A scanner for reading RCFiles into tuples. More...

#include <hdfs-rcfile-scanner.h>

Inheritance diagram for impala::HdfsRCFileScanner:

Collaboration diagram for impala::HdfsRCFileScanner:

Classes
struct	ColumnInfo

struct	RcFileHeader
	Data that is fixed across headers. This struct is shared between scan ranges. More...

Public Member Functions
	HdfsRCFileScanner (HdfsScanNode scan_node, RuntimeState state)

virtual	~HdfsRCFileScanner ()

virtual Status	Prepare (ScannerContext *context)
	One-time initialisation of state that is constant across scan ranges. More...

void	DebugString (int indentation_level, std::stringstream *out) const

virtual void	Close ()

virtual Status	ProcessSplit ()

Static Public Member Functions
static Status	IssueInitialRanges (HdfsScanNode scan_node, const std::vector< HdfsFileDesc > &files)
	Issue the initial ranges for all sequence container files. More...

Static Public Attributes
static const int	FILE_BLOCK_SIZE = 4096

static const char *	LLVM_CLASS_NAME = "class.impala::HdfsScanner"

Protected Types
typedef int(*	WriteTuplesFn )(HdfsScanner , MemPool , TupleRow , int, FieldLocation , int, int, int, int)

Protected Member Functions
Status	ReadSync ()

Status	SkipToSync (const uint8_t *sync, int sync_size)

bool	finished ()

Status	InitializeWriteTuplesFn (HdfsPartitionDescriptor *partition, THdfsFileFormat::type type, const std::string &scanner_name)

void	StartNewRowBatch ()
	Set batch_ to a new row batch and update tuple_mem_ accordingly. More...

int	GetMemory (MemPool pool, Tuple tuple_mem, TupleRow **tuple_row_mem)

Status	CommitRows (int num_rows)

void	AddFinalRowBatch ()

void	AttachPool (MemPool *pool, bool commit_batch)

bool IR_ALWAYS_INLINE	EvalConjuncts (TupleRow *row)

int	WriteEmptyTuples (RowBatch *row_batch, int num_tuples)

int	WriteEmptyTuples (ScannerContext context, TupleRow tuple_row, int num_tuples)
	Write empty tuples and commit them to the context object. More...

int	WriteAlignedTuples (MemPool pool, TupleRow tuple_row_mem, int row_size, FieldLocation *fields, int num_tuples, int max_added_tuples, int slots_per_tuple, int row_start_indx)

Status	UpdateDecompressor (const THdfsCompression::type &compression)

Status	UpdateDecompressor (const std::string &codec)

bool	ReportTupleParseError (FieldLocation fields, uint8_t errors, int row_idx)

virtual void	LogRowParseError (int row_idx, std::stringstream *)

bool	WriteCompleteTuple (MemPool pool, FieldLocation fields, Tuple tuple, TupleRow tuple_row, Tuple template_tuple, uint8_t error_fields, uint8_t *error_in_row)

void	ReportColumnParseError (const SlotDescriptor desc, const char data, int len)

void	InitTuple (Tuple template_tuple, Tuple tuple)

Tuple *	next_tuple (Tuple *t) const

TupleRow *	next_row (TupleRow *r) const

ExprContext *	GetConjunctCtx (int idx) const

Static Protected Member Functions
static llvm::Function *	CodegenWriteCompleteTuple (HdfsScanNode , LlvmCodeGen , const std::vector< ExprContext * > &conjunct_ctxs)

static llvm::Function *	CodegenWriteAlignedTuples (HdfsScanNode , LlvmCodeGen , llvm::Function *write_tuple_fn)

Protected Attributes
FileHeader *	header_
	File header for this scan range. This is not owned by the parent scan node. More...

bool	only_parsing_header_
	If true, this scanner object is only for processing the header. More...

HdfsScanNode *	scan_node_
	The scan node that started this scanner. More...

RuntimeState *	state_
	RuntimeState for error reporting. More...

ScannerContext *	context_
	Context for this scanner. More...

ScannerContext::Stream *	stream_
	The first stream for context_. More...

std::vector< ExprContext * >	conjunct_ctxs_

Tuple *	template_tuple_

int	tuple_byte_size_
	Fixed size of each tuple, in bytes. More...

Tuple *	tuple_
	Current tuple pointer into tuple_mem_. More...

RowBatch *	batch_

uint8_t *	tuple_mem_
	The tuple memory of batch_. More...

int	num_errors_in_file_
	number of errors in current file More...

boost::scoped_ptr< TextConverter >	text_converter_
	Helper class for converting text to other types;. More...

int32_t	num_null_bytes_
	Number of null bytes in the tuple. More...

Status	parse_status_

boost::scoped_ptr< Codec >	decompressor_
	Decompressor class to use, if any. More...

THdfsCompression::type	decompression_type_
	The most recently used decompression type. More...

boost::scoped_ptr< MemPool >	data_buffer_pool_

RuntimeProfile::Counter *	decompress_timer_
	Time spent decompressing bytes. More...

WriteTuplesFn	write_tuples_fn_
	Jitted write tuples function pointer. Null if codegen is disabled. More...

Static Protected Attributes
static const int	SYNC_HASH_SIZE = 16
	Size of the sync hash field. More...

static const int	HEADER_SIZE = 1024

static const int	SYNC_MARKER = -1
	Sync indicator. More...

Private Types
enum	Version { SEQ6, RCF1 }

Private Member Functions
virtual FileHeader *	AllocateFileHeader ()
	Implementation of superclass functions. More...

virtual Status	ReadFileHeader ()

virtual Status	InitNewRange ()
	Reset internal state for a new scan range. More...

virtual Status	ProcessRange ()

virtual THdfsFileFormat::type	file_format () const
	Returns type of scanner: e.g. rcfile, seqfile. More...

Status	ReadNumColumnsMetadata ()

Status	ReadRowGroupHeader ()

Status	ReadKeyBuffers ()

void	GetCurrentKeyBuffer (int col_idx, bool skip_col_data, uint8_t **key_buf_ptr)

Status	ReadColumnBuffers ()

Status	NextField (int col_idx)

Status	ReadRowGroup ()

void	ResetRowGroup ()
	Reset state for a new row group. More...

Status	NextRow ()

Private Attributes
std::vector< ColumnInfo >	columns_

std::vector< uint8_t >	key_buffer_
	Buffer for copying key buffers. This buffer is reused between row groups. More...

int	num_rows_
	number of rows in this rowgroup object More...

int	row_pos_

int	key_length_

int	compressed_key_length_

bool	reuse_row_group_buffer_

uint8_t *	row_group_buffer_

int	row_group_length_

int	row_group_buffer_size_

Static Private Attributes
static const char *const	RCFILE_KEY_CLASS_NAME

static const char *const	RCFILE_VALUE_CLASS_NAME

static const char *const	RCFILE_METADATA_KEY_NUM_COLS

static const uint8_t	RCFILE_VERSION_HEADER [4] = {'R', 'C', 'F', 1}

Detailed Description

A scanner for reading RCFiles into tuples.

Definition at line 231 of file hdfs-rcfile-scanner.h.

Member Typedef Documentation

typedef int(* impala::HdfsScanner::WriteTuplesFn)(HdfsScanner *, MemPool *, TupleRow *, int, FieldLocation *, int, int, int, int)

protectedinherited

Matching typedef for WriteAlignedTuples for codegen. Refer to comments for that function.

Definition at line 212 of file hdfs-scanner.h.

Member Enumeration Documentation

enum impala::HdfsRCFileScanner::Version

private

Enumerator
SEQ6
RCF1

Definition at line 328 of file hdfs-rcfile-scanner.h.

Constructor & Destructor Documentation

HdfsRCFileScanner::HdfsRCFileScanner	(	HdfsScanNode *	scan_node,
		RuntimeState *	state
	)

Definition at line 53 of file hdfs-rcfile-scanner.cc.

HdfsRCFileScanner::~HdfsRCFileScanner ( )

virtual

Definition at line 57 of file hdfs-rcfile-scanner.cc.

Member Function Documentation

void HdfsScanner::AddFinalRowBatch ( )

protectedinherited

Attach all remaining resources from context_ to batch_ and send batch_ to the scan node. This must be called after all rows have been committed and no further resources are needed from context_ (in practice this will happen in each scanner subclass's Close() implementation).

Definition at line 145 of file hdfs-scanner.cc.

References impala::HdfsScanNode::AddMaterializedRowBatch(), impala::HdfsScanner::batch_, impala::HdfsScanner::context_, impala::ScannerContext::ReleaseCompletedResources(), and impala::HdfsScanner::scan_node_.

Referenced by impala::HdfsTextScanner::Close(), impala::BaseSequenceScanner::Close(), and impala::HdfsParquetScanner::Close().

BaseSequenceScanner::FileHeader * HdfsRCFileScanner::AllocateFileHeader ( )

privatevirtual

Implementation of superclass functions.

Implements impala::BaseSequenceScanner.

Definition at line 227 of file hdfs-rcfile-scanner.cc.

void impala::HdfsScanner::AttachPool	(	MemPool *	pool,
		bool	commit_batch
	)

inlineprotectedinherited

Release all memory in 'pool' to batch_. If commit_batch is true, the row batch will be committed. commit_batch should be true if the attached pool is expected to be non-trivial (i.e. a decompression buffer) to minimize scanner mem usage.

Definition at line 256 of file hdfs-scanner.h.

References impala::MemPool::AcquireData(), impala::HdfsScanner::batch_, impala::HdfsScanner::CommitRows(), and impala::RowBatch::tuple_data_pool().

Referenced by impala::HdfsTextScanner::Close(), impala::BaseSequenceScanner::Close(), impala::HdfsParquetScanner::Close(), impala::HdfsTextScanner::FillByteBufferGzip(), impala::HdfsAvroScanner::ProcessRange(), impala::HdfsSequenceScanner::ReadCompressedBlock(), impala::HdfsParquetScanner::BaseColumnReader::ReadDataPage(), and ResetRowGroup().

void BaseSequenceScanner::Close ( )

virtualinherited

Release all resources the scanner has allocated. This is the last chance for the scanner to attach any resources to the ScannerContext object.

Reimplemented from impala::HdfsScanner.

Definition at line 82 of file base-sequence-scanner.cc.

References impala::HdfsScanner::AddFinalRowBatch(), impala::HdfsScanner::AttachPool(), impala::ScannerContext::Stream::bytes_left(), impala::HdfsScanner::Close(), impala::BaseSequenceScanner::FileHeader::compression_type, impala::HdfsScanner::data_buffer_pool_, impala::HdfsScanner::decompressor_, impala::BaseSequenceScanner::file_format(), impala::BaseSequenceScanner::header_, impala::BaseSequenceScanner::num_syncs_, impala::BaseSequenceScanner::only_parsing_header_, impala::HdfsScanNode::RangeComplete(), impala::HdfsScanner::scan_node_, impala::HdfsScanner::stream_, impala::BaseSequenceScanner::total_block_size_, and VLOG_FILE.

Function * HdfsScanner::CodegenWriteAlignedTuples	(	HdfsScanNode *	,
		LlvmCodeGen *	,
		llvm::Function *	write_tuple_fn
	)

staticprotectedinherited

Codegen function to replace WriteAlignedTuples. WriteAlignedTuples is cross compiled to IR. This function loads the precompiled IR function, modifies it and returns the resulting function.

Definition at line 495 of file hdfs-scanner.cc.

References impala::LlvmCodeGen::codegen_timer(), impala::LlvmCodeGen::FinalizeFunction(), impala::LlvmCodeGen::GetFunction(), impala::LlvmCodeGen::ReplaceCallSites(), and SCOPED_TIMER.

Referenced by impala::HdfsTextScanner::Codegen(), and impala::HdfsSequenceScanner::Codegen().

Function * HdfsScanner::CodegenWriteCompleteTuple	(	HdfsScanNode *	,
		LlvmCodeGen *	,
		const std::vector< ExprContext * > &	conjunct_ctxs
	)

staticprotectedinherited

Codegen function to replace WriteCompleteTuple. Should behave identically to WriteCompleteTuple.

Definition at line 296 of file hdfs-scanner.cc.

Referenced by impala::HdfsTextScanner::Codegen(), and impala::HdfsSequenceScanner::Codegen().

Status HdfsScanner::CommitRows ( int num_rows )

protectedinherited

Commit num_rows to the current row batch. If this completes, the row batch is enqueued with the scan node and StartNewRowBatch() is called. Returns Status::OK if the query is not cancelled and hasn't exceeded any mem limits. Scanner can call this with 0 rows to flush any pending resources (attached pools and io buffers) to minimize memory consumption.

Definition at line 124 of file hdfs-scanner.cc.

References impala::HdfsScanNode::AddMaterializedRowBatch(), impala::RowBatch::AtCapacity(), impala::HdfsScanner::batch_, impala::TupleDescriptor::byte_size(), impala::Status::CANCELLED, impala::ScannerContext::cancelled(), impala::RowBatch::capacity(), impala::RuntimeState::CheckQueryState(), impala::RowBatch::CommitRows(), impala::HdfsScanner::conjunct_ctxs_, impala::HdfsScanner::context_, impala::ExprContext::FreeLocalAllocations(), impala::ScannerContext::num_completed_io_buffers(), impala::RowBatch::num_rows(), impala::Status::OK, impala::ScannerContext::ReleaseCompletedResources(), RETURN_IF_ERROR, impala::HdfsScanner::scan_node_, impala::HdfsScanner::StartNewRowBatch(), impala::HdfsScanner::state_, impala::HdfsScanNode::tuple_desc(), and impala::HdfsScanner::tuple_mem_.

Referenced by impala::HdfsParquetScanner::AssembleRows(), impala::HdfsScanner::AttachPool(), impala::HdfsTextScanner::FinishScanRange(), impala::HdfsSequenceScanner::ProcessDecompressedBlock(), impala::HdfsParquetScanner::ProcessFooter(), impala::HdfsTextScanner::ProcessRange(), impala::HdfsAvroScanner::ProcessRange(), impala::HdfsSequenceScanner::ProcessRange(), ProcessRange(), and impala::HdfsParquetScanner::ProcessSplit().

void HdfsRCFileScanner::DebugString	(	int	indentation_level,
		std::stringstream *	out
	)		const

Definition at line 564 of file hdfs-rcfile-scanner.cc.

References impala::ScannerContext::Stream::filename(), impala::HdfsScanner::scan_node_, impala::HdfsScanner::stream_, and impala::HdfsScanNode::tuple_idx().

bool IR_ALWAYS_INLINE impala::HdfsScanner::EvalConjuncts ( TupleRow * row )

inlineprotectedinherited

Convenience function for evaluating conjuncts using this scanner's ExprContexts. This must always be inlined so we can correctly replace the call to ExecNode::EvalConjuncts() during codegen.

Definition at line 266 of file hdfs-scanner.h.

References impala::HdfsScanner::conjunct_ctxs_, and impala::ExecNode::EvalConjuncts().

Referenced by impala::HdfsParquetScanner::AssembleRows(), impala::HdfsAvroScanner::DecodeAvroData(), ProcessRange(), impala::HdfsScanner::WriteCompleteTuple(), impala::HdfsScanner::WriteEmptyTuples(), and impala::HdfsTextScanner::WriteFields().

virtual THdfsFileFormat::type impala::HdfsRCFileScanner::file_format ( ) const

inlineprivatevirtual

Returns type of scanner: e.g. rcfile, seqfile.

Implements impala::BaseSequenceScanner.

Definition at line 263 of file hdfs-rcfile-scanner.h.

bool impala::BaseSequenceScanner::finished ( )

inlineprotectedinherited

Definition at line 117 of file base-sequence-scanner.h.

References impala::BaseSequenceScanner::finished_.

Referenced by impala::HdfsSequenceScanner::ProcessBlockCompressedScanRange(), impala::HdfsAvroScanner::ProcessRange(), impala::HdfsSequenceScanner::ProcessRange(), and ProcessRange().

ExprContext * HdfsScanner::GetConjunctCtx ( int idx ) const

protectedinherited

Simple wrapper around conjunct_ctxs_. Used in the codegen'd version of WriteCompleteTuple() because it's easier than writing IR to access conjunct_ctxs_.

Definition at line 79 of file hdfs-scanner-ir.cc.

References impala::HdfsScanner::conjunct_ctxs_, and gen_ir_descriptions::idx.

void HdfsRCFileScanner::GetCurrentKeyBuffer	(	int	col_idx,
		bool	skip_col_data,
		uint8_t **	key_buf_ptr
	)

private

Process the current key buffer. Inputs: col_idx: column to process skip_col_data: if true, just skip over the key data. Input/Output: key_buf_ptr: Pointer to the buffered file data, this will be moved past the data for this column. Sets: col_buf_len_ col_buf_uncompressed_len_ col_key_bufs_ col_bufs_off_

Definition at line 344 of file hdfs-rcfile-scanner.cc.

References impala::HdfsRCFileScanner::ColumnInfo::buffer_len, columns_, impala::ReadWriteUtil::GetVInt(), impala::HdfsRCFileScanner::ColumnInfo::key_buffer, row_group_length_, impala::HdfsRCFileScanner::ColumnInfo::start_offset, and impala::HdfsRCFileScanner::ColumnInfo::uncompressed_buffer_len.

Referenced by ReadKeyBuffers().

int HdfsScanner::GetMemory	(	MemPool **	pool,
		Tuple **	tuple_mem,
		TupleRow **	tuple_row_mem
	)

protectedinherited

Gets memory for outputting tuples into batch_. *pool is the mem pool that should be used for memory allocated for those tuples. *tuple_mem should be the location to output tuples, and *tuple_row_mem for outputting tuple rows. Returns the maximum number of tuples/tuple rows that can be output (before the current row batch is complete and a new one is allocated). Memory returned from this call is invalidated after calling CommitRows. Callers must call GetMemory again after calling this function.

Definition at line 115 of file hdfs-scanner.cc.

References impala::RowBatch::AddRow(), impala::HdfsScanner::batch_, impala::RowBatch::capacity(), impala::RowBatch::GetRow(), impala::RowBatch::num_rows(), impala::RowBatch::tuple_data_pool(), and impala::HdfsScanner::tuple_mem_.

Referenced by impala::HdfsParquetScanner::AssembleRows(), impala::HdfsTextScanner::FinishScanRange(), impala::HdfsSequenceScanner::ProcessDecompressedBlock(), impala::HdfsParquetScanner::ProcessFooter(), impala::HdfsTextScanner::ProcessRange(), impala::HdfsAvroScanner::ProcessRange(), impala::HdfsSequenceScanner::ProcessRange(), and ProcessRange().

Status HdfsScanner::InitializeWriteTuplesFn	(	HdfsPartitionDescriptor *	partition,
		THdfsFileFormat::type	type,
		const std::string &	scanner_name
	)

protectedinherited

Initializes write_tuples_fn_ to the jitted function if codegen is possible.

partition - partition descriptor for this scanner/scan range
type - type for this scanner
scanner_name - debug string name for this scanner (e.g. HdfsTextScanner)

Definition at line 87 of file hdfs-scanner.cc.

References impala::HdfsPartitionDescriptor::escape_char(), impala::HdfsScanNode::GetCodegenFn(), impala::ExecNode::id(), impala::HdfsScanNode::IncNumScannersCodegenDisabled(), impala::HdfsScanNode::IncNumScannersCodegenEnabled(), impala::Status::OK, impala::HdfsScanner::scan_node_, impala::TupleDescriptor::string_slots(), impala::HdfsScanNode::tuple_desc(), and impala::HdfsScanner::write_tuples_fn_.

Referenced by impala::HdfsSequenceScanner::InitNewRange(), and impala::HdfsTextScanner::ResetScanner().

Status HdfsRCFileScanner::InitNewRange ( )

privatevirtual

Reset internal state for a new scan range.

Implements impala::HdfsScanner.

Definition at line 68 of file hdfs-rcfile-scanner.cc.

References impala::BaseSequenceScanner::FileHeader::codec, columns_, impala::Codec::CreateDecompressor(), impala::HdfsScanner::decompressor_, impala::HdfsScanNode::GetMaterializedSlotIdx(), impala::HdfsScanNode::hdfs_table(), impala::BaseSequenceScanner::header_, impala::BaseSequenceScanner::FileHeader::is_compressed, impala::TableDescriptor::num_cols(), impala::HdfsScanNode::num_partition_keys(), impala::Status::OK, impala::BaseSequenceScanner::only_parsing_header_, RETURN_IF_ERROR, reuse_row_group_buffer_, row_group_buffer_size_, impala::HdfsScanner::scan_node_, impala::ScannerContext::Stream::set_contains_tuple_data(), impala::HdfsScanNode::SKIP_COLUMN, impala::HdfsScanner::stream_, impala::TupleDescriptor::string_slots(), and impala::HdfsScanNode::tuple_desc().

void impala::HdfsScanner::InitTuple	(	Tuple *	template_tuple,
		Tuple *	tuple
	)

inlineprotectedinherited

Initialize a tuple. TODO: only copy over non-null slots. TODO: InitTuple is called frequently, avoid the if, perhaps via templatization.

Definition at line 355 of file hdfs-scanner.h.

References impala::HdfsScanner::num_null_bytes_, and impala::HdfsScanner::tuple_byte_size_.

Referenced by impala::HdfsParquetScanner::AssembleRows(), impala::HdfsAvroScanner::DecodeAvroData(), ProcessRange(), impala::HdfsScanner::WriteCompleteTuple(), and impala::HdfsTextScanner::WriteFields().

Status BaseSequenceScanner::IssueInitialRanges	(	HdfsScanNode *	scan_node,
		const std::vector< HdfsFileDesc * > &	files
	)

staticinherited

Issue the initial ranges for all sequence container files.

Definition at line 40 of file base-sequence-scanner.cc.

References impala::HdfsScanNode::AddDiskIoRanges(), impala::HdfsScanNode::AllocateScanRange(), impala::BaseSequenceScanner::HEADER_SIZE, impala::Status::OK, impala::ScanRangeMetadata::partition_id, and RETURN_IF_ERROR.

Referenced by impala::HdfsScanNode::GetNext().

void HdfsScanner::LogRowParseError	(	int	row_idx,
		std::stringstream *
	)

protectedvirtualinherited

Utility function to append an error message for an invalid row. This is called from ReportTupleParseError() row_idx is the index of the row in the current batch. Subclasses should override this function (i.e. text needs to join boundary rows). Since this is only in the error path, vtable overhead is acceptable.

Reimplemented in impala::HdfsSequenceScanner, and impala::HdfsTextScanner.

Definition at line 572 of file hdfs-scanner.cc.

Referenced by impala::HdfsScanner::ReportTupleParseError().

TupleRow* impala::HdfsScanner::next_row ( TupleRow * r ) const

inlineprotectedinherited

Definition at line 368 of file hdfs-scanner.h.

References impala::HdfsScanner::batch_, and impala::RowBatch::row_byte_size().

Referenced by impala::HdfsParquetScanner::AssembleRows(), impala::HdfsAvroScanner::DecodeAvroData(), ProcessRange(), impala::HdfsScanner::WriteEmptyTuples(), and impala::HdfsTextScanner::WriteFields().

Tuple* impala::HdfsScanner::next_tuple ( Tuple * t ) const

inlineprotectedinherited

Definition at line 363 of file hdfs-scanner.h.

References impala::HdfsScanner::tuple_byte_size_.

Referenced by impala::HdfsParquetScanner::AssembleRows(), impala::HdfsAvroScanner::DecodeAvroData(), ProcessRange(), and impala::HdfsTextScanner::WriteFields().

Status HdfsRCFileScanner::NextField ( int col_idx )

inlineprivate

Look at the next field in the specified column buffer Input: col_idx: Column of the field. Modifies: cur_field_length_rep_[col_idx] key_buf_pos_[col_idx] cur_field_length_rep_[col_idx] cur_field_length_[col_idx]

Definition at line 368 of file hdfs-rcfile-scanner.cc.

References impala::HdfsRCFileScanner::ColumnInfo::buffer_pos, columns_, impala::HdfsRCFileScanner::ColumnInfo::current_field_len, impala::HdfsRCFileScanner::ColumnInfo::current_field_len_rep, impala::ScannerContext::Stream::file_offset(), impala::ReadWriteUtil::GetVLong(), impala::HdfsRCFileScanner::ColumnInfo::key_buffer, impala::HdfsRCFileScanner::ColumnInfo::key_buffer_pos, impala::Status::OK, and impala::HdfsScanner::stream_.

Referenced by NextRow().

Status HdfsRCFileScanner::NextRow ( )

inlineprivate

Move to next row. Calls NextField on each column that we are reading. Modifies: row_pos_

Definition at line 400 of file hdfs-rcfile-scanner.cc.

References columns_, NextField(), num_rows_, impala::Status::OK, RETURN_IF_ERROR, and row_pos_.

Referenced by ProcessRange().

Status HdfsRCFileScanner::Prepare ( ScannerContext * context )

virtual

One-time initialisation of state that is constant across scan ranges.

Reimplemented from impala::BaseSequenceScanner.

Definition at line 60 of file hdfs-rcfile-scanner.cc.

References impala::HdfsScanNode::hdfs_table(), impala::HdfsScanNode::IncNumScannersCodegenDisabled(), impala::HdfsTableDescriptor::null_column_value(), impala::Status::OK, impala::BaseSequenceScanner::Prepare(), RETURN_IF_ERROR, impala::HdfsScanner::scan_node_, and impala::HdfsScanner::text_converter_.

Status HdfsRCFileScanner::ProcessRange ( )

privatevirtual

Process the current range until the end or an error occurred. Note this might be called multiple times if we skip over bad data. This function should read from the underlying ScannerContext materializing tuples to the context. When this function is called, it is guaranteed to be at the start of a data block (i.e. right after the sync marker).

Implements impala::BaseSequenceScanner.

Definition at line 451 of file hdfs-rcfile-scanner.cc.

Status BaseSequenceScanner::ProcessSplit ( )

virtualinherited

Process an entire split, reading bytes from the context's streams. Context is initialized with the split data (e.g. template tuple, partition descriptor, etc). This function should only return on error or end of scan range.

Implements impala::HdfsScanner.

Definition at line 100 of file base-sequence-scanner.cc.

Status HdfsRCFileScanner::ReadColumnBuffers ( )

private

Read the rowgroup column buffers Sets: column_buffer_: Fills the buffer with either file data or decompressed data.

Definition at line 413 of file hdfs-rcfile-scanner.cc.

References impala::HdfsRCFileScanner::ColumnInfo::buffer_len, columns_, impala::HdfsScanner::decompress_timer_, impala::HdfsScanner::decompressor_, impala::BaseSequenceScanner::header_, impala::BaseSequenceScanner::FileHeader::is_compressed, impala::Status::OK, impala::HdfsScanner::parse_status_, impala::ScannerContext::Stream::ReadBytes(), RETURN_IF_ERROR, RETURN_IF_FALSE, row_group_buffer_, row_group_length_, SCOPED_TIMER, impala::ScannerContext::Stream::SkipBytes(), impala::HdfsRCFileScanner::ColumnInfo::start_offset, impala::HdfsScanner::stream_, impala::HdfsRCFileScanner::ColumnInfo::uncompressed_buffer_len, and VLOG_FILE.

Referenced by ReadRowGroup().

Status HdfsRCFileScanner::ReadFileHeader ( )

privatevirtual

Read the file header. The underlying ScannerContext is at the start of the file header. This function must read the file header (which advances context_ past it) and initialize header_.

Implements impala::BaseSequenceScanner.

Definition at line 107 of file hdfs-rcfile-scanner.cc.

References impala::BaseSequenceScanner::FileHeader::codec, impala::Codec::CODEC_MAP, impala::BaseSequenceScanner::FileHeader::compression_type, impala::ScannerContext::Stream::filename(), impala::BaseSequenceScanner::header_, impala::BaseSequenceScanner::FileHeader::header_size, impala::ReadWriteUtil::HexDump(), impala::BaseSequenceScanner::FileHeader::is_compressed, impala::Status::OK, impala::HdfsScanner::parse_status_, RCF1, RCFILE_KEY_CLASS_NAME, RCFILE_VALUE_CLASS_NAME, RCFILE_VERSION_HEADER, impala::ScannerContext::Stream::ReadBoolean(), impala::ScannerContext::Stream::ReadBytes(), ReadNumColumnsMetadata(), impala::ScannerContext::Stream::ReadText(), RETURN_IF_ERROR, RETURN_IF_FALSE, SEQ6, impala::HdfsSequenceScanner::SEQFILE_VERSION_HEADER, impala::HdfsScanner::stream_, impala::BaseSequenceScanner::FileHeader::sync, impala::BaseSequenceScanner::SYNC_HASH_SIZE, impala::ScannerContext::Stream::total_bytes_returned(), impala::HdfsRCFileScanner::RcFileHeader::version, and VLOG_FILE.

Status HdfsRCFileScanner::ReadKeyBuffers ( )

private

Read the rowgroup key buffers, decompress if necessary. The "keys" are really the lengths for the column values. They are read here and then used to decode the values in the column buffer. Calls GetCurrentKeyBuffer for each column to process the key data.

Definition at line 308 of file hdfs-rcfile-scanner.cc.

References columns_, compressed_key_length_, impala::HdfsScanner::decompress_timer_, impala::HdfsScanner::decompressor_, GetCurrentKeyBuffer(), impala::ReadWriteUtil::GetVInt(), impala::BaseSequenceScanner::header_, impala::BaseSequenceScanner::FileHeader::is_compressed, key_buffer_, key_length_, num_rows_, impala::Status::OK, impala::HdfsScanner::parse_status_, impala::ScannerContext::Stream::ReadBytes(), RETURN_IF_ERROR, RETURN_IF_FALSE, row_group_length_, SCOPED_TIMER, impala::HdfsScanner::stream_, and VLOG_FILE.

Referenced by ReadRowGroup().

Status HdfsRCFileScanner::ReadNumColumnsMetadata ( )

private

Reads the RCFile Header Metadata section in the current file to determine the number of columns. Other pieces of the metadata are ignored.

Definition at line 197 of file hdfs-rcfile-scanner.cc.

References impala::ScannerContext::Stream::filename(), impala::BaseSequenceScanner::header_, impala::HdfsRCFileScanner::RcFileHeader::num_cols, impala::Status::OK, impala::StringParser::PARSE_OVERFLOW, impala::HdfsScanner::parse_status_, impala::StringParser::PARSE_SUCCESS, RCFILE_METADATA_KEY_NUM_COLS, impala::ScannerContext::Stream::ReadInt(), impala::ScannerContext::Stream::ReadText(), RETURN_IF_FALSE, and impala::HdfsScanner::stream_.

Referenced by ReadFileHeader().

Status HdfsRCFileScanner::ReadRowGroup ( )

private

Read a row group (except for the sync marker and sync) into buffers. Calls: ReadRowGroupHeader ReadKeyBuffers ReadColumnBuffers

Definition at line 254 of file hdfs-rcfile-scanner.cc.

References impala::HdfsScanner::data_buffer_pool_, impala::ExecNode::mem_tracker(), num_rows_, impala::Status::OK, ReadColumnBuffers(), ReadKeyBuffers(), ReadRowGroupHeader(), ResetRowGroup(), RETURN_IF_ERROR, reuse_row_group_buffer_, row_group_buffer_, row_group_buffer_size_, row_group_length_, impala::HdfsScanner::scan_node_, impala::RuntimeState::SetMemLimitExceeded(), and impala::HdfsScanner::state_.

Referenced by ProcessRange().

Status HdfsRCFileScanner::ReadRowGroupHeader ( )

private

Reads the rowgroup header starting after the sync. Sets: key_length_ compressed_key_length_ num_rows_

Definition at line 278 of file hdfs-rcfile-scanner.cc.

References compressed_key_length_, impala::ScannerContext::Stream::file_offset(), key_length_, impala::Status::OK, impala::HdfsScanner::parse_status_, impala::ScannerContext::Stream::ReadInt(), RETURN_IF_FALSE, and impala::HdfsScanner::stream_.

Referenced by ReadRowGroup().

Status BaseSequenceScanner::ReadSync ( )

protectedinherited

Read and validate sync marker against header_->sync. Returns non-ok if the sync marker did not match. Scanners should always use this function to read sync markers, otherwise finished() might not be updated correctly. If finished() returns true after calling this function, scanners must not process any more records.

Definition at line 170 of file base-sequence-scanner.cc.

References impala::BaseSequenceScanner::block_start_, impala::ScannerContext::Stream::eof(), impala::ScannerContext::Stream::eosr(), impala::ScannerContext::Stream::file_offset(), impala::BaseSequenceScanner::finished_, impala::ScannerContext::Stream::GetBytes(), impala::hash, impala::BaseSequenceScanner::header_, impala::ReadWriteUtil::HexDump(), impala::BaseSequenceScanner::num_syncs_, impala::Status::OK, impala::HdfsScanner::parse_status_, RETURN_IF_FALSE, impala::HdfsScanner::stream_, impala::BaseSequenceScanner::FileHeader::sync, impala::BaseSequenceScanner::SYNC_HASH_SIZE, and impala::BaseSequenceScanner::total_block_size_.

Referenced by impala::HdfsSequenceScanner::ProcessBlockCompressedScanRange(), impala::HdfsAvroScanner::ProcessRange(), impala::HdfsSequenceScanner::ProcessRange(), and ProcessRange().

void HdfsScanner::ReportColumnParseError	(	const SlotDescriptor *	desc,
		const char *	data,
		int	len
	)

protectedinherited

Report parse error for column @ desc. If abort_on_error is true, sets parse_status_ to the error message.

Definition at line 577 of file hdfs-scanner.cc.

References impala::RuntimeState::abort_on_error(), impala::SlotDescriptor::col_pos(), impala::RuntimeState::LogError(), impala::RuntimeState::LogHasSpace(), impala::HdfsScanNode::num_partition_keys(), impala::Status::ok(), impala::HdfsScanner::parse_status_, impala::HdfsScanner::scan_node_, impala::HdfsScanner::state_, and impala::SlotDescriptor::type().

Referenced by ProcessRange(), impala::HdfsScanner::ReportTupleParseError(), and impala::HdfsTextScanner::WritePartialTuple().

bool HdfsScanner::ReportTupleParseError	(	FieldLocation *	fields,
		uint8_t *	errors,
		int	row_idx
	)

protectedinherited

Utility function to report parse errors for each field. If errors[i] is nonzero, fields[i] had a parse error. row_idx is the idx of the row in the current batch that had the parse error Returns false if parsing should be aborted. In this case parse_status_ is set to the error. This is called from WriteAlignedTuples.

Definition at line 546 of file hdfs-scanner.cc.

References impala::RuntimeState::abort_on_error(), impala::ScannerContext::Stream::filename(), impala::RuntimeState::LogError(), impala::RuntimeState::LogHasSpace(), impala::HdfsScanner::LogRowParseError(), impala::HdfsScanNode::materialized_slots(), impala::HdfsScanner::num_errors_in_file_, impala::Status::ok(), impala::HdfsScanner::parse_status_, impala::HdfsScanner::ReportColumnParseError(), impala::RuntimeState::ReportFileErrors(), impala::HdfsScanner::scan_node_, impala::HdfsScanner::state_, and impala::HdfsScanner::stream_.

Referenced by impala::HdfsSequenceScanner::ProcessRange(), and impala::HdfsScanner::WriteAlignedTuples().

void HdfsRCFileScanner::ResetRowGroup ( )

private

Reset state for a new row group.

Definition at line 231 of file hdfs-rcfile-scanner.cc.

References impala::HdfsScanner::AttachPool(), columns_, compressed_key_length_, impala::HdfsScanner::data_buffer_pool_, key_length_, num_rows_, reuse_row_group_buffer_, row_group_buffer_size_, and row_pos_.

Referenced by ProcessRange(), and ReadRowGroup().

Status BaseSequenceScanner::SkipToSync	(	const uint8_t *	sync,
		int	sync_size
	)

protectedinherited

Utility function to advance past the next sync marker, reading bytes from stream_. If no sync is found in the scan range, return Status::OK and sets finished_ to true. It is safe to call this function past eosr.

sync: sync marker to search for (does not include 0xFFFFFFFF prefix)
sync_size: number of bytes for sync

Definition at line 212 of file base-sequence-scanner.cc.

Referenced by impala::BaseSequenceScanner::ProcessSplit().

void HdfsScanner::StartNewRowBatch ( )

protectedinherited

Set batch_ to a new row batch and update tuple_mem_ accordingly.

Definition at line 108 of file hdfs-scanner.cc.

References impala::MemPool::Allocate(), impala::HdfsScanner::batch_, impala::RuntimeState::batch_size(), impala::ExecNode::mem_tracker(), impala::ExecNode::row_desc(), impala::HdfsScanner::scan_node_, impala::HdfsScanner::state_, impala::HdfsScanner::tuple_byte_size_, impala::RowBatch::tuple_data_pool(), and impala::HdfsScanner::tuple_mem_.

Referenced by impala::HdfsScanner::CommitRows(), and impala::HdfsScanner::Prepare().

Status HdfsScanner::UpdateDecompressor ( const THdfsCompression::type & compression )

protectedinherited

Update the decompressor_ object given a compression type or codec name. Depending on the old compression type and the new one, it may close the old decompressor and/or create a new one of different type.

Definition at line 513 of file hdfs-scanner.cc.

References impala::Codec::CreateDecompressor(), impala::HdfsScanner::data_buffer_pool_, impala::HdfsScanner::decompression_type_, impala::HdfsScanner::decompressor_, impala::Status::OK, RETURN_IF_ERROR, impala::HdfsScanner::scan_node_, impala::TupleDescriptor::string_slots(), and impala::HdfsScanNode::tuple_desc().

Referenced by impala::HdfsAvroScanner::InitNewRange(), impala::HdfsSequenceScanner::InitNewRange(), and impala::HdfsTextScanner::ProcessSplit().

Status impala::HdfsScanner::UpdateDecompressor ( const std::string & codec )

protectedinherited

int HdfsScanner::WriteAlignedTuples	(	MemPool *	pool,
		TupleRow *	tuple_row_mem,
		int	row_size,
		FieldLocation *	fields,
		int	num_tuples,
		int	max_added_tuples,
		int	slots_per_tuple,
		int	row_start_indx
	)

protectedinherited

Processes batches of fields and writes them out to tuple_row_mem.

'pool' mempool to allocate from for auxiliary tuple memory
'tuple_row_mem' preallocated tuple_row memory this function must use.
'fields' must start at the beginning of a tuple.
'num_tuples' number of tuples to process
'max_added_tuples' the maximum number of tuples that should be added to the batch.
'row_start_index' is the number of rows that have already been processed as part of WritePartialTuple. Returns the number of tuples added to the row batch. This can be less than num_tuples/tuples_till_limit because of failed conjuncts. Returns -1 if parsing should be aborted due to parse errors.

Definition at line 33 of file hdfs-scanner-ir.cc.

References impala::HdfsScanner::ReportTupleParseError(), impala::HdfsScanner::template_tuple_, impala::HdfsScanner::tuple_, impala::HdfsScanner::tuple_byte_size_, UNLIKELY, and impala::HdfsScanner::WriteCompleteTuple().

Referenced by impala::HdfsSequenceScanner::ProcessDecompressedBlock(), and impala::HdfsTextScanner::WriteFields().

bool HdfsScanner::WriteCompleteTuple	(	MemPool *	pool,
		FieldLocation *	fields,
		Tuple *	tuple,
		TupleRow *	tuple_row,
		Tuple *	template_tuple,
		uint8_t *	error_fields,
		uint8_t *	error_in_row
	)

protectedinherited

Writes out all slots for 'tuple' from 'fields'. 'fields' must be aligned to the start of the tuple (e.g. fields[0] maps to slots[0]). After writing the tuple, it will be evaluated against the conjuncts.

error_fields is an out array. error_fields[i] will be set to true if the ith field had a parse error
error_in_row is an out bool. It is set to true if any field had parse errors Returns whether the resulting tuplerow passed the conjuncts. The parsing of the fields and evaluating against conjuncts is combined in this function. This is done so it can be possible to evaluate conjuncts as slots are materialized (on partial tuples). This function is replaced by a codegen'd function at runtime. This is the reason that the out error parameters are typed uint8_t instead of bool. We need to be able to match this function's signature identically for the codegen'd function. Bool's as out parameters can get converted to bytes by the compiler and rather than implicitly depending on that to happen, we will explicitly type them to bytes. TODO: revisit this

Definition at line 217 of file hdfs-scanner.cc.

References impala::HdfsScanner::EvalConjuncts(), impala::HdfsScanner::InitTuple(), impala::FieldLocation::len, impala::HdfsScanNode::materialized_slots(), impala::HdfsScanner::scan_node_, impala::TupleRow::SetTuple(), impala::HdfsScanner::text_converter_, impala::HdfsScanNode::tuple_idx(), and UNLIKELY.

Referenced by impala::HdfsSequenceScanner::ProcessRange(), and impala::HdfsScanner::WriteAlignedTuples().

int HdfsScanner::WriteEmptyTuples	(	RowBatch *	row_batch,
		int	num_tuples
	)

protectedinherited

Utility method to write out tuples when there are no materialized fields (e.g. select count(*) or only partition keys). num_tuples - Total number of tuples to write out. Returns the number of tuples added to the row batch.

Definition at line 157 of file hdfs-scanner.cc.

References impala::RowBatch::AddRow(), impala::RowBatch::AddRows(), impala::RowBatch::AtCapacity(), impala::RowBatch::capacity(), impala::RowBatch::CommitLastRow(), impala::RowBatch::CommitRows(), impala::HdfsScanner::EvalConjuncts(), impala::RowBatch::GetRow(), impala::RowBatch::INVALID_ROW_INDEX, impala::RowBatch::num_rows(), impala::HdfsScanner::scan_node_, impala::TupleRow::SetTuple(), impala::HdfsScanner::template_tuple_, and impala::HdfsScanNode::tuple_idx().

Referenced by impala::HdfsSequenceScanner::ProcessDecompressedBlock(), impala::HdfsParquetScanner::ProcessFooter(), impala::HdfsTextScanner::ProcessRange(), impala::HdfsAvroScanner::ProcessRange(), impala::HdfsSequenceScanner::ProcessRange(), and ProcessRange().

int HdfsScanner::WriteEmptyTuples	(	ScannerContext *	context,
		TupleRow *	tuple_row,
		int	num_tuples
	)

protectedinherited

Write empty tuples and commit them to the context object.

Definition at line 195 of file hdfs-scanner.cc.

References impala::HdfsScanner::EvalConjuncts(), impala::HdfsScanner::next_row(), impala::HdfsScanner::scan_node_, impala::TupleRow::SetTuple(), impala::HdfsScanner::template_tuple_, and impala::HdfsScanNode::tuple_idx().

Member Data Documentation

RowBatch* impala::HdfsScanner::batch_

protectedinherited

The current row batch being populated. Creating new row batches, attaching context resources, and handing off to the scan node is handled by this class in CommitRows(), but AttachPool() must be called by scanner subclasses to attach any memory allocated by that subclass. All row batches created by this class are transferred to the scan node (i.e., all batches are ultimately owned by the scan node).

Definition at line 177 of file hdfs-scanner.h.

Referenced by impala::HdfsScanner::AddFinalRowBatch(), impala::HdfsScanner::AttachPool(), impala::HdfsScanner::CommitRows(), impala::HdfsScanner::GetMemory(), impala::HdfsScanner::next_row(), impala::HdfsSequenceScanner::ProcessDecompressedBlock(), impala::HdfsParquetScanner::ProcessSplit(), impala::HdfsScanner::StartNewRowBatch(), impala::HdfsTextScanner::WriteFields(), and impala::HdfsScanner::~HdfsScanner().

std::vector<ColumnInfo> impala::HdfsRCFileScanner::columns_

private

Vector of column descriptions for each column in the file (i.e., may contain a different number of non-partition columns than are in the table metadata). Indexed by column index, including non-materialized columns.

Definition at line 376 of file hdfs-rcfile-scanner.h.

Referenced by GetCurrentKeyBuffer(), InitNewRange(), NextField(), NextRow(), ProcessRange(), ReadColumnBuffers(), ReadKeyBuffers(), and ResetRowGroup().

int impala::HdfsRCFileScanner::compressed_key_length_

private

Compressed size of the row group's key buffers. Read from the row group header.

Definition at line 394 of file hdfs-rcfile-scanner.h.

Referenced by ReadKeyBuffers(), ReadRowGroupHeader(), and ResetRowGroup().

std::vector<ExprContext*> impala::HdfsScanner::conjunct_ctxs_

protectedinherited

ExprContext for each conjunct. Each scanner has its own ExprContexts so the conjuncts can be safely evaluated in parallel.

Definition at line 154 of file hdfs-scanner.h.

Referenced by impala::HdfsScanner::Close(), impala::HdfsScanner::CommitRows(), impala::HdfsScanner::EvalConjuncts(), impala::HdfsScanner::GetConjunctCtx(), and impala::HdfsScanner::Prepare().

ScannerContext* impala::HdfsScanner::context_

protectedinherited

Context for this scanner.

Definition at line 147 of file hdfs-scanner.h.

Referenced by impala::HdfsScanner::AddFinalRowBatch(), impala::HdfsParquetScanner::AssembleRows(), impala::HdfsScanner::CommitRows(), impala::HdfsTextScanner::FillByteBufferCompressedFile(), impala::HdfsTextScanner::FillByteBufferGzip(), impala::HdfsParquetScanner::InitColumns(), impala::HdfsTextScanner::InitNewRange(), impala::HdfsSequenceScanner::InitNewRange(), impala::HdfsScanner::Prepare(), impala::HdfsSequenceScanner::ProcessDecompressedBlock(), impala::HdfsParquetScanner::ProcessFooter(), impala::HdfsTextScanner::ProcessRange(), impala::HdfsAvroScanner::ProcessRange(), impala::HdfsSequenceScanner::ProcessRange(), ProcessRange(), impala::HdfsParquetScanner::ProcessSplit(), and impala::HdfsTextScanner::ResetScanner().

boost::scoped_ptr<MemPool> impala::HdfsScanner::data_buffer_pool_

protectedinherited

Pool to allocate per data block memory. This should be used with the decompressor and any other per data block allocations.

Definition at line 205 of file hdfs-scanner.h.

Referenced by impala::HdfsTextScanner::Close(), impala::BaseSequenceScanner::Close(), impala::HdfsTextScanner::FillByteBufferGzip(), impala::HdfsAvroScanner::ProcessRange(), impala::HdfsSequenceScanner::ReadCompressedBlock(), ReadRowGroup(), ResetRowGroup(), impala::HdfsScanner::UpdateDecompressor(), and impala::HdfsTextScanner::WritePartialTuple().

RuntimeProfile::Counter* impala::HdfsScanner::decompress_timer_

protectedinherited

Time spent decompressing bytes.

Definition at line 208 of file hdfs-scanner.h.

Referenced by impala::HdfsTextScanner::FillByteBufferCompressedFile(), impala::HdfsTextScanner::FillByteBufferGzip(), impala::HdfsSequenceScanner::GetRecord(), impala::HdfsScanner::Prepare(), impala::HdfsAvroScanner::ProcessRange(), ReadColumnBuffers(), impala::HdfsSequenceScanner::ReadCompressedBlock(), impala::HdfsParquetScanner::BaseColumnReader::ReadDataPage(), and ReadKeyBuffers().

THdfsCompression::type impala::HdfsScanner::decompression_type_

protectedinherited

The most recently used decompression type.

Definition at line 201 of file hdfs-scanner.h.

Referenced by impala::HdfsTextScanner::FillByteBuffer(), impala::HdfsTextScanner::FillByteBufferCompressedFile(), and impala::HdfsScanner::UpdateDecompressor().

boost::scoped_ptr<Codec> impala::HdfsScanner::decompressor_

protectedinherited

Decompressor class to use, if any.

Definition at line 198 of file hdfs-scanner.h.

Referenced by impala::HdfsTextScanner::Close(), impala::BaseSequenceScanner::Close(), impala::HdfsScanner::Close(), impala::HdfsTextScanner::FillByteBuffer(), impala::HdfsTextScanner::FillByteBufferCompressedFile(), impala::HdfsTextScanner::FillByteBufferGzip(), impala::HdfsTextScanner::FinishScanRange(), impala::HdfsSequenceScanner::GetRecord(), InitNewRange(), impala::HdfsAvroScanner::ProcessRange(), ReadColumnBuffers(), impala::HdfsSequenceScanner::ReadCompressedBlock(), ReadKeyBuffers(), and impala::HdfsScanner::UpdateDecompressor().

const int impala::HdfsScanner::FILE_BLOCK_SIZE = 4096

staticinherited

Assumed size of an OS file block. Used mostly when reading file format headers, etc. This probably ought to be a derived number from the environment.

Definition at line 95 of file hdfs-scanner.h.

FileHeader* impala::BaseSequenceScanner::header_

protectedinherited

File header for this scan range. This is not owned by the parent scan node.

Definition at line 127 of file base-sequence-scanner.h.

Referenced by impala::BaseSequenceScanner::Close(), impala::HdfsSequenceScanner::GetRecord(), impala::HdfsAvroScanner::InitNewRange(), impala::HdfsSequenceScanner::InitNewRange(), InitNewRange(), impala::HdfsAvroScanner::ParseMetadata(), impala::HdfsSequenceScanner::ProcessBlockCompressedScanRange(), impala::HdfsAvroScanner::ProcessRange(), impala::HdfsSequenceScanner::ProcessRange(), impala::BaseSequenceScanner::ProcessSplit(), ReadColumnBuffers(), impala::HdfsAvroScanner::ReadFileHeader(), impala::HdfsSequenceScanner::ReadFileHeader(), ReadFileHeader(), ReadKeyBuffers(), ReadNumColumnsMetadata(), and impala::BaseSequenceScanner::ReadSync().

const int BaseSequenceScanner::HEADER_SIZE = 1024

staticprotectedinherited

Estimate of header size in bytes. This is initial number of bytes to issue per file. If the estimate is too low, more bytes will be read as necessary.

Definition at line 121 of file base-sequence-scanner.h.

Referenced by impala::BaseSequenceScanner::IssueInitialRanges().

std::vector<uint8_t> impala::HdfsRCFileScanner::key_buffer_

private

Buffer for copying key buffers. This buffer is reused between row groups.

Definition at line 379 of file hdfs-rcfile-scanner.h.

Referenced by ReadKeyBuffers().

int impala::HdfsRCFileScanner::key_length_

private

Size of the row group's key buffers. Read from the row group header.

Definition at line 390 of file hdfs-rcfile-scanner.h.

Referenced by ReadKeyBuffers(), ReadRowGroupHeader(), and ResetRowGroup().

const char * HdfsScanner::LLVM_CLASS_NAME = "class.impala::HdfsScanner"

staticinherited

Scanner subclasses must implement these static functions as well. Unfortunately, c++ does not allow static virtual functions. Issue the initial ranges for 'files'. HdfsFileDesc groups all the splits assigned to this scan node by file. This is called before any of the scanner subclasses are created to process splits in 'files'. The strategy on how to parse the scan ranges depends on the file format.

For simple text files, all the splits are simply issued to the io mgr and one split == one scan range.
For formats with a header, the metadata is first parsed, and then the ranges are issued to the io mgr. There is one scan range for the header and one range for each split.
For columnar formats, the header is parsed and only the relevant byte ranges should be issued to the io mgr. This is one range for the metadata and one range for each column, for each split. This function is how scanners can pick their strategy. void IssueInitialRanges(HdfsScanNode* scan_node, const std::vector<HdfsFileDesc*>& files); Codegen all functions for this scanner. The codegen'd function is specific to the scanner subclass but not specific to each scanner object. We don't want to codegen the functions for each scanner object. llvm::Function* Codegen(HdfsScanNode* scan_node);

Definition at line 137 of file hdfs-scanner.h.

Referenced by impala::HdfsScanner::CodegenWriteCompleteTuple().

int impala::HdfsScanner::num_errors_in_file_

protectedinherited

number of errors in current file

Definition at line 183 of file hdfs-scanner.h.

Referenced by impala::HdfsScanner::ReportTupleParseError().

int32_t impala::HdfsScanner::num_null_bytes_

protectedinherited

Number of null bytes in the tuple.

Definition at line 189 of file hdfs-scanner.h.

Referenced by impala::HdfsScanner::InitTuple().

int impala::HdfsRCFileScanner::num_rows_

private

number of rows in this rowgroup object

Definition at line 382 of file hdfs-rcfile-scanner.h.

Referenced by NextRow(), ProcessRange(), ReadKeyBuffers(), ReadRowGroup(), and ResetRowGroup().

bool impala::BaseSequenceScanner::only_parsing_header_

protectedinherited

If true, this scanner object is only for processing the header.

Definition at line 130 of file base-sequence-scanner.h.

Referenced by impala::BaseSequenceScanner::Close(), impala::BaseSequenceScanner::CloseFileRanges(), impala::HdfsAvroScanner::InitNewRange(), impala::HdfsSequenceScanner::InitNewRange(), InitNewRange(), and impala::BaseSequenceScanner::ProcessSplit().

Status impala::HdfsScanner::parse_status_

protectedinherited

Contains current parse status to minimize the number of Status objects returned. This significantly minimizes the cross compile dependencies for llvm since status objects inline a bunch of string functions. Also, status objects aren't extremely cheap to create and destroy.

Definition at line 195 of file hdfs-scanner.h.

Referenced by impala::HdfsTextScanner::FillByteBufferGzip(), impala::HdfsSequenceScanner::GetRecord(), impala::HdfsAvroScanner::ParseMetadata(), impala::HdfsSequenceScanner::ProcessBlockCompressedScanRange(), impala::HdfsSequenceScanner::ProcessDecompressedBlock(), impala::HdfsTextScanner::ProcessRange(), impala::HdfsAvroScanner::ProcessRange(), impala::HdfsSequenceScanner::ProcessRange(), ProcessRange(), impala::BaseSequenceScanner::ProcessSplit(), impala::HdfsSequenceScanner::ReadBlockHeader(), ReadColumnBuffers(), impala::HdfsSequenceScanner::ReadCompressedBlock(), impala::HdfsAvroScanner::ReadFileHeader(), impala::HdfsSequenceScanner::ReadFileHeader(), ReadFileHeader(), ReadKeyBuffers(), ReadNumColumnsMetadata(), ReadRowGroupHeader(), impala::BaseSequenceScanner::ReadSync(), impala::HdfsScanner::ReportColumnParseError(), impala::HdfsScanner::ReportTupleParseError(), impala::BaseSequenceScanner::SkipToSync(), and impala::HdfsTextScanner::WriteFields().

const char *const HdfsRCFileScanner::RCFILE_KEY_CLASS_NAME

staticprivate

Initial value:

=

"org.apache.hadoop.hive.ql.io.RCFile$KeyBuffer"

The key class name located in the RCFile Header. This is always "org.apache.hadoop.hive.ql.io.RCFile$KeyBuffer"

Definition at line 243 of file hdfs-rcfile-scanner.h.

Referenced by ReadFileHeader().

const char *const HdfsRCFileScanner::RCFILE_METADATA_KEY_NUM_COLS

staticprivate

Initial value:

=

"hive.io.rcfile.column.number"

RCFile metadata key for determining the number of columns present in the RCFile: "hive.io.rcfile.column.number"

Definition at line 251 of file hdfs-rcfile-scanner.h.

Referenced by ReadNumColumnsMetadata().

const char *const HdfsRCFileScanner::RCFILE_VALUE_CLASS_NAME

staticprivate

Initial value:

=

"org.apache.hadoop.hive.ql.io.RCFile$ValueBuffer"

The value class name located in the RCFile Header. This is always "org.apache.hadoop.hive.ql.io.RCFile$ValueBuffer"

Definition at line 247 of file hdfs-rcfile-scanner.h.

Referenced by ReadFileHeader().

const uint8_t HdfsRCFileScanner::RCFILE_VERSION_HEADER = {'R', 'C', 'F', 1}

staticprivate

The four byte RCFile unique version header present at the beginning of the file {'R', 'C', 'F' 1}

Definition at line 255 of file hdfs-rcfile-scanner.h.

Referenced by ReadFileHeader().

bool impala::HdfsRCFileScanner::reuse_row_group_buffer_

private

If true, the row_group_buffer_ can be reused across row groups, otherwise, it (more specifically the data_buffer_pool_ that allocated the row_group_buffer_) must be attached to the row batch.

Definition at line 399 of file hdfs-rcfile-scanner.h.

Referenced by InitNewRange(), ReadRowGroup(), and ResetRowGroup().

uint8_t* impala::HdfsRCFileScanner::row_group_buffer_

private

Buffer containing the entire row group. We allocate a buffer for the entire row group, skipping non-materialized columns.

Definition at line 403 of file hdfs-rcfile-scanner.h.

Referenced by ProcessRange(), ReadColumnBuffers(), and ReadRowGroup().

int impala::HdfsRCFileScanner::row_group_buffer_size_

private

This is the allocated size of 'row_group_buffer_'. 'row_group_buffer_' is reused across row groups and will grow as necessary.

Definition at line 411 of file hdfs-rcfile-scanner.h.

Referenced by InitNewRange(), ReadRowGroup(), and ResetRowGroup().

int impala::HdfsRCFileScanner::row_group_length_

private

Sum of the bytes lengths of the materialized columns in the current row group. This is the number of valid bytes in row_group_buffer_.

Definition at line 407 of file hdfs-rcfile-scanner.h.

Referenced by GetCurrentKeyBuffer(), ProcessRange(), ReadColumnBuffers(), ReadKeyBuffers(), and ReadRowGroup().

int impala::HdfsRCFileScanner::row_pos_

private

Current row position in this rowgroup. This value is incremented each time NextRow() is called.

Definition at line 386 of file hdfs-rcfile-scanner.h.

Referenced by NextRow(), ProcessRange(), and ResetRowGroup().

HdfsScanNode* impala::HdfsScanner::scan_node_

protectedinherited

The scan node that started this scanner.

Definition at line 141 of file hdfs-scanner.h.

RuntimeState* impala::HdfsScanner::state_

protectedinherited

RuntimeState for error reporting.

Definition at line 144 of file hdfs-scanner.h.

Referenced by impala::HdfsScanner::Close(), impala::HdfsScanner::CommitRows(), impala::HdfsTextScanner::FillByteBufferGzip(), impala::HdfsTextScanner::FinishScanRange(), impala::HdfsTextScanner::Prepare(), impala::HdfsScanner::Prepare(), impala::HdfsSequenceScanner::Prepare(), impala::HdfsSequenceScanner::ProcessBlockCompressedScanRange(), ProcessRange(), impala::BaseSequenceScanner::ProcessSplit(), impala::HdfsSequenceScanner::ReadCompressedBlock(), impala::BaseSequenceScanner::ReadPastSize(), ReadRowGroup(), impala::HdfsScanner::ReportColumnParseError(), impala::HdfsScanner::ReportTupleParseError(), impala::HdfsAvroScanner::ResolveSchemas(), impala::HdfsScanner::StartNewRowBatch(), impala::HdfsParquetScanner::ValidateColumn(), and impala::HdfsTextScanner::WriteFields().

ScannerContext::Stream* impala::HdfsScanner::stream_

protectedinherited

The first stream for context_.

Definition at line 150 of file hdfs-scanner.h.

Referenced by impala::HdfsTextScanner::Close(), impala::BaseSequenceScanner::Close(), impala::HdfsParquetScanner::CreateColumnReaders(), DebugString(), impala::HdfsTextScanner::FillByteBuffer(), impala::HdfsTextScanner::FillByteBufferCompressedFile(), impala::HdfsTextScanner::FillByteBufferGzip(), impala::HdfsTextScanner::FindFirstTuple(), impala::HdfsTextScanner::FinishScanRange(), impala::HdfsSequenceScanner::GetRecord(), impala::HdfsTextScanner::InitNewRange(), InitNewRange(), NextField(), impala::HdfsAvroScanner::ParseMetadata(), impala::BaseSequenceScanner::Prepare(), impala::HdfsScanner::Prepare(), impala::HdfsSequenceScanner::ProcessBlockCompressedScanRange(), impala::HdfsParquetScanner::ProcessFooter(), impala::HdfsTextScanner::ProcessRange(), impala::HdfsAvroScanner::ProcessRange(), impala::HdfsSequenceScanner::ProcessRange(), ProcessRange(), impala::HdfsTextScanner::ProcessSplit(), impala::BaseSequenceScanner::ProcessSplit(), impala::HdfsParquetScanner::ProcessSplit(), impala::HdfsSequenceScanner::ReadBlockHeader(), ReadColumnBuffers(), impala::HdfsSequenceScanner::ReadCompressedBlock(), impala::HdfsAvroScanner::ReadFileHeader(), impala::HdfsSequenceScanner::ReadFileHeader(), ReadFileHeader(), ReadKeyBuffers(), ReadNumColumnsMetadata(), ReadRowGroupHeader(), impala::BaseSequenceScanner::ReadSync(), impala::HdfsScanner::ReportTupleParseError(), impala::BaseSequenceScanner::SkipToSync(), impala::HdfsParquetScanner::ValidateFileMetadata(), impala::HdfsAvroScanner::VerifyTypesMatch(), and impala::HdfsTextScanner::WriteFields().

const int impala::BaseSequenceScanner::SYNC_HASH_SIZE = 16

staticprotectedinherited

Size of the sync hash field.

Definition at line 49 of file base-sequence-scanner.h.

Referenced by impala::BaseSequenceScanner::ProcessSplit(), impala::HdfsAvroScanner::ReadFileHeader(), impala::HdfsSequenceScanner::ReadFileHeader(), ReadFileHeader(), and impala::BaseSequenceScanner::ReadSync().

const int BaseSequenceScanner::SYNC_MARKER = -1

staticprotectedinherited

Sync indicator.

Definition at line 124 of file base-sequence-scanner.h.

Referenced by impala::HdfsSequenceScanner::ProcessRange(), and ProcessRange().

Tuple* impala::HdfsScanner::template_tuple_

protectedinherited

A partially materialized tuple with only partition key slots set. The non-partition key slots are set to NULL. The template tuple must be copied into tuple_ before any of the other slots are materialized. Pointer is NULL if there are no partition key slots. This template tuple is computed once for each file and valid for the duration of that file. It is owned by the HDFS scan node.

Definition at line 164 of file hdfs-scanner.h.

Referenced by impala::HdfsAvroScanner::AllocateFileHeader(), impala::HdfsParquetScanner::AssembleRows(), impala::HdfsParquetScanner::CreateColumnReaders(), impala::HdfsAvroScanner::DecodeAvroData(), impala::HdfsAvroScanner::InitNewRange(), impala::HdfsScanner::Prepare(), impala::HdfsSequenceScanner::ProcessRange(), ProcessRange(), impala::HdfsAvroScanner::ResolveSchemas(), impala::HdfsScanner::WriteAlignedTuples(), impala::HdfsScanner::WriteEmptyTuples(), and impala::HdfsTextScanner::WriteFields().

boost::scoped_ptr<TextConverter> impala::HdfsScanner::text_converter_

protectedinherited

Helper class for converting text to other types;.

Definition at line 186 of file hdfs-scanner.h.

Referenced by impala::HdfsTextScanner::InitNewRange(), impala::HdfsSequenceScanner::InitNewRange(), Prepare(), ProcessRange(), impala::HdfsScanner::WriteCompleteTuple(), and impala::HdfsTextScanner::WritePartialTuple().

Tuple* impala::HdfsScanner::tuple_

protectedinherited

Current tuple pointer into tuple_mem_.

Definition at line 170 of file hdfs-scanner.h.

Referenced by impala::HdfsTextScanner::FinishScanRange(), impala::HdfsSequenceScanner::ProcessDecompressedBlock(), impala::HdfsTextScanner::ProcessRange(), impala::HdfsSequenceScanner::ProcessRange(), impala::HdfsScanner::WriteAlignedTuples(), and impala::HdfsTextScanner::WriteFields().

int impala::HdfsScanner::tuple_byte_size_

protectedinherited

Fixed size of each tuple, in bytes.

Definition at line 167 of file hdfs-scanner.h.

Referenced by impala::HdfsParquetScanner::AssembleRows(), impala::HdfsScanner::InitTuple(), impala::HdfsScanner::next_tuple(), impala::HdfsScanner::StartNewRowBatch(), and impala::HdfsScanner::WriteAlignedTuples().

uint8_t* impala::HdfsScanner::tuple_mem_

protectedinherited

The tuple memory of batch_.

Definition at line 180 of file hdfs-scanner.h.

Referenced by impala::HdfsScanner::CommitRows(), impala::HdfsScanner::GetMemory(), and impala::HdfsScanner::StartNewRowBatch().

WriteTuplesFn impala::HdfsScanner::write_tuples_fn_

protectedinherited

Jitted write tuples function pointer. Null if codegen is disabled.

Definition at line 215 of file hdfs-scanner.h.

Referenced by impala::HdfsScanner::InitializeWriteTuplesFn(), impala::HdfsSequenceScanner::ProcessDecompressedBlock(), and impala::HdfsTextScanner::WriteFields().

The documentation for this class was generated from the following files:

be/src/exec/hdfs-rcfile-scanner.h
be/src/exec/hdfs-rcfile-scanner.cc

Classes

Public Member Functions

Static Public Member Functions

Static Public Attributes

Protected Types

Protected Member Functions

Static Protected Member Functions

Protected Attributes

Static Protected Attributes

Private Types

Private Member Functions

Private Attributes

Static Private Attributes

Detailed Description

Member Typedef Documentation

Member Enumeration Documentation

Constructor & Destructor Documentation

Member Function Documentation

Member Data Documentation