#include <hdfs-sequence-scanner.h>

Inheritance diagram for impala::HdfsSequenceScanner:

Collaboration diagram for impala::HdfsSequenceScanner:

Classes
struct	RecordLocation
	Struct for record locations and lens in compressed blocks. More...

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

Public Member Functions
	HdfsSequenceScanner (HdfsScanNode scan_node, RuntimeState state)

virtual	~HdfsSequenceScanner ()

virtual Status	Prepare (ScannerContext *context)
	Implementation of HdfsScanner interface. More...

virtual void	Close ()

virtual Status	ProcessSplit ()

Static Public Member Functions
static llvm::Function *	Codegen (HdfsScanNode , const std::vector< ExprContext > &conjunct_ctxs)
	Codegen writing tuples and evaluating predicates. More...

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 uint8_t	SEQFILE_VERSION_HEADER [4] = {'S', 'E', 'Q', 6}

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
virtual FileHeader *	AllocateFileHeader ()
	Implementation of sequence container super class methods. 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	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)

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 Member Functions
Status	ReadBlockHeader ()

Status	ProcessBlockCompressedScanRange ()

Status	ReadCompressedBlock ()

Status	ProcessDecompressedBlock ()

Status	GetRecord (uint8_t *record_ptr, int64_t record_len)

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

Private Attributes
boost::scoped_ptr < DelimitedTextParser >	delimited_text_parser_
	Helper class for picking fields and rows from delimited text. More...

std::vector< FieldLocation >	field_locations_

std::vector< RecordLocation >	record_locations_

int	current_block_length_
	Length of the current sequence file block (or record). More...

int	current_key_length_
	Length of the current key. This is specified as 4 bytes in the format description. More...

uint8_t *	unparsed_data_buffer_
	Buffer for data read from HDFS or from decompressing the HDFS data. More...

int64_t	num_buffered_records_in_compressed_block_
	Number of buffered records unparsed_data_buffer_ from block compressed data. More...

uint8_t *	next_record_in_compressed_block_
	Next record from block compressed data. More...

Static Private Attributes
static const int	MAX_BLOCK_SIZE = (1024 * 1024 * 1024)

static const char *const	SEQFILE_VALUE_CLASS_NAME

Detailed Description

Definition at line 156 of file hdfs-sequence-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.

Constructor & Destructor Documentation

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

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

HdfsSequenceScanner::~HdfsSequenceScanner ( )

virtual

Definition at line 46 of file hdfs-sequence-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 * HdfsSequenceScanner::AllocateFileHeader ( )

protectedvirtual

Implementation of sequence container super class methods.

Implements impala::BaseSequenceScanner.

Definition at line 99 of file hdfs-sequence-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(), ReadCompressedBlock(), impala::HdfsParquetScanner::BaseColumnReader::ReadDataPage(), and impala::HdfsRCFileScanner::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 * HdfsSequenceScanner::Codegen	(	HdfsScanNode *	,
		const std::vector< ExprContext * > &	conjunct_ctxs
	)

static

Codegen writing tuples and evaluating predicates.

Definition at line 50 of file hdfs-sequence-scanner.cc.

References impala::RuntimeState::codegen_enabled(), impala::HdfsScanner::CodegenWriteAlignedTuples(), impala::HdfsScanner::CodegenWriteCompleteTuple(), impala::RuntimeState::GetCodegen(), impala::Status::ok(), and impala::HdfsScanNode::runtime_state().

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

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 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 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(), ProcessDecompressedBlock(), impala::HdfsParquetScanner::ProcessFooter(), impala::HdfsTextScanner::ProcessRange(), impala::HdfsAvroScanner::ProcessRange(), ProcessRange(), impala::HdfsRCFileScanner::ProcessRange(), and impala::HdfsParquetScanner::ProcessSplit().

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(), impala::HdfsRCFileScanner::ProcessRange(), impala::HdfsScanner::WriteCompleteTuple(), impala::HdfsScanner::WriteEmptyTuples(), and impala::HdfsTextScanner::WriteFields().

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

inlineprotectedvirtual

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

Implements impala::BaseSequenceScanner.

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

bool impala::BaseSequenceScanner::finished ( )

inlineprotectedinherited

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

References impala::BaseSequenceScanner::finished_.

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

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(), ProcessDecompressedBlock(), impala::HdfsParquetScanner::ProcessFooter(), impala::HdfsTextScanner::ProcessRange(), impala::HdfsAvroScanner::ProcessRange(), ProcessRange(), and impala::HdfsRCFileScanner::ProcessRange().

Status HdfsSequenceScanner::GetRecord	(	uint8_t **	record_ptr,
		int64_t *	record_len
	)

inlineprivate

Read compressed or uncompressed records from the byte stream into memory in unparsed_data_buffer_pool_. Not used for block compressed files. Output: record_ptr: ponter to the record. record_len: length of the record

Definition at line 103 of file hdfs-sequence-scanner.cc.

References current_block_length_, current_key_length_, impala::HdfsScanner::decompress_timer_, impala::HdfsScanner::decompressor_, impala::ReadWriteUtil::GetVLong(), impala::BaseSequenceScanner::header_, impala::BaseSequenceScanner::FileHeader::is_compressed, impala::Status::OK, impala::HdfsScanner::parse_status_, ReadBlockHeader(), impala::ScannerContext::Stream::ReadBytes(), impala::ScannerContext::Stream::ReadVLong(), RETURN_IF_ERROR, RETURN_IF_FALSE, SCOPED_TIMER, impala::ScannerContext::Stream::SkipBytes(), impala::HdfsScanner::stream_, unparsed_data_buffer_, and VLOG_FILE.

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

Status HdfsSequenceScanner::InitNewRange ( )

protectedvirtual

Reset internal state for a new scan range.

Implements impala::HdfsScanner.

Definition at line 61 of file hdfs-sequence-scanner.cc.

References impala::BaseSequenceScanner::FileHeader::codec, impala::HdfsPartitionDescriptor::collection_delim(), impala::HdfsScanner::context_, delimited_text_parser_, impala::HdfsPartitionDescriptor::escape_char(), impala::HdfsPartitionDescriptor::field_delim(), impala::HdfsScanNode::hdfs_table(), impala::BaseSequenceScanner::header_, impala::HdfsScanner::InitializeWriteTuplesFn(), impala::BaseSequenceScanner::FileHeader::is_compressed, impala::HdfsScanNode::is_materialized_col(), impala::HdfsTableDescriptor::null_column_value(), num_buffered_records_in_compressed_block_, impala::TableDescriptor::num_cols(), impala::HdfsScanNode::num_partition_keys(), impala::Status::OK, impala::BaseSequenceScanner::only_parsing_header_, impala::ScannerContext::partition_descriptor(), RETURN_IF_ERROR, impala::HdfsScanner::scan_node_, impala::HdfsScanner::text_converter_, and impala::HdfsScanner::UpdateDecompressor().

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(), impala::HdfsRCFileScanner::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 HdfsSequenceScanner::LogRowParseError	(	int	row_idx,
		std::stringstream *
	)

privatevirtual

Appends the current file and line to the RuntimeState's error log. row_idx is 0-based (in current batch) where the parse error occurred.

Reimplemented from impala::HdfsScanner.

Definition at line 493 of file hdfs-sequence-scanner.cc.

References record_locations_.

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(), impala::HdfsRCFileScanner::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(), impala::HdfsRCFileScanner::ProcessRange(), and impala::HdfsTextScanner::WriteFields().

Status HdfsSequenceScanner::Prepare ( ScannerContext * context )

virtual

Implementation of HdfsScanner interface.

Reimplemented from impala::BaseSequenceScanner.

Definition at line 88 of file hdfs-sequence-scanner.cc.

References impala::RuntimeState::batch_size(), field_locations_, impala::HdfsScanNode::materialized_slots(), impala::Status::OK, impala::BaseSequenceScanner::Prepare(), record_locations_, RETURN_IF_ERROR, impala::HdfsScanner::scan_node_, and impala::HdfsScanner::state_.

Status HdfsSequenceScanner::ProcessBlockCompressedScanRange ( )

private

Process an entire block compressed scan range. Block compressed ranges are more common and can be parsed more efficiently in larger pieces.

Definition at line 161 of file hdfs-sequence-scanner.cc.

References impala::ScannerContext::Stream::eof(), impala::ScannerContext::Stream::file_offset(), impala::BaseSequenceScanner::finished(), impala::BaseSequenceScanner::header_, impala::BaseSequenceScanner::FileHeader::is_compressed, impala::RuntimeState::LogError(), impala::RuntimeState::LogHasSpace(), num_buffered_records_in_compressed_block_, impala::Status::OK, impala::HdfsScanner::parse_status_, ProcessDecompressedBlock(), impala::ExecNode::ReachedLimit(), ReadCompressedBlock(), impala::ScannerContext::Stream::ReadInt(), impala::BaseSequenceScanner::ReadSync(), RETURN_IF_ERROR, RETURN_IF_FALSE, impala::HdfsScanner::scan_node_, impala::HdfsScanner::state_, and impala::HdfsScanner::stream_.

Referenced by ProcessRange().

Status HdfsSequenceScanner::ProcessDecompressedBlock ( )

private

Utility function for parsing next_record_in_compressed_block_. Called by ProcessBlockCompressedScanRange.

Definition at line 198 of file hdfs-sequence-scanner.cc.

Referenced by ProcessBlockCompressedScanRange().

Status HdfsSequenceScanner::ProcessRange ( )

protectedvirtual

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 271 of file hdfs-sequence-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 HdfsSequenceScanner::ReadBlockHeader ( )

private

Read the record header. Sets: current_block_length_

Definition at line 419 of file hdfs-sequence-scanner.cc.

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

Referenced by GetRecord().

Status HdfsSequenceScanner::ReadCompressedBlock ( )

private

Read a compressed block. Does NOT read sync or -1 marker preceding sync. Decompress to unparsed_data_buffer_ allocated from unparsed_data_buffer_pool_.

Definition at line 441 of file hdfs-sequence-scanner.cc.

References impala::HdfsScanner::AttachPool(), impala::HdfsScanner::data_buffer_pool_, impala::HdfsScanner::decompress_timer_, impala::HdfsScanner::decompressor_, impala::RuntimeState::LogError(), impala::RuntimeState::LogHasSpace(), MAX_BLOCK_SIZE, next_record_in_compressed_block_, num_buffered_records_in_compressed_block_, impala::Status::OK, impala::HdfsScanner::parse_status_, impala::ScannerContext::Stream::ReadBytes(), impala::ScannerContext::Stream::ReadVLong(), RETURN_IF_ERROR, RETURN_IF_FALSE, SCOPED_TIMER, impala::ScannerContext::Stream::SkipText(), impala::HdfsScanner::state_, impala::HdfsScanner::stream_, unparsed_data_buffer_, and VLOG_FILE.

Referenced by ProcessBlockCompressedScanRange().

Status HdfsSequenceScanner::ReadFileHeader ( )

protectedvirtual

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 345 of file hdfs-sequence-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::HdfsSequenceScanner::SeqFileHeader::is_row_compressed, impala::Status::OK, impala::HdfsScanner::parse_status_, impala::ScannerContext::Stream::ReadBoolean(), impala::ScannerContext::Stream::ReadBytes(), impala::ScannerContext::Stream::ReadInt(), impala::ScannerContext::Stream::ReadText(), RETURN_IF_FALSE, SEQFILE_VALUE_CLASS_NAME, SEQFILE_VERSION_HEADER, impala::ScannerContext::Stream::SkipText(), impala::HdfsScanner::stream_, impala::BaseSequenceScanner::FileHeader::sync, impala::BaseSequenceScanner::SYNC_HASH_SIZE, impala::ScannerContext::Stream::total_bytes_returned(), and VLOG_FILE.

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 ProcessBlockCompressedScanRange(), impala::HdfsAvroScanner::ProcessRange(), ProcessRange(), and impala::HdfsRCFileScanner::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 impala::HdfsRCFileScanner::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 ProcessRange(), and impala::HdfsScanner::WriteAlignedTuples().

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(), 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 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 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 ProcessDecompressedBlock(), impala::HdfsParquetScanner::ProcessFooter(), impala::HdfsTextScanner::ProcessRange(), impala::HdfsAvroScanner::ProcessRange(), ProcessRange(), and impala::HdfsRCFileScanner::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(), ProcessDecompressedBlock(), impala::HdfsParquetScanner::ProcessSplit(), impala::HdfsScanner::StartNewRowBatch(), impala::HdfsTextScanner::WriteFields(), and impala::HdfsScanner::~HdfsScanner().

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(), InitNewRange(), impala::HdfsScanner::Prepare(), ProcessDecompressedBlock(), impala::HdfsParquetScanner::ProcessFooter(), impala::HdfsTextScanner::ProcessRange(), impala::HdfsAvroScanner::ProcessRange(), ProcessRange(), impala::HdfsRCFileScanner::ProcessRange(), impala::HdfsParquetScanner::ProcessSplit(), and impala::HdfsTextScanner::ResetScanner().

int impala::HdfsSequenceScanner::current_block_length_

private

Length of the current sequence file block (or record).

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

Referenced by GetRecord(), and ReadBlockHeader().

int impala::HdfsSequenceScanner::current_key_length_

private

Length of the current key. This is specified as 4 bytes in the format description.

Definition at line 246 of file hdfs-sequence-scanner.h.

Referenced by GetRecord(), and ReadBlockHeader().

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(), ReadCompressedBlock(), impala::HdfsRCFileScanner::ReadRowGroup(), impala::HdfsRCFileScanner::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(), GetRecord(), impala::HdfsScanner::Prepare(), impala::HdfsAvroScanner::ProcessRange(), impala::HdfsRCFileScanner::ReadColumnBuffers(), ReadCompressedBlock(), impala::HdfsParquetScanner::BaseColumnReader::ReadDataPage(), and impala::HdfsRCFileScanner::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(), GetRecord(), impala::HdfsRCFileScanner::InitNewRange(), impala::HdfsAvroScanner::ProcessRange(), impala::HdfsRCFileScanner::ReadColumnBuffers(), ReadCompressedBlock(), impala::HdfsRCFileScanner::ReadKeyBuffers(), and impala::HdfsScanner::UpdateDecompressor().

boost::scoped_ptr<DelimitedTextParser> impala::HdfsSequenceScanner::delimited_text_parser_

private

Helper class for picking fields and rows from delimited text.

Definition at line 222 of file hdfs-sequence-scanner.h.

Referenced by InitNewRange(), ProcessDecompressedBlock(), and ProcessRange().

std::vector<FieldLocation> impala::HdfsSequenceScanner::field_locations_

private

Definition at line 223 of file hdfs-sequence-scanner.h.

Referenced by Prepare(), ProcessDecompressedBlock(), and ProcessRange().

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(), GetRecord(), impala::HdfsAvroScanner::InitNewRange(), InitNewRange(), impala::HdfsRCFileScanner::InitNewRange(), impala::HdfsAvroScanner::ParseMetadata(), ProcessBlockCompressedScanRange(), impala::HdfsAvroScanner::ProcessRange(), ProcessRange(), impala::BaseSequenceScanner::ProcessSplit(), impala::HdfsRCFileScanner::ReadColumnBuffers(), impala::HdfsAvroScanner::ReadFileHeader(), ReadFileHeader(), impala::HdfsRCFileScanner::ReadFileHeader(), impala::HdfsRCFileScanner::ReadKeyBuffers(), impala::HdfsRCFileScanner::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().

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().

const int impala::HdfsSequenceScanner::MAX_BLOCK_SIZE = (1024 * 1024 * 1024)

staticprivate

Maximum size of a compressed block. This is used to check for corrupted block size so we do not read the whole file before we detect the error.

Definition at line 187 of file hdfs-sequence-scanner.h.

Referenced by ReadCompressedBlock().

uint8_t* impala::HdfsSequenceScanner::next_record_in_compressed_block_

private

Next record from block compressed data.

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

Referenced by ProcessDecompressedBlock(), and ReadCompressedBlock().

int64_t impala::HdfsSequenceScanner::num_buffered_records_in_compressed_block_

private

Number of buffered records unparsed_data_buffer_ from block compressed data.

Definition at line 252 of file hdfs-sequence-scanner.h.

Referenced by InitNewRange(), ProcessBlockCompressedScanRange(), ProcessDecompressedBlock(), ProcessRange(), and ReadCompressedBlock().

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().

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

std::vector<RecordLocation> impala::HdfsSequenceScanner::record_locations_

private

Records are processed in batches. This vector stores batches of record locations that are being processed. TODO: better perf not to use vector?

Definition at line 240 of file hdfs-sequence-scanner.h.

Referenced by LogRowParseError(), Prepare(), ProcessDecompressedBlock(), and ProcessRange().

HdfsScanNode* impala::HdfsScanner::scan_node_

protectedinherited

The scan node that started this scanner.

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

const char *const HdfsSequenceScanner::SEQFILE_VALUE_CLASS_NAME

staticprivate

Initial value:

=

"org.apache.hadoop.io.Text"

The value class name located in the SeqFile Header. This is always "org.apache.hadoop.io.Text"

Definition at line 191 of file hdfs-sequence-scanner.h.

Referenced by ReadFileHeader().

const uint8_t HdfsSequenceScanner::SEQFILE_VERSION_HEADER = {'S', 'E', 'Q', 6}

static

The four byte SeqFile version header present at the beginning of every SeqFile file: {'S', 'E', 'Q', 6}

Definition at line 160 of file hdfs-sequence-scanner.h.

Referenced by ReadFileHeader(), and impala::HdfsRCFileScanner::ReadFileHeader().

RuntimeState* impala::HdfsScanner::state_

protectedinherited

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

protectedinherited

The first stream for context_.

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

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(), ReadFileHeader(), impala::HdfsRCFileScanner::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 ProcessRange(), and impala::HdfsRCFileScanner::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(), ProcessRange(), impala::HdfsRCFileScanner::ProcessRange(), impala::HdfsAvroScanner::ResolveSchemas(), impala::HdfsScanner::WriteAlignedTuples(), impala::HdfsScanner::WriteEmptyTuples(), and impala::HdfsTextScanner::WriteFields().