#include "exec/hdfs-parquet-scanner.h"
#include <limits>
#include <boost/algorithm/string.hpp>
#include <gflags/gflags.h>
#include <gutil/strings/substitute.h>
#include "common/object-pool.h"
#include "common/logging.h"
#include "exec/hdfs-scan-node.h"
#include "exec/scanner-context.inline.h"
#include "exec/read-write-util.h"
#include "exprs/expr.h"
#include "runtime/descriptors.h"
#include "runtime/runtime-state.h"
#include "runtime/mem-pool.h"
#include "runtime/raw-value.h"
#include "runtime/row-batch.h"
#include "runtime/tuple-row.h"
#include "runtime/tuple.h"
#include "runtime/string-value.h"
#include "util/bitmap.h"
#include "util/bit-util.h"
#include "util/decompress.h"
#include "util/debug-util.h"
#include "util/error-util.h"
#include "util/dict-encoding.h"
#include "util/rle-encoding.h"
#include "util/runtime-profile.h"
#include "rpc/thrift-util.h"
#include "common/names.h"

Include dependency graph for hdfs-parquet-scanner.cc:

Classes
class	impala::HdfsParquetScanner::BaseColumnReader

class	impala::HdfsParquetScanner::ColumnReader< T >

class	impala::HdfsParquetScanner::BoolColumnReader

Namespaces
	impala
	This file contains type definitions that are used throughout the code base.

Macros
#define	LOG_OR_ABORT(error_msg, runtime_state)

#define	LOG_OR_RETURN_ON_ERROR(error_msg, runtime_state)

Functions
	DEFINE_bool (convert_legacy_hive_parquet_utc_timestamps, false,"When true, TIMESTAMPs read from files written by Parquet-MR (used by Hive) will ""be converted from UTC to local time. Writes are unaffected.")

static bool	RequiresSkippedDictionaryHeaderCheck (const HdfsParquetScanner::FileVersion &v)

bool	IsEncodingSupported (parquet::Encoding::type e)

string	PrintRepetitionType (const parquet::FieldRepetitionType::type &t)

string	PrintParquetType (const parquet::Type::type &t)

Variables
const int	MAX_PAGE_HEADER_SIZE = 8 * 1024 * 1024

const int	MAX_DICT_HEADER_SIZE = 100

Macro Definition Documentation

#define LOG_OR_ABORT	(	error_msg,
		runtime_state
	)

Value:

if (runtime_state->abort_on_error()) {                                \
    return Status(error_msg);                                           \
  } else {                                                              \
    runtime_state->LogError(error_msg);                                 \
    return Status::OK;                                                  \
  }

Definition at line 69 of file hdfs-parquet-scanner.cc.

Referenced by impala::HdfsParquetScanner::BaseColumnReader::ReadDataPage().

#define LOG_OR_RETURN_ON_ERROR	(	error_msg,
		runtime_state
	)

Value:

if (runtime_state->abort_on_error()) {                                \
    return Status(error_msg.msg());                                     \
  }                                                                     \
  runtime_state->LogError(error_msg);

Definition at line 77 of file hdfs-parquet-scanner.cc.

Referenced by impala::HdfsParquetScanner::AssembleRows(), and impala::HdfsParquetScanner::ValidateColumn().

Function Documentation

DEFINE_bool	(	convert_legacy_hive_parquet_utc_timestamps	,
		false	,
		"When	true,
		TIMESTAMPs read from files written by Parquet-MR(used by Hive) will""be converted from UTC to local time.Writes are unaffected."
	)