const sal_uInt8* pInBuffer, sal_uInt32 nLen,
 265                   sal_uInt8* pOutBuffer,
 266                   unsigned int nObject, unsigned int nGeneration ) const;
 267 };
 268
 269 struct PDFObject : public PDFContainer
 270 {
 271     PDFEntry*       m_pObject;
 272     PDFStream*      m_pStream;
 273     unsigned int    m_nNumber;
 274     unsigned int    m_nGeneration;
 275
 276     PDFObject( unsigned int nNr, unsigned int nGen )
 277     : m_pObject( NULL ), m_pStream( NULL ), m_nNumber( nNr ), m_nGeneration( nGen ) {}
 278     virtual ~PDFObject();
 279     virtual bool emit( EmitContext& rWriteContext ) const;
 280     virtual PDFEntry* clone() const;
 281
 282     // writes only the contained stream, deflated if necessary
 283     bool writeStream( EmitContext& rContext, const PDFFile* pPDFFile ) const;
 284
 285     private:
 286     // returns true if stream is deflated
 287     // fills *ppStream and *pBytes with start of stream and count of bytes
 288     // memory returned in *ppStream must be freed with rtl_freeMemory afterwards
 289     // fills in NULL and 0 in case of error
 290     bool getDeflatedStream( char** ppStream, unsigned int* pBytes, const PDFContainer* pObjectContainer, EmitContext& rContext ) const;
 291 };
 292
 293 struct PDFPart : public PDFContainer
 294 {
 295     PDFPart() : PDFContainer() {}
 296     virtual ~PDFPart();
 297     virtual bool emit( EmitContext& rWriteContext ) const;
 298     virtual PDFEntry* clone() const;
 299 };
 300
 301 class PDFReader
 302 {
 303     public:
 304     PDFReader() {}
 305     ~PDFReader() {}
 306
 307     PDFEntry* read( const char* pFileName );
 308     PDFEntry* read( const char* pBuffer, unsigned int nLen );
 309 };
 310
 311 } // namespace
 312
 313 #endif