Update Kuzu from branch master (#156)

Co-authored-by: mewim <14037726+mewim@users.noreply.github.com>

Author: mewim

Sources/cxx-kuzu/kuzu/src/storage/table/column_chunk_data.cpp

@@ -334,9 +334,11 @@ void ColumnChunkData::setToOnDisk(const ColumnChunkMetadata& otherMetadata) {
 
 ColumnChunkMetadata ColumnChunkData::flushBuffer(PageAllocator& pageAllocator,
  const PageRange& entry, const ColumnChunkMetadata& otherMetadata) const {
- if (!otherMetadata.compMeta.isConstant() && getBufferSize(numValues) != 0) {
- return flushBufferFunction(buffer->getBuffer(), pageAllocator.getDataFH(), entry,
- otherMetadata);
+ const auto bufferSizeToFlush = getBufferSize(numValues);
+ if (!otherMetadata.compMeta.isConstant() && bufferSizeToFlush != 0) {
+ KU_ASSERT(bufferSizeToFlush <= buffer->getBuffer().size_bytes());
+ const auto bufferToFlush = buffer->getBuffer().subspan(0, bufferSizeToFlush);
+ return flushBufferFunction(bufferToFlush, pageAllocator.getDataFH(), entry, otherMetadata);
  }
  KU_ASSERT(otherMetadata.getNumPages() == 0);
  return otherMetadata;

Sources/cxx-kuzu/kuzu/src/storage/table/compression_flush_buffer.cpp

@@ -15,6 +15,7 @@ using namespace transaction;
 ColumnChunkMetadata uncompressedFlushBuffer(std::span<const uint8_t> buffer, FileHandle* dataFH,
  const PageRange& entry, const ColumnChunkMetadata& metadata) {
  KU_ASSERT(dataFH->getNumPages() >= entry.startPageIdx + entry.numPages);
+ KU_ASSERT(buffer.size_bytes() <= entry.numPages * KUZU_PAGE_SIZE);
  dataFH->writePagesToFile(buffer.data(), buffer.size(), entry.startPageIdx);
  return ColumnChunkMetadata(entry.startPageIdx, entry.numPages, metadata.numValues,
  metadata.compMeta);

Sources/cxx-kuzu/kuzu/src/storage/table/string_column.cpp

@@ -149,13 +149,14 @@ void StringColumn::scanSegment(const SegmentState& state, ColumnChunkData* resul
  auto* indexChunk = stringResultChunk->getIndexColumnChunk();
  indexColumn->scanSegment(getChildState(state, ChildStateIndex::INDEX), indexChunk,
  startOffsetInSegment, numValuesToScan);
+
+ const auto initialDictSize =
+ stringResultChunk->getDictionaryChunk().getOffsetChunk()->getNumValues();
  if (numValuesToScan == state.metadata.numValues) {
  // Append the entire dictionary into the chunk
  // Since the resultChunk may be non-empty, each index needs to be incremented by the initial
  // size of the dictionary so that the indices line up with the values that will be scanned
  // into the dictionary chunk
- auto initialDictSize =
- stringResultChunk->getDictionaryChunk().getOffsetChunk()->getNumValues();
  for (row_idx_t i = 0; i < numValuesToScan; i++) {
  indexChunk->setValue<string_index_t>(
  indexChunk->getValue<string_index_t>(startOffsetInResult + i) + initialDictSize,
@@ -172,11 +173,10 @@ void StringColumn::scanSegment(const SegmentState& state, ColumnChunkData* resul
  auto index = indexChunk->getValue<string_index_t>(startOffsetInResult + i);
  auto element = indexMap.find(index);
  if (element == indexMap.end()) {
- indexMap.insert(
- std::make_pair(index, startOffsetInResult + offsetsToScan.size()));
- indexChunk->setValue<string_index_t>(startOffsetInResult + offsetsToScan.size(),
+ indexMap.insert(std::make_pair(index, initialDictSize + offsetsToScan.size()));
+ indexChunk->setValue<string_index_t>(initialDictSize + offsetsToScan.size(),
  startOffsetInResult + i);
- offsetsToScan.emplace_back(index, startOffsetInResult + offsetsToScan.size());
+ offsetsToScan.emplace_back(index, initialDictSize + offsetsToScan.size());
  } else {
  indexChunk->setValue<string_index_t>(element->second, startOffsetInResult + i);
  }