CompressedVectorWriterImpl.cpp

Go to the documentation of this file.

/*
 * Original work Copyright 2009 - 2010 Kevin Ackley (kackley@gwi.net)
 * Modified work Copyright 2018 - 2020 Andy Maloney <asmaloney@gmail.com>
 *
 * Permission is hereby granted, free of charge, to any person or organization
 * obtaining a copy of the software and accompanying documentation covered by
 * this license (the "Software") to use, reproduce, display, distribute,
 * execute, and transmit the Software, and to prepare derivative works of the
 * Software, and to permit third-parties to whom the Software is furnished to
 * do so, all subject to the following:
 *
 * The copyright notices in the Software and this entire statement, including
 * the above license grant, this restriction and the following disclaimer,
 * must be included in all copies of the Software, in whole or in part, and
 * all derivative works of the Software, unless such copies or derivative
 * works are solely in the form of machine-executable object code generated by
 * a source language processor.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE, TITLE AND NON-INFRINGEMENT. IN NO EVENT
 * SHALL THE COPYRIGHT HOLDERS OR ANYONE DISTRIBUTING THE SOFTWARE BE LIABLE
 * FOR ANY DAMAGES OR OTHER LIABILITY, WHETHER IN CONTRACT, TORT OR OTHERWISE,
 * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
 * DEALINGS IN THE SOFTWARE.
 */
 
#include <cmath>
#include <numeric>
 
#include "CheckedFile.h"
#include "CompressedVectorNodeImpl.h"
#include "CompressedVectorWriterImpl.h"
#include "ImageFileImpl.h"
#include "SectionHeaders.h"
#include "SourceDestBufferImpl.h"
 
namespace e57
{
   struct SortByBytestreamNumber
   {
      bool operator()( const std::shared_ptr<Encoder> &lhs, const std::shared_ptr<Encoder> &rhs ) const
      {
         return ( lhs->bytestreamNumber() < rhs->bytestreamNumber() );
      }
   };
 
   CompressedVectorWriterImpl::CompressedVectorWriterImpl( std::shared_ptr<CompressedVectorNodeImpl> ni,
                                                           std::vector<SourceDestBuffer> &sbufs ) :
      cVector_( ni ), isOpen_( false ) // set to true when succeed below
   {
      //???  check if cvector already been written (can't write twice)
 
      if ( sbufs.empty() )
      {
         throw E57_EXCEPTION2( E57_ERROR_BAD_API_ARGUMENT,
                               "imageFileName=" + cVector_->imageFileName() + " cvPathName=" + cVector_->pathName() );
      }
 
      proto_ = cVector_->getPrototype();
 
      setBuffers( sbufs ); //??? copy code here?
 
      for ( unsigned i = 0; i < sbufs_.size(); i++ )
      {
         std::vector<SourceDestBuffer> vTemp;
         vTemp.push_back( sbufs_.at( i ) );
 
         ustring codecPath = sbufs_.at( i ).pathName();
 
         NodeImplSharedPtr readNode = proto_->get( sbufs.at( i ).pathName() );
         uint64_t bytestreamNumber = 0;
         if ( !proto_->findTerminalPosition( readNode, bytestreamNumber ) )
         {
            throw E57_EXCEPTION2( E57_ERROR_INTERNAL, "sbufIndex=" + toString( i ) );
         }
 
         bytestreams_.push_back(
            Encoder::EncoderFactory( static_cast<unsigned>( bytestreamNumber ), cVector_, vTemp, codecPath ) );
      }
 
      sort( bytestreams_.begin(), bytestreams_.end(), SortByBytestreamNumber() );
#ifdef E57_MAX_DEBUG
      for ( unsigned i = 0; i < bytestreams_.size(); i++ )
      {
         if ( bytestreams_.at( i )->bytestreamNumber() != i )
         {
            throw E57_EXCEPTION2( E57_ERROR_INTERNAL, "bytestreamIndex=" + toString( i ) + " bytestreamNumber=" +
                                                         toString( bytestreams_.at( i )->bytestreamNumber() ) );
         }
      }
#endif
 
      ImageFileImplSharedPtr imf( ni->destImageFile_ );
 
      sectionHeaderLogicalStart_ = imf->allocateSpace( sizeof( CompressedVectorSectionHeader ), true );
 
      sectionLogicalLength_ = 0;
      dataPhysicalOffset_ = 0;
      topIndexPhysicalOffset_ = 0;
      recordCount_ = 0;
      dataPacketsCount_ = 0;
      indexPacketsCount_ = 0;
 
      imf->incrWriterCount();
 
      isOpen_ = true;
   }
 
   CompressedVectorWriterImpl::~CompressedVectorWriterImpl()
   {
#ifdef E57_MAX_VERBOSE
      std::cout << "~CompressedVectorWriterImpl() called" << std::endl; //???
#endif
 
      try
      {
         if ( isOpen_ )
         {
            close();
         }
      }
      catch ( ... )
      {
         //??? report?
      }
   }
 
   void CompressedVectorWriterImpl::close()
   {
#ifdef E57_MAX_VERBOSE
      std::cout << "CompressedVectorWriterImpl::close() called" << std::endl; //???
#endif
      ImageFileImplSharedPtr imf( cVector_->destImageFile_ );
 
      imf->decrWriterCount();
 
      checkImageFileOpen( __FILE__, __LINE__, static_cast<const char *>( __FUNCTION__ ) );
 
      if ( !isOpen_ )
      {
         return;
      }
 
      isOpen_ = false;
 
      flush();
      while ( totalOutputAvailable() > 0 )
      {
         packetWrite();
         flush();
      }
 
      sectionLogicalLength_ = imf->unusedLogicalStart_ - sectionHeaderLogicalStart_;
#ifdef E57_MAX_VERBOSE
      std::cout << "  sectionLogicalLength_=" << sectionLogicalLength_ << std::endl; //???
#endif
 
      CompressedVectorSectionHeader header;
      header.sectionLogicalLength = sectionLogicalLength_;
      header.dataPhysicalOffset = dataPhysicalOffset_;      
      header.indexPhysicalOffset = topIndexPhysicalOffset_; 
#ifdef E57_MAX_VERBOSE
      std::cout << "  CompressedVectorSectionHeader:" << std::endl;
      header.dump( 4 ); //???
#endif
#ifdef E57_DEBUG
      header.verify( imf->file_->length( CheckedFile::Physical ) );
#endif
 
      imf->file_->seek( sectionHeaderLogicalStart_ );
      imf->file_->write( reinterpret_cast<char *>( &header ), sizeof( header ) );
 
      cVector_->setRecordCount( recordCount_ );
      cVector_->setBinarySectionLogicalStart( sectionHeaderLogicalStart_ );
 
      bytestreams_.clear();
 
#ifdef E57_MAX_VERBOSE
      std::cout << "  CompressedVectorWriter:" << std::endl;
      dump( 4 );
#endif
   }
 
   bool CompressedVectorWriterImpl::isOpen() const
   {
      return isOpen_;
   }
 
   std::shared_ptr<CompressedVectorNodeImpl> CompressedVectorWriterImpl::compressedVectorNode() const
   {
      return cVector_;
   }
 
   void CompressedVectorWriterImpl::setBuffers( std::vector<SourceDestBuffer> &sbufs )
   {
 
      if ( !sbufs_.empty() )
      {
         if ( sbufs_.size() != sbufs.size() )
         {
            throw E57_EXCEPTION2( E57_ERROR_BUFFERS_NOT_COMPATIBLE,
                                  "oldSize=" + toString( sbufs_.size() ) + " newSize=" + toString( sbufs.size() ) );
         }
 
         for ( size_t i = 0; i < sbufs_.size(); ++i )
         {
            std::shared_ptr<SourceDestBufferImpl> oldbuf = sbufs_[i].impl();
            std::shared_ptr<SourceDestBufferImpl> newBuf = sbufs[i].impl();
 
            oldbuf->checkCompatible( newBuf );
         }
      }
 
      proto_->checkBuffers( sbufs, false );
 
      sbufs_ = sbufs;
   }
 
   void CompressedVectorWriterImpl::write( std::vector<SourceDestBuffer> &sbufs, const size_t requestedRecordCount )
   {
 
      setBuffers( sbufs );
      write( requestedRecordCount );
   }
 
   void CompressedVectorWriterImpl::write( const size_t requestedRecordCount )
   {
#ifdef E57_MAX_VERBOSE
      std::cout << "CompressedVectorWriterImpl::write() called" << std::endl; //???
#endif
      checkImageFileOpen( __FILE__, __LINE__, static_cast<const char *>( __FUNCTION__ ) );
      checkWriterOpen( __FILE__, __LINE__, static_cast<const char *>( __FUNCTION__ ) );
 
      if ( requestedRecordCount > sbufs_.at( 0 ).impl()->capacity() )
      {
         throw E57_EXCEPTION2( E57_ERROR_BAD_API_ARGUMENT,
                               "requested=" + toString( requestedRecordCount ) +
                                  " capacity=" + toString( sbufs_.at( 0 ).impl()->capacity() ) + " imageFileName=" +
                                  cVector_->imageFileName() + " cvPathName=" + cVector_->pathName() );
      }
 
      for ( auto &sbuf : sbufs_ )
      {
         sbuf.impl()->rewind();
      }
 
      uint64_t endRecordIndex = recordCount_ + requestedRecordCount;
      while ( true )
      {
         uint64_t totalRecordCount = 0;
         for ( auto &bytestream : bytestreams_ )
         {
            totalRecordCount += endRecordIndex - bytestream->currentRecordIndex();
         }
#ifdef E57_MAX_VERBOSE
         std::cout << "  totalRecordCount=" << totalRecordCount << std::endl; //???
#endif
 
         if ( totalRecordCount == 0 )
         {
            break;
         }
 
 
#ifdef E57_MAX_VERBOSE
         std::cout << "  currentPacketSize()=" << currentPacketSize() << std::endl; //???
#endif
 
#ifdef E57_WRITE_CRAZY_PACKET_MODE
         constexpr size_t E57_TARGET_PACKET_SIZE = 500;
#else
         constexpr size_t E57_TARGET_PACKET_SIZE = ( DATA_PACKET_MAX * 3 / 4 );
#endif
         if ( currentPacketSize() >= E57_TARGET_PACKET_SIZE )
         { //???
            packetWrite();
            continue; 
         }
 
         float totalBitsPerRecord = 0; // an estimate of future performance
         for ( auto &bytestream : bytestreams_ )
         {
            totalBitsPerRecord += bytestream->bitsPerRecord();
         }
 
#ifdef E57_MAX_VERBOSE
         const float totalBytesPerRecord = std::max( totalBitsPerRecord / 8, 0.1F ); //??? trust
 
         std::cout << "  totalBytesPerRecord=" << totalBytesPerRecord << std::endl; //???
#endif
 
 
         for ( auto &bytestream : bytestreams_ )
         {
            if ( bytestream->currentRecordIndex() < endRecordIndex )
            {
               uint64_t recordCount = endRecordIndex - bytestream->currentRecordIndex();
               recordCount = ( recordCount < 50ULL ) ? recordCount : 50ULL; // min(recordCount, 50ULL);
               bytestream->processRecords( static_cast<unsigned>( recordCount ) );
            }
         }
      }
 
      recordCount_ += requestedRecordCount;
 
   }
 
   size_t CompressedVectorWriterImpl::totalOutputAvailable() const
   {
      size_t total = 0;
 
      for ( const auto &bytestream : bytestreams_ )
      {
         total += bytestream->outputAvailable();
      }
 
      return total;
   }
 
   size_t CompressedVectorWriterImpl::currentPacketSize() const
   {
      return ( sizeof( DataPacketHeader ) + bytestreams_.size() * sizeof( uint16_t ) + totalOutputAvailable() );
   }
 
   uint64_t CompressedVectorWriterImpl::packetWrite()
   {
#ifdef E57_MAX_VERBOSE
      std::cout << "CompressedVectorWriterImpl::packetWrite() called" << std::endl; //???
#endif
 
      size_t totalOutput = totalOutputAvailable();
      if ( totalOutput == 0 )
      {
         return ( 0 );
      }
#ifdef E57_MAX_VERBOSE
      std::cout << "  totalOutput=" << totalOutput << std::endl; //???
#endif
 
      size_t packetMaxPayloadBytes =
         DATA_PACKET_MAX - sizeof( DataPacketHeader ) - bytestreams_.size() * sizeof( uint16_t );
#ifdef E57_MAX_VERBOSE
      std::cout << "  packetMaxPayloadBytes=" << packetMaxPayloadBytes << std::endl; //???
#endif
 
      std::vector<size_t> count( bytestreams_.size() );
 
      if ( totalOutput < packetMaxPayloadBytes )
      {
         for ( unsigned i = 0; i < bytestreams_.size(); i++ )
         {
            count.at( i ) = bytestreams_.at( i )->outputAvailable();
         }
      }
      else
      {
         float fractionToSend = ( packetMaxPayloadBytes - 1 ) / static_cast<float>( totalOutput );
         for ( unsigned i = 0; i < bytestreams_.size(); i++ )
         {
            count.at( i ) =
               static_cast<unsigned>( std::floor( fractionToSend * bytestreams_.at( i )->outputAvailable() ) );
         }
      }
#ifdef E57_MAX_VERBOSE
      for ( unsigned i = 0; i < bytestreams_.size(); i++ )
      {
         std::cout << "  count[" << i << "]=" << count.at( i ) << std::endl; //???
      }
#endif
 
#ifdef E57_DEBUG
      const size_t totalByteCount = std::accumulate( count.begin(), count.end(), 0 );
 
      if ( totalByteCount > packetMaxPayloadBytes )
      {
         throw E57_EXCEPTION2( E57_ERROR_INTERNAL, "totalByteCount=" + toString( totalByteCount ) +
                                                      " packetMaxPayloadBytes=" + toString( packetMaxPayloadBytes ) );
      }
#endif
 
      ImageFileImplSharedPtr imf( cVector_->destImageFile_ );
 
      char *packet = reinterpret_cast<char *>( &dataPacket_ );
#ifdef E57_MAX_VERBOSE
      std::cout << "  packet=" << packet << std::endl; //???
#endif
 
      dataPacket_.header.reset();
 
      auto bsbLength = reinterpret_cast<uint16_t *>( &packet[sizeof( DataPacketHeader )] );
#ifdef E57_MAX_VERBOSE
      std::cout << "  bsbLength=" << bsbLength << std::endl; //???
#endif
      for ( unsigned i = 0; i < bytestreams_.size(); i++ )
      {
         bsbLength[i] = static_cast<uint16_t>( count.at( i ) ); // %%% Truncation
#ifdef E57_MAX_VERBOSE
         std::cout << "  Writing " << bsbLength[i] << " bytes into bytestream " << i << std::endl; //???
#endif
      }
 
      char *p = reinterpret_cast<char *>( &bsbLength[bytestreams_.size()] );
#ifdef E57_MAX_VERBOSE
      std::cout << "  after bsbLength, p=" << p << std::endl; //???
#endif
 
      for ( size_t i = 0; i < bytestreams_.size(); i++ )
      {
         size_t n = count.at( i );
 
#ifdef E57_DEBUG
         if ( &p[n] > &packet[DATA_PACKET_MAX] )
         {
            throw E57_EXCEPTION2( E57_ERROR_INTERNAL, "n=" + toString( n ) );
         }
#endif
 
         bytestreams_.at( i )->outputRead( p, n );
 
         p += n;
      }
 
      auto packetLength = static_cast<unsigned>( p - packet ); 
#ifdef E57_MAX_VERBOSE
      std::cout << "  packetLength=" << packetLength << std::endl; //???
#endif
 
#ifdef E57_DEBUG
      if ( packetLength != sizeof( DataPacketHeader ) + bytestreams_.size() * sizeof( uint16_t ) + totalByteCount )
      {
         throw E57_EXCEPTION2( E57_ERROR_INTERNAL, "packetLength=" + toString( packetLength ) + " bytestreamSize=" +
                                                      toString( bytestreams_.size() * sizeof( uint16_t ) ) +
                                                      " totalByteCount=" + toString( totalByteCount ) );
      }
#endif
 
      while ( packetLength % 4 )
      {
         if ( p >= &packet[DATA_PACKET_MAX - 1] )
         {
            throw E57_EXCEPTION1( E57_ERROR_INTERNAL );
         }
         *p++ = 0;
         packetLength++;
#ifdef E57_MAX_VERBOSE
         std::cout << "  padding with zero byte, new packetLength=" << packetLength << std::endl; //???
#endif
      }
 
      dataPacket_.header.packetLogicalLengthMinus1 = static_cast<uint16_t>( packetLength - 1 ); // %%% Truncation
      dataPacket_.header.bytestreamCount = static_cast<uint16_t>( bytestreams_.size() );        // %%% Truncation
 
      dataPacket_.verify( packetLength );
 
      uint64_t packetLogicalOffset = imf->allocateSpace( packetLength, false );
      uint64_t packetPhysicalOffset = imf->file_->logicalToPhysical( packetLogicalOffset );
      imf->file_->seek( packetLogicalOffset ); //??? have seekLogical and seekPhysical instead?
                                               // more explicit
      imf->file_->write( packet, packetLength );
 
#ifdef E57_MAX_VERBOSE
//  std::cout << "data packet:" << std::endl;
//  dataPacket_.dump(4);
#endif
 
      if ( dataPacketsCount_ == 0 )
      {
         dataPhysicalOffset_ = packetPhysicalOffset;
      }
      dataPacketsCount_++;
 
 
      return ( packetPhysicalOffset ); //??? needed
   }
 
   void CompressedVectorWriterImpl::flush()
   {
      for ( auto &bytestream : bytestreams_ )
      {
         bytestream->registerFlushToOutput();
      }
   }
 
   void CompressedVectorWriterImpl::checkImageFileOpen( const char *srcFileName, int srcLineNumber,
                                                        const char *srcFunctionName ) const
   {
      // unimplemented...
   }
 
   void CompressedVectorWriterImpl::checkWriterOpen( const char *srcFileName, int srcLineNumber,
                                                     const char *srcFunctionName ) const
   {
      if ( !isOpen_ )
      {
         throw E57Exception( E57_ERROR_WRITER_NOT_OPEN,
                             "imageFileName=" + cVector_->imageFileName() + " cvPathName=" + cVector_->pathName(),
                             srcFileName, srcLineNumber, srcFunctionName );
      }
   }
 
   void CompressedVectorWriterImpl::dump( int indent, std::ostream &os )
   {
      os << space( indent ) << "isOpen:" << isOpen_ << std::endl;
 
      for ( unsigned i = 0; i < sbufs_.size(); i++ )
      {
         os << space( indent ) << "sbufs[" << i << "]:" << std::endl;
         sbufs_.at( i ).dump( indent + 4, os );
      }
 
      os << space( indent ) << "cVector:" << std::endl;
      cVector_->dump( indent + 4, os );
 
      os << space( indent ) << "proto:" << std::endl;
      proto_->dump( indent + 4, os );
 
      for ( unsigned i = 0; i < bytestreams_.size(); i++ )
      {
         os << space( indent ) << "bytestreams[" << i << "]:" << std::endl;
         bytestreams_.at( i )->dump( indent + 4, os );
      }
 
      os << space( indent ) << "dataPacket:" << std::endl;
      auto p = reinterpret_cast<uint8_t *>( &dataPacket_ );
 
      for ( unsigned i = 0; i < 40; ++i )
      {
         os << space( indent + 4 ) << "dataPacket[" << i << "]: " << static_cast<unsigned>( p[i] ) << std::endl;
      }
      os << space( indent + 4 ) << "more unprinted..." << std::endl;
 
      os << space( indent ) << "sectionHeaderLogicalStart: " << sectionHeaderLogicalStart_ << std::endl;
      os << space( indent ) << "sectionLogicalLength:      " << sectionLogicalLength_ << std::endl;
      os << space( indent ) << "dataPhysicalOffset:        " << dataPhysicalOffset_ << std::endl;
      os << space( indent ) << "topIndexPhysicalOffset:    " << topIndexPhysicalOffset_ << std::endl;
      os << space( indent ) << "recordCount:               " << recordCount_ << std::endl;
      os << space( indent ) << "dataPacketsCount:          " << dataPacketsCount_ << std::endl;
      os << space( indent ) << "indexPacketsCount:         " << indexPacketsCount_ << std::endl;
   }
}