1
0
mirror of https://github.com/mariadb-corporation/mariadb-columnstore-engine.git synced 2025-07-30 19:23:07 +03:00

MCOL-987 Add LZ4 compression.

* Adds CompressInterfaceLZ4 which uses LZ4 API for compress/uncompress.
* Adds CMake machinery to search LZ4 on running host.
* All methods which use static data and do not modify any internal data - become `static`,
  so we can use them without creation of the specific object. This is possible, because
  the header specification has not been modified. We still use 2 sections in header, first
  one with file meta data, the second one with pointers for compressed chunks.
* Methods `compress`, `uncompress`, `maxCompressedSize`, `getUncompressedSize` - become
  pure virtual, so we can override them for the other compression algos.
* Adds method `getChunkMagicNumber`, so we can verify chunk magic number
  for each compression algo.
* Renames "s/IDBCompressInterface/CompressInterface/g" according to requirement.
This commit is contained in:
Denis Khalikov
2021-04-01 17:26:38 +03:00
parent dd12bd3cd0
commit cc1c3629c5
45 changed files with 1311 additions and 549 deletions

View File

@ -89,7 +89,7 @@ int32_t EMReBuilder::collectExtent(const std::string& fullFileName)
}
// Read and verify header.
char fileHeader[compress::IDBCompressInterface::HDR_BUF_LEN * 2];
char fileHeader[compress::CompressInterface::HDR_BUF_LEN * 2];
rc = fileOp.readHeaders(dbFile.get(), fileHeader);
if (rc != 0)
{
@ -116,8 +116,8 @@ int32_t EMReBuilder::collectExtent(const std::string& fullFileName)
}
// Read the `colDataType` and `colWidth` from the given header.
compress::IDBCompressInterface compressor;
const auto versionNumber = compressor.getVersionNumber(fileHeader);
const auto versionNumber =
compress::CompressInterface::getVersionNumber(fileHeader);
// Verify header number.
if (versionNumber < 3)
{
@ -129,10 +129,11 @@ int32_t EMReBuilder::collectExtent(const std::string& fullFileName)
return -1;
}
auto colDataType = compressor.getColDataType(fileHeader);
auto colWidth = compressor.getColumnWidth(fileHeader);
auto blockCount = compressor.getBlockCount(fileHeader);
auto lbidCount = compressor.getLBIDCount(fileHeader);
auto colDataType = compress::CompressInterface::getColDataType(fileHeader);
auto colWidth = compress::CompressInterface::getColumnWidth(fileHeader);
auto blockCount = compress::CompressInterface::getBlockCount(fileHeader);
auto lbidCount = compress::CompressInterface::getLBIDCount(fileHeader);
auto compressionType = compress::CompressInterface::getCompressionType(fileHeader);
if (colDataType == execplan::CalpontSystemCatalog::UNDEFINED)
{
@ -155,7 +156,7 @@ int32_t EMReBuilder::collectExtent(const std::string& fullFileName)
uint64_t hwm = 0;
rc = searchHWMInSegmentFile(oid, getDBRoot(), partition, segment, colDataType, colWidth,
blockCount, isDict, hwm);
blockCount, isDict, compressionType, hwm);
if (rc != 0)
{
return rc;
@ -172,13 +173,13 @@ int32_t EMReBuilder::collectExtent(const std::string& fullFileName)
{
for (uint32_t lbidIndex = 0; lbidIndex < lbidCount - 1; ++lbidIndex)
{
auto lbid = compressor.getLBIDByIndex(fileHeader, lbidIndex);
auto lbid = compress::CompressInterface::getLBIDByIndex(fileHeader, lbidIndex);
FileId fileId(oid, partition, segment, colWidth, colDataType, lbid, /*hwm*/ 0, isDict);
extentMap.push_back(fileId);
}
// Last one has an actual HWM.
auto lbid = compressor.getLBIDByIndex(fileHeader, lbidCount - 1);
auto lbid = compress::CompressInterface::getLBIDByIndex(fileHeader, lbidCount - 1);
FileId fileId(oid, partition, segment, colWidth, colDataType, lbid, hwm, isDict);
extentMap.push_back(fileId);
@ -192,7 +193,7 @@ int32_t EMReBuilder::collectExtent(const std::string& fullFileName)
else
{
// One extent per segment file.
auto lbid = compressor.getLBIDByIndex(fileHeader, 0);
auto lbid = compress::CompressInterface::getLBIDByIndex(fileHeader, 0);
FileId fileId(oid, partition, segment, colWidth, colDataType, lbid, hwm, isDict);
extentMap.push_back(fileId);
@ -293,7 +294,7 @@ int32_t EMReBuilder::rebuildExtentMap()
int32_t EMReBuilder::searchHWMInSegmentFile(
uint32_t oid, uint32_t dbRoot, uint32_t partition, uint32_t segment,
execplan::CalpontSystemCatalog::ColDataType colDataType, uint32_t colWidth,
uint64_t blockCount, bool isDict, uint64_t& hwm)
uint64_t blockCount, bool isDict, uint32_t compressionType, uint64_t& hwm)
{
std::unique_ptr<ChunkManagerWrapper> chunkManagerWrapper;
try
@ -302,13 +303,15 @@ int32_t EMReBuilder::searchHWMInSegmentFile(
{
chunkManagerWrapper = std::unique_ptr<ChunkManagerWrapperDict>(
new ChunkManagerWrapperDict(oid, dbRoot, partition, segment,
colDataType, colWidth));
colDataType, colWidth,
compressionType));
}
else
{
chunkManagerWrapper = std::unique_ptr<ChunkManagerWrapperColumn>(
new ChunkManagerWrapperColumn(oid, dbRoot, partition, segment,
colDataType, colWidth));
colDataType, colWidth,
compressionType));
}
}
catch (...)
@ -401,12 +404,13 @@ int32_t ChunkManagerWrapper::readBlock(uint32_t blockNumber)
ChunkManagerWrapperColumn::ChunkManagerWrapperColumn(
uint32_t oid, uint32_t dbRoot, uint32_t partition, uint32_t segment,
execplan::CalpontSystemCatalog::ColDataType colDataType, uint32_t colWidth)
execplan::CalpontSystemCatalog::ColDataType colDataType, uint32_t colWidth,
uint32_t compressionType)
: ChunkManagerWrapper(oid, dbRoot, partition, segment, colDataType,
colWidth)
{
pFileOp = std::unique_ptr<WriteEngine::ColumnOpCompress1>(
new WriteEngine::ColumnOpCompress1());
new WriteEngine::ColumnOpCompress1(compressionType));
chunkManager.fileOp(pFileOp.get());
// Open compressed column segment file. We will read block by block
// from the compressed chunks.
@ -463,12 +467,13 @@ bool ChunkManagerWrapperColumn::isEmptyValue(const uint8_t* value) const
ChunkManagerWrapperDict::ChunkManagerWrapperDict(
uint32_t oid, uint32_t dbRoot, uint32_t partition, uint32_t segment,
execplan::CalpontSystemCatalog::ColDataType colDataType, uint32_t colWidth)
execplan::CalpontSystemCatalog::ColDataType colDataType, uint32_t colWidth,
uint32_t compressionType)
: ChunkManagerWrapper(oid, dbRoot, partition, segment, colDataType,
colWidth)
{
pFileOp = std::unique_ptr<WriteEngine::DctnryCompress1>(
new WriteEngine::DctnryCompress1());
new WriteEngine::DctnryCompress1(compressionType));
chunkManager.fileOp(pFileOp.get());
// Open compressed dict segment file.
pFile = chunkManager.getSegmentFilePtr(oid, dbRoot, partition, segment,